Result 16 of 127 [Go To Full Text] [Tips]
Aşkar, Petek; B.Akkoyunlu*Kolb Öğrenme Stili Envanteri*Eğitim ve Bilim. 87,1993. 37-47.
Title: Are Learning Approaches and Thinking Styles Related? A Study in Two Chinese Populations.
Subject(s): LEARNING strategies; THOUGHT & thinking -- Psychological aspects
Source: Journal of Psychology, Sep2000, Vol. 134 Issue 5, p469, 21p
Author(s): Zhang, Li-Fang; Sternberg, Robert J.
ARE LEARNING APPROACHES AND THINKING STYLES RELATED? A STUDY IN TWO CHINESE POPULATIONS
ABSTRACT. This article presents the results of an investigation of the construct validity of J. B. Biggs's (1987) theory of learning approaches and of R. J. Sternberg's (1988) theory of thinking styles in two Chinese populations. The study is also an examination of the nature of the relations between the two theories. University students from Hong Kong (n = 854) and from Nanjing, mainland China (n = 215), completed the Study Process Questionnaire (J. B. Biggs, 1992) and the Thinking Styles Inventory (R. J. Sternberg & R. K. Wagner, 1992). Results indicated that both inventories were reliable and valid for assessing the constructs underlying their respective theories among both Hong Kong and Nanjing university students. Results also showed that the learning approaches and thinking styles are related in the hypothesized ways: The surface approach was hypothesized to be positively and significantly correlated with styles associated with less complexity, and negatively and significantly correlated with the legislative, judicial, liberal, and hierarchical styles. The deep approach was hypothesized to be positively and significantly correlated with styles associated with more complexity, and negatively and significantly correlated with the executive, conservative, local, and monarchic styles. Implications of these relations are discussed.
THINKING AND LEARNING STYLES are sources of individual differences in academic performance that are related not to abilities but to how people prefer to use their abilities. There are alternative theories of thinking and learning styles, all of which share a common goal--that is, to explain individual differences in performance that are not explained by abilities (Sternberg, 1994, 1997).
Given the differences among theories of thinking and learning styles, a question that arises is whether such theories relate to different constructs, using a common root word "style," or rather if they are theories of the same construct but have different names for overlapping styles. Psychologists and educators need to understand whether the various theories--and the measures associated with them--provide insights into different constructs or the same constructs under different labels. Following this view, the primary goal of the present study was to verify the nature of the relations between Biggs's (1987, 1992) theory of approaches to learning and Sternberg's (1988, 1990, 1994, 1997) theory of mental self-government, a theory of thinking styles.
What, exactly, is a style? How does a style differ from an ability? A style is a preferred way of thinking or of doing things. A style is not an ability, but rather a preference in the use of the abilities people have. It is an interface between ability and personality (Sternberg, 1994, 1997).
Since the beginning of the cognitive-styles movement in the 1950s and early 1960s, different theories of thinking styles have been constructed. Because there are many more than we could address here (for more extensive reviews, see Grigorenko & Sternberg, 1995; Kogan, 1983; Sternberg, 1997), we review only selected theories.
Myers (1980; Myers & McCaulley, 1988) proposed a series of psychological types based on Jung's (1923) theory of types. According to Myers, there are 16 types, resulting from all possible combinations of (a) two ways of perceiving (sensing vs. intuiting), (b) two ways of judging (thinking vs. feeling), (c) two ways of dealing with self and others (being introverted vs. being extraverted), and (d) two ways of dealing with the outer world (judging vs. perceiving). Gregorc (1985) proposed four main types of styles, based on all possible combinations of two dimensions (concrete vs. abstract and sequential vs. random). Renzulli and Smith (1978) suggested various learning styles, with each corresponding to a method of teaching (e.g., projects, drill and recitation, and discussion), and Holland (1973) proposed six styles (realistic, investigative, artistic, social, enterprising, and conventional) that have been used as a basis for understanding career interests. Some other theories of styles are not general theories; rather, they are theories of specific aspects of cognitive-stylistic functioning (Grigorenko & Sternberg, 1995). For example, Kagan (1976) studied individual differences between impulsive and reflective persons, and Witkin (1978) examined the differences between field-independent and field-dependent individuals.
Recently, two theories have been proposed that are fairly general. One is Biggs's (1987, 1992) theory of students' approaches to learning, also known as the 3P model; the other is Sternberg's (1988, 1990, 1997) theory of mental self-government.
Biggs's Theory of Approaches to Learning
Adapted from Dunkin and Biddle's (1974) presage-process-product model, Biggs's model addresses those three components in the classroom. Presage concerns components before learning takes place; process pertains to components while learning is taking place; product pertains to outcomes after learning has taken place. In the present study we focus on the process of learning. According to the 3P model, there are three common approaches to learning: surface, which involves a reproduction of what is taught to meet the minimum requirements; deep, which involves a real understanding of what is learned; and achieving, which involves using a strategy that will maximize one's grades. Each approach is composed of two elements: motive and strategy (see Biggs, 1987, 1992, for a description of the Study Process Questionnaire [SPQ]). Motive describes why students choose to learn, whereas strategy describes how students go about their learning.
An alternative theory is that of Marton (e.g., Marton & Booth, 1997), who proposed surface and deep but not achieving strategies. A question in need of resolution, therefore, is whether the achieving style truly is distinguishable from the other two as a third style, or is a variant of one or both of them. Related work has been done by Entwistle and his colleagues (e.g., Entwistle, 1988, 1990; Entwistle, Koseki, & Politt, 1987; Entwistle & Marton, 1994), who have considered both the two-style and three-style models.
One of the instruments used to assess learning approaches among university students is the SPQ (Biggs, 1987, 1992), which was originally designed to assess the learning approaches of Canadian and Australian students. Many studies have been undertaken with the SPQ. Focusing on students' motives and strategies for learning, Biggs (1992) summarized major endeavors regarding the 3P model using the SPQ before 1992. These motives and strategies for learning have been examined in the following contexts: cross-cultural comparisons, the language medium of instruction, teaching/learning environments, student characteristics, professional and staff development, and factor structure and dimensionality of subscales.
More recent work examining learning approaches as defined by the 3P model have as their foci the investigation of the relationships between learning approaches and academic achievement (e.g., Albaili, 1995; Rose, Hall, Bolen, & Webster, 1996) and the construction of other versions of the SPQ (e.g., Albaili; Watkins & Murphy, 1994). Investigation of the factorial structure of the SPQ continues to be one of the major approaches to examining the instrument and its underlying 3P model (e.g., Bolen, Wurm, & Hall, 1994; Niles, 1995; O'Neil & Child, 1984). In addition, individual differences based on age and gender (e.g., Sadler-Smith & Tsang, 1998; Watkins & Hattie, 1981; Wilson, Smart, & Watson, 1996) also have been of major interest to scholars using the SPQ in their investigations of student learning.
In their study of the relationship between SPQ scores and overall grade point average (GPA) among 202 U.S. undergraduate students, Rose et al. (1996) found that only scores on the achieving approach contributed to prediction (negative correlation) of GPA. Albaili (1995), in his study of 246 United Arab Emirates undergraduate students, found that GPAs were negatively correlated with the surface approach and positively correlated with the deep and achieving approaches.
As mentioned earlier, the SPQ was originally constructed to measure Australian and Canadian university students' learning approaches. Other versions of the SPQ, however, also have been constructed. For example, in 1992, a Hong Kong version was established (Biggs, 1992). In 1994, when they studied Brunei university students, Watkins and Murphy came up with a simplified English as a Second Language (ESL) version and a Malay version. In 1995, Albaili established an Arabic version of the SPQ in his study of university students in the United Arab Emirates. All of the versions of the SPQ proved to be reliable and valid measures for assessing students' learning approaches.
The study of the validity of the SPQ has taken two forms. One is the examination of its internal structure. The other is the examination of the SPQ compared with other instruments. Many studies have had a focus on the examination of the internal structure of the SPQ. Although some studies supported Biggs's original argument that there are three factors in the SPQ, other studies have shown that there are only two factors. For example, in their study of a sample of U.S. university students' approaches to learning, Bolen et al. (1994) identified three factors--Surface Approach, Deep Approach, and Achieving Approach. Similarly, O'Neil and Child (1984), studying British university students, also identified three factors in the SPQ. However, in a study in Australia by Niles (1995) of overseas and Australian university students, and in Watkins and Dahlin's (1997) study of university students in Sweden, a two-factor model was identified (see also Entwistle, 1981; Marton & Booth, 1997). The two factors were Deep Approach and Surface Approach to learning, and the two Achieving subscales were split between the two factors.
Few researchers have investigated the relations between the SPQ and other instruments. We identified three such studies in the literature. A first study was conducted by Kember and Gow (1990), who administered both the SPQ and the Approaches to Studying Inventory (ASI; Entwistle, 1981) to Hong Kong university students. Although all three factors (Surface, Deep, and Achieving) appeared in the SPQ, only two factors (Deep and Achieving) appeared in the ASI. Surface learning was replaced by a factor labeled "narrow orientation" (Harper & Kember, 1989), which has been variously called "operation learning" by Watkins (1982, p. 80) and "disorganized study" by Ramsden and Entwistle (1981, p. 372).
A second study was carried out by Murray-Harvey (1994), who conducted a factor analysis on the Productivity Environmental Preference survey and the SPQ data collected from 400 Australian university students. Results indicated that the two inventories measure two quite different conceptualizations of student learning. It was concluded that learning approach is relatively stable over time and that learning style is not quite as stable.
A third study was conducted by Wilson et al. (1996), who also studied the relationship between the SPQ and the ASI. Analyzing the data collected from 283 Australian university students, the authors found significant correlations between the scales in the two inventories. They concluded that the two inventories measure similar constructs.
Age and gender are two of the variables that scholars have investigated in relation to the SPQ. Findings are, again, varied. For example, Sadler-Smith and Tsang (1998), studying British and Hong Kong university students, did not find any age or gender difference in the British sample; they did, however, observe an interaction of age and gender in their effects on deep and strategic (see Entwistle, 1981) approaches. That is, mature male students reported higher scores on the deep approach than did the non-mature male students; however, for female students, this pattern was reversed. Sadler-Smith and Tsang specified 23 years as the cutoff age between non-mature and mature participants.
By the same token, Watkins and Hattie (1981) also observed age and gender differences. They found that male students scored significantly higher on the scales measuring surface learning than did female students, whereas female students scored significantly higher on the scales measuring deep learning than did their male counterparts. They also found that older students scored significantly higher on the scales measuring deep learning than did their younger counterparts. On the contrary, Wilson et al. (1996) found no gender differences.
In summary, there is strong evidence that the SPQ is a reliable and valid instrument for assessing the learning approaches of university students, including Chinese university students.
Sternberg's Theory of Mental Self-Government
Sternberg's (1988, 1990, 1997) theory of mental self-government addresses people's thinking styles, which may be used in many settings, including university, home, and community. At the heart of this theory is the notion that people need somehow to govern or manage their everyday activities. There are many ways of doing so; whenever possible, people choose styles of managing themselves with which they are comfortable. Still, people are at least somewhat flexible in their use of styles and try with varying degrees of success to adapt themselves to the stylistic demands of a given situation. Thus, an individual with one preference in one situation may have a different preference in another situation. Moreover, styles may change with time and with life demands. Thinking styles are at least partly socialized (Sternberg, 1994, 1997), a fact that suggests that, to some extent, they can be modified by the environment in which people reside. As applied to individuals, the theory of mental self-government posits 13 thinking styles that fall along five dimensions of mental self-government: (a) functions, (b) forms, (c) levels, (d) scope, and (e) leanings.
Functions
As in government, there are three functions in human beings' mental self-government: legislative, executive, and judicial. An individual with a legislative style enjoys being engaged in tasks that require creative strategies. These individuals prefer to choose their own activities, or at least to do the activities chosen for them in their own way. An individual with an executive style is more concerned with implementation of tasks with set guidelines. Such an individual prefers more direction or guidance in structuring tasks. An individual with a judicial style focuses attention on evaluating the products of others' activities.
Forms
Also as in government, a human being's mental self-government takes four different forms: monarchic, hierarchic, oligarchic, and anarchic. An individual with a monarchic style enjoys being engaged in tasks that allow complete focus on one thing at a time. In contrast, an individual with a hierarchic style prefers to distribute attention to several tasks that are given priority according to their value to the individual in achieving his or her goals. An individual with an oligarchic style also likes to work on multiple activities in the service of multiple objectives, but may not enjoy setting priorities. Finally, an individual with an anarchic style enjoys working on tasks that allow flexibility as to what, where, when, and how one works, but he or she eschews systems of almost any kind.
Levels
As with governments, human beings' mental self-government functions at two different levels: local and global. An individual with a local style enjoys being engaged in tasks that require working with concrete details. In contrast, an individual with a global style prefers to pay more attention to the overall picture of an issue and to abstract ideas.
Scope
Mental self-government can deal with internal and external matters. An individual with an internal style enjoys being engaged in tasks that allow that individual to work independently. In contrast, an individual with an external style likes being engaged in tasks that allow for collaborative ventures with other people.
Leanings
Finally, in mental self-government, there are two leanings: liberal and conservative. An individual with a liberal style enjoys engaging in tasks that involve novelty and ambiguity, whereas an individual with a conservative style prefers adhering to the existing rules and procedures in performing tasks.
The theory of mental self-government has been operationalized through inventories, including the Thinking Styles Inventory (TSI; Sternberg & Wagner, 1992), which have been shown to be reliable and valid for U.S. and Hong Kong samples. Furthermore, results from such research have shown some value of the theory and have generated a number of implications for teaching and learning in educational settings. In the United States, Sternberg and Grigorenko conducted a series of studies. In one such study, Sternberg and Grigorenko (1995) reported significant relationships between teaching styles and grade taught, length of teaching experience, and subject area taught. Specifically, teachers teaching at lower grade levels were more legislative than teachers teaching at higher grade levels; complementarily, teachers teaching at lower grade levels were less executive than teachers at higher grade levels. It was shown that teachers with more teaching experience were more executive, local, and conservative than were those teachers with less teaching experience. Furthermore, it was found that humanities teachers were more liberal than were science teachers.
A second set of findings indicated significant relationships between students' learning styles and such demographic data as students' socioeconomic status (SES) and birth order (Sternberg & Grigorenko, 1995). Specifically, participants of higher SES status tended to score higher on the legislative style. Likewise, participants who were later-borns in their family scored higher on the legislative style than did participants who were earlier-borns. A third data set indicated that teachers inadvertently favored those students who had thinking styles similar to their own (Sternberg & Grigorenko). In a more recent study, Grigorenko and Sternberg (1997) found that certain thinking styles contribute significantly to prediction of academic performance over and above prediction of scores on ability tests. Their study also indicated that students with particular thinking styles fared better on some forms of evaluation than on others.
Three studies concerning the theory of mental self-government have been carried out in Hong Kong (Zhang, 1999; Zhang & Sachs, 1997; Zhang & Sternberg, 1998). These studies indicate that the thinking styles defined by Sternberg's theory also can be identified among university students in Hong Kong. The internal consistency reliabilities and validity data are generally satisfactory (see description in the Method section, under Inventories). Furthermore, results from these studies have suggested that students' thinking styles are statistically different based on such variables as age, sex, college class, teaching experience, college major, school subject taught, and travel experience. For example, male participants scored higher on the global style than did their female counterparts. Participants who had had more teaching experience (as measured by the length of teaching) and those who had had more travel experience scored higher on the creativity-promoting thinking styles, such as legislative and liberal. In our recent study (Zhang & Sternberg, 1998) of 622 Hong Kong university students, we found that thinking styles (as defined by the theory of mental self-government) could serve as reasonable predictors of academic achievement over and above self-rated abilities. For example, higher achievement was positively correlated with the use of conservative, hierarchic, and internal styles of thinking; yet, higher achievement was negatively correlated with the use of the legislative, liberal, and external styles of thinking.
Although both the SPQ and the TSI and their underlying theories have been well researched, the present study is the first to investigate the relationships among the scales in the two inventories and the connections between the two theories. In the present study, we examined the relations between the two theories and corresponding measures of styles in two Chinese populations--university students from Nanjing, mainland China, and university students from Hong Kong. The means to achieve this goal was to determine the reliability and validity of the SPQ and of the TSI, to examine the relations between the scales in the two inventories, and to determine whether the hypothesized relationships between the SPQ and the TSI exist among more than one sample. These two inventories were studied together because they are based on similar theoretical constructs. By nature, both Biggs's theory of learning approaches and Sternberg's theory of mental self-government concern two types of mental functioning and thus, two ways of processing information: more simple and more complex.
We proposed two sets of hypotheses, drawn in part on past work in the field by Beishuizen, Stoutjesdijk, and Van-Putten (1994), who studied the relation between cognitive levels of task accomplishment and deep versus surface processing of material. Beishuizen et al. expected deep-processing students to benefit from metacognitive support and surface-processing students to benefit from cognitive support. They found that students who processed at a surface level tended to benefit from cognitive support. Students who combined self-regulation with deep processing and students who combined external regulation with surface processing outperformed students who showed the opposite pairings of type of regulation with type of processing.
We expected students who take a surface approach to learning and those who use executive, monarchic, local, and conservative styles to be individuals who want to get things done with given structures, who do not want to make mistakes, and who want to "play it safe." We expected students who take a deep approach to learning and those who tend toward legislative, judicial, hierarchic, anarchic, global, and liberal styles to want to make up their own minds and use their own judgments in learning. We expected these students to want to work more in situations in which their creativity and imagination would be allowed free rein. Furthermore, we expected them to be less afraid of making mistakes.
Thus, we proposed the following: First, the surface approach should be positively and significantly correlated with styles associated with less complexity--executive, monarchic, local, and conservative styles. Complementarily, this approach should be negatively and significantly correlated with the legislative, judicial, liberal, and hierarchic styles. Second, the deep approach should be positively and significantly correlated with styles associated with more complexityNlegislative, judicial, hierarchic, anarchic, global, and liberal styles. Complementarily, this approach should be negatively and significantly correlated with the executive, conservative, local, and monarchic styles.
No specific predictions were made regarding the relations between the achieving approach subscales of the SPQ and the subscales of the TSI, because previous research (e.g., Niles, 1995; Watkins & Dahlin, 1997; Wong, Lin, & Watkins, 1996) has yielded conflicting results. In particular, the achieving motive and strategy subscales of the SPQ (which assess the achieving approach) may be either clustered with one of the two scales (Deep and Surface) or split between the two. In other words, like Marton and Booth's (1997) theory, Biggs's theory conceptually addresses two approaches to learning: deep and surface.
Method
Participants
Hong Kong sample. A total of 854 (362 male and 492 female) students were selected randomly from about 4,000 entering students at the University of Hong Kong during the orientation week of the fall semester of 1997. These participants were from all of the nine faculties (Architecture, Arts, Dentistry, Education, Engineering, Law, Medicine, Science, and Social Sciences) and the School of Business at the university. Of these students, 501 were in social sciences/humanities, 349 were in natural sciences, and 4 were not identifiable. Of all the participants, 702 were undergraduate freshmen, 66 were beginning to pursue their post-graduate certificates, and 86 were starting their education for a master's degree. The average age of the participants was 21 years; 66% were 19 years old or younger, 20% were between the ages of 20 and 25, and 14% were between 26 and 57 years of age. At the time the study was conducted, 535 of the participants were not holding any job, 110 were working full-time, and 198 were working part-time. Eleven did not indicate their employment status.
Nanjing sample. A total of 215 (114 male, 101 female) entering freshmen from two big universities in Nanjing, mainland China, participated in the study at the beginning of the fall semester of 1997. Ten teachers were trained in the administration of the questionnaires. Each of the 10 teachers informed his or her class about the nature of the study. Those students who were not willing to participate in the study were not required to participate. Those who volunteered (98% of the students) to participate were from several areas of study, including chemistry, computer science, education, finance, history, law, management, mathematics, medicine, and political science. Classified into the two broad fields of study, 126 were from social sciences/humanities and 89 were from natural sciences. The average age of the participants was 19 years, with a range from 15 to 23. In all, 75% of the participants were 19 years old or younger.
Inventories
Two inventories and a demographic questionnaire were used in the study. The first inventory was Biggs's SPQ (1992; Chinese version normed on Hong Kong university students). The second was Sternberg and Wagner's (1992) TSI. Both of the inventories were developed originally in English and were later translated and back-translated between Chinese and English.
The SPQ is a self-report questionnaire consisting of 42 items. This questionnaire has 6 subscales, with 7 items on each subscale. For each item, the respondents are asked to rate themselves on a 5-point scale anchored by 1 (never or only rarely true of you) and 5 (always or almost always true of you). The 6 subscales are Surface Motive, Surface Strategy, Deep Motive, Deep Strategy, Achieving Motive, and Achieving Strategy. Therefore, the 3 scales based on the three approaches to learning are Surface (Motive and Strategy), Deep (Motive and Strategy), and Achieving (Motive and Strategy). As described earlier, motive describes why students choose to learn, whereas strategy describes how students go about their learning.
As mentioned earlier, numerous studies involving the use of the SPQ have been conducted all over the world (e.g., Albaili, 1995; Bolen et al., 1994; Kember & Gow, 1990; Murray-Harvey, 1994; Watkins & Akande, 1992; Watkins & Regmi, 1990). Most of those studies have resulted in internal consistencies ranging from the mid .50s to the low or mid .70s for the 6 subscales and from the low .70s to the low .80s for the three scales (see Albaili, 1995, for details).
The TSI (Sternberg & Wagner, 1992) is a self-report questionnaire consisting of 65 items. The inventory has 13 subscales, with 5 items on each subscale. For each item, respondents are asked to rate themselves on a 7-point scale anchored by 1 (does not characterize you at all) and 7 (characterizes you extremely well). These 13 subscales correspond to the 13 thinking styles described in Sternberg's theory of mental self-government.
Sternberg and Wagner (1992) collected norms for various age groups on the long version of the TSI (which contains 104 items, 8 for each of the 13 subscales). For Sternberg and Wagner's college sample, subscale reliabilities ranged from .42 (monarchic) to .88 (external), with a median reliability of .78. In another study using the TSI, Sternberg (1994) found a five-factor model corresponding to the five dimensions of mental self-government described in his theory of thinking styles. These five factors accounted for 77% of the variance in the data.
The TSI also has been validated against instruments based on other theories of styles (e.g., Myers-Briggs Type Indicator, Gregorc's measure of mind styles), as well as a standard IQ test, the Scholastic Assessment Test (SAT), and GPA. Results from these construct-validity studies indicated that, among U.S. students, the TSI is a reliable and valid instrument for studying thinking styles as defined by the theory of mental self-government.
The TSI also has proved to be reliable and valid for identifying thinking styles of university students in Hong Kong. The statistics from three studies (Zhang, 1999; Zhang & Sachs, 1997; Zhang & Sternberg, 1998) conducted in Hong Kong are similar in magnitude to those obtained by Sternberg (1988, 1990, 1994, 1997). For example, the alpha coefficients in Sternberg's (1994) study ranged from .44 to .88; those in Zhang and Sachs's (1997) study ranged from .53 to .87 (from .46 to .89 in Zhang, 1999, and from .43 to .78 in Zhang & Sternberg, 1998). Although Zhang and Sachs's (1997) study extracted only three factors corresponding to the constructs in the theory of mental self-government, both Sternberg's (1994) and Zhang's (1999) studies extracted five factors (the former accounted for 77% of the variance and the latter, 78%). In these studies, each factor roughly corresponded to one of the five dimensions delineated in the theory. In our recent study (Zhang & Sternberg, 1998), the validity of the TSI was tested through an interscale correlation matrix. It was shown that the scales were, in general, correlated in the predicted directions. For example, the correlation between the executive and conservative styles was .63 (p < .001); that between the legislative and liberal styles was .41 (p < .001); and that between the internal and external styles was -.30 (p < .001).
Data Analysis
The following analyses were conducted both separately for men and women and for the sexes combined. The reliability of each of the 6 subscales in the SPQ and the 13 subscales in the TSI was estimated by Cronbach's alpha. The validity of each of the two inventories was examined through the relations shown among the subscales by its respective intercorrelation matrix. The relations between the two theories were examined via a correlation matrix, with the subscales of the SPQ providing one set of variables and those of the TSI providing another.
Results
In both the Hong Kong and Nanjing samples, t tests on the 6 subscales of the SPQ and the 13 subscales of the TSI resulted in a few pairs of statistically significant (p < .05) means for men and women. On a 5-point Likert-type scale (of the SPQ), the statistically significant mean differences were (a) .19 on Achieving Motive, (b) .11 on Deep Strategy, and (c) .11 on Surface Motive for the Hong Kong sample; and (a) .24 on Deep Motive and (b) .31 on Deep Strategy for the Nanjing sample. On a 7-point Likert-type scale, the statistically significant mean differences were (a) .12 on the legislative style, (b) .20 on the judicial style, (c) .15 on the global style, (d) .38 on the liberal style, and (e) .23 on the internal style for the Hong Kong sample; and (a) .39 on the legislative style, (b) .74 on the liberal style, and (c) .35 on the internal style for the Nanjing sample. In all cases, men scored higher than women. These differences, although statistically significant, were small in magnitude. Furthermore, none of the remaining statistical analyses conducted for men and women separately indicated significant gender differences. These analyses included (a) a correlational analysis on the 13 subscales of the TSI, (b) a factor analysis on the SPQ, and (c) a correlational analysis between the subscales of the two inventories. Because of the lack of gender differences in the previous three statistical procedures, the results are reported with combined gender analyses.
Subscale Reliabilities for the SPQ
The alpha estimates of internal consistency for the Deep and Achieving Motive and Strategy subscales for both the Hong Kong and Nanjing samples are in line with those obtained by Biggs (1987) for his Australian norming sample (see Table 1). The findings also are in line with estimates obtained by other authors, such as Watkins and Dahlin (1997), in their study of Swedish university students. However, the alpha coefficients of the Surface Motive and Surface Strategy subscales are higher for the samples in this study (in the mid .60s and low .70s) than for the aforementioned Australian and Swedish samples (low .40s for Surface Motive and mid .50s for Surface Strategy). The alpha coefficients for the 6 subscales ranged from .65 to .80, with a median of .73, for the Hong Kong students, and from .64 to .74, with a median of .70, for the Nanjing students. The alpha coefficients for the Surface, Deep, and Achieving scales were .80, .82, and .83, respectively, for the Hong Kong sample, and .78, .78, and .76, respectively, for the Nanjing sample. These alpha coefficients were considered sufficiently high to allow further statistical analyses.
Subscale Reliabilities for the TSI
The magnitudes of the estimates of internal consistency for the TSI for the Hong Kong sample and the Nanjing sample were similar (see Table 2). Furthermore, these results are comparable to those obtained by Sternberg (1994) in his study of U.S. participants, by Zhang and Sachs (1997), and by Zhang (1999). Notice that 3 subscales were less internally consistent in those respective studies. These subscales were local, monarchic, and anarchic. Even so, the estimates of internal consistency obtained in the present study were considered to be adequate to allow further statistical analyses.
Subscale Intercorrelations for the SPQ
In accordance with Biggs's theory, we predicted that the Deep Motive and Deep Strategy subscales would be significantly negatively correlated with the Surface Motive and Surface Strategy subscales. Furthermore, as mentioned earlier, no prediction was made on the Achieving Motive and Achieving Strategy subscales because these subscales may be positively and significantly correlated with either the Deep subscales or the Surface subscales, or split between the two (Watkins & Dahlin, 1997; Wong et al., 1996). The predictions were fully supported by the results from the Nanjing sample. Results from the Hong Kong sample, however, did not support these predictions, in that three of the correlations were in the direction opposite from what was expected from the theory. These correlation coefficients were (a) Surface Motive with Deep Motive (r = .17, p < .01), (b) Surface Motive with Deep Strategy (r = .16, p < .01), and (c) Surface Strategy with Deep Strategy (r = .10, p < .01). These three correlations indicate that students who took a surface approach to learning also tended to take a deep approach, a pattern not consistent with Biggs's theory, according to which surface subscales presumably should be negatively correlated with the deep subscales.
Because of the presence of the three unexpected correlations, we conducted a principal-axis factor analysis with a varimax rotation, to examine further the validity of the SPQ for the Hong Kong sample. A scree test (Cattell, 1966) indicated that a two-factor solution would be appropriate. Furthermore, there were two factors with eigenvalues greater than 1. Thus, a two-factor model was retained (see Table 3 for details). The analysis yielded a clear factor for a deep approach (factor loadings: .86 for Deep Motive; .89 for Deep Strategy; .76 for Achieving Strategy) and one for a surface approach (factor loadings: .88 for Surface Motivation; .87 for Surface Strategy; .71 for Achieving Motive). The Achieving Motive and Achieving Strategy subscales thus were split between the Deep and Surface subscales, as expected (Niles, 1995; Watkins & Dahlin, 1997; Wong et al., 1996).
A principal-axis factor analysis with a varimax rotation also was conducted with the Nanjing participants' data to confirm the validity of the SPQ for the Nanjing sample. Results from this analysis revealed the same two factors as those for the Hong Kong data (see Table 3). The first factor corresponded to the deep approach (factor loadings: .81 for Deep Motive; .81 for Deep Strategy; .77 for Achieving Strategy). The second factor corresponded to the surface approach (factor loadings: .86 for Surface Motive; .86 for Surface Strategy; .61 for Achieving Motive).
Consequently, the SPQ, when conceptualized as a two--rather than three-factor instrument, appeared to be valid for assessing the learning approaches of the two Chinese samples. These results from factor analyses supported not only previous studies using the SPQ (e.g., Niles, 1995; Watkins & Dahlin, 1997; Wong et al., 1996) but also Marton and Booth's (1997) findings regarding learning approaches.
Subscale Intercorrelations for the TSI
In general, for both the Hong Kong and Nanjing samples, the correlations among the 13 subscales were in the direction predicted by the theory of mental self-government (see Table 4 for details). Some of the examples are (a) Executive with Conservative (r = .65 for Hong Kong; r = .66 for Nanjing), (b) Legislative with Liberal (r = .42 for Hong Kong; r = .50 for Nanjing), (c) Conservative with Liberal (r = -.10 for Hong Kong; r = -.42 for Nanjing), (d) Global with Local (r = .08 for Hong Kong; r = -.35 for Nanjing), and (e) Internal versus External (r = -.23 for Hong Kong; r = -.28 for Nanjing). Except for the correlation between Global and Local for Nanjing, these correlations were significant at the .01 level. Furthermore, the magnitudes of these correlations were generally stronger for the Nanjing sample than for the Hong Kong sample.
Correlations Among Subscales in the Two Inventories
In general, the hypotheses were supported by the data from both samples (see Table 5). The majority of the correlations were in the expected directions. Some of the examples are (a) Surface Motive with executive style (r = .24 for Hong Kong; r = .23 for Nanjing), (b) Surface Strategy with liberal style (r = -.03 for Hong Kong; r = -.31 for Nanjing), (c) Deep Motive with judicial style (r = .40 for Hong Kong; r = .31 for Nanjing), and (d) Surface Strategy with judicial style (r = -.13 for Hong Kong; r = -.11 for Nanjing). These correlations varied from being statistically insignificant to being significant at the .01 level. Achieving subscales were inconsistently correlated positively with either the Deep or the Surface subscales. These correlations indicated that students who took a surface approach to learning tended to use an executive thinking style, but not judicial or liberal thinking styles. In addition, students who took a deep approach to learning tended to use the judicial thinking style.
There were a few correlations that clearly did not support the predictions. First, for the Hong Kong sample, the correlation between Deep Strategy and executive style was significantly positive (r = .18, p < .001), meaning that the Hong Kong students in this sample who used a deep strategy to learn also preferred using an executive thinking style. Second, our prediction about the relations between learning approach subscales and the global and local styles were only partially supported (see Table 5). Results of this study suggested that regardless of their level of mental functioning (global or local), students could take either a deep or surface approach to learning. Finally, all learning approach subscales were positively and significantly correlated with the monarchic style, which probably means that students with a monarchic thinking style may take either a deep or a surface approach to learning. These unexpected correlations were mostly from the Hong Kong sample, however. These results perhaps can be explained by Pask's (1976) concept of the "versatile learner." For example, the deep learners in Hong Kong may be creative (using the legislative and liberal styles) in their learning; meanwhile, they may also follow closely their teachers' instructions (using the executive and conservative styles).
Discussion
The major goal of this study was to establish the relations between the constructs in Biggs's theory of learning approaches and Sternberg's theory of thinking styles in two Chinese populations. Results indicated that the two inventories were reliable and valid (there are two factors in the SPQ--Deep Approach and Surface Approach) for assessing the underlying theoretical constructs for these two populations and that the subscales in the two inventories were related in largely predicted ways. Our study suggests that the SPQ and the TSI measure similar constructs. Students who reported taking a surface approach to learning preferred using executive, local, and conservative thinking styles (which are more traditional, norm favoring, and task oriented), whereas students who reported taking a deep approach to learning preferred using legislative, judicial, and liberal thinking styles (which are more creative, norm questioning, and meaning seeking). Although most of the correlations between the scales of the two inventories were low, they were statistically significant. In addition, these results both supported our own hypotheses (based on the study of Beishuizen et al., 1994) about the relationships between the two inventories and confirmed previous research findings of similar studies (e.g., Wilson et al., 1996). Therefore, we believe that these correlations, although weak, revealed true relationships between the two inventories.
The contributions of this study may be considered from two perspectives: research and practice.
From a research viewpoint, the results of this study have enhanced our knowledge about theories of styles. The question raised earlier was whether theories of styles are different theories of different things, using a common root word ("style") or theories of the same thing but with different names for overlapping styles. Sternberg (1997) suggested that alternative theories of styles cover roughly similar attributes, but with different labels. The relations indicated by the subscales in the two inventories used in this study suggest that Biggs's (1987, 1992) theory of students' approaches to learning and Sternberg's (1988, 1990, 1994, 1997) theory of mental self-government cover similar but not identical ground, with different names for overlapping styles. This finding is also consistent with previous construct-validity studies of measures derived from the theory of mental self-government (e.g., compared with the Myers-Briggs Type Indicator and with Gregorc's measures of mind styles; Sternberg, 1994). Of further theoretical importance is the finding that the two-learning-style approach of Marton and Booth (deep and surface; 1997) appears to capture better the structure of the data than does the three learning-style approach of Biggs (deep, surface, achieving).
From a practical viewpoint, we believe that there are three implications. First, both teachers and students should be aware that people approach learning differently and use their abilities in a variety of ways.
Second, but equally important, teachers and students should understand the relations between learning approaches and thinking styles. An understanding of the existence of different learning approaches and different thinking styles can assist teachers in using several measures to facilitate effective learning. Teachers should try to teach via a variety of styles so that all students, regardless of their preferred ways of dealing with learning tasks, can benefit from teachers' instructions. Alternatively, because learning styles can be modified (Saracho, 1993; Sternberg, 1988, 1990, 1997), awareness of the different learning styles can make students more in tune with how they usually approach their learning tasks and help them identify their preferred, as well as their nonpreferred, learning styles. As a result, students may learn not only how to capitalize on their strengths and compensate for their weaknesses but also how to adapt to those learning environments with which their own styles may not be compatible.
Third, a teacher can use different assessment techniques to allow for different learning and thinking styles (Sternberg, 1988, 1990, 1994, 1997). Recognizing this fact, Biggs (1995) coined the term "backwash effect." In particular, he argued that assessment drives the ways in which students learn and think, the content of the curriculum, and how teachers teach. Therefore, among other things, assessment links Biggs's and Sternberg's theories--it has a common impact on both learning approaches and thinking styles. Learning approaches and thinking styles as implemented at a given time may vary as a function of the assessment measures used. For example, if student performance is measured by a multiple-choice test, students may tend to take a surface approach to learning and use executive, conservative, internal, and local thinking styles. In contrast, if student performance is assessed by a group project, it is more likely that students will take a deep approach to learning and use such thinking styles as judicial, legislative, liberal, and external.
An awareness of the interrelations between the two theories also can be helpful in teachers' efforts toward the enhancement of effective learning. Each of the learning approaches discussed by Biggs (1987, 1992), as mentioned earlier, contains two concepts, motivation and strategy. Students' learning motivations, learning strategies, and thinking styles are intertwined. Given this intertwining, teachers can facilitate the students' efforts to be flexible in their implementations of styles. For example, teachers may wish to motivate students to take a deep approach to learning more important material, but a surface approach to learning less important material. The significant positive correlations manifested in this study indicate that when students are deeply motivated to learn, they will think critically and creatively, and certainly, also will use a deep strategy in performing their learning tasks. Alternatively, teachers may allow for different thinking styles by using the aforementioned strategies, such as teaching about styles, instructing in different ways, and using varied assessment tools.
TABLE 1 Study Process Questionnaire Subscales: Means, Standard Deviations, and Alpha Coefficients
Legend for Chart:
A - Subscale
B - Items
C - M HK
D - M NJ
E - SD HK
F - SD NJ
G - alpha HK
H - alpha NJ
A B C D E
F G H
Achieving Motive 3, 9, 15, 21, 27, 33, 39 3.04 3.51 .74
.73 .78 .72
Achieving Strategy 6, 12, 18, 24, 30, 36, 42 3.16 3.49 .69
.66 .80 .73
Deep Motive 2, 8, 14, 20, 26, 32, 38 3.26 3.42 .58
.64 .65 .64
Deep Strategy 5, 11, 17, 23, 29, 35, 41 3.33 3.60 .58
.62 .75 .74
Surface Motive 1, 7, 13, 19, 25, 31, 37 2.96 2.80 .66
.73 .68 .67
Surface Strategy 4, 10, 16, 22, 28, 34, 40 2.74 2.47 .60
.58 .70 .64
Note. HK = Hong Kong. NJ = Nanjing. Hong Kong n = 854. Nanjing
n = 215.
TABLE 2 Thinking Styles Inventory Subscales: Means, Standard Deviations, and Alpha Coefficients
Legend for Chart:
A - Subscale
B - Items
C - M HK
D - M NJ
E - SD HK
F - SD NJ
G - alpha HK
H - alpha NJ
A B C D E
F G H
Legislative 5, 10, 14, 32, 49 4.91 5.45 .81
.86 .71 .65
Executive 8, 11, 12, 31, 39 4.91 4.68 .79
.97 .66 .61
Judicial 20, 23, 42, 51, 57 4.67 4.87 .85
.92 .72 .62
Global 7, 18, 38, 48, 61 4.28 4.59 .76
.95 .58 .60
Local 1, 6, 24, 44, 62 4.35 4.35 .72
.90 .48 .49
Liberal 45, 53, 58, 64, 65 4.20 4.74 .94
1.0 .80 .81
Conservative 13, 22, 26, 28, 36 4.50 3.96 .86
1.12 .72 .74
Hierarchical 4, 19, 25, 33, 56 4.87 5.01 .88
1.06 .76 .78
Monarchic 2, 43, 50, 54, 60 4.59 4.98 .76
.86 .48 .43
Oligarchic 27, 29, 30, 52, 59 4.57 4.62 .80
.95 .64 .66
Anarchic 16, 21, 35, 40, 47 4.45 4.48 .73
.76 .44 .13
Internal 9, 15, 37, 55, 63 4.35 4.71 .99
.97 .77 .67
External 3, 17, 34, 41, 46 4.83 5.12 .89
1.06 .74 .72
Note. HK = Hong Kong. NJ = Nanjing. Hong Kong n = 854.
Nanjing n = 215.
TABLE 3 Oblimin-Rotated Two-Factor Model for the Study Process Questionnaire
Legend for Chart:
A - Subscale/Item
B - Hong Kong Factor 1
C - Hong Kong Factor 2
D - Nanjing Factor 1
E - Nanjing Factor 2
A B C D E
Surface Motive -.04 .89 -.12 .86
Surface Strategy -.10 .89 -.10 .86
Deep Motive .88 -.07 .81 -.04
Deep Strategy .90 -.07 .82 -.16
Achieving Motive .32 .67 .50 .60
Achieving Strategy .74 .15 .77 .04
% of variance 48.2 24.7 36.5 31.0
Cumulative % 48.2 72.9 36.5 67.5
Eigenvalue 2.89 1.48 2.19 1.86
Note. Hong Kong n = 854. Nanjing n = 215.
TABLE 4 Interscale Pearson Correlation Matrix for 13 Subscales of the Thinking Styles Inventory
Legend for Chart:
A - Subscale
B - 1
C - 2
D - 3
E - 4
F - 5
G - 6
H - 7
I - 8
J - 9
K - 10
L - 11
M - 12
N - 13
A B C D E F G H
I J K L M N
1. Legislative -.09 .34 .24 .06 .50 -.14
.22 .10 .02 .22 .54 -.10
2. Executive .33 .05 .04 .34 -.20 .66
.23 .31 .31 .22 -.01 .29
3. Judicial .44 .20 .18 .13 .51 -.11
.37 .27 .11 .29 .20 .22
4. Global .36 .25 .34 -.35 .20 .06
.18 .24 .11 .13 .24 .03
5. Local .22 .34 .32 .08 .09 .30
.15 .17 .22 .24 .04 .17
6. Liberal .42 .03 .52 .37 .26 -.42
.33 .19 .05 .33 .37 .07
7. Conservative .23 .65 .05 .16 .33 -.10
-.05 .21 .29 .12 -.08 .15
8. Hierarchical .30 .32 .45 .18 .41 .28 .20
.23 .22 .33 .26 .14
9. Monarchic .40 .42 .28 .34 .35 .25 .41
.30 .31 .31 .20 .13
10. Oligarchic .21 .41 .18 .26 .33 .14 .44
.18 .35 .30 -.06 .37
11. Anarchic .35 .25 .33 .24 .39 .34 .25
.34 .31 .33 .27 .23
12. Internal .64 .21 .34 .28 .27 .40 .16
.23 .34 .05 .23 -.28
13. External .07 .21 .35 .20 .28 .21 .07
.36 .15 .31 .35 -.23
Note. Numbers above the diagonal are for the Nanjing sample.
Numbers below the diagonal are for the Hong Kong sample. Hong
Kong n = 854. Nanjing n = 215.
TABLE 5 Pearson Correlation Matrix for the Subscales in the Study Process Questionnaire and Thinking Styles Inventory
Legend for Chart:
A - Subscale
B - SM HK
C - SM NJ
D - DM HK
E - DM NJ
F - AM HK
G - AM NJ
H - SS HK
I - SS NJ
J - DS HK
K - DS NJ
L - AS HK
M - AS NJ
A B C D E
F G H I
J K L M
Legislative .05 -.09 .28(*) .24(*)
.21(*) .20 -.02 -.12
.26(*) .33(*) .10(*) .02
Executive .24(*) .23(*) .17(*) .08
.20(*) .20 .26(*) .34(*)
.18(*) -.04 .20(*) .20
Judicial -.00 -.02 .40(*) .31(*)
.17(*) .15 -.13(*) -.11
.38(*) .49(*) .26(*) .18
Global .17(*) .05 .24(*) .04
.18(*) .13 .13(*) .02
.25(*) .13 .13(*) .00
Local .17(*) .18 .24(*) .15
.21(*) .14 .17(*) .23(*)
.26(*) .10 .30(*) .23(*)
Liberal .07 -.15 .37(*) .31(*)
.20(*) .08 -.03 -.31(*)
.37(*) .53(*) .19(*) .18
Conservative .25(*) .36(*) .07 .00
.19(*) .19 .36(*) .47(*)
.07 -.16 .19(*) .07
Hierarchical -.01 -.13 .32(*) .35(*)
.13(*) .23(*) -.04 -.14
.36(*) .39(*) .39(*) .49(*)
Monarchic .22(*) .20 .28(*) .23(*)
.26(*) .30(*) .22(*) .18
.24(*) .21 .29(*) .31(*)
Oligarchic .18(*) .23(*) .13(*) .23(*)
.10 .24(*) .19(*) .23
.13(*) .14 .12(*) .25(*)
Anarchic .04 .14 .25(*) .26(*)
.10 .28(*) .08 .08
.24(*) .27(*) .18(*) .30(*)
Internal .07 -.02 .24(*) .13
.24(*) .36(*) .05 -.02
.20(*) .30(*) .07 .10
External .02 -.02 .22(*) .07
.02 -.06 -.02 .02
.24(*) .09 .20(*) .22(*)
Note. HK = Hong Kong. NJ = Nanjing. Hong Kong n = 854. Nanjing
n = 215. SM = Surface Motivation. DM = Deep Motivation.
AM = Achieving Motivation. SS = Surface Strategy. DS = Deep
Strategy. AS = Achieving Strategy.
(*) p < .01.
REFERENCES
Albaili, M. A. (1995). An Arabic version of the Study Process Questionnaire: Reliability and validity. Psychological Reports, 77, 1083-1089.
Beishuizen, J., Stoutjesdijk, E., & Van-Putten, K. (1994). Studying textbooks: Effects of learning styles, study tasks, and instructions. Learning and Instruction, 4, 151-174.
Biggs, J. B. (1987). Student approaches to learning and studying. Hawthorn, Australia: Australian Council for Educational Research.
Biggs, J. B. (1992). Why and how do Hong Kong students learn? Using the Learning and Study Process Questionnaires (Education Paper No. 14). Hong Kong: Faculty of Education, The University of Hong Kong.
Biggs, J. B. (1995). Assessment of learning. In J. Biggs & D. Watkins (Eds.), Classroom learning: Educational psychology for the Asian teacher (pp. 167-191). Singapore: Prentice-Hall.
Bolen, L. M., Wurm, T. R., & Hall, C. W. (1994). Factorial structure of the Study Process Questionnaire. Psychological Reports, 75, 1235-1241.
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276.
Dunkin, M. J., & Biddle, B. J. (1974). The study of teaching. New York: Holt, Rinehart & Winston.
Entwistle, N. (1981). Styles of learning and teaching. New York: Wiley.
Entwistle, N. (1988). Motivation and learning strategies. Educational and Child Psychology, 59, 326-339.
Entwistle, N. (1990). Student learning and classroom environment. London: Falmer.
Entwistle, N., Koseki, B., & Politt, A. (1987). Measuring styles of learning and motivation. European Journal of Psychology of Education, 2, 183-203.
Entwistle, N., & Marton, F. (1994). Knowledge objects: Understandings constituted through intensive academic study. British Journal of Educational Psychology, 64, 161-178.
Gregorc, A. F. (1985). Inside styles: Beyond the basics. Maynard, MA: Gabriel Systems.
Grigorenko, E., & Sternberg, R. J. (1995). Thinking styles. In D. Saklosfske & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 205-229). New York: Plenum.
Grigorenko, E., & Sternberg, R. J. (1997). Styles of thinking, abilities, and academic performance. Exceptional Children, 63, 295-312.
Harper, G., & Kember, D. (1989). Interpretation of factor analyses from the Approaches to Studying Inventory. British Journal of Educational Psychology, 59, 66-74.
Holland, J. L. (1973). Making vocational choices: A theory of careers. Englewood Cliffs, NJ: Prentice-Hall.
Jung, C. (1923). Psychological types. New York: Harcourt Brace.
Kember, D., & Gow, D. (1990). Cultural specificity of approaches to study. British Journal of Educational Psychology, 60, 356-363.
Kagan, J. (1976). Commentary on reflective and impulsive children: Strategies of information processing underlying differences in problem solving. Monographs of the Society for Research in Child Development, 41(No. 5, Serial No. 168).
Kogan, N. (1983). Stylistic variation in childhood and adolescence: Creativity, metaphor, and cognitive styles. In P. H. Mussen (Ed.), J. H. Flavell & E. M. Markman (Vol. Eds.), Handbook of child psychology: Vol. 3. Cognitive development (4th ed., pp. 630-706). New York: Wiley.
Marton, F., & Booth, S. A. (1997). Learning and awareness. Mahwah, NJ: Erlbaum.
Murray-Harvey, R. (1994). Learning styles and approaches to learning: Distinguishing between concept and instruments. British Journal of Educational Psychology, 64, 373-388.
Myers, I. B. (1980). Gifts differing. Palo Alto, CA: Consulting Psychologists Press.
Myers, I. B., & McCaulley, M. H. (1988). Manual: A guide to the development and use of the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press.
Niles, F. S. (1995). Cultural differences in learning motivation and learning strategies: A comparison of overseas and Australian students at an Australian university. International Journal of Intercultural Relations, 19, 369-385.
O'Neil, M. J., & Child, D. (1984). Biggs' SPQ: A British study of its internal structure. British Journal of Educational Psychology, 54, 228-234.
Pask, G. (1976). Styles and strategies of learning. British Journal of Educational Psychology, 46, 128-148.
Ramsden, P., & Entwistle, N. J. (1981). Effects of academic departments on students' approaches to studying. British Journal of Educational Psychology, 51, 368-383.
Renzulli, J. S., & Smith, L. H. (1978). Learning Styles Inventory. Mansfield Center, CT: Creative Learning Press.
Rose, R. J., Hall, C. W., Bolen, L. M., & Webster, R. E. (1996). Locus of control and college students' approaches to learning. Psychological Reports, 79, 163-171.
Sadler-Smith, E., & Tsang, F. (1998). A comparative study of approaches to studying in Hong Kong and the United Kingdom. British Journal of Educational Psychology, 68, 81-93.
Saracho, O. N. (1993). Sociocultural perspectives in the cognitive styles of young students and teachers. Early Child Development and Care, 84, 1-17.
Sternberg, R. J. (1988). Mental self-government: A theory of intellectual styles and their development. Human Development, 31, 197-224.
Sternberg, R. J. (1990). Metaphors of mind: Conceptions of the nature of intelligence. New York: Cambridge University Press.
Sternberg, R. J. (1994). Thinking styles: Theory and assessment at the interface between intelligence and personality. In R. J. Sternberg & P. Ruzgis (Eds.), Personality and intelligence (pp. 169-187). New York: Cambridge University Press.
Sternberg, R. J. (1997). Thinking styles. New York: Cambridge University Press.
Sternberg, R. J., & Grigorenko, E. (1995). Styles of thinking in the school. European Journal for High Ability, 6, 201-219.
Sternberg, R. J., & Wagner, R. K. (1992). Thinking Styles Inventory (unpublished test). New Haven, CT: Yale University.
Watkins, D. (1982). Identifying the study process dimensions of Australian university students. Australian Journal of Education, 26, 76-85.
Watkins, D., & Akande, A. (1992). Assessing the approaches to learning of Nigerian students. Assessment and Evaluation in Higher Education, 17, 11-20.
Watkins, D., & Dahlin, B. (1997). Assessing study approaches in Sweden. Psychological Reports, 81, 131-136.
Watkins, D., & Hattie, J. (1981). The learning processes of Australian university students: Investigations of contextual and personological factors. British Journal of Educational Psychology, 51, 384-393.
Watkins, D., & Murphy, J. (1994). Modifying the Study Process Questionnaire for students learning English as a second language: Comparisons of reliability and factor structure. Psychological Reports, 74, 1023-1026.
Watkins, D., & Regmi, M. (1990). An investigation of the approach to learning of Nepalese tertiary students. Higher Education, 20, 459-469.
Wilson, K. L., Smart, R. M., & Watson, R. J. (1996). Gender differences in approaches to learning in first year psychology students. British Journal of Educational Psychology, 66, 59-71.
Witkin, H. A. (1978). Cognitive styles in personal and cultural adaptation: The 1977 Heinz Werner lectures. Worcester, MA: Clark University Press.
Wong, N. Y., Lin, W. Y., & Watkins, D. (1996). Cross-cultural validation of models of student learning. Educational Psychology, 16, 317-327.
Zhang, L. F. (1999). Further cross-cultural validation of the theory of mental self-government. The Journal of Psychology, 133, 165-181.
Zhang, L. F., & Sachs, J. (1997). Assessing thinking styles in the theory of mental self-government: A Hong Kong validity study. Psychological Reports, 81, 915-928.
Zhang, L. F., & Sternberg, R. J. (1998). Thinking styles, abilities, and academic achievement among Hong Kong university students. Educational Research Journal, 13, 41-62.
Received July 6, 1999
Research for this project was supported in part by the Committee on Research and Conference Grants as administered by The University of Hong Kong.
Preparation of this article was supported in part under the Javits Act Program (Grant No. R206R50001) as administered by the Office of Educational Research and Improvement, U.S. Department of Education. Grantees undertaking such projects are encouraged to express freely their professional judgment. This article, therefore, does not necessarily represent the position or policies of the Office of Educational Research and Improvement or the U.S. Department of Education, and no official endorsement should be inferred.
Address correspondence to Li-fang Zhang, Department of Education, The University of Hong Kong, Pokfulam Road, Hong Kong; lfzhang@hkucc.hku.hk (e-mail).
~~~~~~~~
By Li-Fang Zhang, Department of Education The University of Hong Kong and Robert J. Sternberg, Department of Psychology Yale University
--------------------------------------------------------------------------------
Copyright of Journal of Psychology is the property of Heldref Publications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: Journal of Psychology, Sep2000, Vol. 134 Issue 5, p469, 21p.
Item Number: 3644717 Result 16 of 127 [Go To Full Text] [Tips]
Title: The psychometric properties of the Learning Style Inventory and the Learning Style Questionnaire: Two normative measures of learning styles.
Subject(s): LEARNING strategies; LEARNING ability
Source: South African Journal of Psychology, Jun2000, Vol. 30 Issue 2, p44, 9p, 14 charts, 1 diagram
Author(s): Pickworth[*], Glynis E.; Schoeman, Willem J.
THE PSYCHOMETRIC PROPERTIES OF THE LEARNING STYLE INVENTORY AND THE LEARNING STYLE QUESTIONNAIRE: TWO NORMATIVE MEASURES OF LEARNING STYLES
David Kolb has provided a detailed, useful and widely accepted theory of experiential learning and learning styles. He developed the Learning Styles Inventory (LSI) to assess four learning abilities and four learning styles. Kolb's work is viewed favourably for establishing the existence of individual differences in learning styles, but the major criticism against his work is focused on his method of measuring learning styles and more specifically on the psychometric properties of the LSI. The LSI is an ipsative instrument and the limitations placed on the statistical analysis of data of ipsative measures makes it inappropriate for reliability and validity evaluation of the instrument. In this study the psychometric properties of two normative measures of learning styles, a normative version of the LSI (referred to as the LSI-Likert) and the Learning Style Questionnaire (LSQ), are investigated. A review of the literature on the LSI is presented and the development of normative versions of the LSI is reviewed. First-year university students registered for either a science or human sciences degree completed the two normative instruments. The internal reliability of the four learning ability scales was determined using alpha coefficient. The internal reliability of the LSI-Likert and LSQ was found to be relatively high. The presence of a response bias for both instruments was suspected. It appeared that the LSI-Likert was more successful than the LSQ in differentiating learning abilities and styles in the sample used. Item factor analysis demonstrated two bipolar factors in line with Kolb's theory for the LSQ. The four-factor solution for the LSI-Likert produced four factors which to some extent represented the four learning abilities.
* To whom correspondence should be addressed.
Kolb's theory of experiential learning
The experiential learning movement emerged through the theories and work of John Dewey, Kurt Lewin and Jean Piaget. The work of these three theorists form the foundation of Kolb's theory of experiential learning (Hickcox, 1990). According to Kolb (1984) learning is a continuous process through which knowledge is derived from, and modified through, testing out the experiences of the learner. Kolb also postulated that learning requires the resolution of conflicts between dialectically opposed modes of adaptation to the world. On the prehension (perceiving) dimension the process of apprehension (concrete experience) opposes the process of comprehension (abstract conceptualisation). Kolb referred in this regard to research on brain hemisphere dominance that provides evidence that "there are two distinct, coequal, and dialectically opposed ways of understanding the world" (Kolb, 1984, p. 48), the right-brain mode corresponding to apprehension and the left-brain mode corresponding to comprehension. On the transformation (processing) dimension the process of intention (reflective observation) opposes the process of extension (active experimentation). Kolb stated that Carl Jung's concepts of introversion (intention) and extraversion (extension) best describe this transformation dimension. Learning results from the resolution of conflicts between involvement in new experiences versus conceptualising, and between acting versus reflecting.
The process of experiential learning is described as a four-stage cycle involving four learning abilities: Concrete Experience (CE), Reflective Observation (RO), Abstract Conceptualization (AC), and Active Experimentation (AE) (see Figure 1). It is theorized that one learns best by going through the CE, RO, AC, AE sequence of the cycle and that people learn more effectively as they develop learning abilities in their areas of weakness. In the experiential learning process concrete experience is followed by observation and reflection, leading to the formation of abstract concepts that result in hypotheses to be tested in future actions and this in turn leads to new experiences. The learning cycle is continuously recurring and is directed by individual needs and goals (Kolb, Rubin & McIntyre, 1984).
The learning abilities are represented on a two dimensional learning styles plane by two bipolar dimensions, one a vertical axis running from CE to AC, and the other a horizontal axis running from AE to RO. The four quadrants formed by the intersection of the two bipolar axes represent the four learning styles derived from the combination of two preferred learning abilities: Diverger (CE and RO), Assimilator (AC and RO), Converger (AC and AE), and Accommodator (CE and AE). (See Figure 1). Kolb (1984) took a contextualist view of learning styles and stated that "psychological types or styles are not fixed traits but stable states" (p. 63). These stable states or enduring patterns of individual human behaviour arise from consistent patterns of transactions between the person and the environment. However, an individual can adjust their learning style according to the demands of the task at hand.
Sugarman (1985) pointed out that as Kolb's theory combines a theory of learning and a theory of learning styles there are at least three components that must be addressed in an evaluation of his work:
(a) establishing the existence of individual differences in learning styles;
(b) effectively measuring these differences, if they are found to exist; and
(c) validating the cyclical model of learning.
Kolb's work is viewed favourably for aspects (a) and (c), but the major criticism against his work is focused on his method of measuring learning styles and more specifically on the psychometric properties of the Learning Styles inventory (LSI). in the following section literature reporting on the psychometric properties of the LSI is summarized.
Assessment of learning styles: The Learning Styles Inventory (LSI)
The LSI was developed by Kolb and takes the form of a self-description, self-scoring test that aims to help an individual to identify their relative emphasis on the four learning abilities within the learning cycle (CE, RO, AC and AE) as well as their predominant learning style (Diverger, Assimilator, Converger or Accommodator).
The LSI-1976
According to Hickcox (1990) Kolb published the first version of the LSI in 1971. However, the inventory is generally referred to in the literature as the 1976 version. This version will be referred to as the LSI-1976. The LSI-1976 consists of nine sets of words, each set consisting of four words. The four words, each representing one of the four learning abilities, are presented in the same order (CE, RO, AC, AE) throughout so that the words associated with each of the four learning abilities are grouped in columns to facilitate scoring for the self-scoring format of the inventory. A respondent rank orders the four words in each of the nine sets according to how well he/she perceives each word as describing his/her individual learning style. The rankings for only six of the nine items, that is, for only 24 of the 36 words contribute to the scores for the four learning abilities CE, RO, AC and AE. The other twelve words serve as distracters. Two combination scores AC-CE (that indicates the extent to which an individual emphasizes abstractness over concreteness) and AE-RO (the extent that an individual emphasizes action over reflection) are calculated.
By plotting these scores on the vertical and horizontal axes respectively, the respondent is positioned in one of the four quadrants representing one of the four learning styles. Due to the ranking format the instrument is an ipsative measure (Kerlinger, 1973). Research results pertaining to the psychometric properties of the LSI-1976 are summarized in Table 1.
The LSI-1985
A revised version of the LSI was published in 1985. This version of the LSI will be referred to as the LSI-1985. The format was changed and the LSI-1985 consists of 12 sentence-completion items and therefore has more items than the LSI-1976. Each sentence has four word endings corresponding to the four learning abilities. As for the LSI-1976 the four words are presented in the same order (CE, RO, AC, AE) throughout to facilitate the scoring of the self-scoring inventory. A respondent rank orders the four words for each sentence or item. The ratings for all 12 words are summed for each of the learning abilities CE, RO, AC, AE. These scores are used to calculate the combination scores AC-CE and AE-RO and by plotting these two scores on the corresponding bipolar axes, the respondent is assigned to one of the four quadrants, each representing one of the four learning styles. Kolb thus increased the number of items and placed the words in the context of a sentence, but remained committed to the ranking format and the instrument remains an ipsative measure. Research results pertaining to the psychometric properties of the LSI-1985 are summarized in Table 2.
The LSI IIA
In 1993 a further revised version of the LSI was published. The instrument is called the LSI IIA and the following information is given in the publishers McBer & Company's catalogue: "The LSI IIA has a revised questionnaire format and scoring key. The twelve-question inventory now has scrambled sentence endings and new scoring instructions that have proved to have high test-retest reliability in recent studies." (p. 11). The same 12 items are presented in the same order, but the four word endings for each item have been randomized. Kolb remains committed to the ranking format and the instrument remains an ipsative measure. To date no literature has been found relating to the LSI IIA.
Problems associated with ipsative measures
Hicks (1970) defined an ipsative measure as follows: "A format in which respondents compare or rank items will always yield purely ipsative scores if respondents rank all alternatives per item, if all these rankings are scored, and if alternatives representing all assessed variables are compared with each other and presented for preferential choice by the respondent." (p. 170). According to this definition the LSI-1985 and LSI IIA are purely ipsative instruments. The LSI-1976 is a partially ipsative instrument as not all alternatives ranked by respondents are scored.
Kolb remained committed to the ranking format of the inventory thus making it an ipsative measure. An ipsative measure is designed to measure within-individual differences, and this creates difficulties when researchers try to make between-subjects analyses. Statistically the ipsative measure results in a between-subjects sum of squares of zero and one individual's preferences cannot be compared with another's (Merritt & Marshall, 1984). Cornwell and Dunlap (1994) stated that ipsative scores cannot be factored and that correlation-based analysis of ipsative data produced uninterpretable and invalid results. As ipsative scores contain only categorical information across individuals multinomial statistical techniques are appropriate. Instead of using the sum of the rank ordered ipsative scores, Cornwell and Dunlap suggested rank ordering the summed responses across the four learning modes for each individual and then applying multinomial techniques to this categorical data. Cornwell, Manfredo and Dunlap (1991) recommended the use of non-ipsative scores for evaluating the construct validity of the LSI.
The minimum requirement for an instrument's scores to be amenable to construct interpretations is that the instrument must yield internally consistent scores (Tenopyr, 1988). Tenopyr states that the internal consistencies of scales of ipsative inventories are interdependent and that there is a possibility for artifactual internal consistency to be generated in such inventories. This places limitations on the usefulness of reliability data for ipsative inventories and such instruments are not suitable for psychometric evaluation and should not be used for making important decisions concerning individuals. ipsative scores are also not suitable for theory building (Hicks, 1970). The usual statistics are not applicable to ipsative measures because of the lack of independence and negative correlations among items and analysis of correlations, as in factor analysis, could be seriously distorted by the negative correlations (Kerlinger, 1973). Many of the studies tabled in this article have treated ipsative data normatively and the results of such studies are of little value. Although an ipsative measure is designed to measure intra-individual differences, the limitations placed on the statistical analysis' of data of ipsative measures makes it inappropriate for reliability and validity evaluation of the instrument.
Normative versions of the LSI
From previous studies using normative forms of the LSI-1976 Marshall and Merritt (1986) concluded that a semantic differential format could be used to develop a reliable and valid normative assessment instrument to assess individual's preferences for ways of learning as proposed by Kolb. They developed the Learning Style Questionnaire (LSQ). In the experimental phase 100 semantic differential word pairs were compiled, with 25 word pairs for each of the four scales. A five-point scale was used by respondents to rate the consistency with which the opposing words characterized their particular learning style. This experimental form of the LSQ was administered to 543 university students from randomly selected classes at two universities. Thirty-seven different majors were represented. About three-fourths of the subjects were under 23 years of age; two-thirds were female and about two-thirds had completed at least two years of college. The 100 items were analyzed and 40 items were selected for the final instrument, 10 items for each of the four scales (CE, RO, AC, AE). The internal consistency reliabilities based on alpha coefficient for the finalized 40-item LSQ were: CE = .78, RO = .86, AC = .85, AE = .88, CE-AC = .90 and RO-AC = .93. Least squares factor analysis of the items was used to examine the construct validity of the instrument. All 40 items loaded on bipolar factors in accordance with Kolb's proposed learning abilities and styles. The authors concluded that the reliability estimates for both bipolar dimensions were very high and that the construct validity for these dimensions had been demonstrated. They recommended that the instrument be used to determine individual learning styles as well as for research purposes.
Romero, Tepper and Tetrault (1992) developed a normative, two-dimensional instrument to measure learning style. Rather than construct an instrument that assesses the four learning abilities, the authors constructed an instrument that assessed the two dimensions concreteness/abstractness and reflection/action directly. The instrument consists of 14 pairs of self-descriptive anchor statements, each pair on a six-point Likert scale. Seven bipolar items assess concreteness versus abstractness, and seven bipolar items assess reflection versus action. The instrument was administered to two independent samples. The one sample consisted of 507 undergraduate students in the fields of liberal arts, business and engineering. The average age was about 21 years and 53% were male. The instrument was administered once to this sample. The second sample consisted of 153 MBA students and the instrument was administered twice with a six week interval. The average age was 28 years and 65% were male. The internal consistency alpha coefficient for the concreteness/abstract scale was .84 for sample I and .78 for sample 2. The alpha coefficient for the reflective/action scale was .86 for sample 1 and .80 for sample 2. The test-retest stability for sample 2 was .75 for the concreteness/abstract scale and .73 for the reflection/action scale. The authors reported that the internal consistency and test-retest stability were acceptable. The two dimensional structure of the instrument was confirmed by factor analysis of both samples using LISREL. Validity support was obtained by comparing student majors with learning style for sample 1.
Geiger, Boyle and Pinto (1993) constructed a normative version of the LSI-1985 that was scored on n seven-point Likert scale consisting of 48 (12 sentence items X four word endings) separate items randomly presented. The standard LSI-1985 and the normative versions were administered to 455 business administration students (first, second and third year students). The age range was from 18 to 47 years (mean age = 21.4 years) and 281 were male and 174 female. Alpha coefficient internal consistency reliability measures for the ipsative version were as follows: CE = .83, RO = .81, AC = .85 and AE = .84. Alpha coefficient reliabilities for the normative version were as follows: CE = .83, RO = .77, AC = .86 and AE =.84. Correlations of the four scale scores were used to determine the equivalence of the ipsative and normative versions. Correlations ranged from .368 to .526 indicating a moderate amount of agreement. Adjusted scale correlations ranged from .466 to .615 with three of the four coefficients exceeding .50. Separate factor analyses were performed on the two versions. For the ipsative version two strong bipolar dimensions were identified running from CE to RO and from AE to AC. These dimensions are not congruent with Kolb's theorized bipolar dimensions. Analysis of the normative version did not produce any bipolar dimensions, but strong support for the four separate learning abilities was obtained. The comments made in the previous section on the inappropriate use of the statistical analysis of data of ipsative measures pertain to these findings.
Method
Measures
Due to the problems relating to ipsative measures described previously it was decided to investigate the psychometric properties of two normative measures of learning style. The Learning Style Questionnaire (LSQ) developed by Marshall and Merritt (1986), described previously, and the Likert-scale form of the LSI-1985 developed by Geiger et al. (1993), described previously, were used. The five-point Likert-scale version of the LSI used in this study will be referred to as the LSI-Likert. These two instruments were obtained from their American authors and were available only in English. This study can be seen as a pilot study to start investigating the reliability and construct validity of the two instruments in a South African population.
Sample
First-year students registered for full-time courses presented in English at the University of Pretoria in the fields of science (BSc) and the human sciences (BA) participated in the study. A total of 464 students were tested at the beginning of the 1995 and 1996 academic years. Due to some incomplete answer sheets 419 answer sheets could be scored for the LSI-Likert and 415 for the LSQ. In scoring the four ability scales (CE, RO, AC, AE) missing or ambiguous responses were substituted with the group average score for an item, for a maximum of two items per questionnaire. For the LSI-Likert the group average score was substituted for one item in 43 cases and for two items in 9 cases. For the LSQ the group average score was substituted for one item in 39 cases and for two items in 4 cases. The allocation of a learning style to a subject is determined by the composite scores ACCE and AE-RO. If a zero score was obtained for either of these composite scores a subject was not allocated a learning style.
There was a higher proportion of females than males in the sample with approximately two-thirds females. Regarding home language, 35% of the students were English first language speakers and for the rest English was a second language. The african cultural group comprised 50% of the sample and the white group 38%. The rest were from the Coloured, Indian and Asian cultural groups. The two fields of study (BSc and BA) were fairly evenly represented. The BSc field of study represented 46% of the sample and comprised students studying mainly for degrees in the biological and agricultural sciences, and engineering fields. The BA field of study represented 54% of the sample and comprised first-year Psychology students.
The order in which the LSI-Likert and LSQ was completed was varied with 51% of the students completing the LSI-Liken followed by the LSQ, and 49% completing the LSQ followed by the LSI-Likert. Hotteling's T test indicated that the four scales for the LSI-Likert and LSQ did not have equal vector of means for these two test groups and it was concluded that the order in which the LSI-Likert and the LSQ were completed did not affect the scores obtained for the two instruments (Pickworth, 1997).
Results and discussion
Item analysis and internal reliability
Item analysis was done for the LSI-Likert and the LSQ using the ITEMAN Conventional Item analysis Program (Assessment Systems Corporation, 1993). Intercorrelations and alpha coefficient reliabilities for the four scales of the two instruments were also calculated using the ITEMAN program. The LSI-Likert item-scale correlations for the CE scale ranged from .29 to .61 (mean = .47), for the RO scale from .30 to .59 (mean = .46), for the AC scale from .34 to .61 (mean = .52), and for the AE scale from .33 to .62 (mean = .49). Intercorrelations for the four scales are given in Table 3 and the alpha coefficients in Table 4.
The LSQ item-scale correlations for the CE scale ranged from .47 to .69 (mean = .58), for the RO scale from .48 to .72 (mean = .59), for the AC scale from .41 to .68 (mean = .56), and for the AE scale from .44 to .70 (mean = .58). Intercorrelations for the four scales are given in Table 3 and the alpha coefficients in Table 4.
Response bias
A five-point Likert scale was used for the LSI-Likert. Options 1 and 2 (Not at all like me and Somewhat unlike me) were endorsed at most by 35% of respondents. For 28 out of the 48 items options 1 and 2 where used by 10% or less of respondents. Relatively high item means, ranging from 3.0 to 4.7, reflect this. This could indicate a response bias.
Each item of the LSQ consists of a word pair on a five-point semantic differential scale. Each of the two words in an item represent opposite learning abilities. In the list of word pairs below the item number is given and the word highlighted was endorsed by less than 20% of the respondents using one of the two response options Generally (Most of the time) or Over half the time:
The Abstract Conceptualisation scale
15 consider impulsive
17 reason hunch
26 careful emotional
27 logical sentimental
29 thinking instinctive
34 resolving feeling
36 intellectual emotional
The Concrete Experience scale
4 sensing thinking
5 premonition reason
12 perceptual intellectual
18 impulsive planning
25 intuitive reasoning
30 hunch logical
The Active Experimentation scale
6 active reserved
23 involved distant
39 solve reflect
40 exercise view
The Reflective Observation scale
31 passive active
37 reflective productive
The above could reflect a response bias in which "logical" (Abstract Conceptualization) words are favoured over "feelings" (Concrete Experience) words, and "active" (Active Experimentation) words are favoured over "passive/reflective" (Reflective Observation) words. The "logical" and "active" words may be perceived to be more socially correct in a learning context. It must also be remembered that the majority of the students are not English first language speakers and may have experienced difficulty with the meanings of the words. In some cases the words more commonly endorsed may be words they are more familiar with.
Learning style frequency
The distributions of learning styles for the BSc and BA groups as measured by the LSI-Likert and LSQ were calculated using the FREQ procedure of the SAS statistical package (SAS Institute Inc., 1990). This procedure was also used to calculate the Chi-square test of significance for the frequencies. The frequency distributions of learning styles as measured by the LSI-Likert are given in Table 5 and as measured by the LSQ in Table 6.
For the LSI-Likert the Chi-square statistic had a value of 27.49 with three degrees of freedom which was significant at the 5% level of significance. There was thus a strong association between field of study and learning style as measured by the LSI-Likert. There were more Divergers in the BA group, more Convergers in the BSc group and more Accommodators in the BA group. Assimilators were fairly equally represented in the BSc and BA groups. Except that one would expect more Assimilators in the BSc than the BA group, these results are in line with the descriptions of the learning styles (Kolb, 1984) and thus provide some evidence of construct validity for the learning style constructs for the LSI-Likert.
For the LSQ the Chi-square statistic had a value of 7.556 with three degrees of freedom which is not significant at the 5% level of significance. There is thus no strong association between field of study and learning style as measured by the LSQ.
Item factor analysis
Factor analysis of the items of the LSI-Likert and the LSQ was performed using the principal factor method to extract factors, followed by a direct quartimin (oblique) rotation of factors. The BMDP4M factor analysis statistical package (BMDP Statistical Software Inc., 1993) was used. The factor loadings for the two-factor and four-factor solution for the LSI-Likert are reported in Tables 7.1 and 7.2. The factor loadings for the two-factor and four-factor solution for the LSQ are reported in Tables 8.1 and 8.2. The two-factor solution was expected to yield the CE -- AC and AE -- RO bipolar axes and the four-factor solution was expected to yield the four learning abilities (CE, RO, AC, AE).
The first factor for the two-factor solution for the LSI-Likert (see Table 7.1) combines mainly AC and RO items and would appear to represent the Assimilator learning style. The second factor combines mainly CE and AE items and would appear to represent the Accommodator learning style. The anticipated bipolar axes did not emerge and the LSI therefore does not support the bipolar axes theorized by Kolb.
For the four-factor solution of the LSI-Likert (see Table 7.2) the four factors appear to represent to some extent each of the four learning abilities with the first factor representing AC, the second AE, the third CE and the fourth RO. However, the first two factors combine items representing other learning abilities.
Two bipolar factors, AE -- RO and AC -- CE, emerge for the two-factor solution for the item factor analysis for the LSQ (see Table 8.1). The LSQ therefore supports the bipolar axes as theorized by Kolb.
For the four-factor solution of the LSQ (see Table 8.2) the first factor is bipolar representing the AE -- RO axis and supports the construct validity of the AE and RO learning abilities. The second factor appears to represent the CE learning ability, the third factor the AC learning ability and the fourth factor incorporates RO, AC and one CE item. The four learning abilities are therefore supported for the LSQ. Both the delineation of the two bipolar AE -- RO and AC -- CE axes, as well as the four learning abilities (CE, RO, AC, AE) reflects the careful developmental work done by Marshall and Merritt to produce an instrument that measures the constructs proposed by Kolb in his experiential learning theory.
Conclusion
The results indicate that the internal reliability of the LSQ is somewhat higher than for the LSI-Likert (see Table 4). The presence of a response bias on both instruments is suspected, it would appear that the LSI-Likert was more successful than the LSQ in differentiating learning abilities and styles in the sample used. Frequency distributions of learning style demonstrated more differentiated patterns for the LSI-Likert than for the LSQ (see Tables 5 and 6). The
Chi-square statistic was significant only for the LSI-Likert. Except that one would expect a higher percentage of Assimilators in the BSc group than the BA group, the distributions of learning styles as measured by the LSI-Likert were in accordance with Kolb's theory. Item factor analysis of the LSI-Likert and the LSQ demonstrates that the LSQ produces two bipolar factors in line with Kolb's proposed theoretical constructs whereas the LSI-Likert did not (see Tables 7.1 and 8.1). The four-factor solution for the LSI-Likert and the LSQ produces evidence for the four learning abilities (see Tables 7.2 and 8.2).
From the results of this study it would appear that the normative measures of learning style used in this study show promise for use in counselling, academic advising and for research purposes. This study did not make comparisons between gender, different cultural groups and English speaking versus non-English speakers. The effect of these variables needs to be investigated. The reliability and construct validity of the two instruments should also be investigated further.
Acknowledgement
The financial assistance of the Centre for Science Development of the Human Sciences Research Council towards this research is hereby acknowledged. Opinions expressed in this article and conclusions arrived at, are those of the authors and are not necessarily to be attributed to the Centre for Science Development of the Human Sciences Research Council.
Table 1
Summary of reliability and validity findings for the LSI-1976
Legend for Chart:
A - Author(s)
B - N
C - Reliability
D - Bipolar theory
E - Validity
A
B
C
D
E
Plovnick (1974)
N = 27 Test-retest (3-4 month interval)
Pearson product-moment correlations
CE = .48 RO = .73 AC = .65 AE = .64
CE-AC = .61 RO-AE = .71
--
Supported
1976 LSI Technical Manual
(cited in Geller, 1979)
N = 687 Internal consistency
Spearman-Brown split-half
correlations
CE = .55 RO = .62 AC = .75 AE = .66
AC-CE = .74 AE-RO = .82
Test-retest (3, 6 and 7 month
intervals)
--
--
N = 23 CE = .48 RO = .51 AC = .73 AE = .43
AC-CE = .51 AE-RO = .48
N = 18 CE = .46 RO = .34 AC = .64 AE = .50
AC-CE = .53 AE-RO = .51
N = 42 CE = .49 RO = .40 AC = .40 AE = .33
AC-CE = .30 AE-RO = .43
Freedman & Stumpf (1978, 1980)
N = 412 Internal consistency: Alpha
coefficient
Moderate support
Limited support
N = 1179 CE = .40 RO = .57 AC = .70 AE = .47
Test-retest (five-week interval)
Moderate support
Limited support
N = 101 Pearson product-moment correlations
CE = .39 RO = .49 AC = .63 AE = .47
AC-CE = .58 AE-RO = .51
Whitney & Caplan (1978)
N = 111 --
--
Not supported
Wunderlich & Gjerder (1978)
N = 24 Test-retest (six-week interval)
Correlations ranged from .44 to .72
--
Not supported
Geller (1979)
N = 50 Test-retest (31 day interval)
Pearson product-moment correlations
CE = .56 RO = .52 AC = .59 AE = .61
AC-CE = .70 AE-RO = .55
--
--
West (1982)
N = 42 --
--
Not supported
Fox (1984)
N = 54 --
--
Not supported
Garvey, Bootman, Mc Ghan & Meredith (1984)
N = 501 Internal consistency
Alpha coefficient
CE= .30 RO = .58 AC = .60 AE = .36
Spearman-Brown Prophecy Formula
AC-CE = .72 AE-RO = .79
--
Partial support
Merritt & Marshall (1984)
N = 187 Internal consistency: Alpha
coefficient
CE = .29 RO = .58 AC = .52 AE = .41
Supported
--
Sims, Veres, Watson & Buckner (1986)
N = 438 Internal consistency: Alpha
coefficient
CE = .48 RO = .58 AC = .52 AE = .23
--
--
N = 309 Test-retest (3 applications at
five-week intervals)
N = 132 Zero-order correlation coefficients
CE = .45 to .60 RO = .46 to .57
AC = .51 to .60 AE = .42 to .46
Katz (1986)
N = 739 --
Supported
Supported
Wilson (1986)
N = 102 Internal consistency
Split-half correlation coefficients
CE = .15 RO = .53 ac = .49 ae = .41
AC-CE = .45 AE-RO = .52
Not supported
--
N = 51 Test-retest (six-week interval)
CE = .40 RO = .77 AC = .63 AE = .40
AC-CE = .53 AE=RO = .61
Green, Snell & Parimanath (1990)
N = 147 --
--
Supported
Lashinger (1990)
-- --
--
Partial support
Review of experiential learning theory research in the nursing
profession.
Welman & Huysamen (1993)
N = 573 Internal consistency: Alpha
coefficient
AC-CE = .63 AE-RO = .55
--
Partial support
Table 2
Summary of reliability and validity findings for the LSI-1985
Sims, Veres, Watson & Buckner (1986)
N = 181 Internal consistency: Alpha
coefficient
CE = .76 RO = .84 AC = .85 AE = .82
--
--
N = 131 Test-retest (3 applications at
five-week intervals)
N = 94 Zero-order correlation coefficients
CE = .24 to .44 RO = .39 to .66
AC = .42 to .50 AE = .56 to .62
Highhouse & Doverspike (1987)
N = 111 --
--
Partial support
Veres, Sims & Shake (1987)
N = 230 Internal consistency: Alpha
coefficient
CE = .82 RO = .85 AC = .83 AE = .84
--
--
N = 230 Test-retest (3 applications at
three-week intervals)
Zero-order correlation coefficients
CE = .30 to .52 RO = .36 to .46
AC = .45 to .56 AE = .28 to .44
--
--
Atkinson (1988)
N = 26 Test-retest (nine-day interval)
Pearson product-moment coefficient
CE = .57 RO = .40 AC = .54 AE = .59
AC-CE = .69 AE-RO = .24
--
--
Cornwell, Manfredo & Dunlap (1991)
N = 317 --
Not supported
Not supported
Ruble & Stout (1991)
N = 231 Internal consistency: Alpha
coefficient
CE = .82 RO = .79 AC = .81 AE = .82
--
--
N - 139 Test-retest (five-week interval)
Pearson product-moment correlations
CE = .18 RO = .46 AC = .36 AE = .47
AC-CE = .22 AE-RO = .54
Veres, Sims & Locklear (1991) Random ordering of the four
sentence endings of the LSI-1985
-- Internal consistency
Mean Alpha coefficients for three
applications
--
--
N = 711 CE = .56 RO = .67 AC = .71 AE = .52
N = 1042 CE = .67 RO = .67 AC = .74 AE = .58
Test-retest (3 applications at
eight-week intervals)
Zero-order correlation coefficients
N = 711 CE = .92 to .96 RO = .93 to .97
AC = .94 to .97 AE = .91 to .96
N = 1042 CE = .97 to .99 RO = .97 to .98
AC = .97 to .99 AE = .96 to .99
Geiger, Boyle & Pinto (1992)
N = 718 --
Not supported
Not supported
Cornwell & Manfredo (1994)
N = 292 --
--
Not supported
Table 3
Intercorrelations for the scales of the LSI-Likert and the LSQ
CE RO AC AE
CE -- .254 .335 .454
RO .252 -- .411 .305
AC -.424 .033 -- .439
AE .077 -.521 .265 --
Correlations above the diagonal are for the LSI-Likert and those below the diagonal are for the LSQ.
Table 4
Alpha coefficients for the scales of the LSI-Likert and the LSQ
CE RO AC AE
LSI-Likert .741 .717 .799 .799
Geiger et al. .83 .77 .86 .84
LSQ .801 .812 .823 .839
Marshall & Merritt .78 .86 .85 .88
The alpha coefficients reported by Geiger et al. (1993) and Marshall & Merritt (1986) are included for comparison.
Table 5
Frequency of learning styles as measured by the LSI-Likert for the BSc and BA fields of study
Legend for Chart:
A - Learning style
B - BSc
C - BA
D - Total
A B C D
Diverger
Frequency 3 26 29
Column % 1.78% 13.47% 8.01%
Assimilator
Frequency 45 47 92
Column % 26.63% 24.35% 25.41%
Converger
Frequency 96 72 168
Column % 56.80% 37.31% 46.41%
Accommodator
Frequency 25 48 73
Column % 14.79% 24.87% 20.17%
Total 169 193 362
46.69% 53.31% 100%
The Chi-square has a value of 27.49 with three degrees of freedom which is significant at the 5% level.
Table 6
Frequency of learning styles as measured by the LSQ for the BSc and BA fields of study
Legend for Chart:
A - Learning style
B - BSc
C - BA
D - Total
A B C D
Diverger
Frequency 8 9 17
Column % 4.32% 4.57% 4.45%
Assimilator
Frequency 42 40 82
Column % 22.70% 20.30% 21.47%
Converger
Frequency 123 118 241
Column % 66.49% 59.90% 63.09%
Accommodator
Frequency 12 30 42
Column % 6.49% 15.23% 10.99%
Total 185 197 382
48.43% 51.57% 100%
The Chi-square has a value of 7.556 with three degrees of freedom which is not significant at the 5% level.
Table 7.1
Item factor analysis for the LSI-Likert: Oblique rotated factor loadings for a two factor solution
Legend for Chart:
A - Scale
B - Item
C - Factor 1
D - Factor 2
A B C D
Abstract Conceptualisation 4 .435 -.045
6 .501 .026
10 .158 .130
11 .457 -.170
19 .492 .153
24 .245 .142
25 .429 -.085
26 .388 .255
29 .579 -.008
32 .186 .358
43 .562 .166
47 .462 .235
Concrete Experience 1 -.028 .309
7 -.011 .231
14 -.119 .307
15 .080 .440
18 .406 .169
22 .030 .122
28 -.047 .212
31 -.039 .237
33 0.048 .445
38 .208 .474
42 .153 .580
45 .216 .179
Active Experimentation 5 .072 .441
12 .418 .243
13 -.160 .553
17 .265 .102
20 .077 .330
34 .321 .294
35 -.093 .561
37 .048 .458
39 .104 .423
41 .137 .512
44 .040 .446
48 .291 .279
Reflective Observation 2 .377 -.046
3 .301 -.223
8 .258 .260
9 .195 .094
16 .384 -.002
21 .194 .120
23 .196 .147
27 .336 -.079
30 .360 -.056
36 .280 .034
40 .364 -.050
46 .304 -.204
VP[*] 4.130 3.974
* The VP is the variance explained by the factor. It is computed as the sum of squares for the elements of the factor's column in the factor loading matrix.
Table 7.2
Item factor analysis for the LSI-Likert: Oblique rotated factor loadings for a four factor solution
Legend for Chart:
A - Scale
B - Item
C - Factor 1
D - Factor 2
E - Factor 3
F - Factor 4
A B C D E
Abstract Conceptualisation
4 .316 -.081 .000 .205
6 .485 -.002 -.132 .101
10 .187 -.040 .192 .026
11 .280 -.172 -.038 .248
19 .551 -.228 .037 .050
24 .447 -.166 .165 -.153
25 .394 -.247 .073 .110
26 .476 .167 -.054 .027
29 .486 -.066 -.035 .188
32 .395 .112 .168 -.143
43 .625 .088 -.135 .074
47 .640 .028 -.010 -.062
Concrete Experience
1 .054 .016 .500 -.081
7 -.082 .147 .281 .120
14 -.127 -.039 .756 -.025
15 .212 .210 .294 -.060
18 .432 .051 .014 .083
22 .029 .037 .135 .046
28 -.202 -.025 .678 .167
- .202 -.025 .678 .167
31 -.062 -.066 .628 .014
33 .178 .069 .526 -.243
38 .353 .351 .077 -.037
42 .356 .454 .066 -.093
45 .187 -.008 .277 .103
Active Experimentation
5 .240 .398 -.031 -.084
12 .546 .063 .003 -.010
13 .093 .438 .075 -.189
17 .279 .061 -.048 .066
20 .060 .417 -.041 .109
34 .506 .046 .108 -.099
35 .188 .404 .073 -.207
37 .031 .541 .034 .122
39 -.045 .563 .074 .296
41 .056 .614 .083 .230
44 .014 .468 .132 .124
48 .401 .217 -.063 -.019
Reflective Observation
2 .149 .043 -.087 .382
3 .111 -.195 -.035 .259
8 .354 .130 .054 .002
9 .181 .069 -.030 .102
16 .171 .045 -.025 .372
21 .226 .035 .029 .039
23 .050 .093 .177 .278
27 -.087 .023 .111 .632
30 -.011 .062 .020 .581
36 .030 .057 .074 .443
40 .016 .041 .026 .572
46 .150 -.310 .139 .208
VP[*] 4.350 2.719 2.508 2.387
* The VP is the variance explained by the factor. It is computed as the sum of squares for the elements of the factor's column in the factor loading matrix.
Table 8.1
Item factor analysis for the LSQ: Oblique rotated factor loadings for a two factor solution
Legend for Chart:
A - Scale
B - Item
C - Factor 1
D - Factor 2
A B C D
Abstract Conceptualisation 10 .159 -.315
15 .083 -.414
17 -.101 -.505
24 -.059 -.515
26 -.023 -.533
27 -.011 -.609
29 .041 -.481
34 -.099 -.425
36 -.008 -.546
38 .012 -.288
Concrete Experience 1 -.107 .330
4 -.053 .527
5 .071 .495
12 .059 .412
14 .009 .491
18 -.081 .463
21 -.078 .612
25 .094 .524
28 .117 .398
30 .074 .593
Active Experimentation 6 -.602 .003
7 -.599 .075
11 -.507 .138
13 -.399 .268
16 -.588 .075
19 -.458 -.181
23 -.520 -.197
32 -.647 -.035
39 -.236 -.307
40 -.441 -.071
Reflective Observation 2 .611 -.056
3 .512 -.022
8 .667 -.102
9 .666 -.045
20 .363 .055
22 .338 .248
31 .549 .183
33 .349 -.070
35 .503 .039
37 .434 .230
VP[*] 5.388 5.102
* The VP is the variance explained by the factor. It is computed as the sum of squares for the elements of the factor's column in the factor loading matrix.
Table 8.2
Item factor analysis for the LSQ: Oblique rotated factor loadings for a four factor solution
Legend for Chart:
A - Scale
B - Item
C - Factor 1
D - Factor 2
E - Factor 3
F - Factor 4
A B C D E
Abstract Conceptualisation
10 -.111 -.169 .163 .171
15 .007 -.276 .137 .296
17 .217 -.369 .136 .371
24 .010 -.038 .707 -.094
26 .072 -.149 .486 .199
27 .027 -.197 .563 .109
29 .067 -.267 .216 .355
34 .103 .009 .569 .035
36 -.003 -.104 .620 .025
38 -.024 .130 .536 -.011
Concrete Experience
1 .153 .366 -.037 .089
4 .115 .561 -.122 .097
5 -.020 .607 -.034 .075
12 -.028 .538 -.023 .035
14 .125 .193 -.586 .315
18 .054 .505 -.004 -.145
21 .123 .501 -.279 .035
25 -.118 .671 .071 -.136
28 -.126 .499 .033 -.060
30 -.092 .534 -.154 -.120
Active Experimentation
6 .638 -.026 -.058 .067
7 .649 .094 -.028 .085
11 .603 .086 -.163 .204
13 .450 .310 -.051 .084
16 .597 .195 .135 -.045
19 .425 .005 .280 -.111
23 .552 -.063 .159 .082
32 .668 .128 .185 -.007
39 .331 -.049 .250 .275
40 .470 .156 .243 .051
Reflective Observation
2 -.532 .142 .133 .270
3 -.436 .173 .130 .239
8 -.538 .081 .055 .471
9 -.543 .190 .106 .454
20 -.284 .048 -.104 .227
22 -.358 .316 .053 -.088
31 -.513 .321 .070 .104
33 -.183 -.098 -.196 .513
35 -.459 .171 .083 .137
37 -.375 .194 -.147 .155
VP[*] 5.119 3.608 3.077 1.718
* The VP is the variance explained by the factor. It is computed as the sum of squares for the elements of the factor's column in the factor loading matrix.
DIAGRAM: Figure 1 Kolb's model of experiential learning, learning abilities and learning styles.
References
Assessment Systems Corporation. (1993). User's manual for the ITEMAN Conventional Item Analysis Program. Minnesota: Assessment Systems Corporation.
Atkinson, G. (1988). Reliability of the Learning Style Inventory-1985. Psychological Reports, 62, 755-758. BMDP Statistical Software, Inc. (1993).
BMDP Manual. Cork, Ireland: BMDP Statistical Software.
Cornwell, J.M. & Dunlap, W.P. (1994). On the questionable soundness of factoring ipsative data: a response to Saville & Willson (1991). Journal of Occupational and Organizational Psychology, 67, 89-100.
Cornwell, J.M., & Manfredo, P.A. (1994). Kolb's learning style theory revisited. Educational and Psychological Measurement, 54(2), 317-327.
Cornwell, J.M., Manfredo, P.A. & Dunlap, W.P. (1991). Factor analysis of the 1985 revision of Kolb's Learning Style Inventory. Educational and Psychological Measurement, 51, 455-462.
Fox, R.D. (1984). Learning styles and instructional preferences in continuing education for health professionals: a validity study of the LSI. Adult Education Quarterly, 35(2), 72-85.
Freedman, R.D. & Stumpf, S.A. (1978). What can one learn from the Learning Style Inventory? Academy of Management Journal, 21(2), 275-282.
Freedman, R.D. & Stumpf, S.A. (1980). Learning style theory: less than meets the eye. Academy of Management Journal, 5(3), 445-447.
Garvey, M, Bootman, J.L., McGhan, W.F., & Meredith, K. (1984). An assessment of learning styles among pharmacy students. American Journal of Pharmaceutical Education, 48, 134-140.
Geiger, M.A., Boyle, E.J. & Pinto, J.K. (1992). A factor analysis of Kolb's revised Learning Style Inventory. Educational and Psychological Measurement, 52, 753-759.
Geiger, M.A., Boyle, E.J. & Pinto, J.K. (1993). An examination of ipsative and normative versions of Kolb's revised Learning Style Inventory. Educational and Psychological Measurement, 53, 717-726.
Geller, L.M. (1979). Reliability of the Learning Style Inventory. Psychological Reports, 44, 555-561.
Green, D.W., Snell, J.C. & Parimanath, A.R. (1990). Learning styles in assessment of students. Perceptual and Motor Skills, 70, 363-369.
Hickcox, L.K. (1990). An historical review of Kolb's formulation of experiential learning theory. Unpublished D. Ed. thesis. Oregon State University.
Hicks, L.E. (1970). Some properties of ipsative, normative and forced-choice normative measures. Psychological Bulletin, 74(3), 167-184.
Highhouse, S. & Doverspike, D. (1987). The validity of the Learning Style Inventory 1985 as a predictor of cognitive style and occupational preference. Educational and Psychological Measurement, 47, 749-753.
Katz, N. (1986). Construct validity of Kolb's Learning Style Inventory, using factor analysis and Guttman's smallest space analysis. Perceptual and Motor Skills, 63, 1323-1326.
Kerlinger, F.N. (1973). Foundations of behavioral research (2nd ed.). New York: Holt, Rinehart & Winston.
Kolb, D.A. (1984). Experiential learning: experience as the source of learning and development. Englewood Cliffs, N.J.: Prentice-Hall.
Kolb, D.A., Rubin, I.M. & McIntyre, J.M. (1984). Organizational psychology: an experiential approach to organizational behavior (4th ed.). Englewood Cliffs, N.J.: Prentice-Hall.
Laschinger, H.K.S. (1990). Review of experiential learning theory research in the nursing profession. Journal of Advanced Nursing, 15, 985-993.
Marshall, J.C. & Merritt, S.L. (1986). Reliability and construct validity of the Learning Style Questionnaire. Educational and Psychological Measurement, 46(1), 257-262.
Merritt, S.L. & Marshall, J.C. (1984). Reliability and construct validity of ipsative and normative forms of the Learning Style Inventory. Educational and Psychological Measurement, 44(2), 463-472.
Pickworth, G.E. (1997). An integration of the theories of JL Holland and DA Kolb: a theoretical and empirical study of vocational personality and learning style types. Unpublished DPhil. thesis. University of Pretoria.
Plovnick, M.S. (1974). Individual learning styles and the process of career choice in medical students. Unpublished doctoral thesis. Cambridge, Massachusetts: Massachusetts Institute of Technology.
Plovnick, M.S. (1975). Primary care career choices and medical student learning styles. Journal of Medical Education, 50, 849-855.
Romero, J.E., Tepper, B.J. & Tetrault, L.A. (1992). Development and validation of new scales to measure Kolb's (1985) Learning Style dimensions. Educational and Psychological Measurement, 52, 171-180.
Ruble, T.L. & Stout, D.E. (1991). Reliability, classification stability and response-set bias of alternate forms of the Learning Style Inventory (LSI-1985). Educational and Psychological Measurement, 51, 481-489.
SAS Institute Inc. (1990). SAS/STAT User's Guide. Carey, N.C.: SAS Institute Inc.
Sims, R.R., Veres, J.G., Watson, P. & Buckner, K.E. (1986). The reliability and classification stability of the Learning Style Inventory. Educational and Psychological Measurement, 46, 753-760.
Sugarman, L. (1985). Kolb's model of experiential learning: touchstone for trainers, students, counselors and clients, Journal of Counseling and Development, 64(4), 264-268.
Tenopyr, M.L. (1988). Artifactual reliability of forced-choice scales. Journal of Applied Psychology, 73(4), 749-751.
Veres, J.G., Sims, R.R. & Locklear, T.S. (1991). Improving the reliability of Kolb's revised Learning Style Inventory. Educational and Psychological Measurement, 51, 143-150.
Veres, J.G., Sims, R.R. & Shake, L.G. (1987). The reliability and classification stability of the Learning Style Inventory in corporate settings. Educational and Psychological Measurement, 47, 1127-1133.
Welman, J.C. & Huysamen, G.K. (1993). Die geldigheid van 'n leerstylvraelys in die onderskeiding tussen studierigtings. Suid-Afrikaanse Tydskrif vir Hoer Onderwys, 7(3), 258-269.
West, R.F. (1982). A construct validity study of Kolb's learning style types in medical education. Journal of Medical Education, 57. 794-796.
Whitney, M.A. & Caplan, R.M. (1978). Learning styles and instructional preferences of family practice physicians. Journal of Medical Education, 53, 684-686.
Wilson, D.K. (1986). An investigation of the properties of Kolb's Learning Style Inventory. Leadership & Organization Development Journal, 7(3), 3-15.
Wunderlich, R. & Gjerde, C.L. (1978). Another look at Learning Style Inventory and medical career choice. Journal of Medical Education, 53, 45-54.
~~~~~~~~
By Glynis E. Pickworth[*], Faculty of Medicine, University of Pretoria, P.O. Box 667, Pretoria 0001, South Africa, E-mail: glynis@medic.up.ac.za and Willem J. Schoeman, Department of Psychology, Rand Afrikaans University, P.O. Box 524, Auckland Park 2006, South Africa
--------------------------------------------------------------------------------
Copyright of South African Journal of Psychology is the property of Foundation for Education Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: South African Journal of Psychology, Jun2000, Vol. 30 Issue 2, p44, 9p, 14 charts, 1 diagram.
Item Number: 3510857 Result 25 of 127 [Go To Full Text] [Tips]
Result 28 of 127 [Go To Full Text] [Tips]
Title: A Strategy For Helping Students Learn How to Learn.
Subject(s): LEARNING, Psychology of; LEARNING strategies -- Study & teaching; MARYGROVE College (Detroit, Mich.); MYERS-Briggs Type Indicator; PERSONALITY & academic achievement
Source: Education, Spring2000, Vol. 120 Issue 3, p479, 8p, 4 charts
Author(s): McClanaghan, Mary Ellen
A STRATEGY FOR HELPING STUDENTS LEARN HOW TO LEARN
Engaging in the process of learning how to learn must include awareness of how one perceives and processes material to be learned. Instructors can enhance students' awareness by calling their attention to the ways and means by which they are approaching their subject. Varying teaching methods in each component of the instructional cycle on a regular basis and then discussing what each student finds most compelling and most challenging provides opportunities to raise awareness.
Introduction
Successful people know how to learn. This key to success has never been more important than it is today in our information-saturated society. Marygrove College faculty recognize the importance of learning how to learn and have made "learning to learn," one of seven across the curriculum emphases. This emphasis is begun in the First Year Seminar, an introductory course with the overall goal of student success. To achieve this goal we emphasize self-awareness and learning how to learn. Although there are several approaches to these objectives, one approach is attention to learning styles. In his book Powerful Learning, (1998) Ron Brandt states that attention to individual learning styles is an avenue that leads to learning how to learn. There are several definitions of learning style but basically it is an individual's characteristic means of perceiving and processing information. It is important to first validate a student's dominant means of learning if we hope to challenge them to work in a style in which they feel less competent.
Learning Styles
The First Year Seminar or Mg 102 as it is referred to at Marygrove introduces students to the theory of learning styles. Students take a short form of the Myers/Briggs Type Indicator (MBTI). This is one of the best known psychological instruments in the world today. It is based on the theory of Swiss psychologist, Carl Jung. According to Jung, a persons attitude or readiness to act is determined by a preference for either extraversion, which focuses on the external world, or introversion, which focuses on the internal world. He also identified four behavioral functions that, in various combinations constitute personality type: sensing, intuition, thinking, and feeling. Myers and Briggs built on Jung's theory by adding a judging / perceiving scale. The judging and perceiving scale indicates if a person has a stronger attraction toward one of the perceiving functions (sensing and intuition) or one the judging functions (thinking and feeling). One component of Jung's theory that has a parallel to the teaching and learning process comes from Jung's theory of human development, which identifies two major objectives of psychic development' perfection and completion. The perfection objective involves the human need to develop one's own natural strengths and abilities to the maximum. Completion continues the development process to strengthen also the less dominant but potential abilities (Hanson and Silver, 1996). The value of understanding one's learning style is first to develop one's natural approaches to learning and then to develop the capacity to learn in ways that may require more attention and effort. Learning how to learn in different ways will assist students to be life long learners who are capable of learning in various settings and situations. If students can be successful by learning in ways that are most natural to them they are more likely to take on the challenge to move toward Jung's concept of completion.
Action Research
The faculty who teach The First Year Seminar, Mg 102, commit to meet together several times during the semester to discuss student progress and to continually seek to improve our service to our new students. As a result of these meetings the idea surfaced to track our students' learning styles in order to offer better instruction and support services.
Objectives of the study were to:
• work with the Student Support Service tutors and Career Services to empower them to build on the introduction that the students receive regarding learning style theory in the First Year Seminar course.
• provide workshops for students with very strong learning preferences to assist them in developing their weaker styles.
• study and report the results of the learning style profiles of new Mary grove students to identify any possible clustering of styles in our population.
• Offer on-going support to the First Year Seminar faculty to help them make better use of the information to assist their students' in monitoring their own learning style development.
Several researchers and educators have adapted the theory of the MBTI and developed instruments for specific uses. One example is "The Thoughtful Education Model"' developed by J. Robert Hanson and Harvey F. Silver. Their work centers on four learning skills identified by Jung: Sensing/Intuition and Thinking/Feeling and offers very valuable application for educators. To test the consistency of the short form of the Myers/Briggs instrument four groups of students completed either "Learning Style Instrument for Adults," developed by Hanson and Silver or the Form G of the MBTI in addition to the short form.
The following characteristics of the learning styles is based upon the research of Hanson and Silver (1996) as reported in course materials produced by Canter Educational Associates for Marygrove College's Master in the Art of Teaching Program (1996).
ST / Sensing-Thinking Learning Style In the sensing-thinking learning style (ST), students want concrete, specific information and need to know what is right and wrong. They need a structured environment and lose interest if things move too slow or don't seem practical. They learn best from repetition, drill, memorization and actual experience. They need immediate feedback.
NT / Intuitive- Thinking Learning Style In the intuitive-thinking learning style (NT), students are skeptical, analytical and logical. They trust hard evidence and reason. They prefer to work independently; they understand things and ideas by breaking them down into their component parts. They want to be challenged and allowed to be creative, and are concerned with relevance and meaning. They have great patience and persistence if their attention is captured.
SF / Sensing-Feeling Learning Style In the sensing-feeling learning style (SF), students process information based on their personal experience. They respond to collegiality, trust, respect and learning cooperatively. They view content mastery as secondary to harmonious relationships. They are very sensitive to approval or disapproval. They learn best by talking and like group activities.
NF/Intuitive-Feeling Learning Style - In the intuitive-feeling learning style (NF), students are looking for possibilities and patterns, and connections with prior learning. They look for uniqueness, originality and aestheticism. They learn best in a flexible and innovative atmosphere. They have difficulty planning and organizing their time. They need to see the big picture. They are bored by routine and rote assignments.
With these categories in mind information was collected on 207 Marygrove students between the years 1995-1998. Of these 207 students 167 were female. Although the age of the students was not documented it should be noted that the average age of Marygrove's undergraduate student is 32.
These results are quite different from other studies. According to the research of Hanson and Silver which does not specify age level:
Intuitive Feelers make up about 10% of all students.
Sensing Feelers make up about 35% of all students.
Intuitive Thinkers make up about 20% of all students.
Sensing Thinkers make up about 35% of all students,
The Marygrove research was compared to another study of college students conducted by Mary Todd and Daniel Robinson at Bunker Hill Community College in Boston, Massachusetts in 1995. Bunker Hill has approximately 6,000 day and evening students. The MBTI preferences of 1007 students were collected over a ten-year period (1985-1995). This research was reported at the Center for Application of Psychological Type Conference held in Orlando, Florida in March of 1998. The study reported scores by racial identification.
The racial composition in the Mary grove study was the opposite of the Bunker Hill study. Of 725 students in the Bunker Hill study 63% were Caucasian and 20% were African American. In the Marygrove study, of the 204 students whose racial background was reported 87% were African American and 11% were Caucasian.
The most obvious difference in the Marygrove results is in the high percent age of Intuitive Feelers (NF) students. Keirsey and Bates (1984) report that only 12% of the general population are NFs. This is much closer to what Hanson and Silver and Todd and Robinson (Bunker Hill) report.
Insight into why Marygrove may attract a higher percentage of NF students may be in the characteristics of the college itself. Marygrove is a small (approximately 1,000 undergraduate students) Catholic liberal arts college. The literature boasts of small class size and a warm and personal atmosphere. Fairhurst and Fairhurst (1999) describe NF students as preferring small group discussions and one on one instruction. They want a personalized learning setting. They seek harmony and demonstrate sensitivity and caring for others. Personal values are very important to them. If we place the stated characteristics of Mary grove College, as described by the mission statement and college catalog, along side the characteristics of the Intuitive Feeler learner the results may not be so surprising.
Conclusions and Future Action
Adult students bring a consumer mentality to higher education. They will seek out learning environments that offer them the best chance at success. This study pro vides Marygrove faculty and support staff with a closer look at the students who choose Marygrove as their ticket to the future. A great percentage of these students are looking for a personal environment that will allow them to unleash their unique creative potential. Affirming this natural preference for learning provides an important variable that contributes to success. Successful students are more likely to develop abilities that might not have been tapped. These students bring with them many years of life experience in which they have developed habits and attitudes toward learning. Some of these habits and attitudes must be transformed if these students are to graduate and move on to a successful future.
This study was undertaken as an action research project that does not seek correlation beyond the population studied. However, faculty at other institutions could easily conduct their own study to ascertain the profile of their student body. The value is in determining if there is a dominant student profile at the institution. If there is, faculty and support staff have a better opportunity to begin working with the students" most natural style. Research has suggested that knowing one's preferred learning style enhances a student's ability to achieve academic success. The knowledge that there are different styles for achieving .success is in itself an eye opener for many students.
Some studies have indicated that academically successful students have fewer strong learning style preferences than do low achievers. The challenge is to assist students in perfecting their natural learning style while providing the incentive to develop less dominant styles they will need in the workforce and other areas of their lives. Engaging in the process of learning how to learn must include awareness of how one perceives and processes material to be learned. Instructors can enhance students' awareness by calling their attention to the ways and means by which they are approaching their subject. Varying teaching methods in each component of the instructional cycle on a regular basis and then discussing what each student finds most compelling and most challenging pro vides opportunities to raise awareness.
Hanson and Silver offer the following suggestions for what they call teaching around the wheel. Each aspect of instruction offers opportunities to reach the variety of styles by changing teaching methods on a regular basis.
Anticipatory Sets (Introductions that prepare for the lesson or a unit)
ST Give facts, details
NT Raise issues & potential problems
SF Relate to students' experiences, feelings & prior knowledge
NF Suggest new and original possibilities
Questions
ST Who, what, where, when
NT Explain, compare, identify cause and effect
SF Ask: What has been your experience? What do you know about ?
NF Ask: What might happen if or ask for an application
Tasks
ST Organize factual information, practice for recall
NT Create a problem solving mode where students must sort out data, analyze and draw conclusions
SF Provide for group work or a task that involves the affect
NF Provide choices for completing assignments and projects or assign tasks that involve imagination, innovation
Setting
ST Traditional rows or pairs; teacher at focus
NT Teams that will create a debating atmosphere; teacher moves from team to team.
SF Groups or pairs for collaboration; teacher meets students at eye level
NF Learning centers, student arranged for interest; teacher is a resource
Feedback
ST Frequent, quick, short/need to know if they are right NT Infrequent but with explanation of why they received the grade they did
SF Frequent, quick with an emphasis on the amount of effort that is evidenced
NF Infrequent but with emphasis on its value' its uniqueness and creativity
Homework
ST Provide a model of what a complete and accurate assignment will look like, practice and drill
NT Problem solving, analyzing work, it too must be modeled
SF Opportunities for articulating ideas, learning from others, develop skills of collaboration designed to convince students they have knowledge
NF Projects or opportunities to create new or different ways of looking at material, important to set criteria
Assessment
ST True and false, fill in the blanks, any measure that allows students to recall factual material
NT Critical essays, debates, research projects which mea sure the ability to see relationships
SF Interviews in and out of class. Let the students question you
NF Anything that can show what the student can do with what they have learned
Helping students learn how to learn may be the most important lesson faculty can teach students. Life-long learners, capable of learning and working in diverse settings, are vital to the 21century society. Assisting students in achievement of this goal puts a demand on faculty to take the time to teach around the learning style wheel. The reward for this effort will be more students who are engaged in at least some aspect of the learning process. Going a step further and talking with students about how they experience learning when instruction or tasks call on styles that are not natural for them, raises awareness of their own approach to learning. Students may believe that what comes natural to them is all that they can do well and they are doomed to failure in all other areas. Unless we sup port students to develop under developed aspects of their styles they are unlikely to have lifelong success. An important task of learning how to learn is to develop an awareness of oneself as a learner. Students need to reflect on their experience of learning in order to take charge of the full development of their abilities. The ultimate goal of higher education can not be content learning alone. Content may become obsolete. The U. S. Department of Labor has identified the ability of knowing how to learn as the most fundamental skill for the next century (Camevale, 1988). Self-awareness and then self-monitoring are essentials for learning how to learn. Faculty and support staffs who nurture this type of learning are helping develop tomorrow's workers. The kind of workers who are needed for the learning organizations that will fuel our global economy.
Results of Marygrove study
Introverts 68%
Feelers 66%
Intuitive Feelers (NF) 41%
Sensing Feelers (SF) 25%
Intuitive Thinkers (NT) 18%
Sensing Thinkers (ST) 16%
Bunker Hill results reported by racial group
Style African American (20%) Caucasian (63%)
Introvert (I) 63% 46%
Feeler (F) 29% 50%
Intuitive Feeler (NF) 07% 24%
Sensing Feeler (SF) 21% 26%
Intuitive Thinker (NT) 12% 19%
Sensing Thinker (ST) 60% 31%
Marygrove results reported by racial group
Style African American (87%) Caucasian (11%)
Introvert (I) 69% 43%
Feeler (F) 63% 78%
Intuitive Feeler (NF) 40% 39%
Sensing Feeler (SF) 23% 39%
Intuitive Thinker (NT) 19% 13%
Sensing Thinker (ST) 18% 09%
Comparison of Intuitive Feeler (NF) data
Keirsey & Bates Hanson & Silver Bunker Hill Marygrove
12% 10% African Amer African Amer
07% 40%
Caucasian Caucasian
24% 39%
Marygrove College Characteristics
Small, warm and friendly environment
Religiously oriented: Catholic and very ecumenical in approach
Committed to the liberal arts
Especially noted for the helping professions and the arts; committed to valuing diversity
Strives to develop graduates who exemplify competence, commitment and compassion
Intuitive-Feeler (NF) Learner
Learn best in a nurturing environment
Have a keen interest in other belief systems and enjoy discussing moral dilemmas
Needs to explore creative potential and find ways to express her/his ideas and beliefs and share this inspiration
Are inspired be sensitive, supportive, humanistic teachers who show them they care about them as individuals
Looks for similarities among people and encourages cooperation and harmony
References
Brandt, R. (1998). Powerful learning. Alexandria, VA: ASCD.
Canter Educational Productions (1996). Learning styles and multiple intelligences. Santa Monica, CA.
Carnevale, A., Gainer, L & Meltzer, A. (1988). Workplace basics: the skills employers want. Washington, D.C.: U.S. Department of Labor.
Fairhurst, A. & Fairhurst ' L. (1995). Effective teaching and effective learning: making the personality connection in your classroom. Palo Alto, CA: Davis-Black Publishing.
Hanson, J.R. & Silver, H.F. (1995). Learning styles & strategies. Princeton, NJ: Hanson Silver Strong & Associates.
Keirsey, D. & Bates, M. (1984). Please understand me: character and temperament types. Del Mar, CA: Prometheus Nemesis.
Lawrence, G.D. (1979). People types and tiger stripes: a practical guide to learning styles.. Gainsville, FL: Center for Application of Psychological Type.
Todd, M & Robinson, D. (1998). "Students of Color at an Urban Community College." Gainsville, FL: Center for Application of Psychological Type.
~~~~~~~~
By Mary Ellen McClanaghan, PH.D., Marygrove College 22461 Revere St. Clair Shores, Michigan 48080
Adapted by PH.D.
--------------------------------------------------------------------------------
Copyright of Education is the property of Project Innovation and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: Education, Spring2000, Vol. 120 Issue 3, p479, 8p, 4 charts.
Item Number: 2990111 Result 28 of 127 [Go To Full Text] [Tips]
Result 30 of 127 [Go To Full Text] [Tips]
Title: Students' Learning Styles in Two Classes.
Subject(s): LEARNING strategies; EDUCATION
Source: College Teaching, Fall99, Vol. 47 Issue 4, p130, 6p, 1 chart, 1 graph
Author(s): Diaz, David P.; Cartnal, Ryan B.
STUDENTS' LEARNING STYLES IN TWO CLASSES
Online Distance Learning and Equivalent On-Campus
The idea that people learn differently is venerable and probably had its origin with the ancient Greeks (Wratcher et al. 1997). Educators have, for many years, noticed that some students prefer certain methods of learning more than others. These dispositions, re ferred to as learning styles, form a student's unique learning preference and aid teachers in the planning of small-group and individualized instruction (Kemp, Morrison and Ross 1998, 40). Grasha (1996) has defined learning styles as "personal qualities that influence a student's ability to acquire information, to interact with peers and the teacher, and otherwise to participate in learning experiences" (41).
Blackmore (1996) suggested that one of the first things we teachers can do to aid the learning process is simply to be aware that there are diverse learning styles in the student population:
There are probably as many ways to "teach" as there are to learn. Perhaps the most important thing is to be aware that people do not all see the world in the same way. They may have very different preferences than you for how, when, where and how often to learn. [online]
Although many of us are aware that different learning styles exist, the application of this knowledge is often inconsequential. Some faculty simply opt to use a wide variety of teaching activities, hoping that they will cover most student learning preferences along the way. This method, though expedient, may not be the most effective way to address student learning preferences. Further, many teachers think that the same teaching methods that work in their traditional classes will also work for distance learning. The underlying assumption is that students who enroll in distance education classes will have the same learning preferences as those in traditional classes. Faculty often assume that teaching styles, and accompanying classroom processes, are like a "master key" and thus appropriate for any setting.
There is not an overabundance of re search on learning styles and distance education. Most of the studies focus on the discovery of relationships between learn ing styles and specific student achievement outcomes: drop rate, completion rate, attitudes about learning, and predictors of high risk.
One of the most popular learning style inventories, which is often used in distance learning research, is the Kolb Learning Style Inventory (LSI) (Kolb 1986). Kolb's LSI measures student learning style preference in two bipolar dimensions. Over time, learners develop a preference for either concrete experiences when learning or a preference for engaging in abstract or conceptual analyses when acquiring skills and knowledge. They also may emphasize interest in turning theory into practice by active experimentation, or they may prefer to think about their experiences by reflective observation (Dille and Mezack 1991, 27).
James and Gardner (1995) described Kolb's LSI as a cognitive learning style mode. Cognitive processes include storage and retrieval of information in the brain and represent the learner's ways of perceiving, thinking, problem solving, and remembering (20).
Dille and Mezack (1991) used Kolb's LSI to identify predictors of high risk among community college telecourse students. Successful students had lower scores on their preferences for concrete experiences than did the unsuccessful students. Thus, because distance learning courses often lead to social isolation and require greater reliance on independent learning skills, students with less need for concrete experience in learning may be expected to be better suited to the distance format. People with higher scores on concrete experience tend to exhibit a greater sensitivity to feelings and thus would be expected to require more interactions with peers and the teacher.
Successful telecourse students also preferred to look for abstract concepts to help explain the concrete experiences associated with their learning. That is, they wanted to know "why" certain things happened in conceptual or theoretical terms. This more abstract approach clearly favored success in the telecourse. Dille and Mezack concluded that students who needed concrete experience and were not able to think abstractly were more high-risk in a telecourse.
Gee (1990) studied the impact of learning style variables in a live teleconference distance education class. The study ex amined the influence of learning style preferences of students in an on-campus or remote classroom on their achievement in the following: course content, course completion rates, and attitudes about learning. Both distance and on-campus groups were taught simultaneously by the same teacher, received identical course content, and met weekly. Gee administered the Canfield Learning Styles Inventory (CLSI) (Canfield 1980).
Students in the distance learning class who possessed a more independent and conceptual learning style had the highest average scores in all of the student achievement areas. People with the lowest scores in the distance learning course had a more social and conceptual learning style. Students with both a social and applied learning style performed much better in the on-campus class. The outcomes of the Gee study suggested that successful distance education students favored an independent learning environment, and successful on-campus students preferred working with others. The relatively small sample of twenty-six students suggested that additional research is needed.
An important question, however, is raised by such research: Are there differences in learning styles between students who enroll in a distance education class and their on-campus counterparts? That question, no matter how it is answered, is vital for anyone interested in students' success. If there are no differences in learning styles, it is likely that faculty can transfer the same types of teaching/learning activities that have worked in the traditional environment into the distance setting with similar success. That is probably true, if enough sensitivity and thought have been given to learning styles and to how these methods will be transferred to the distance education en vironment using current communications technologies.
On the other hand, if there are differences in learning styles between groups of students, then faculty must use learning style information to aid their planning and preparation for distance education ac tivities. Sarasin (1998) noted that professors should be willing to change their teaching strategies and techniques based on an appreciation of the variety of student learning styles. "[Teachers] should try to ensure that their methods, materials, and resources fit the ways in which their students learn and maximize the learning potential of each student" (2).
If optimal learning is dependent on learning styles, and these styles vary be tween distance and equivalent on-campus students, then faculty should be aware of these differences and alter their preparation and instructional methods accordingly. In any case, the first step in using learning style information in distance education is to determine students' learning styles.
Selecting a Learning Style Instrument
As educators consider transplanting their traditional courses into distance learning, they should assess the learning styles of the students who enroll. With a variety of learning style instruments in use, it is important to select one according to the unique requirements of the distance learning context. Three important factors to consider when selecting a learning style instrument are defining the intended use of the data to be collected, matching the instrument to the intended use, and finally, selecting the most appropriate instrument (James and Gardner 1995). Other concerns include the underlying concepts and design of the instrument, validity and reliability issues, administration difficulties, and cost (22).
One of the distinguishing features of most distance education classes is the ab sence of face-to-face social interaction between students and teacher. Thus, an inventory used in that setting should address the impact of different social dynamics on the learning preferences of the students. An example of this can be seen in Gee (1990), who employed the Canfield Learning Styles Inventory (CLSI). The CLSI demonstrated merit in distance learning studies because it at tempted to measure students' preferences in environmental conditions, such as the need for affiliation with other students and instructor, and for independence or structure.
Those varied social dynamics are one of the main differences between distance learning and equivalent on-campus environments. However, in our opinion, both the Canfield Inventory and Kolb's LSI create a narrow range of applicability for learning styles by limiting learning preferences to one or two dimensions. Al though this learning style "stereotyping" may be convenient for statistical analysis, it is less helpful in terms of teaching students about weaker or unused learning preferences. Further, the Kolb LSI, which has been widely used, is primarily a cognitive learning preference instrument, which does not specifically take into a ccount social preferences that are the key distinction between distance and traditional classrooms.
Of the different learning style instruments, the Grasha-Reichmann Student Learning Style Scales (GRSLSS) seem ideal for assessing student learning preferences in a college-level distance learning setting. The GRSLSS (Hruska-Riechmann and Grasha 1982; Grasha 1996) was chosen as the tool for determining student learning styles in the present study based on criteria suggested by James and Gardner (1995). First, the GRSLSS is one of the few instruments de signed specifically to be used with senior high school and college students (Hruska-Riechmann and Grasha, 1982). Second, the GRSLSS focuses on how students interact with the instructor, other students, and with learning in general. Thus, the scales address one of the key distinguishing features of a distance class, the relative absence of social interaction between instructor and student and among students. Third, the GRSLSS promotes an optimal teaching/ learning environment by helping faculty design courses and develop sensitivity to students' needs.
Finally, the GRSLSS promotes understanding of learning styles in a broad context, spanning six categories. Students possess all six learning styles, to a greater or lesser extent. This type of understanding prevents simplistic views of learning styles and provides a rationale for teachers to encourage students to pursue personal growth and development in their underused learning styles.
Only a brief definition of each is provided here in order to assist the reader with the interpretation of the information from this study.
1. Independent students prefer independent study and self-paced instruction and would prefer to work alone rather than with other students on course projects.
2. Dependent learners look to the teacher and to peers as a source of structure and guidance and prefer an authority figure to tell them what to do.
3. Competitive students learn in order to perform better than their peers and to receive recognition for their academic ac complishments.
4. Collaborative learners acquire in formation by sharing and cooperating with teacher and peers. They prefer lectures with small-group discussions and group projects.
5. Avoidant learners are not enthusiastic about attending class or acquiring class content. They are typically uninterested and are sometimes overwhelmed by class activities.
6. Participant learners are interested in class activities and discussion and are eager to do as much class work as possible. They are keenly aware of, and have a desire to meet, the teacher's expectations.
The styles described by the GRSLSS refer to a blend of characteristics that apply to all students (Grasha 1996, 127). Each person possesses some of each of the learning styles. Ideally, one would have a balance of all the learning styles; however, most people gravitate toward one or two styles. Learning preferences are likely to change as one matures and encounters new educational experiences. Dowdall (1991) and Grasha (1996) also have suggested that particular teaching styles might encourage students to adopt certain learning styles.
Problem and Purpose
Students' performance may be related to their learning preferences or styles. Students may also self-select into or away from distance learning classes. As a result, success in distance learning classes may ultimately depend on understanding the learning styles of the students who enroll.
Because more online courses will in variably be offered in the future, some as surance must be provided to the college, the faculty, and the students, that distance education will meet expectations for a good education. Not only will students expect an education that is equal in quality to that provided by traditional offerings, they will expect a student-centered learning environment, designed to meet their individual needs.
There have been few studies on the relationship of learning styles to student success in a distance learning environment, and none that we are aware of have used the GRSLSS. The purpose of this study was to compare the student learning styles of online and equivalent on-campus, health education classes, by using the GRSLSS.
The population for the current study in cluded health education students in a medium-sized (8,000--9,000 enrollment) community college on the central coast of California. The distance education sample included students in two sections of health education offered in an online format (N = 68). The comparison class was selected from four equivalent on-campus sections of health education (N = 40) taught by the lead author.
The online distance students were taught according to the same course outline, used the same textbook, covered the same lecture material, and took the same tests as the on-campus students. Three main differences between on-campus and online groups were the delivery mode for the lectures, the mode of teacher/student and student/student communication, and the mode for the assignments.
The distance classes reviewed multimedia slides (Power Point presentations converted to HTML) and lecture notes online, while the equivalent classes heard the teacher's lectures and participated in face-to-face discussion. The distance class made heavy use of a class Web site and used a listserv and e-mail for communication/discussion with other students and the instructor. Assignments for the distance class were almost entirely Internet-based and independent, while the equivalent class completed some online assignments but participated most frequently in classroom discussions and other traditional assignments.
All 108 participants first reviewed the student cover letter that explained the na ture of the research and provided opportunity for informed consent. Next, the authors distributed the GRSLSS and re viewed the instructions for completion of the inventory. The GRSLSS was administered in a group setting during the second week of classes. Thus, we used the General Class Form to assess the initial learning styles of the students. Students self-scored the inventory, and we obtained raw scores for each of the learning style categories. Inventories were reviewed by the researchers for compliance with di rections and for accuracy of scoring.
Research Outcomes
The present study compared social learning styles between distance education and equivalent on-campus classes using the GRSLSS. The average or mean scores of the distance learning class and the equivalent health education class on each of the six categories are shown in figure 1. Relatively larger differences in the average scores of the two classrooms oc curred for the independent and the de pendent learning styles. Compared with those students enrolled in the traditional classroom, the students in the distance class had higher scores on the independent learning style scale and lower scores on the dependent scale. A statistical test (a t test) was used to determine if the differences in the scores between the independent and dependent learning styles were due to chance.
The variations in average scores between the two styles were found to be statistically significant and thus not likely due to chance (p < .01). The variations in average scores between the two classrooms on the avoidant, competitive, collaborative, and participant learning styles were relatively small, and a statistical analysis using a t test revealed that they were not statistically significant.
To ascertain the patterns in the relationships among the learning styles within each class, we examined the associations among different combinations of styles. This was done by calculating the correlation coefficients associated with the combinations of the six learning styles. The outcomes of this analysis are shown in table 1 for the distance learning and traditional classroom groups. For reading this table, we remind the reader that a correlation coefficient varies from -1, 0, to +1, and that the degree to which it deviates from zero in either direction reflects the strength of the relationship between the two variables. The asterisks with some of the values indicate that the size of the correlation was statistically significant and thus not due to chance.
Correlational analysis within the on line group showed a negative relationship between the independent learning style and the collaborative and dependent styles. In other words, people who were more independent in their learning styles also tended to be less collaborative and dependent. A second important relationship (positive correlation) was found be tween the collaborative learning style and the dependent and participant learning styles. That is, students who were more collaborative in their learning styles also were more dependent and participatory in their approach to learning.
In the equivalent on-campus group, significant positive correlations were found between the collaborative learning style and the competitive and participant styles. That is, on-campus students who were collaborative also tended to be competitive and participatory in the classroom. Finally, a positive correlation be tween the competitive and participant styles of learning also was observed. Students who tended to compete also were "good classroom citizens" and were more willing to do what the teacher wanted them to do.
Discussion
Gibson (1998) has challenged distance education instructors to "know the learner" (140). She noted that distance learners are a heterogeneous group and that in structors should design learning activities to capitalize on this diversity (141). Be cause the dy namic nature of the distance population pre cludes a "typical" student profile (Thomp son 1998, 9), we should continually assess students' characteristics.
A professor using the present data could plan learning opportunities that would emphasize the learning preferences with each of the commonly preferred learning styles (independent, de pendent, collaborative, and participant), thus matching teaching strategies with learning styles.
Of particular interest were the significant differences between the groups in the independent and dependent categories. The distance students more strongly favored independent learning styles. It is not surprising that students who prefer independent, self-paced in struction would self-select into an online class. It may be that they are well suited to the relative isolation of the distance learning environment. In his research, Gee (1990) noted that successful telecourse students fa vored an independent learning style. James and Gardner (1995) suggested that students who favored reliance on independent learning skills would be more suited to a distance format.
As a result of these significant differences, teaching strategies in the distance class should emphasize relatively more independent and fewer dependent learning opportunities. This approach has practical significance given that professors often complain of too little class time to devote to learning objectives. Armed with learning style data, we can more efficiently allocate instructional time to various learning types.
Not only were online students more independent than the on-campus students, but their independent learning preferences were displayed in a way that was negatively related to how dependent and collaborative they were. That is, the independence of online learners was not tied to needs for external structure and guidance from their teacher (dependence) or a need to collaborate with their classmates. The online students can be described as "strongly independent," in that they match the stereotype of the independent learner in terms of autonomy and the ability to be self-directed.
Self-direction and independence were facilitated in the online course by offering students flexible options to shape their learning environment. The lead author, Diaz, used self-paced, independent learning activities that allowed students to choose from a menu of online "cyber as signments" based on their personal interests and the relevance of the assignments. Students completed their chosen assignments by deadlines posted at the class Web site.
In contrast, students in the equivalent on-campus class were significantly more dependent learners than the distance group. Because dependent learners prefer structure and guidance, it is not difficult to understand why they might view the isolation and need for self-reliance in a distance education environment with some apprehension. The low level of in dependence displayed by on-campus students was not related to any other aspects of their styles as learners. Thus, independence was clearly a weaker learning preference for traditional class students.
The online students also displayed collaborative qualities related to their need for structure (dependence) and their willingness to participate as good class citizens (participant dimension). Thus, although online students prefer independent learning situations, they are willing and able to participate in collaborative work if they have structure from the teacher to initiate it. In his online class, Diaz has used listservs and "threaded discussion" areas to promote collaboration among distance students.
In the past, he designed collaborative activities that required students to initiate peer contact and conduct the collaboration with a minimum of structure and support from him. Based on the findings of the current study, it is apparent why this strategy failed: Online students will apparently respond well to collaborative activities, but only if the teacher provides enough structure and guidance. Diaz's mistake was that he assumed that online students would be self-directed, and autonomous, regardless of the type of learning activity.
In contrast, the traditional class students had collaborative tendencies related to their needs to be competitive, and good class citizens. In other words, they were interested in collaboration to the extent that it helped them to compete favorably in the class and to meet the expectations of their teachers. Thus, collaboration was tied to obtaining the rewards of the class, not to an inherent interest in collaboration.
Average avoidant and competitive learning style scores indicated that these learning preferences were favored to a lesser degree by both groups. It was interesting that, though we live in a highly competitive society, neither the online or equivalent on-campus students really preferred a competitive learning environment. However, the on-campus students ap peared to favor competitiveness if it was clear that it was expected (i.e., thus the relationship of competitive and participant styles).
We can also use learning style data to help design "creative mismatches" in which students can experience their less-dominant learning style characteristics in a less-threatening environment (Grasha 1996, 172). Designing collaborative as signments for independent learners, or independent assignments for dependent or collaborative learners, is appropriate and even necessary. Strengthening less-preferred learning styles helps students to expand the scope of their learning, be come more versatile learners, and adapt to the requisites of the real world (Sarasin 1998, 38).
Learning styles were not the only differences between the distance and comparison groups in this study. Demographic data indicated that the distance group had a higher percentage of females (59 percent, 49 percent), students currently enrolled in under 12 units (66 percent, 50 percent), students who had completed 60 or more college units (12 percent, 1 percent), students who had completed a degree (12 percent, 7 percent), and students above 26 years of age (36 percent, 6 percent). These characteristics agree with the general profile of distance students as reported by Thompson (1998). Although it is tempting to identify and depend on a "typical" distance student profile, it is likely that the dynamic nature of distance education in general will keep student characteristics fluid. Thus, distance education instructors should continually monitor students' characteristics.
Conclusions
We have concluded that local health education students enrolled in an online class are likely to have different learning styles than equivalent on-campus students. We found that online students were more independent, and on-campus students were more dependent, in their styles as learners. The on-campus students seemed to match the profile of traditional students who are willing to work in class provided they can obtain rewards for working with others and for meeting teacher expectations. Online students ap peared to be driven more by intrinsic mo tives and clearly not by the reward structure of the class.
One of the limitations of this study was the use of a non-probability (convenience) sampling technique. Non-probability sampling is used when it is impossible or impractical to use random sampling techniques. That is the case in a large portion of educational research. Although still valid, the results should not be overgeneralized. We have demonstrated a real and substantial difference in learning styles between distance and equivalent on-campus health education students at our college.
Before faculty rush to find out the effects of learning styles on student outcomes, they should first address the issue of whether learning style differences exist at all. The results of this study should send an important notice to faculty who are teaching their traditional courses in a distance mode, that there may be drastic differences in learning styles, as well as other characteristic differences, between distance and traditional students.
As the World Wide Web becomes an important medium for education delivery, more and more courses will be offered in an online format. Though faculty may attempt to use the same teaching methods in a distance environment that they would employ in an on-campus class, the data from the current study suggest that faculty will encounter significantly different learning preferences as well as other different student characteristics. Professors may want to employ learning style inventories, as well as collect relevant demographic data, to better prepare for distance classes and to adapt their teaching methods to the preferences of the learners.
Faculty should use social learning style inventories and resulting data for help in class preparation, designing class delivery methods, choosing educational technologies, and developing sensitivity to differing student learning preferences within the distance education environment. Future field-based research should replicate the current study in different in stitutions and disciplines.
ACKNOWLEDGMENT
The authors would like to express their thanks to Tony Grasha, whose encouragement, guidance, and ediorial comments were in strumental in bringing this article to fruition.
Table 1.--Intercorrelations between Learning Style Scales for Online and Equivalent On-Campus Students
Legend for Chart:
A - Scale
B - 1
C - 2
D - 3
E - 4
F - 5
G - 6
A B C D E F
G
Online students (N = 68)
1. Independent -- -.08 -.36(**) -.37(**) .07
-.12
2. Avoidant -- -.03 .12 -.02
-.58(**)
3. Collaborative -- .37(**) -.04
.28(*)
4. Dependent -- .08
.24
5. Competitive --
-.12
6. Participant --
Equivalent on-campus students (N = 40)
1. Independent -- -.20 .10 -.12 .13
.09
2. Avoidant -- -.37(*) -.12 -.01
-.67(**)
3. Collaborative -- .27 .51(**)
.52(**)
4. Dependent -- .15
.31
5. Competitive --
.46(**)
6. Participant --
Note: (*) p < .05, two-tailed. (**) p < .01, two-tailed.
Figure 1. Comparison of Average Group Ratings for Each Learning Style
Legend for Chart:
B - Control group
C - Distance group
A B C
Independent 3.25 3.56(*)
Avoidant 2.49 2.57
Collaborative 3.80 3.58
Dependent 3.84(*) 3.55
Competitive 2.46 2.38
Participant 3.79 3.77
(*) Significant at .01 level
Grasha-Riechmann Learning Styles
REFERENCES
Blackmore, J. 1996. Pedagogy: Learning styles. Retrieved September 10, 1997 from the World Wide Web: . net/~jblackmo/diglib/styl-a.html
Canfield, A. 1980. Learning styles inventory manual. Ann Arbor, Mich.: Humanics Media.
Dille, B., and M. Mezack. 1991. Identifying predictors of high risk among community college telecourse students. The American Journal of Distance Education, 5(1), 24-35.
Dowdall, R. J. 1991. Learning style and the distant learner. Consortium project extending the concept and practice of classroom based research report. (ERIC Document Reproduction Service No. ED 348 117)
Gee, D. G. 1990. The impact of students' preferred learning style variables in a distance education course: A case study. Portales: Eastern New Mexico University. (ERIC Document Reproduction Service No. ED 358 836)
Gibson, C. C. 1998. The distance learners academic self-concept. In Distance learners in higher education: Institutional re sponses for quality outcomes, ed. C. Gibson, 65-76. Madison, Wisc.: Atwood.
Grasha, A. F. 1996. Teaching with style. Pittsburgh, Pa.: Alliance.
Hruska-Riechmann, S., and A. F. Grasha. 1982. The Grasha-Riechmann student learning style scales. In Student learning styles and brain behavior ed. J. Keefe 81-86. Reston, Va.: National Association of Secondary School Principals.
James, W. B. and D. L.Gardner. 1995. Learning styles: Implications for distance learning. (ERIC Document Reproduction Service No. EJ 514 356)
Kemp, J. E., G. R. Morrison, and S. M. Ross. 1998. Designing effective instruction (2nd ed.). Upper Saddle River, N.J.: Prentice-Hall.
Kolb, D. A. 1986. Learning style inventory: Technical manual (Rev. ed.). Boston, Mass.: McBer.
Sarasin, L. C. 1998. Learning style perspectives: Impact in the classroom. Madison, Wisc.: Atwood.
Thompson, M. M. 1998. Distance learners in higher education. In Distance learners in higher education: Institutional responses for quality outcomes, ed. C. Gibson, 9-24. Madison, Wisc.: Atwood.
Wratcher, M. A., E. E. Morrison, V. L. Riley, and L. S. Scheirton. 1997. Curriculum and program planning: A study guide for the core seminar. Fort Lauderdale, Fla.: Nova Southeastern University. Programs for higher education.
~~~~~~~~
By David P. Diaz and Ryan B. Cartnal
David P. Diaz is a professor of health education, and Ryan B. Cartnal is a re search analyst at Cuesta College, San Luis Obispo, California.
--------------------------------------------------------------------------------
Copyright of College Teaching is the property of Heldref Publications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: College Teaching, Fall99, Vol. 47 Issue 4, p130, 6p, 1 chart, 1 graph.
Item Number: 2516393 Result 30 of 127 [Go To Full Text] [Tips]
Title: Can computer-aided instruction accommodate all learners equally?
Subject(s): LEARNING strategies; HUMAN-computer interaction; EDUCATIONAL technology
Source: British Journal of Educational Technology, Jan99, Vol. 30 Issue 1, p5, 20p, 1 graph
Author(s): Ross, Jonathan; Schulz, Robert
CAN COMPUTER-AIDED INSTRUCTION ACCOMMODATE ALL LEARNERS EQUALLY?
Abstract
This exploratory study investigated the impact of learning styles on human-computer interaction. Seventy learners who were enrolled in a large urban post-secondary institution participated in the study. The Gregorc Style Delineator Trademark was used to obtain subjects' dominant learning style scores. Results indicated that patterns of learning indices did not differ significantly based on subjects' dominant learning style. Five of the six measures indicating human-computer interaction behavior were not significant at the p < 0.05 level. However, learning styles significantly affected learning outcomes, as indicated by a significant main effect, as well as an interaction effect between dominant learning style and achievement scores. It would appear that Abstract Random learners may be at-risk for doing poorly with certain forms of computer-aided instruction. Based on the review of literature and results found in this study, it was concluded that computer-aided instruction may not be the most appropriate method of learning for all students.
One of the most powerful features of computer-aided instruction (CAI) is its capacity to individualize instruction to meet the specific needs of the learner (Rasmussen and Davidson, 1996). Self-paced instruction, the ability to present content in a variety of ways (eg, text, video, sound, graphics), and features such as hypertext make CAI an effective learning medium.
The use of CAI in education has burgeoned in recent years (Price, 1991; Nelson and Palumbo, 1992; Hawkridge, 1995). Faced with increasing class sizes and heavier work loads, teachers are looking towards CAI as a means of supplementing classroom instruction. In addition, CAI software continues to improve in its ability to engage learners and provide realistic and stimulating learning environments (Price, 1991). Learners can now choose from a variety of educational software packages designed to augment the curriculum (Dwyer, 1996).
As the use of CAI systems continues to grow, research in the area of human-computer interaction is becoming increasingly important. Currently, a select few studies examine individual differences and their effects on CAI (Marquez and Lehman, 1992; Nelson and Palumbo, 1992; Reed, 1996). Findings generally indicate that while CAI has tremendous potential to individualize instruction, a number of learner characteristics such as motivation, learning styles, and background knowledge may affect the quality and effectiveness of a CAI instructional session.
This exploratory study examines the influences of cognitive learning styles on both achievement levels and human-computer interaction behaviors. Findings from this study indicate that certain forms of CAI may not accommodate all learners equally (see Ross, 1997). Educators should, therefore, remain cautious when using the computer as a learning tool. Just as teachers need to use a variety of approaches to meet the diverse needs of their students, so educators should be aware that CAI may not be the learning medium of choice for all students.
Literature review
The Gregorc Style Delineator Trademark
According to Gregorc (1979): "Learning style consists of distinctive behaviors which serve as indicators of how a person learns from and adapts to his environment. It also gives clues as to how a person's mind operates" (p. 234). Designed to assess learning styles, The Gregorc Style Delineator Trademark is a self-scoring battery which focuses on two types of mediation abilities in adult individuals: perception (the means through which one is able to grasp information), and ordering (the means in which one arranges, systematizes and disposes of information). The two dimensions of ordering are referred to as sequential and random; the two qualities of perception are known as abstractness and concreteness (Gregorc, 1982 a).
Abstractness allows the individual to comprehend that which is not visible to the senses. Data can be mentally visualized, grasped, and conceived through the faculty of reason. Individuals who are strong in concreteness use the physical senses to comprehend and mentally register data.
Sequential individuals perceive and organize data in a linear, methodical fashion, and can express themselves in a precise manner. Furthermore, discrete pieces of information can be categorized naturally. In contrast, randomness disposes the mind to organize information in a nonlinear and multidimensional fashion. This quality enables individuals to deal with, and process, multiple data simultaneously.
Gregorc combines these abilities to create four mediation channels of mind styles: concrete sequential (CS), concrete random (CR), abstract sequential (AS) and abstract random (AR). Gregorc believes that individuals have, to a certain degree, characteristics of each category, but most individuals tend to show a stronger orientation toward specific channels.
The inventory's scores are obtained by ranking four words at a time ('1' indicating "least like me", '4' indicating "most like me"). Ten categories of four words determine the scores for each of the four mind-styles. Each word corresponds to a particular mediation channel, and when summed, give a measure of a person's propensity for operating within specific learning channels.
Gregorc (1982a) divides the scores received on The Style Delineator Trademark into three levels:
1) Strong orientation towards qualities associated with the particular channel (or pointy-headedness), indicated by a score of 27-40;
2) Moderate ability, indicated by a score range of 16-26 on any one mediation channel: and
3) Minimal capacity (stubby pointedness), indicated by a score of 10-15 in a specific channel. According to Gregorc (1985) approximately 60% of the channel's characteristics are observed in people with a score of 27 or over; hence, 27 has been selected as the cut-off point for "pointy-headedness". Another major cut-off point, 15, has been identified as an indication of "stubby pointedness" because very few of the channel's characteristics are observed in people with scores below 15 (Gregorc, 1982a).
Learner characteristics
(Unless otherwise stated, information presented in this section is cited from Gregorc's book An Adult's Guide to Style, 1982b)
People who are dominant CS are usually practical, thorough, well-organized and prefer quiet, structured environments. CS individuals tend to perceive reality as the concrete world of the physical senses, and think in a sequential and orderly fashion. The CS can detect the most minute details, working with the exactitude of a machine (Gregorc, 1982a). The CS student is a perfectionist and prefers being told what to do. These learners do not like to go against the norm, view work as a job assignment, and enjoy being physically involved and active in lessons.
AS people consider themselves as evaluative, analytical, and logical individuals with a preference for mentally stimulating, orderly, and quiet environments. The AS has an academic-type mind which is driven by a thirst for knowledge. To an AS, "knowledge is power", and the ability to synthesize and relate concepts enables the AS to transmit ideas (both through the spoken and written word) intelligibly and eloquently. AS learners thrive on teachers who are experts in their area of interest, learning well through lecture-style teaching.
AR individuals are highly focused on the world of feeling and emotion, and are sensitive, spontaneous, attuned, person-oriented people. Thought processes of AR individuals tend to be nonlinear, multidimensional, emotional, perceptive, and critical. AR people prefer active, free, and colorful environments. ARs thrive on building relationships with others and, as learners, dislike extremely structured assignments.
Finally, CR individuals process information in three-dimensional patterns and think intuitively, instinctively, impulsively, and independently. CR people prefer competitive, unrestricted, and stimulus-rich environments. CRs can be risk-takers and can easily jump to conclusions, often correctly. Such individuals are divergent thinkers, thriving in environments which engender exploration. CR learners do not need many details to solve a problem, instead operating according to personally constructed standards.
Overall, everyone has the capacity to learn within each of the above channels; no one is a "pure type" (Gregorc, 1982b, 41). Therefore, The Style Delineator Trademark is a tool which:
"provides an individual with a key to understand better the subtle and potent qualities of the mind, (their) behavior, the behavior of others and the demands placed upon individuals by their environment." (Gregorc, 1982b, 41)
Learning styles and CAI: An overview
CAI and learner profiles
A study conducted by Friend and Cole (1990) discovered that sensing-thinking individuals (dimensions correlated with CS and AS) responded more favorably to CAI than did intuitive-feeling types (dimensions which are correlated with AR). Friend and Cole postulated that intuitive-feeling types require more human interaction to achieve desired learning outcomes, and that CAI may not be suitable for all learners.
Enochs et al. (1985) found that concrete learners (as determined by Kolb's Learning Style Inventory) learned better from a CAI session than did abstract learners. Pritchard (1982) gives further support for the claim that CAI may not accommodate all learning styles equally. In his article on educational computing, Pritchard explained that CAI is suited best for individuals with an affinity for accuracy and attending to detail. Moreover, the researcher claims individuals with certain learning styles may be more partial to learning from computers than would others, and that people who have a preference for CAI usually enjoy working alone (see also Wood et al., 1996).
In keeping with CAI and learner profiles, Hoffman and Waters (1982) stated that CAI is suited best for individuals who: "...have the ability to quietly concentrate, are able to pay attention to details, have an affinity for memorizing facts, and can stay with a single track until completion" (p. 51).
Dunn and Dunn (1979) reported that certain students may only achieve through selected instructional methods (eg, CAI, whole-group instruction, etc.), and that matching can significantly improve academic achievement. Dunn and Dunn asserted that students who are motivated, require specific instructions, are sequential, and enjoy frequent feedback generally do well with programmed learning such as CAI. However, students who are kinesthetic, peer-oriented learners (ie, AR learners) may not be engaged adequately by the same method of instruction.
The computer as a matching tool
Although the idea of matching instruction to students' learning styles has been supported in the literature (eg, Butler, 1984; Hettiger, 1988), it can be difficult for educators to match teaching and learning styles in the traditional classroom. It has been argued that effective CAI can correct for many teachers' inability to meet the needs of all learners (Schlechter, 1991). Yet, CAI may not be the preferred mode of learning for all students. According to Gregorc (1985), sequential students (CS and AS) tend to prefer CAI because the computer is seen as an extension of the sequential person's mind. Random individuals (CR and AR) require environments which are flexible and provide opportunities for multidimensional thinking (Butler, 1984). AR individuals, in particular, are inherently social and enjoy learning with others (Butler, 1984). It is apparent that traditional CAI does not always provide such an environment for this group of learners.
Unlike the teacher who may be able to troubleshoot and modify lessons to meet the specific learning needs of the student, the computer is only as good as the program that has been created for it; and, as Gregorc (1985) wrote:
"Students who cannot adapt to the demands of the medium are 1) denied access to the content and goals, and 2) are vulnerable to possible psychological damage if they cannot free themselves of the medium .... Children can therefore become victims of a medium which is offensive to them. They are at the mercy of the machine." (p. 168)
Moreover, because a computer requires sequential thinking in order to gain access to its content (Gregorc, 1985), many CR and AR individuals may become flustered and agitated when problems arise with the medium. Gregorc (1985) warns that problems such as "burnout" and other mental and physical ailments can arise if individuals are made to accept certain media which are seen as adversive.
Butler (1984) claimed that technology, in general, places demands on the learner. The computer is often not inherently flexible, intuitive or adaptive, and may therefore restrict the behaviors and responses of the user. As a result, "learners can master such equipment only when they have mastered its invisible demands" (p. 27). The author concluded that "an all out movement towards computer-aided instruction is bound to leave many students behind" (p. 29).
In an effort to ensure that all learners can benefit from computer technology, Gregorc (1985) recommended that leaders (eg, teachers, administrators, employers, professors) provide human mediators who can correct for matching problems that may arise from using an inappropriate and potentially invasive learning medium.
Further support for the notion of instructional matching was voiced by Burger (1985). In her opinion, CAI may be overused to a certain degree:
"Requiring all students to use [Computer-Aided Instruction[ may not be in the best interest of the student. The matching of the teaching style of the specific computer program and the learning style of the student must be considered." (p. 21)
Inasmuch as the computer can be a powerful learning medium, the machine is limited in its capacity to modify instruction to meet individual needs (Enochs et al., 1985; Gregorc, 1985). While there have been advances in the area of intelligent tutoring and adaptive interfaces (see Steinberg and Gitomer, 1992; Mills and Ragan, 1994), some of the software interfaces that are currently available are unintuitive and unnecessarily complex (Mitta and Packebusch, 1995). Wallace and Anderson (1993) explained "designing good computer interfaces has proven a formidable challenge" (p. 259).
Hence, many students may be forced to adapt and harmonize with the computer (ie, style flex) in order to attain desired learning goals.
"These inanimate objects lack empathy. Machines cannot sense the opportunities, qualifications, fears or problems. Nor can they sense the pressures from the forced intimacy we demand between learners and the media. Without compassion, there are no adjustments or alternative approaches offered. There is no sense of harm or restraint as the frozen medium makes its learning demands for sympathetic resonance. School personnel must recognize these facts when purchasing machines." (Gregorc, 1985; 168)
Butler (1984) elucidated the notion of mismatching learning styles and media discussed by Gregorc (1985). "Instructional technology biases the way information is presented, and demands, to varying degrees, that we use certain mediation channels" (p. 237). In other words, the use of technology may systematically discriminate against certain learners who are unable to match learning styles with the medium. Just as the lecture approach in education is best suited to AS learners (Gregorc, 1982b), so the computer may be better suited to certain learning styles.
Method
Problem
Research suggests that CAI may have a limited ability to accommodate users with varying learning styles (eg, Butler, 1984; Gregorc, 1985; Hettiger, 1988; Cordell, 1991). Based on the limited number of studies examining the learning styles and CAI, it would appear that sequential students fare better with most CAI applications than do random students.
Yet, in any given classroom, one half of students have a propensity for learning best in the random mediation channel (O'Brien, 1994). When coupled with the fact that the use of hypermedia information systems with little or no teacher guidance is increasing in education (Small and Grabowski, 1992), the need for continuing research in the area becomes apparent. Specifically, further research in the area of learning styles and human-computer interaction is needed in order to understand better the influences of individual differences and CAI.
Research questions
Since this appears to be the first study to investigate the Gregorc mediation channels and their impact on learning from, and interacting with, a CAI program, no hypotheses have been made. I, instead, explored the following research questions:
1. Will learning outcomes differ significantly based on student cognitive learning styles as measured by The Gregorc Style Delineator Trademark?
2. Will human-computer interaction behaviors (ie, time spent on the program, navigation, events recorded, video, tools and lesson preference) differ significantly based on student cognitive learning styles as measured by The Gregorc Style Delineator Trademark?
3. Will differences in entry level domain knowledge affect learning outcomes above and beyond that of learning style?
Subjects
Seventy University of Calgary undergraduate volunteers (26 males, 44 females) participated in the study. The following is a breakdown of students by Faculty: Nursing = 18; Kinesiology = 20; Education = 13; Other = 19.
Treatment
To investigate differences between participants, learning style groups received the same treatment. For the purposes of this study, the one-rescuer adult CPR procedure was used to collect data. Content for the CAI program was vetted for accuracy and validated by a three member committee comprised of experienced CPR Instructor Trainers.
The entire experimental sessions took two hours to complete for each group of approximately 15 participants. One hour was devoted to assessing and interpreting learning style scores. The second hour was dedicated to the CAI session.
Following completion of the workshop, the researcher explained the interface to the participants so that each learner would be familiar with the features and options available to them during the CAI session. Participants then completed the on-line questionnaire (comprised of six demographical questions measuring participants' age, year of program, gender, comfort level with CAI, and CPR confidence level), the 20 question pre-test and the tutorial program.
No time restriction was imposed on the learners during the CAI session, as time was a variable under investigation. It was imperative that learners did not feel rushed to complete their learning in a stipulated time limit; similarly, a time restriction may have forced quicker learners to stretch out the CAI session to meet the time restriction. Participants worked independently on the computer, using headphones to listen to audio information.
CPR is a psychomotor skill requiring knowledge of theoretical principles, procedural steps and performance principles and practices. The computer program instructs and tests both theory and understanding of procedures necessary to perform CPR, leaving motor performance instruction and evaluation to a certified CPR instructor.
Both the pre-test and the post-test were comprised of ten knowledge-type questions, five comprehension questions and five application questions (20 multiple choice questions in total). Questions covered one-rescuer Cardiopulmonary Resuscitation (CPR) guidelines and procedures as stipulated by the Heart and Stroke Foundation of Canada's Emergency Cardiac Care Committee.
Construct validity for the test items was determined by a three member CPR instructor trainer review committee. The test-retest reliability alpha coefficient for the examination was determined to be 0.86 for the pre-test and 0.89 for the post-test.
Instrument
The Gregorc Style Delineator Trademark is a widely-used measure of assessing cognitive learning styles (O'Brien, 1994). The assessment tool was selected, in part, for the following reasons:
• Easy to administer
• Easy to interpret
• Self-scoring battery
• Relatively quick to administer and complete
• Inexpensive
• Discrete, easily reportable scales
• The only inventory available with a technical manual for administrators
• Validity and reliability measures have been supported by research (eg, Gregorc, 1982a)
Joniak and Isaksen (1988) examined the internal consistency of The Style Delineator Trademark. The data revealed alpha coefficients raging from 0.23 to 0.66, below that which was reported by Gregorc (1982 a). O'Brien (1990) found similar results. Using a sample size of 263 undergraduate students, O'Brien reported alpha coefficients ranging from O. 51 for the AS scale to 0.64 for the CS scale, but concluded that internal consistency scales meet minimal requirements for factor definition (O'Brien, 1990).
Gregorc (1982a) reported test-retest alpha coefficients of 0.85 to 0.88. In addition, Gregorc (1982a) published internal consistency reliability coefficients ranging from 0.89 for the AS scale to 0.93 for the AR scale, and predictive validity correlations ranging from 0.55 to 0.76 (all figures significant at the p < 0.001 level). Results were based on a sample size of 110 participants.
Quality of material
With any CAI program, quality of material presented is always an issue. According to Rushby (1997), three factors are essential to ensuring that a CAI program meets acceptable standards: content is accurate and up to date, the program is rigorously tested to ensure minimal running errors, and the program is free from typographical errors.
The program used for the study meets all three quality standards. A CPR committee verified the accuracy of content ensuring that it was up to date and a reflection of current practices in CPR. The program went through a lengthy six month beta testing stage at which time all errors were found by a focus group and addressed by the programmer. To this date, we have had no error reports from any Nursing schools who have purchased the program. In terms of typographical accuracy, an editor with 10 years in the area was used to verify the textual consistency and grammatical structure of the program lessons and narration tracks.
Independent measures
Learning styles--Subjects' highest learning style scores (as determined by The Style Delineator Trademark) were treated as a measure of dominant learning style. The following is a breakdown of subjects by dominant learning style score: CS = 20; CR = 20; AS = 14; AR = 16.
Domain knowledge level--Pre-test scores were used as a measure of domain knowledge levels and for learning outcome achievement analysis.
Dependent measures
A program audit trail file was created for the purposes of this study to track participants' patterns of learning. Together with the pre-test score, learning style scores, and the preliminary survey information, the audit trail file also stored detailed information (eg, which tools and video options were accessed on which screens, and continuous time reports). The term "patterns of learning", referred to in a study by Liu and Reed (1994), are used in this study to describe human-computer interaction indices. These indicators are listed below:
Total time (in minutes) to complete the tutorial--Participants were given no time restriction to move through the tutorial program; hence, time scores varied from subject to subject.
Navigation trend--Participants' patterns of movement through the tutorial were determined by a numerical score. The tutorial consisted of 15 instructional screens detailing the discrete steps in performing CPR. The audit trail recorded navigation by assigning a '+1' value when the next screen button was selected, and a '-1' value when the previous screen button was selected. For example, if a subject were to move through the procedures in a linear fashion, a score of 14 would be assigned (14 'next screen' selections x 1). If a subject were to go back three screens while covering all 14 steps, a net score of + 11 would be assigned by the audit trail file ({14 'next screen' selections x 1} + {3 'back screen' selections x -1}).
Total number of tools used--The frequency with which the subject accessed the program tools (note pad, search tool, index tool and glossary) was reflected by this measure.
Total number of video events--The number of times the user accessed video controls 'Play', 'Pause', 'Rewind', and 'Volume' was indicated by this total. It is important to note that the video, by default, played automatically upon moving to a new screen; hence, the score reflected in this category indicated the number of video events above and beyond the standard score of 15 (or 15 video play options).
User preference for instructional sequence--The tutorial program was comprised both of a 15 step tutorial sequence and a video review section. The video review section summarized all video steps covered in the tutorial. Learners could choose to watch the review video prior to, or following, the tutorial. A code of '1' was assigned to those participants who chose the 'Review Video' option first; an indicator of '2' was assigned for learners who chose to move through the tutorial first.
Total number of events--This measure indicated the level of user interaction. This number was derived by adding the total number of tools used, videos accessed, and navigational events. A low number reflects user passivity.
Post-test results--Learners completed a 20-question multiple-choice post-test. The results from the post-test were used as a dependent variable for the purposes of achievement analysis.
Results
The alpha level representing statistical significance was set at the p < 0.05 level. Results that have lower or higher p values will be reported as such. Data were analysed using SPSS 6 and BMDP IV.
Learning outcomes
To explore whether learning outcomes were influenced by dominant learning style groups, a two-way ANOVA (2 x 4 factorial analysis) was conducted. The data revealed a significant main effect for the pre-test and post-test means over time (F[sub (1,66)] = 57.91, p < 0.001). There was also a significant interaction between learning style and learning outcome (F[sub (3.66)] = 20.11, p < 0.001). Figure 1 depicts the interaction between dominant learning style and learning outcome.
The mean test scores reveal that three of the four dominant learning style groups showed gains from the pre-test to the post-test. The AS group increased an average of 3.64 points (or 18%), displaying the highest gain of the three groups. CS and CR groups increased an average of about 2 points (or 10%). Interestingly, the AR group decreased from pre-test to post-test by an average of just over 2 points (or 10%).
In summary, the results indicate that there were significant differences in achievement between the four dominant learning style groups. Dominant learning styles, it would appear, affected the magnitude and direction of the differences in the pre-test and posttest results.
Human-computer interaction
To investigate the effects of dominant learning styles on human-computer interaction, a MANOVA was conducted using six patterns of learning as the dependent variables and dominant learning style as the independent variable. Results indicated that there was not a significant effect for patterns of learning by dominant learning style (1 = 0.6 6, F = 1.51, p = 0.09).
The data suggest that only one pattern of learning, navigation style, differed significantly at the p < 0.01 level. Results of a post-hoe Scheffe test indicated that the AS and CR group means were significantly different from the AR group mean. It would appear that the AR group was the least linear of the four dominant learning style groups, recording a mean score of just over 10 points. Table 1 delineates the mean scores for the six patterns of learning.
Results suggested that AR participants spent less time with the program, used less video and made fewer interactions with the computer than did the other three dominant learning style groups. In contrast, AS subjects tended to spend more time with the program, used a higher number of tools and interacted to a higher degree with the computer than did the other three groups. Although not statistically significant, mean scores do suggest some interesting differences between dominant learning style, to be explored in future studies.
The overall lack of significant differences between dominant learning style and patterns of learning measures of time, total events, tools, video and lesson preference suggests that learning styles, as measured by The Gregorc Style Delineator Trademark, did not significantly affect the way in which learners interacted with the computer-aided instructional software.
Domain knowledge
Examination was also conducted to ascertain whether content knowledge affected learning outcomes above and beyond that of learning styles.
Pre-test results indicated that there were disparities in the entry-level domain knowledge participants possessed. An ANCOVA was conducted to identify the influences of learning style on post-test scores, while controlling for (or equalizing) differences in pretest knowledge demonstrated by the four learning style groups.
The ANCOVA showed a significant effect for pre1test (b= 0.79; t = 8.4: sig t = 0.001). However, learning styles still retained a significant influence on post-test scores (F[sub 14, 65] = 19.58, p < 0.001). Furthermore, the adjusted r2 value of O. 52 suggested that dominant learning styles alone explained 52% of the variance in post-test scores, after controlling for the influences of pre-test scores.
Discussion
Learning outcomes
In terms of learning outcomes, the data suggests that, as a group, participants showed an increase from pre-test to post-test, statistically significant at the p < 0.001 level. This would suggest that the tutorial program led to gains in Cardiopulmonary Resuscitation (CPR) knowledge.
Once subjects were distilled into their dominant learning style groups, however, the data revealed a significant interaction effect between learning style group and achievement levels. In short, learning styles significantly affected both the magnitude and direction of achievement levels. The AS group made a gain of close to four points, while the CS and CR groups made modest gains of about two points from pre-test to post-test. Interestingly, the AR group decreased an average of more than two points from pre-test to post-test, a result which has significant implications for CAI if findings are supported by future studies. The question of why the AR group decreased from pretest to post-test will be discussed further.
Theoretical explanations for achievement differences
According to Gregorc (1982b), individual learning styles influence preference for method of instruction. Butler (1984) and Gregorc (1985) believe that dominance in CS and AS mediation channels predisposes the individual to having a preference for working with computers (be it in the capacity as a computer programmer, or as a learner using CAI software). Randoms are said to find working with computers frustrating (see Gregorc, 1985, 202,203). In this exploratory study, CR learners did well with the computer program, but ARs did not. Although some of the content covered in the tutorial program required linear processing, CR individuals did well compared with AR learners.
Hence, one cannot argue successfully the point that the content, and not the computer medium, was responsible for the differences in learning outcomes. It would appear that possessing the abstract and random qualities together made for a less successful computer-aided learning session. Butler (1984) explained that AR individuals prefer human contact throughout the learning process, and enjoy tasks requiring verbal, multidimensional responses; certain forms of CAI, therefore, may be unsuitable for these learners.
Results from the present study are consistent with results reported by Davidson et al. (1992). The researchers found that AR individuals, enrolled in a computer applications university course, showed significantly lower achievement levels than did the other three Style Delineator Trademark groups. AS individuals showed the highest gains in the course, indicating their ability to work well with computer technology. The only significant differences in methodology between the two studies are that Davidson et al. utilized course assignments as a measure of success, whereas this study used pre-tutorial and post-tutorial results as an indicator of achievement levels.
Further exploration: differences in pre-test group scores
It is interesting to note that the pre-test means were different between the four learning style groups. While the AS group had a mean pre-test score of just 10, the AR group had a mean pre-test score of 15. Such a sharp contrast may be explained by a number of factors.
a) Varied CPR background: It appears that the CPR course background varied between the four learning style groups (CS = 2, CR = 1.4, AS = 1.5, and AR = 2.8). An ANOVA was conducted to investigate whether the differences between groups were significant. Results from the ANOVA indicate that differences in CPR course backgrounds were statistically significant (F[sub (3.66)] = 5.16, p < 0.01). A post-hoc Scheffe test was used to ascertain which groups were significantly different. The AR group's mean CPR course background score was deemed significantly different from the CR and AS groups' score. Such contrastingly different group course backgrounds may explain why the AR group had such a high pre-test score.
b) CPR confidence: The data suggest that the four dominant learning style groups differed in CPR perceived confidence. As may be recalled, the preliminary survey asked participants to rate their CPR level of confidence (using a Likert-Type Scale; 1 being very confident, 5 being not confident at all). Group mean scores (CS = 3.3, CR = 3.4, AS = 4.1, AR = 2.7) were significantly different (F[sub (3.66)], = 2.71, p < 0.05), indicating differences in CPR confidence by learning style groups. Scheffe post-hoc analysis shows that the AS group mean was significantly different from the three other groups.
The majority of CS and All group participants indicated that they were pursuing Nursing degrees; hence, regular CPR certification is required, and may explain the variation in learning style groups' scores. It is not surprising, then, to see disparities in the mean pre-test scores across groups. The AS group, the majority of whom were from Education or other faculties, had taken the least number of CPR courses of the four groups, and had the lowest confidence level in their skills. This group also displayed the lowest pre-test score. Similarly, the All group recorded taking the most number of CPR courses of the four groups, and reported the highest CPR confidence levels.
The question of entry level differences in domain knowledge
It can be argued that there were obvious background disparities between the four groups upon entering the study. While it is true that groups did differ based on their pre-test scores, the ANCOVA showed that groups still differed significantly on the posttest when controlling for pre-test differences in knowledge. Hence, it would appear that achievement in the CAI session was affected most significantly by cognitive learning styles. Although three of the four dominant learning style groups learned from the CAI lesson, AR learners consistently did not (two dominant AR subjects increased scores from pre-test to post-test, nine decreased scores, and four showed no change).
Patterns of learning
Patterns of learning, indicating human-computer interaction behaviors, were not significantly different between dominant groups. Although five of six patterns of learning were not significant at the p < 0.05 level, three indices showed some interesting between-group differences.
The mean scores revealed in Table I show some interesting differences between groups. It would appear that the All group spent, on average, less time in the program, used less video, and recorded fewer events than did the other three dominant learning style groups. The AS group showed diametrically opposite behaviors, spending more time in the program, interacting to a higher degree, and using more video than did the other three groups.
Furthermore, the AR group recorded a significantly different mean navigation value (significant at the p < 0.001 level) from the other groups. AR participants, on average, recorded a mean value of around ten points, indicating some degree of non-linear movement (either moving backward to review previous screens, or using the index tool to jump from step to step). The other learning style groups showed values which hovered around the expected level of 14.
Patterns of learning as indicators of achievement
Upon closer inspection of the audit trail print-outs, it would appear that many AR participants missed entire screens while traversing from step to step. One participant missed five screens, jumping from step 3 to step 9, moving through the remaining six steps, and then finishing back with step 8. It is not known if the subject knew the content covered by the missed screens; however, it is clear that such an approach to learning CPR--a procedure that requires linear movement through a pre-determined sequence of steps--may interfere with current learning, and may very well interfere with previous learning.
One knowledge-type test question, for example, asked participants to put the steps of CPR in order. A correct response for this question required the learner to have moved through the program in a linear fashion. Skipping steps and moving to previous screens may have interfered with the learning required for a correct answer to these types of questions.
When teaching CPR in the traditional classroom setting, it would be detrimental for the instructor to move from step 1 to step 12, and then back to step 2. Regardless of the type of learning style one has, certain materials require sequential processing. Excessive and inappropriate use of the index tool--a tool that allows the user to jump from any given step to another--may have contributed to cognitive interference in many of the AR subjects.
According to Milheim and Azbell (1988) cited in Small and Grabowski (1992), systems that give the user control over the learning process are empowering for some and destructive for others. Small and Grabowski warn that too much user control can lead to navigation decisions resulting in either skipping pertinent content or leaving the tutorial program before all content has been thoroughly covered (also see Schroeder, 1994). Castelli et al. (1996) discovered that many users of hypermedia "get lost" in hyperspace. The notion of becoming disoriented due to incessant "jumping around" is consistent with findings from Hammond (1989).
The overall lack of interaction recorded by AR subjects (based on low events score, video use and time in program) may have resulted from a lack of interest in the CAI session. Attitudes towards computers can be a significant indicator of student achievement with the computer (Brudenell and Stewart, 1990). A breakdown of the computer attitudes survey question by dominant learning styles indicated that close to 60% of AS subjects reported being comfortable with using the computer. In contrast, only 36% of AR subjects felt comfortable with computer technology. Over 50% of CS subjects and 55% of CR subjects felt comfortable with using the computer. Hence, AR subjects were less likely to be comfortable with using the computer than were the other learning style groups.
Motivation is also the key to any type of self-paced CAI session, according to findings from Keller (1968). Keller, in his essay on computers in the school, warned of the dangers of leaving important instructional decisions to students. Students may neither have the metacognitive abilities nor the motivation to select appropriate paths for achieving desired learning goals. Small and Grabowski (1992) found that high motivation levels led to subjects spending more time with the computer program, and subsequently contributed to higher learning outcomes. Low motivation levels had an inverse effect.
In direct contrast to the AR group, the data revealed that AS subjects were highly engaged in the CAI lesson. Although not statistically significant, all patterns of learning appeared to indicate that these subjects interacted to a high degree with the program. Such enthusiasm and diligence may have contributed to the higher achievement levels observed.
In terms of patterns of learning, Liu and Reed (1994) also found that, overall, human-computer interaction measures were not significantly affected by learning styles under investigation in their study. However, field independence (a propensity for thinking analytically and logically) was linked to using the index tool, and field dependence (thinking in a more global way) was correlated with using more video. In addition, field dependent subjects used the courseware significantly more than did field independent participants. (It should be noted that comparisons cannot be made between field dependence/independence and Gregorc's mediation channels. There is no research to support relationships between these dimensions of cognitive learning styles.)
Recommendations for the responsible use of technology in education
The following recommendations are meant to be used as guidelines for the successful implementation of computer technology, and are based on findings from this study. It remains essential for a clearly stated list of recommendations, outlining proper computer use, to be published. In this way, all individuals are afforded the right to learn in the way that suits them best.
1. Educators should closely monitor--and mediate where necessary--all computer instruction. Students should have clear and identifiable tasks to complete, and learning outcomes should be measured periodically. This is consistent with the views expressed by Greenberg and Pengelby (1989).
2. Students should be asked to express their views towards CAI through the use of a teacher-constructed survey. Furthermore, if teachers have an interest, they should ascertain the learning styles of their students, and provide insight on how learning styles influence students preferences for instruction. Learning style scores could be used in conjunction with preference surveys to identify potential matching problems.
3. Opportunities for group work should be given to those students who are hesitant to work on the computer alone. Research shows that AR students enjoy working with others and sharing ideas during the learning process (Ross, 1998). Since the focus shifts from being intimate with a machine to working collaboratively with a group, the potentially negative effects of CAL for these individuals may be masked and/or lessened (Ross, 1998).
4. Government Departments of Education should remain cautious with sweeping decisions to convert entire curricula onto electronic media (as was mentioned in the article by Dwyer, 1996). The goals of such a process should be weighed against the potential problems (eg, alienating certain learners).
5. To avoid alienating a certain learning style group, educators should continue to incorporate a number of different teaching strategies into their lessons. If a particular student is unable to learn from the computer, instructors should provide alternative ways for content to be delivered.
Conclusion
CAI is rapidly becoming one of the most influential media of instruction in educational environments. However, findings from this exploratory study indicate that CAI, as an instructional methodology, may not be suitable for all learners. While computer-aided instruction has tremendous potential to provide teachers and industry with a powerful educational tool, educators must be cognizant of inherent differences which exist between learners--differences such as cognitive learning styles. Results from this exploratory study suggest that some learners (AR learners in particular) may have difficulty adapting to certain forms of computer-mediated learning.
Studies continue to support the need to critically evaluate this ubiquitous tool which has permeated the classroom and homes more quickly than most other technologies have in the past (see Schlechter, 1991). As more research is conducted in the area of CAI, information regarding appropriate and educationally sound uses for the CAI will become available.
It remains essential, then, that the computer continue to be used as a tool for supplementing classroom instruction. Some learners may need greater support and guidance from the teacher, while others may be able to learn from the computer relatively independently. Thus, teachers should not assume every student will automatically benefit from computers in the classroom. There remains the need for interpersonal contact and guidance to ensure that all students attain their learning potential.
Limitations of study and opportunities for further research
Results from this study have some significant implications for computer-aided instruction, if supported by further research. If replicated, a number of considerations should be followed to improve the generalizability of the results.
1. This study used the traditional goals--tutor--test approach to gather data from participants. A study should be conducted with a computer program adhering to a different learning model (eg, discrimination learning, simulation, intelligent tutoring system, etc.). If results prove to be consistent with those of this study, then it can be more conclusively argued that CAI may not sufficiently accommodate all learners equally.
2. It is not clear whether low motivation, as indicated by AR subjects' patterns of learning, was due to the computer or to the content presented in the program. Since the majority of AR subjects were in a Nursing program--and are assumed to enjoy medical procedures and training--it is questionable that the CPR content, in and of itself, led to disparities in pre and post-test scores. However, further research may shed some light on the question of subject matter versus computer instruction.
3. This study used content that was familiar to most, if not all, subjects (as indicated by the relatively high pre-test mean). Inasmuch as it was desirable to have subjects who had varying levels of domain knowledge for the purposes of exploring one research question (namely, domain knowledge as a measure of individual differences), further research should be conducted using a subject area that is unfamiliar to all participants. In this way, learning outcomes could be more accurately measured.
4. Subjects were expected to interact with the CAI tutorial program for a relatively short period of time (about 30-45 minutes). Further research should explore the effects of learning styles and other individual differences on CAI using a one week to one month study time frame.
5. This study used a program requiring the learner to move though content in a somewhat linear way. This study should be replicated using a program with content that can be learned in a non-linear manner.
6. Further research should explore the impact of group learning on learners who may be pre-disposed to encountering difficulty with the computer. AR learners matched with other AR learners may be more successful when using the computer to learn. Research should determine what type of cooperative groupings work best for "imperiled" computer users.
The previous recommendations for future research have a common theme: there remains a need for more research in the area of learning styles and human-computer interaction. The literature suggests that there are definite learning preferences which are consistent with learning style profiles. It follows, then, that CAI may not be suitable for all learners. Unfortunately, the relationship between learning styles and computer-mediated learning needs to be explored in greater detail before more conclusive statements can be made.
Table 1: Mean pattern of learning scores by dominant learning style score
Legend for Chart:
A - Learning
B - Style
C - N
D - Mean
E - SD
A B C D E
Time
CS 20 28.95 7.06
CR 20 26.85 9.27
AS 14 30.64 13.37
AR 16 23.31 9.10
Navigation
CS 20 11.95 3.56
CR 20 14.05 3.87
AS 14 14.42 4.14
AR 16 10.06 4.74
Events
CS 20 53.70 20.86
CR 20 60.60 32.27
AS 14 60.71 28.00
AR 16 44.06 15.08
Lesson
preference CS 20 1.35 0.49
CR 20 1.35 0.49
AS 14 1.64 0.50
AR 16 1.37 0.50
Tools
CS 20 5.10 5.92
CR 20 5.80 7.43
AS 14 7.00 8.18
AR 16 4.00 3.06
Video
CS 20 8.00 8.78
CR 20 16.65 21.64
AS 14 11.57 10.97
AR 16 7.81 6.97
GRAPH: Figure 1: Interaction between tutorial effect and dominant learning style group
References
Brudenell I and Stewart C (1990) Adult learning styles and attitudes towards computer-assisted instruction Journal of Nursing Education 29 (2) 79-83.
Butler K (1984) Learning and teaching styles in theory and practise Gabriel Systems Inc., Maynard, MA.
Castelli C, Colazzo L and Molinari A (1996) Getting lost in hyperspace: Lessons learned and future directions CD-ROM Proceedings from the annual ED-MEDIA/ED-TELECOM conference Article No 208.
Cordell B J (1991) A study of learning styles and computer-assisted instruction Computers Education 16 (2) 175-183.
Davidson G V, Savenye W C and Orr K B (1992) How do learning styles relate to performance in a computer application course? Journal of Research on Computers in Education 24 (3) 349-358.
Dunn R D and Dunn K J (1979) Learning/teaching styles: Should they/can they be matched? Educational Leadership 36 (4) 238-244.
Dwyer V (1996) Surfing back to school Madean's Magazine 26 August 40-46.
Enochs J R, Handley H M and Wollenberg J P (1984) The relationship of learning style, reading vocabulary, reading comprehension and aptitude for learning to achievement in the sell-paced computer-assisted instructional modes of the yeoman "A" school at the Navel Technical Training Center. Meridian. Paper presented at the annual meeting of the Mid-South ERA, New Orleans.
Friend C L and Cole C L (1990)Learner control in computer-based instruction: A current literature review Educational Technology November 47-49.
Greenberg HI J and Pengelby R M (1989) A conceptual basis for the role of the microcomputer in the teaching and learning of college math in Maurer H (ed) Computer-Aided Learning: International Conference ICCAL (second ed) Springer-Verlag.
Gregorc A F (1979) Learning/teaching styles: potent forces behind them Educational Leadership 36 (4) 234-236.
Gregorc A F (1982a) Gregorc Style Delineator: Development Technical and Administration Manual Gregorc Associates Inc., Columbia, CT.
Gregorc A F (1982b) An adult's guide to style Gregorc Associates Inc., Columbia, CT.
Gregorc A F (1984) Learning is a matter of style Vocational Education 59 (3) 27-29.
Gregorc A F (1985) Inside Style: Beyond the Basics Gregorc Associates Inc., Columbia, CT.
Hammond N (1989) Hypermedia and learning: who guides whom? in Maurer H (ed) Computer-Aided Learning: International Conference ICCAL (second ed) Springer-Verlag, 167-181.
Hawkridge D (1995) Do companies need technology-based training? in Heap N, Thomas R, Einon G, Mason R and Makay H (eds) Information Technology and Society Sage, London, 182-210.
Hettiger G A (1988) Operationalizing cognitive constructs in the design of computer-based instruction Annual Meeting of the Association for Educational Communications and Technology H) 295 645.
Hoffman J L and Waters K (1982) Some effects of student personality on success with computer-assisted instruction Educational Technology 47-48.
Joniak A J and Isaksen S G (1988) The Gregorc style delineator: Internal consistency and its relationship to Kirton's adaptive-innovative distinction Educational and Psychological Measurement 48 1043-1049.
Keller R S (1968) Goodbye teacher Journal of Applied Behavior Analysis 1 79-89.
Liu M and Reed W M (1994) The relationship between the learning strategies and learning styles in a hypermedia environment Computers in Human Behavior 10 (4) 419-434.
Marquez M and Lehman J D (1992) Hypermedia user interface design: The role of individual differences in the placement of icon buttons Journal of Educational Multimedia and Hypermedia 1 (4) 417-429.
Milheim W D and Azbell J W (1988) How past research on learner control can aid in the design of interactive video materials. Paper presented at the National Conference for the Association for Educational Communications and Technology, New Orleans, LA.
Mills S C and Ragan T J (1994) Adapting instruction to individualize learner differences: a research paradigm for computer-based instruction Paper presented at the 1994 National Convention of the Association for Educational Communications and Technology ED 373 740.
Mitta D and Packebusch S J (1995) Improving interface quality: an investigation of human-computer interaction task learning Ergonomics 38 (7) 1307-1325.
Nelson W A and Palumbo D B (1992) Learning instruction and hypermedia Journal of Educational Multimedia and Hypermedia 1 287-299.
O'Brien T P (1990) Construct validation of the Gregorc style delineator: an application of LISREL 7 Education and Psychological Measurement 50 631-636.
O'Brien T P (1994) Cognitive learning styles and academic achievement in secondary education Journal of Research and Development in Education 28 (1) 11-21.
Price R V (1991) Computer-aided Instruction: A Guide Jot Authors Brooks/Cole, Pacific Grove, CA.
Rasmussen K and Davidson G V (1996) Dimensions of learning styles and their influence on perform-ante in hypermedia lessons CD-ROM Proceedings from the annual ED-MEDIA/ED-TELECOM conference Article No 385.
Reed W M (1996) A review of the research on the effect of learning styles on hypermedia-related performance and attitudes CD-ROM Proceedings from the annual ED-MEDIA/ED-TELECOM conference Article No 491.
Ross J L (1997) The effects of cognitive learning styles on human-computer interaction: Implications for computer-aided learning Unpublished Master of Science Thesis, The University of Calgary, Alberta, Canada.
Ross J L (1998) On-line but off course: a wish list for distance educators. International Electronic Journal for Leadership in Learning 2 (3).
Rushby N J (1997) Quality criteria for multimedia Association for Learning Technology Journal 5(2) 18-30.
Schroeder E E (1994) Navigating through hypertext: Navigational techniques, individual differences and learning Proceedings of Selected Research and Development Presentations at the 1994 National Convention of the Association for Educational Communications and Technology ED 373 760.
Schlechter T M (1991) Problems and Promises of Computer-based Training Army Research Institute for Behavioral and Social Sciences Ablex Publishing Corporation, Noorwood, NJ.
Small R V and Grabowski B L (1992) An exploratory study of information-seeking behaviors and learning with hypermedia information systems Journal of Educational Multimedia and Hypermedia 1 (4) 445-464.
Steinberg L S and Gitomer D H (1992) Cognitive task analysis interface design and technical troubleshooting Educational Testing Services, Princeton, NJ (Ed 384 677). Wallace M D and Anderson T J (1993) Approaches to interface design Interacting with Computers (3) 259-278.
Wood F, Ford IN, Miller D, Sobczyk G and Duffin R (1996) Information skills searching behaviour and cognitive styles for student-centered learning: a computer assisted learning approach Journal of Information Sciences 22 (2) 79-92.
~~~~~~~~
By Jonathan Ross and Robert Schulz
Jonathan L. Ross is a doctoral candidate in Educational Technology at the University of Calgary. He is also a senior instructional designer with Media Learning Systems in the Faculty of Education. His web site address is . Robert Schulz is a Professor in the Faculty of Management at the same university. Address for correspondence: Faculty of Education, The University of Calgary, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4. Tel: + 1 403 220 6490; email: jross@acs.ucalgary.ca
--------------------------------------------------------------------------------
Copyright of British Journal of Educational Technology is the property of Blackwell Publishers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: British Journal of Educational Technology, Jan99, Vol. 30 Issue 1, p5, 20p, 1 graph.
Item Number: 3251744 Result 44 of 127 [Go To Full Text] [Tips]
Result 46 of 127 [Go To Full Text] [Tips]
Title: Intuition-analysis cognitive style and learning preferences of business and management students.
Subject(s): LEARNING strategies; STUDENTS -- Attitudes; COGNITIVE styles; INDUSTRIAL management
Source: Journal of Managerial Psychology, 1999, Vol. 14 Issue 1/2, p26, 13p, 4 charts, 1 diagram
Author(s): Sadler-Smith, Eugene
Abstract: Studies the cognitive style and learning preferences of business and management education students. Methodology used on the study; Results and discussion.
AN: 1675008
ISSN: 0268-3946
Database: Academic Search Premier
Print: Click here to mark for print.
View Item: Full Page Image
View Links: Check linked full text sources
--------------------------------------------------------------------------------
[Go To Citation]
INTUITION-ANALYSIS COGNITIVE STYLE AND LEARNING PREFERENCES OF BUSINESS AND MANAGEMENT STUDENTS
A UK exploratory study
Keywords Human resource developmenl, Learning styles, Managemenl education, Training, United Kingdom
Abstract The study is an attempt to provide empirical elaboration, in the context of business and management education, for the "onion" and cognitive control models of cognitive style. Using, a sample of 226 business and management undergraduates the research explored the relationship between cognitive style measured using the cognitive style, index and learning preference. Using principal components analysis, three, categories of learning preference were discerned (active, reflective and individual. Correlational analysis and one way analysis of variance revealed statistically significant relationships between preferences for reflective and individual methods and cognitive style. The results provide some support for the "onion" and cognitive control models; the implications for business and management education, training and development are discussed.
Background
Introduction
Curry (1983), in her "onion" model, argued that learning style and cognitive style constructs may be grouped into three main types or layers resembling the skin of an onion. At the onion's core is the "central personality" dimension, remote from external influences and stable over time. Overlying this central core are:
(1) "cognitive personality style": a relatively permanent and stable characteristic measured by instruments such as the embedded figures test (Witkin, 1962);
(2) "information processing style"' a relatively stable set of responses to acquiring and assimilating information in a given learning situation (measured by means of instruments such as the learning styles Inventory (Kolb, 1984));
(3) the outer layer of the onion represents the behavioural manifestations of the interaction between these inner layers and the external environment through the expression of, for example, preferences for particular types of teaching and learning methods, such as self-direction, collaboration and dependence (Grasha and Reichmann, 1975) and specific approaches to learning in given environments and within particular assessment regimes, such as deep versus surface approaches to studying (Entwistle, 1988; Marton and Saljo, 1976).
Riding (1997) presents a "cognitive control" model (a theoretical elaboration of Curry) consisting of primary sources (knowledge, personality, gender and cognitive history), cognitive control (the wholist-analytical and verbal-imagery dimensions of cognitive style) and cognitive input (perception) and output (learning strategies). Like the onion model, it is an attempt to unify the relationship between apparently similar constructs. The aim of this paper is to examine, in the context of business and management education, the implicit proposition in the onion and cognitive control models, that learning preference is related to cognitive style. This has clear implications for:
• the planning and design of business and management education;
• training and development in organisational contexts through the matching (or mismatching) of teaching and learning methods to the cognitive style of the learner;
• the development of stylistic versatility (by complementing style with strategies).
Learning preferences
Learning preferences may be defined as an individual's propensity to choose or express a liking for a particular teaching or learning technique or combination of techniques (Sadler-Smith, 1996). From the work of Reichmann and Grasha, (1974) and Renzulli and Smith (1978) it is possible to synthesise three groups of learning preference:
(1) dependence: preference for teacher-directed, highly structured programmes with explicit assignments set and assessed by the teacher;
(2) collaboration: discussion-orientation and favouring group projects, collaborative assignments and social interaction;
(3) independence: preference for exercising an influence on the content and structure of learning programmes within which the teacher or instructor is a resource (Sadler-Smith and Riding, 1999).
The learning preference construct has not been as widely researched as learning style, approaches to studying or cognitive style. Learning preferences represent the outer skin of the "onion" and as such they are the most easily accessible but least stable of the constructs and are the interface between the internal world and external learning environment. Like "learning styles" and "approaches" (which may be considered as varieties of learning strategy), preferences are ways of dealing with the external world (see Figure 1). They differ from learning strategies in that the latter are ways in which the individual acquires and assimilates information, whereas the expression and operationalisation of learning preferences are the ways by which the learner attempts (by accommodating her/his preferences) to adapt to or cope with the demands of the external learning environment.
Figure 1 Learning preferences, styles, strategies and cognitive style
Sadler-Smith (1997) found statistically significant correlations between learning preferences and learning style (learning styles questionnaire-Honey and Mumford, 1992) and approaches to studying (revised approaches to studying inventory - Entwistle and Tait, 1994) but not between learning preference and cognitive style (cognitive styles analysis -Riding, 1991). The present study will explore the latter, using an alternative model of cognitive style.
Cognitive style
Messick (1984, p. 5) described cognitive style as "consistent individual differences in preferred ways of organising and processing information and experience". Steinberg and Grigorenko described it as representing "a bridge between what might seem to be two fairly distinct areas of psychological investigation: cognition and personality" (Steinberg and Grigorenko, 1997, p. 701). A number of assumptions relating to cognitive style may be identified:
(1) it is concerned with the form rather than the content of information processing;
(2) it is a pervasive dimension that can be assessed using psychometric techniques;
(3) it is stable over time;
(4) it is bipolar;
(5) it may be value differentiated (i.e. styles describe "different" rather than "better" thinking processes) (Sadler-Smith and Badger, 1998).
One model of cognitive style which satisfies these criteria for a "cognitive style" and lends itself to research in a business and management context is the intuition-analysis dimension (Allinson and Hayes, 1996). The style models of Allinson and Hayes and Honey and Mumford may be traced back to their origins in Jungian psychological types. Hurst et al. (1989) in a useful, but concise, summary, described the "types" in terms of information gathering modes (intuition versus sensation) and information evaluation modes (feeling versus thinking) to give four basic types (intuiting-feeling; intuiting-thinking; sensing-feeling; sensing-thinking). Intuition was defined by Bunge (1983, p. 2A8) as "that ill-defined ability to spot problems or errors, to "perceive" relations or similarities ... in short to imagine, conceive, reason or act in novel ways". Analysis, on the other hand, is often presented as the antithesis of intuition: "to analyse ... is to exhibit [an object or system's] components, environment (or context) and structure (organisation)" (Bunge, 1983, p. 219). Hurst et al. (1989) went on to speculate that differences in preferences for each type of thinking may be related to hemispherical differences in the brain: "sensing and thinking are left hemisphere related and intuition and feeling right hemisphere related" (Hurst et al., p. 91). This echoed the views of Mintzberg (1976): "in the left hemisphere of most people's brains the logical thinking processes are to be found ... in contrast the right hemisphere is specialised for simultaneous processing; that is it operates in a more holistic ... way". More recently, Mintzberg (1994a, p. 114) has re-stated these ideas in the context of strategic planning, arguing that the planning function in organisations is populated by two types of person: the analytic ("left-brained') thinker and the creative ("right- brained") thinker. He expressed the view that organisations need both types in "appropriate proportions" (see also Leonard and Strauss (1997) on the "whole-brained organisation"). Like Hurst et al. (1989) and Mintzberg (1976), Allinson and Hayes (1996) speculated on hemispherical differences in the brain as a possible basis for cognitive style differences (stemming from the work of Sperry and others - see Nebes and Sperry (1971); they too use the term "intuition" to describe "right brain" thinking (i.e. immediate judgement based on feeling and the adoption of a global perspective) and "analysis" for "left brain" thinking (i.e. judgement based on mental reasoning and a focus on detail). "Style" in this context is the dominance of one mode of thinking over the other and describes "different" rather than "better" approaches to learning, problem solving, etc.
It should be noted that the attribution of differences in analytical versus intuitive behaviour to hemispherical differences in brain functioning should, in the absence of firm neuro-physiological evidence, be treated metaphorically rather than literally (see Riding et al. (1997) for a neuro-physiological study of cognitive style). Finally, Allinson and Hayes' intuition-analysis dimension of style may be considered to be broadly equivalent to the wholist-analytical dimension (Riding, 1991) and the adaptor-innovator dimension (Kirton, 1994), though there is a pressing need for concurrent validity studies.
Style, preferences and performance
Hayes and Allinson (1996) reviewed 19 studies which investigated the effects of matching styles to learning method and found that in 12 studies there was some support for the proposition that matching style and method contributed to improved learning performance. Fox (1984, p. 72) argued that "continuing educators must develop programmes that meet the needs of learners" and suggested that some participants do not "fit" with certain activities. Smith and Renzulli (1984) argue that congruence of style and method can have an effect on learner motivation and "investment" in the learning material. Equally important, matching can "help eliminate barriers to learning which arise when we [educators] fail to address the affective response various teaching modalities elicit from students" (Smith and Renzulli p. 74). Dunn (1984) reviewed several studies in which she found that where students were placed in academic situations where they were taught and/or tested in ways that matched or mismatched their self-reported preferences, those who were matched performed better than those who were mismatched. This led her to conclude that "their preferences must be their strength" (Dunn, 1984, p. 13). Miller (1991) took a somewhat different view: he argued that the analytic-holist model of style allows the possibility of individuals who are skilled at both analytical and holistic functioning- referred to as "versatile". He went on to discuss the issues surrounding attempts to engender "versatility" in those already not predisposed towards it. However, his conclusions are that to do so in all students is a waste of time and is potentially damaging and dangerous (given that styles may be forms of psychological defence). He argued that extremely specialised students should be left alone but that teaching should be accommodated to these styles and that versatility is a reasonable goal in those who are already disposed to it. The challenge as far as Miller was concerned was to identify the specialised and the "proto-versatile" (Miller, 1991, p. 236). The "versatility" argument (perhaps through the mismatch) is echoed in the pleas from Mintzberg (1994) for balance in strategic planning teams and Leonard and Strauss (1997) to harness the "energy released by the intersection of different thought processes" to propel innovation (p. 121). The challenge, therefore, for business schools and human resource development practitioners, is to acknowledge the differences that exist between individuals and use the differences constructively, for example, by giving careful consideration to when to "match", when to "mismatch" and how to engender cognitive "versatility".
At a more superficial level, the onion and cognitive control models suggest that cognitive style may exert some influence over preferences for different learning methods (for example role play versus lectures). Riding (1991) has argued that style may affect social behaviour, which may suggest that intuitives will tend to be dependent and gregarious and prefer collaborative ' learning situations, while analysts may be isolated and self-reliant. Hence, it may be expected that different business and management teaching and learning methods, with their varying degrees of social interaction and autonomy, would be viewed more or less favourably by different cognitive style groups. Similarly, with respect to the cognitive aspects of learning, Allinson and Hayes (1996) argued that analysts may prefer to pay attention to detail, focus on "hard" data, adopt a step-by-step approach to learning and are self-reliant. This suggests that analysts may prefer learning methods which allow opportunities for independent work with the opportunity to analyse data and reflect on information and experiences. Leonard and Strauss (1997) suggested that abstract thinkers (who share some of the attributes of analysts) will prefer to assimilate information from a variety of sources such as books, reports, videos, etc. Conversely, Allinson and Hayes (1996) argued that intuitives are less concerned with detail, adopt a global perspective and take an action-oriented approach to learning and problem solving. These "experiential" individuals will prefer to get information from "direct interaction with people and things" (Leonard and Strauss, 1997, p. 113). This may lead one to suggest that intuitives may prefer learning methods which are active, participatory and gregarious rather than analytical, reflective and self-referential. Sadler-Smith and Riding (1999) in a study of learning preferences and cognitive style (using the cognitive styles analysis (Riding, 1991)), found that wholists expressed a stronger preference for collaborative methods (role play and discussion groups) than did analytics. They attributed this to the gregarious nature and social dependence of the wholists. Clearly, one challenge for research in this field is to build on a growing empirical base.
The study
The study aimed to investigate the relationship between learning preferences and the intuition-analysis dimension of cognitive style in the context of business and management education and provide empirical elaboration for the onion and cognitive control models.
Sample and data collection
The sample consisted of 226 undergraduates studying a range of business and management degree programmes at a university business school in the UK. The sample was an opportunity sample and participation in the research was voluntary. Data were collected by means of a questionnaire which consisted of three sections:
(1) the cognitive style index (Allinson and Hayes, 1996);
(2) a learning preferences inventory;
(3) respondent data.
Cognitive style. This was measured by means of the cognitive style index (CSI) (Allinson and Hayes, 1996). The CSI is a paper and pencil inventory consisting of 38 questions scored on a three point "true-uncertain-false" scale. The theoretical maximum score is 76; the higher the score the more analytical is the respondent's style.
Learning preferences. Because of the limitations of existing measures a new questionnaire, the learning preferences inventory (LPI), was developed for the purposes of this study and is an extension of exploratory work reported in Sadler-Smith (1997) and Sadler-Smith and Riding (1999). The Reichmann Grasha (1974) instrument, the Rezler and Resmovic (1981) and Dunn et al. (1989) questionnaires appear to conflate notions of style and preference. The LPI consists of 13 items (see Table I); respondents are requested to indicate which teaching and learning methods they prefer in general according to a fivepoint Likert scale ranging from "definitely like" (scored five), through "neither like nor dislike" (scored three) to "definitely dislike" (scored one). The instrument's psychometric properties are discussed below.
Respondent data
Respondents' were requested to give their age, gender and programme of study and were assured of anonymity and confidentiality.
Results
Characteristics of the sample
The sample consisted of 128 (56.64 percent) males and 98 (43.36 percent) females; the mean age was 21.00. Respondents were a second year cohort in a single higher education institution in the UK; it is acknowledged therefore, that the characteristics of the sample are likely to introduce severe bias. This is compounded from an international perspective since the subjects have in the main experienced the UK's primary and secondary educational systems, which are likely to exert a considerable influence over their learning preferences (see Figure 1).
Item and factor analysis
The CSI has previously demonstrated construct validity through confirmatory factor analysis and correlational studies (see Allinson and Hayes, 1996). Its level of internal consistency is high, ranging from 0.84 to 0.92 and Allinson and Hayes (1996) report test re-test reliabilities of 0.90.
The LPI's factor structure was investigated by mean of a principal components analysis. Examination of the scree plot (Cattell, 1966) suggested that three factors (accounting for 42.2 percent of the variance) should be extracted. The three extracted factors were rotated to simple structure by means of a varimax rotation (the three factors were not inter-correlated). The resultant factor matrix with loadings of less than 0.4 suppressed is shown in Table I.
Factor I consists of methods which are active (for example role play exercises, workshops and practical classes) and participatory (for example giving presentations and seminars). Factor I was labelled "active". Factor II consists of methods which are reflective and didactic (for example, lectures) and self-directed (for example computer based and self-study methods). Factor II was labelled "reflective". Two items had high loadings (> 0.5) on Factor III individual work loaded positively and group work loaded negatively. Factor III was labelled "individual".
Descriptive statistics
Cognitive styles. The level of internal reliability for the CSI was high (see Table m). CSI scores by gender are shown in Table II. Hayes et al. (1998) argue that gendered stereotypic thinking "suggests that intuition is a feminine characteristic whereas analysis is a masculine characteristic" and go on to test this view. In a comparison of style and gender, using a sample of under-graduate business and management students, they found highly significant gender differences (p < 0.001) in cognitive style, with females (43.84; SD, 14.02) being more analytical than males (M, M, 36.33; SD, 15.56). This was the converse of the stereotypical view of "female intuition". Although in the present study females did generally score higher than males the differences were only marginally significant and hence style and gender may be considered independent in this context (see Table II).
There is some ambiguity in gender-related style differences. For example, Riding and Rayner (1998) argued that style is independent of gender. Complementary work using the CSI in a professional development context appears to suggest that while style and gender are independent they appear to interact in their effect on learning preferences. There is a need for further research into the relationship between style and gender and their combined effect on learning and workplace behaviours.
Learning preferences. The mean scores for each of the three learning preference scales identified were computed and are shown along with their inter-correlations in Table III. The levels of internal consistency (coefficient x) were as follows:
(1) active (0.50);
(2) reflective (0.59);
(3) individual orientation (0.81).
While the latter was satisfactory, the x's for active and reflective were low but considered acceptable for use in this exploratory study.
The three factors were not correlated among themselves. The general preference was in favour of reflective methods (M = 3.53; SD = 0.63), while individually-oriented methods were least preferred (M = 3.32; SD = 0.74), however, the observed differences were small.
Cognitive style and learning preferences
The lack of any important differences in the preferences expressed by the sample as a whole compounds the potential importance of any style-related differences, especially from the point of view of the planning and design of business and management education. The relationship between CSI score and learning preferences was explored by means of simple linear correlations. There were statistically significant correlations between CSI score and:
(1) reflective methods (r = 0.32; p < 0.001);
(2) individually oriented methods (r = 0.25; p < 0.001) see Table III.
The effect of style was further investigated by means of a one way analysis of variance in order to test for any non-liner relationships. The sample was divided into three cognitive style groupings: intuitives (0 < CSI < 39); intermediates (39 CSI 48); analysts (48 < CSI < 76). Mean preferences for the three methods for each of the style groups are shown in Table IV.
The intuition-analysis model of style leads one to anticipate stronger preferences for active methods on the part of the intuitives. However, there were no significant differences in this regard, therefore the assertion that intuitives will prefer active/participatory methods was not supported. The model also leads one to anticipate that for:
(1) reflective methods the analysts would express the strongest preferences and the intuitives the least strongest;
(2) individually-oriented methods the analysts would express the strongest preferences and the intuitives the least strongest
These data support both of these assertions (see Table IV). A two way analysis of variance (style by gender) did not reveal any statistically significant main effects for gender or interactions of gender and style in their effect on learning preferences.
Discussion
The onion model and cognitive control models (Curry, 1983; Riding, 1991) infer a relationship between cognitive style and learning preferences, albeit with the latter influenced by the learning environment and context. The present study has lent some support to the notion of learning preference being a correlate of cognitive style. With respect to analysts, the assertion that they would prefer reflective and individually oriented methods received support. With respect to the intuitives the assertion that they would express a dis-preference for reflective and individually-oriented methods also received support. Therefore, these data would suggest that there is a relationship between cognitive style and preferences for reflective and individually oriented methods. This may suggest that cognitive style manifests itself in learning situations as a preference for those methods which the learner unconsciously or consciously perceives as matching their preferred way of organising and processing information. Under such circumstances the learner may anticipate a benefit which may have a concomitant effect on motivation. The majority of empirical studies (Dunn, 1984; Hayes and Allinson, 1996) present evidence in favour of matching style and method. However, as noted earlier, some have argued that it is beneficial for the learner to consciously expose themselves to methods which do not match their preferred style in order to develop a wider range of learning skills ("learning-to-learn") (Entwistle, 1988; Honey and Mumford, 1992) and gain a "meta-cognitive advantage". The empirical evidence in favour of the mismatch of method and style is less robust than that which supports the concept of matching (Hayes and Allinson, 1996), although the latter is hardly unequivocal. It could be argued that mismatching learner and learning method is potentially valuable in the hands of a skilled facilitator with clearly formulated objectives and is perhaps one way in which learning-to-learn may be engendered.
The anticipated preference for participatory/active methods on the part of the intuitives did not receive support. This may suggest that: there is no simple and direct relationship between style and preference with respect to participatory/active methods; there are idiosyncrasies in the participatory/ active methods used in the institution concerned which intervened to confound any relationship with style; the relevant scale of the LPI may be a crude and underdeveloped measure (it had the lowest level of internal reliability) of preference for participatory/active methods. The latter could be improved by the exclusion of those items which loaded ambiguously (i.e. "seminars") or had the lowest factor loadings (i.e. "giving presentations")- see Table I. The relationship between style and preference is worthy of further investigation, using undergraduate samples from a broader range of educational institutions, post-graduate and professional development students and, most importantly, randomly selected work-based samples. The extension of this work into international contexts (given the UK bias in the present study) in order to explore the cross cultural validity of the style and preferences constructs and their inter-relationships would also be potentially valuable.
Conclusion
The aim of this study was to examine the validity of the onion and cognitive control models and it is argued that limited support has been provided. Two central issues may be identified: the status and validity of the "matching hypothesis"; and the notion of learning-to-learn. The two issues are related in that if individuals achieve the latter the former becomes a redundant concept. A key aspect of learning-how-to-learn is strategy development. Riding and Sadler-Smith (1997, pp. 204-5) argued that individuals may adopt a three-stage approach to strategy development based on the fit between their cognitive style and the demands of the learning situation. The first stage is sensing the extent to which the learner feels comfortable with the situation in terms of their own preferences. The second stage involves them, as they become more metacognitively aware, in selecting the most appropriate learning methods. The third stage is strategy development in which individuals attempt to make learning "easier" by translating, adapting or reducing the processing load imposed on them by the situation. This suggests that explicit acknowledgement of cognitive style and learning preferences (along with learning styles and approaches to studying), perhaps through comprehensive "profiling" of these attributes, may be an important step forward in bringing learners and management educators together in an understanding of each other's styles and their mutual interdependence. This is crucial since one of the keys to efficient and effective performance in both the classroom and the workplace is the ability to balance intuition and analysis, since neither is sufficient by itself.
Table I. Factor Matrix for the LPI
Item Factor I Factor II Factor III
Group work -0.72
Role play exercises 0.59 -0.47
Lecturer presenting
facts and theories 0.46
Lecturer presenting
Examples 0.47
Self-study 0.76
Texts and journals 0.60
Computer-based methods 0.60
Analysis of cases 0.55
Workshops and
practical cases 0.78
Problem solving
Exercises 0.64
Giving presentations 0.42
Individual work 0.72
Seminars 0.59 0.44
Table II. Cognitive styles scores by gender (*p = 0.05, one tailed test)
Males Females
N M SD n M SD df t
CSI 128 43.27 9.56 98 45.41 9.69 224 -1.67*
Table III. Learning preferences means, standard deviations, inter-correlations, reliabilities and relationship with cognitive style
CSI Active Reflective Individual M SD
CSI 0.89 0.05 0.32*** 0.25*** 44.25 9.66
Active 0.50 0.10 -0.16* 3.43 0.60
Reflective 0.59 0.12 3.53 0.62
Individual 0.81 3.32 0.74
Note: Coefficient alphas are shown in bold along the diagonal.
223 [< or equal to] n [< or equal to] 226 ; *p < 0.05; **p <
0.01; ***p < 0.001
Table IV Cognitive style and learning method preferences
Intuitives Intermediates Analysts
(n=71) (n=65) (n=89)
M SD M SD M SD df F
Active 3.29 0.64 3.54 0.56 3.44 0.59 221 2.77
Reflective 3.31 0.66 3.53 0.56 3.70 0.58 221 8.32**
Individual 3.17 0.76 3.23 0.66 3.49 0.75 221 4.21*
Note: *P < 0.05; **p < 0.01
References
Allinson, C.W. and Hayes, J. (1996), "The cognitive style index: a measure of intuition-analysis for organisational research", Journal of Management Studies, Vol. 33 No. 1, pp. 119-35.
Bunge, M.A. (1983), Exploring the World: Epistemology and Methodology (treatise on basic philosophy), D. Reidel, Dordrecht.
Cattell, R.B. (1966), "The scree test for the number of factors", Multivariate Behavioural Research, Vol. 1, pp. 245-76.
Curry, L. (1983), Learning Styles in Continuing Medical Education, Canadian Medical Association, Ottowa.
Dunn, R. (1984), "Learning style: state of the science", Theory into Practice, Vol. 23 No. 1, pp. 10-19.
Dunn, R., Dunn, K. and Price, G.E. (1989), Learning Styles Inventory, Price Systems, Lawrence, KS.
Entwistle, NJ. (1988), Styles of Learning and Teaching, David Fulton, London.
Entwistle, NJ. and Tait, H. (1994), The Revised Approaches to Studying Inventory, Centre for Research into Learning and Instruction, University of Edinburgh, Edinburgh.
Fox, R.D. (1984), "Learning styles and instructional preferences in continuing education for health professionals: a validity study of the LSI", Adult Education Quarterly, Vol. 35 No. 2, pp. 72-85.
Grasha, A.F. and Reichmann, S.W. (1975), Student Learning Styles Questionnaire, University of Cincinnati Faculty Resource Centre, Cincinnati, OH.
Hayes, J. and Allinson, C.W. (1996), "The implications of learning styles for training and development: a discussion of the matching hypothesis", British Journal of Management, Vol. 7 No. 1, pp. 63-73.
Hayes, J., Allinson, C.W. and Taggart, W.M. (1998), "Intuition, women managers and gendered stereotypes" (under review).
Honey, P. and Mumford, A. (1992), The Manual of Learning Styles, Peter Honey, Maidenhead.
Hurst, D.K., Rush, J.C. and White, J.E. (1989), "Top management teams and organisational renewal", Strategic Management Journal, Vol. 10, pp. 87-105.
Kirton, M.J. (1994), Adaptors and Innovators: Styles of Crcativity and Problem Solving. Routledge, London.
Kolb, D.A. (1984), Experiental Learning, Prentice Hall, Englewood Cliff, NJ.
Leonard, D. and Strauss, S. (1997), "Putting your company's whole brain to work", Harvard Business Review, July-August, pp. 111-21.
Matron, F. and Saljo, R. (1976), "On qualitative differences in learning. 1' outcomes and processes", British Journal of Educational Psychology, Vol. 46, pp. 4-11
Miller, A. (1991), "Personality types, learning styles and educational goals", Educational Psychology, Vol. 11 No. 34, pp. 217-37.
Mintzberg, H. (1976), "Planning on the left side and managing on ihe right", Harvard Business Review, July-August, pp. 49-58.
Mintzberg, H. (1994), "The fall and rise of strategic planning", Harvard Business Review, January-February, pp. 107-14.
Messick, S. (1984), "The nature of cognitive styles: problems and promises in educational research", Educational Psychologist, Vol. 19, pp. 59-74.
Nebes, R.D. and Sperry, R.W. (1971), "Cerebral dominance in perception", Neurophsychologica, Vol. 9 No. 247, p. 53.
Reichmann, S.W. and Grasha, A.F. (1974), "A rational approach to developing and assessing the construct validity ()f a study learning styles scale inventory", Journal of Psychology, Vol. 87, pp. 213-23.
Renzulli, J.S. and Smith, L.H. (1978), The Learning Styles Inventory: a Measure of Student Preference for lnstructional Techniques. Creative Learning Press, Mansfield Centre, CT.
Rezler, A.G. And Rezmovic, V. (1981), "The learning preferences inventory", Journal of Allied Health, February, pp. 28-34.
Riding, R.J. (1991), Cognitive Styles Analysis, Learning and Training Technology, Birmingham.
Riding, R.J. (1997), "On the nature of cognitive style", Educational Psychology Vol. 17 Nos 1-2, pp. 29-50.
Riding, RJ. and Rayner, S.G. (1998), Cognitive Styles and Learning Strategies, Fulton, London.
Riding, R.J. and Sadler-Smith, E. (1997), "Cognitive style and learning strategies: some implications for training design", International Journal of Training and Development, Vol. 1 No. 3, pp. 199-208.
Riding, R.J., Glass, A., Butler, S.R. and Pleydell-Pearce, C.W. (1997), "Cognitive style and individual differences in EEG alpha during information processing", Educational Psychology, Vol. 17 Nos 1-2, pp. 219-34.
Sadler-Smith, E. (1996), "Learning styles: a holistic approach", Journal of European Industrial Training, Vol. 20 No. 7, pp. 29-36.
Sadler-Smith, E. (1997), "Learning style: frameworks and instruments", Educational Psychology, VoI. 17 Nos 1 and 2, pp. 51-63.
Sadler-Smith, E. and Badger, B. (1998), "Cognitive style, learning and innovation", Technology Analysis and Strategic Management, Vol. 10 N(). 2, pp. 247-65.
Sadler-Smith, E. and Riding, R.J. (1999), "Cognitive style and instructional preferences", lnstructional Science, in press.
Smith, L.H. and Renzulli, J.S. (1984), "Learning style preferences: a practical approach for classroom teachers", Theory into Practice, Vol. 23 N(). 1, pp. 44-50.
Sternberg, R.J. and Grigorenko, E.L. (1997), "Are cognitive styles still in style?", American Psychologist, July, pp. 700-12.
Witkin, H.A. (1962), Psychological Differentiation: Studies of Development, Wiley, New York, NY.
~~~~~~~~
By Eugene, University of Plymouth Business School, Plymouth, UK
--------------------------------------------------------------------------------
Copyright of Journal of Managerial Psychology is the property of Emerald and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Source: Journal of Managerial Psychology, 1999, Vol. 14 Issue 1/2, p26, 13p, 4 charts, 1 diagram.
Item Number: 1675008 Result 46 of 127 [Go To Full Text] [Tips]
Result 50 of 127 [Go To Full Text] [Tips]
Title: Can we generalize about the learning style characteristics...
Subject(s): LEARNING strategies; GIFTED children -- Education; ACADEMIC achievement
Source: Roeper Review, May/Jun98, Vol. 20 Issue 4, p276, 6p, 3 charts
Author(s): Burns, Deborah E.; Johnson, Scott E.; et al
CAN WE GENERALIZE ABOUT THE LEARNING STYLE CHARACTERISTICS OF HIGH ACADEMIC ACHIEVERS?
In 1980 Dunn and Price used their Learning Style Inventory (Dunn, Dunn & Price, 1975) to investigate differences between the learning style preferences of high academic achieving students and the preferences expressed by same-age students with average or below average academic achievement The purpose of the study described in this article was to determine if and how the learning style preferences of a different group of high academic achieving students, inventoried at a later date, but with the same instrument, differed from those identified in the original study. A discriminant function analysis analyzed the learning styles data obtained from 500 students in grades 4 - 8. While significant differences (p ................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
- joshua yopp capstone
- chapter 8 business crisis and continuity management and
- result 16 of 127 go to full text tips
- abd e news iowa
- ann b stanford university
- to understand the economics of contemporary college
- kansas city v a medical center psychology fellowship va
- midwest orthopaedics at rush best chicagoland orthopedic
Related searches
- free full text journal articles
- free full text research articles
- free full text scholarly articles
- full text peer reviewed journals
- sec full text search
- free full text articles
- full text of huck finn
- full text journal articles
- ifrs 9 full text pdf
- unabomber manifesto full text pdf
- full text of the constitution
- us constitution full text pdf