Poster Session 1
Poster Session 1 July 3, 10.45-12.45 Room: Rotonde
Poster 01-01
DIF investigation of TIMSS 2007 science items in terms of test language and gender
Hayri Eren Suna (Baskent University, Turkey)
In today’s educational assessment process, international studies that evaluate the educational achievement of students are becoming more important and in order to decide their educational strategies more accurately, the number of participant countries of these studies is increasing. International assessment studies, like TIMSS, PIRLS and PISA, which focus on educational achievement across different cultures put great emphasis on fairness and comparable test results. The purpose of this study is to investigate whether science items of TIMSS 2007 include Differential Item Functioning (DIF) in terms of test language and gender or not. In order to determine potentially biased items, Mantel-Haenszel (MH), Logistic Regression (LR) and Simultaneous Item Bias Test (SIBTEST) will be used. The analyses have been carring out with respect to 4,498 Turkish and 4,025 English students’ responses. All of science items in TIMSS 2007 are included in 14 booklets; so each booklet has a characteristic item block. In this study, each item block, presented in 14 different booklets, will be investigated across different test language and gender groups. Results will be presented and discussed under the light of broader results of the previous studies on international assessment studies considering the factors in the Turkish educational system.
Poster 01-02
A Bayesian model comparison approach in mixture IRT models: An empirical analysis using the TIMSS mathematics assessment
Yoon Soo Park (University of Illinois at Chicago, USA)
Young-Sun Lee (Columbia University, USA)
Mixture item response theory (IRT) has become a common measurement model to filter heterogeneous populations into homogeneous subpopulations. Empirical evidence has shown that unidimensional IRT models often misfit data when examinees have underlying differences; mixture IRT models have been developed to handle such qualitative differences (Mik-de Meij et al., 2008). Applications of mixture IRT models have been prevalent in differential item functioning analyses, where the advantage of mixture IRT models is its ability to classify a priori unknown grouping based on each examinee’s response pattern and does not rely on manifest information about the group (Cohen & Bolt, 2005). This study adopts the Carlin-Chib method (Carlin & Chib, 1995) in a Bayesian framework to select the best fitting model among the 1PL, 2PL, and the 3PL mixture IRT models using the TIMSS 2007 8th grade mathematics data. Unlike traditional methods for model selection (e.g., AIC and BIC), an advantage of this technique includes the ability to specify and fine-tune pseudopriors, which allows the sampler to move efficiently across models of different dimensions. This study presents a method to demonstrate an efficient alternative to incorporate additional parameters in the model that can be useful for both theoretical and applied purposes.
Poster 01-03
An application of a multiple-group CDM to compare cognitive attribute distributions based on eighth grade TIMSS mathematics
Jung Yeon Park (Teachers College, Columbia University, USA)
Matthew Johnson (Teachers College, Columbia University, USA)
Young-Sun Lee (Teachers College, Columbia University, USA) (presenting author)
Ruchi Sachdeva (Teachers College, Columbia University, USA)
Jianzhou Zhang (Teachers College, Columbia University, USA)
Cognitive Diagnosis Models (CDMs) have received increasing attention in recent years. CDMs were developed to provide information to researchers and educators on cognitive skills or attributes that are required to solve an item. Consequently, diagnostic information can be applied to various instructional practices by identifying the presence or the absence (i.e., mastery and non-mastery) of specific, fine-grained skills or attributes. The purpose of the study is to develop novel extensions to CDMs to model the distributions of attribute profiles for multiple groups of students on the TIMSS 2007 mathematics assessment. To handle the complex sampling design used in TIMSS, we develop the multiple group CDM to account for the clustering of students within schools. This study will allow us to 1) gain better understanding of what skills eighth grade students lack or possess, and 2) compare the distributions of attributes across countries and within countries. We will analyze 11 countries including 9 top-performing countries and two U.S. benchmark states. This study will provide a rich source of information for educational policy makers, educational researchers and teachers to direct their attention to specific areas where students need further instruction by identifying strengths and weaknesses in the performance of the students on the TIMSS mathematics assessment.
Poster 01-04
The effect of assessment as learning components training on the self-concept and mathematic academic achievement
Mohammad Asgari (Malayer University, Iran)
The aim of this study was to determine the effect of assessment as learning components training on the self-concept and mathematic academic achievement of third grade guidance school students in Arak, Iran. A sample of 60 students (including 2 classes), selected randomly, was assigned in both experimental and control groups. Ahlovalia self-concept and teacher- made mathematic achievement pretests were administered to both groups. Then, the experimental group attended an assessment as learning components training for 12 sessions, but the control group received no treatment. Finally, both groups were administered Ahlovalia self-concept and teacher- made mathematic achievement post-tests. Data analysis through MANOVA showed that the assessment as learning components training significantly increased the self-concept and the mathematic academic achievement of the subjects at p< 0.01. Also, assessment as learning components training positively affected all sub-scales related to self-concept including fame, anxiety, physical appearance and attitude, satisfaction and happiness, and intellectual and academic states except behavior subscale.
Poster 01-05
Invariance of Predicting Mathematics Achievement by Motivation and Self-Efficacy Across Gender and Achievement Levels
AbdelAziz Sartawi (United Arab Emirates University, United Arab Emirates)
Othman Alsawaie (United Arab Emirates University, United Arab Emirates)
Hamzeh Dodeen (United Arab Emirates University, United Arab Emirates)(presenting author)
This study aims at investigating the extent to which self-efficacy and motivation can predict mathematics achievement across gender and achievement levels. Self-efficacy was measured by two scales that differ in the level of specificity which are Category Specific and Task Specific. Motivation was measured by four sub-constructs of motivation: Amotivation, External Regulation, Introjected Regulation, and Intrinsic Motivation. A total of 287 fifth grade students from UAE with an average age of 10.3 years participated in this study. Responses of the two genders on the six subscales were compared and discussed. The multiple regression model showed that the six predictors were able to explain 32% of the variance of mathematics achievement. Also the results indicated that the best three predictors were Task Specific, External Regulation, and Intrinsic Regulation. When conducting the regression model across gender, the results showed that 30% of the variance in math achievement was explained by the six predictors for the males group while only 21% of the variance was explained for the females group. The performance of the regression model and importance of each predictor were discussed and compared between male and female students. Finally, the regression model was found to be variant across achievement levels. Specifically, while the model predicted around 20% of the variance of math achievement for each of the low and high achieving students, the model was not statistically appropriate for the medium achievement students.
Poster 01-06
The moderation of school academic optimism effect on student achievement: DIF and multi-group SEM analysis
Liang-Cheng Zhang (National Taiwan Normal University, Taiwan)
Jia-Jia Syu (National Chengchi University, Taiwan)
Tian-Ming Sheu (National Taiwan Normal University, Taiwan)
Academic optimism represents different facets of schools: collective efficacy, faculty trust in students and parents, and academic emphasis. It is one of the most important school properties due to the fact that it has salient impact on student achievement even holding the socioeconomic status. Location, however, could moderate the effect of academic optimism on student achievement because economic, geographic, physical and social disadvantages in rural schools would result in the isolation from other areas. Therefore, the main purpose of this paper is to test the second-order structure model of the academic optimism influence on math achievement between rural and non-rural schools. The data is gathered from Taiwan 294 junior high schools where there are 146 schools in the rural area and the others are in the non-rural area. We used DIF to verify validity of academic optimism scale and decrease the item number from 27 to 18. Furthermore, with multi-group SEM, we found that rurality moderated the effects of academic optimism on math scores, and the academic optimism effect favors non-rural schools. These results could imply that the location could be a key variable when designing the school teaching policy and developing academic optimism scale.
Poster 01-07
Need for achievement influences test taking motivation influences achievement test performance
Ludwig Kreuzpointner (Universität Regensburg, Germany)
Ines Giermann (Universität Regensburg, Germany)
In this study the influence of the general need for achievement on the situational test taking motivation and on an intelligence test performance was examined. 337 pupils (9th and 10th grade of all German school types) performed three questionnaires (RLMI-KJ, Lukesch, in prep.; AMG-S K-J, Schmalt, 2005; and a self-constructed questionnaire for test taking motivation, TTM) and the Leistungsprüfsystem 2, (LPS-2, Kreuzpointner, 2010), a German intelligence and performance test. Statistically significant bivariate correlations between the motivational variables and the test performance were only found for the fear of failure active (.17) and passive (.21) and the TTM measured after the test execution (.18). A SEM was used to show to what extend the TTM depends on the need for achievement and to what extent both variables influence the performance test result. As expected, there is a statistically significant path coefficient for the influence of the need for achievement on the TTM (.56). The path coefficient for the influence of the TTM on the performance test achievement was rather unexpected in its extent (.29). Cronbach (1960) seems to be right when he writes: Test performance in a given test is a joint function of ability and motivation.
Poster 01-08
Validating the Filipino version of Academic Delay of Gratification Scale (ADOG)
Fraide A. Ganotice, Jr. (Palawan State University, Philippines)
Ronnel B. King (The University of Hong Kong, Hong Kong SAR, China)
The objective of this study was to examine the acceptability of psychometric properties of the Filipino version of the Academic Delay of Gratification Scale (ADOGS; Bembenutty & Karabenick, 1998). This scale measures the degree to which students postpone immediately available opportunities to satisfy their impulses in order to pursue important goals that are temporally remote but ostensibly more import. In this study, we developed a conversational Filipino version of the ADOGS using combined forward and back-translation techniques. Filipino adolescent students from Philippines (N = 494) completed the ADOGS. Both within-network and between-network approaches to construct validation were adopted in the study. Responses to this questionnaire are shown to have good internal consistency and support is provided for its construct validity in terms of its correlation with other educational outcomes including cognitive and metacognitive strategies. Implications for cross-cultural research are discussed.
Poster 01-09
Cancelled.
Poster 01-10
Online Learning Strategies Scale adaptation to a Colombian population: Factor analysis and Rasch model
Meneses Alba (Universidad El Bosque, Spain)
Garzón Angélica (Universidad El Bosque, Spain)
Organista Pedro (Universidad El Bosque, Spain)
Masias Jacquelinne y Sabogal Guiovanna (Universidad El Bosque, Spain)
The Online Learning Strategies Scale (OLSS) (Tsai, 2009) was adapted to Colombian university students. The OLSS assesses the e-learning strategies used by the students from metacognitive perspectives. Five experts conducted back and forth translation translation of the scale and two different versions of OLSS were applied to 194 college students. Data were analyzed using factor analysis and the Masters Partial Credit Model for polytomous items, which are based on Rasch models. The results showed that 13 of the 20 items of the scale are adequate in terms of there psychometric properties. Seven items were discarded and five of them consisted of inverse questions. High reliability was obtained for person (0.89), items (0.98) and Cronbach's Alpha (0.91). Acceptable results were obtained for the unidimensional model, for the total explained variance (62.7%) and for the first component (19%). Factor analysis showed a lack of evidence for five construct factors that were considered in the original model. The results for the Colombian university students outline only two factors: (1) anxiety, motivation, self-monitoring and concentration subscales, and (2) literacy subscale. This study discusses the advantages of using TCT and Rasch models for test adaptation to other cultures.
Poster 01-11
The uniform and non-uniform DIF of Personal Epistemology Beliefs Scale for senior high school students
Jia-Jia Syu (National Chengchi University, Taiwan)
The Personal Epistemological Belief Scale for Senior High school students is comprised by the estimation of the essentials of knowledge and ability. It has been verified to relate to effective learning. A scale with good psychometric characteristics would be the premise. As the overall model fit of first order measurement model of Epistemology Beliefs is adequate (chi2 = 877.75, RMSEA=0.051), Cronbach’s α’s are between 0.67 to 0.90, we tend to assert the scale is ample to estimate the construct under the normal validity procedure; however, this is not sufficient in the scope of modern response theory. The results of DIF analysis show that some items in the plausible scale would favor certain group, even the researchers have controlled the potential confounding by experimental design, such as sampling equally. For Different approach schools level comparison, the scale shows non-uniform DIF effects. The ability-related questions favor normal school students and first-year student, and knowledge-related questions favors prestigious schools students and second-year students. As for different area of school, the scale is more unfavorable for the students in the south of Taiwan than the middle and north counterparts. The results suggest the value of precise detections and the possible effect of learning experience.
Poster 01-12
Evaluation of stress level and related factors using DASS-21
Majid Akbari (Mashhad University of Medical Sciences, Iran)
AmirHossein Nejat (Mashhad University of Medical Sciences, Iran)
SeyedMehdi Dastorani (Mashhad University of Medical Sciences, Iran)
Armita Rouhani (Mashhad University of Medical Sciences, Iran)
Dental students are exposed to clinical environmental stresses in addition to educational environmental stresses. If psychological stressors become chronic and/or remain untreated, they result in psychological and physical symptoms. There are few studies about stress and dental students in our country. The aim of this study was to assess stress level and related factors sorted according to the academic year and gender, among dental students of Mashhad University of Medical Sciences during 2008-2009. This study was a description-analytical and cross-sectional study. Two questionnaires were used concurrently. One was DASS-21 standard test of which the validity for Iranian population was confirmed and the other was a researcher made tool for evaluation of related factors. It has 20 sentences that explain possible stressful situation which have to be rated on a Likert scale. This check list was validate by a focus group of students and its reliability confirmed by tets retest. Data was analyzed by the SPSS software. 52% of students showed abnormal levels of stress. Students of the fourth academic year showed the highest level of stress. Stress levels in the two genders were not significantly different. It was also found that university stressors played a more significant role in stress level than non-university stressors. Based on the results of this study, greater attentions must be given to this issue by the authorities and they should have more attention to raise copying skills in the students. Furthermore, the fourth year students who have recently entered the clinic are in the greatest need of help.
Poster 01-13
The effect of item presentation style on a Japanese language reading test in a high school admissions examination
Kazuhiro Yasunaga (Nagoya University, Japan / Japan Society for the Promotion of Science, Japan)
Makoto Saitoh (Nagoya University, Japan)
Hidetoki Ishii (Nagoya University, Japan)
The purpose of this study was to examine if the manner in which item is presented on a Japanese language comprehension test of a high school admissions examination affects item difficulty and item discrimination. A total of 493 students were administered this test. Students were given 50 minutes to complete the test, and they were graded based on an answer rubric. Four variations of item presentation were used: 1) choosing a sentence from the text (the first 5 words or free-response), 2) the presentation or non-presentation of concrete examples in the original text, 3) varying the style of filling in blanks (the same form throughout or varied), 4) limiting length of answer column. It was discovered that 1) did not affect difficulty and discrimination. Further, 2) the presentation of concrete examples in text lead to higher difficulty and discrimination. In addition, 3) blank filling variation contributed in part to higher discrimination. Finally, 4) not limiting the answer length lead to the highest difficulty and discrimination. In conclusion, the results of this study demonstrate the importance in examining item presentation style in tests.
Poster 01-14
A psycholinguistic approach to foreign language proficiency testing
Eleonora Nosenko (Oles Honchar Dnipropetrovsk National University, Russia)
Iryna Arshava (Oles Honchar Dnipropetrovsk National University, Russia) (presenting author)
Liliia Ponomarova (Oles Honchar Dnipropetrovsk National University, Russia)
Victoriya Kornienko (Oles Honchar Dnipropetrovsk National University, Russia)
The objective of the presentation was to substantiate the rationale of a psycholinguistic approach to the computer-assisted testing of foreign-languages oral proficiency. The approach is based on detecting differences in the oral the oral speech generation processes pertinent to the native and foreign-language speakers, which reflect the degrees of conscious appraisal of the two basic operations performed in speech: the choice of words and ascribing grammatically relevant characteristics to the utterances. While either category of speakers consciously appraise the choice of words, foreign language speakers, unlike the native speakers, ascribe grammatical characteristics to the utterances either on the subconscious, or even on the conscious level. These differences are reflected in the temporal characteristics of oral speech, which can be automatically assessed in terms of the continuity of speech utterance, uninterrupted by hesitation pauses (of 250 msec an longer), articulation time / total pause time ratio and the like. This allows to conclude that the approach described can be used both in the foreign language proficiency testing and in lie detection, speech impairment diagnostics of various genesis etc.
Poster 01-15
About trying to catch water with two fingers
Almut Thomas (University College of Teacher Education Carinthia, Austria)
Doris Latschen (Viktor Frankl Hochschule, Austria)
Erik Frank (Viktor Frankl Hochschule, Austria)
Annette Weishaupt (Viktor Frankl Hochschule, Austria)
Johann Weishaupt (Viktor Frankl Hochschule, Austria)
In the context of the item generation for the Austrian standards of education in listening comprehension for L1 learners at grades 4 we sought to assess listening as a multidimensional construct. Although theoretical considerations suggest that listening comprehension is multidimensional (Buck, 2001), common tests usually only assess the understanding of basic ideas and specific information of a text. We aimed at developing an instrument comprising additionally the assessment of the following dimensions: understanding of paraverbal information, evaluation of the appropriateness of speech acts, differentiation of phonologically similar words, and comprehension monitoring. We tested the new instrument with a sample of 239 children in grade 4 and analyzed the data using confirmatory factor analysis and item response theory. Results provided restrained support for the factor structure of the subscales; homogeneity of the subscales was corroborated using Andersen’s Likelihood Ration tests. Quality of the item structure was probably influenced by the nature of the response option as well as fluctuation in childrens’ attention.
Poster 01-16
DIF analysis for medical education students on the ICDL Test
Hassan Gholami (Mashhas University of Medical Sciences, Iran)
Hassan Ghavami (Mashhas University of Medical Sciences, Iran)
Hosien Karimi Mounaghi (Mashhas University of Medical Sciences, Iran)
Hadi Abbassi (Mashhas University of Medical Sciences, Iran)
Sara Samadi (Mashhas University of Medical Sciences, Iran)
Test scores show evidence of construct validity if the test (or items) measures the intended construct. One method used to ensure that test items have construct validity is to identify and flag items that could be biased against a group of examinees on the basis of demographic characteristics such as gender, language or geographical location. This study is a study to determine the differential item functioning analysis (test bias) of ICDL’s test medical education students. We selected one final test that includes 75 items and we determined differential item functioning (DIF) based on gender. We computed DIF indexes with Winsteps software. We found that 53 items out of 75 items (71%) showed significant DIF (p< 0.05) based on gender in two groups. The results revealed that DIF resulted in gender bias. This study may be a starting point for determining tests’ bias. The goal of this research was to provide insight regarding the nature of DIF with respect to accommodations in general as well as specific practices used in real testing situations. The DIF index can show item bias and precision of measurement when used for detecting gender bias.
Poster 01-17
Factorial validity of a measure of motivation toward career decision-making alternatives
José Tomás da Silva (University of Coimbra, Portugal)
José Pacheco Miguel (University of Coimbra, Portugal)
The career decision making (CDM) process is a key aspect of career construction. Traditionally, CDM models emphasized the cognitive dimensions rather than the affective-motivational processes involved in career selection. However, recently scholars began to study the role played by volitional processes in the choice of vocational alternatives. The goal of the present study is to present some preliminary reliability and dimensionality data about a measure of motivation toward career decision making activities, the Career Decision-Making Autonomy Scale (CDMAS). After the technical translation of the items to Portuguese the CDMAS was administered to a sample (N = 152) of undergraduate psychology students enrolled at a major Portuguese university. The results showed that the internal consistency of the four motivation subscales (external, introjected, identified, and intrinsic) were all above .85. An exploratory factor analysis (with PAF extraction, and Varimax rotation) was also carried out on the 32 items of the CDMAS. Only the four factors deemed important by a Parallel Analysis (PA) based on minimum rank factor analysis, were rotated. In general the items loaded on the target factor, thus supporting the hypothesized structure of the CDMAS. Future directions for advancing the psychometric analysis of the measure are discussed.
Poster 01-18
Obedient and worrier: A portrait of the successful university student in Russia
Ekaterina Orel (National Research University Higher School of Economics, Russia)
Ekaterina Kochergina (National Research University Higher School of Economics, Russia)
Meta-analytic research in psychology of academic performance proved that Big Five Conscientiousness and Openness to Experience predict scholastic achievements of university students (O’Connor, Paunonen, 2007; Poropat, 2009). But we claim that psychological predictiors of academic success depend on educational environment and can be culture-related. We examined 176 2nd and 3rd year economy and computer science university students in Russia with the Big Five – Ipsative version test (Shmelyov, 2010) and discovered that GPA and USE (United State Examination in Russia) scores are significantly correlated with Agreeableness (r = 0.15; p< 0.01 for GPA and r = 0.22 p< 0.01 for USE math) and Neuroticism (r = 0.2, p< 0.01 for GPA and r = -0,17; p< 0,01 for USE math). We suppose that the difference between our result and results provided by the meta-analyses mentioned above can be explained by the differences in educational environment in Russia and other countries. We assume that big number of classes and relatively small amount of individual and analytical assignments create the environment where Agreeableness and Neuroticism are important for the academic success.
Poster 01-19
Measuring identification with career roles
Nicole de Jong (University of Groningen, The Netherlands)
Hans A. Hoekstra (University of Groningen, The Netherlands)
Karin I. van Oudenhoven- van der Zee (University of Groningen, The Netherlands)
A model of six universal Career Roles is presented as a content model of career development. This model specifies six Career Roles thought to be applicable in any job (Maker, Expert, Director, Guide, Inspirator and Presenter). The career role model guided the development of the "Career Role Identification Questionnaire (CRIQ)” measuring career role identification with the six career roles. In identification measures for positive targets self-presentation tactics and acquiescence are to be expected. To address such problems and maintain scale independence, we used a new measurement technique called "Comparison Awareness Induction Technique (CAIT)". This technique is aimed at enhancing the awareness of differences between the career roles by suggesting comparison through the graphic form of the presented scales. A study among 212 employees broadly sampled from the working population in the Netherlands. The paper describes instrument development and evaluation of the CRIQ. To test our instrument confirmatory factor analysis was conducted. Psychometric properties of the Career Role Identification Questionnaire using the CAIT-technique, as well as implications, are discussed.
Poster 01-20
Creating through course assessments for the Advanced Placement Program®
Vytas Laitusis (The College Board, USA)
Rosemary Reshetar (The College Board, USA) (presenting author)
James Daubs (The College Board, USA)
Elizabeth Daniel (The College Board, USA)
Edward Hardin (The College Board, USA)
Brett Mayhan (The College Board, USA)
The Advanced Placement Program® (AP) is considering a new assessment model that would feature a set of Performance Based Assessment (PBA) tasks delivered via a through course testing model in addition to the existing three hour summative exam. Development work has ensued for the English Language and Composition Exam and piloting of two tasks created during this process is taking place in the 2011 academic year. The current summative exam is administered annually to over 400,000 students in over 10,000 schools. It includes a 60 minute multiple choice section and a 135 minute section which requires students to write three essays in response to each of a passage analysis, argument and synthesis prompt. Two tasks have been defined and developed for piloting. The first is a researched argument paper and the second a recorded oral presentation that develops and supports a position through the rhetorical analysis of two visual images. This paper describes the assessment design, development process, and psychometric issues considered and presents results of piloting in selected AP classrooms. This work informs established large scale assessment programs that wish to move beyond single summative exam models and will guide the AP program’s future developments.
Poster 01-21
Are serious games a reliable approach to the assessment of vocational interests?
Pierre Vrignaud (Université Paris Ouest Nanterre La Défense, France)
Even Loarer (Université Paris Ouest Nanterre La Défense, France)
Odile Dosnon (Université Paris Ouest Nanterre La Défense, France)
Thierry Boy (Université Paris Ouest Nanterre La Défense, France)
In the field of training the use of games opened new opportunities to develop training tools. In this project the approach of serious games was applied to the assessment of vocational interests, using Holland’s hexagonal model (based on six vocational types) as a theoretical framework. The game takes place in a village (18 buildings). The subject has to play six mini-games, each game corresponding to a Holland’s type (RIASEC). Five types of information are recorded: the number of times the subject played the game, the score obtained to the game, the time played, the game finished, a mark for the liking or disliking of a game by the subject, and at the end of each day, the subject is invited to rank the six mini-games from 1 to 6. The use of this serious game approach raises very fundamental and heuristic questions: 1) Allows this game situation a reliable assessment? 2) Which in the recorded information are the most reliable? To answer these questions we gathered data on several samples of subjects playing the game and responding to a classical vocational interest blank. The statistical and psychometrical analysis aims at estimating the reliability of the game.
……………………………………………………………………………………………………………………………………………………....
Poster Session 2a July 3, 13.30-15.30 Room: Rotonde
Scholarship Session
Poster 02-01 Sponsored by APA and TestGroup Consulting
Urdu adaptation and validation of State Self Esteem Scale in young adults
Sobia Aftab (University of Karachi-Pakistan, Pakistan)
Sofia Aslam (University of Karachi-Pakistan, Pakistan)
The objective of the present study is to adapt and validate the State Self-Esteem Scale in National language (i.e., Urdu). After adaptation, reliability and validity estimates were made on a randomly selected sample of 237 young adults with an age range of 18-25 years (with a mean age of 22.04; SD = 1.78 ) belonging to all three socioeconomic status (i.e., lower, middle, and upper). The measures which were used in the study include: Demographic Form; State Self-Esteem Scale (SSES; Heatherton & Polivy, 1991); Rosenberg Self-Esteem Scale – Urdu Version (Sardar, 1998); The Aga Khan University Anxiety and Depression Scale (AKUADS; Ali, Reza, Khan, & Jehan, 1998), and Beck Scale for Suicidal Ideation-Urdu Version (Ayub, 2008). Results indicate good internal consistency (i.e., α = .801), test-retest reliability (i.e., r = .968); and Guttman split-half reliability coefficient (i.e., .856). Moreover, State Self-Esteem Scale-Urdu Version is found to be positively correlated with RSE (r = .497) and negatively correlated with AKUADS (r = -.512) and BSSI (r = -.290) which indicates good convergent validity of the scale. Implications, limitations and avenues for future research are discussed.
Poster 02-02 Sponsored byGITP and Profiles International
Good intentions, difficult challenges: Feeding grade 8 continuous assessment scores into national examinations’ scores in Malawi
Ella Gift Banda (Malawi)
Continuous Assessment (CA) is defined as the on-going process of gathering valid and reliable evidence of a pupil’s learning achievements against clearly defined criteria using a variety of tools, methods and techniques in different contexts (Ministry of Education 2007). The introduction of CA, through a process known as the Primary Curriculum and Assessment Reform (PCAR), is critical to the implementation of the new Malawian Outcomes Based Education (OBE) curriculum. The implementation of the OBE curriculum entails that the final assessment of learners in the primary schools, including grade 8 (which is the exit class) includes scores generated by the learners from continuous assessment. Unlike summative assessment, CA creates more assessment opportunities and the use of more assessment methods, creating opportunities for learners to demonstrate their mastery of a variety of learning objectives. In Malawi however, there are a number of challenges associated with CA which have resulted in non-inclusion of CA scores to the final grade 8 scores, four years after the implementation of the OBE. This paper highlights the main challenges being faced and suggests some solutions to the challenges.
Poster 02-03 Sponsored byMeasured Progress and Meurs HRM
Predicting school achievement beyond intelligence: Does metacognition matter?
Hudson F. Golino (Universidade Federal de Minas Gerais, Brazil)
Cristiano Mauro (Universidade Federal de Minas Gerais, Brazil)
This study examines the extent to which metacognition contributes to the prediction of school achievement beyond intelligence. Concomitantly, it investigates if general metacognitive ability (GMA) and specific metacognitive ability (SMA) predict different academic achievements. Intelligence (Gf) was measured by the Conjunto de Testes de Inteligência Fluida (CTIF - Fluid Intelligence Test Kit). Metacognition were measured through the Read Monitoring Test and the Appraisals Ability on Mathematics Expressions. General academic achievement (GAA) was measured by annual grades in Math, Brazilian Portuguese, Geography and History. Specific academic achievement (SAA) was measured by an arithmetic test. The sample was composed of 684 students (6th to 12th graders) of a private school in Belo Horizonte, Minas Gerais, Brazil. Two hypotheses were raised: (1) GMA explains general academic achievement (GAA) beyond intelligence, and does not explain, or even weakly explain, specific academic achievement (SAA); (2) SMA explains SAA beyond intelligence, and does not explain, or even weakly explain, GAA. A model based on both hypotheses was generated and tested through SEM. The model presented adequate data fit (χ²= 51.18; df= 19; CFI= 1.00; RMSEA= 0.05). GMA and Gf explain GAA, while SAA was predicted by SMA, Gf and Math achievement. The result shows that metacognition does matter in the prediction of general and specific academic achievement, beyond intelligence.
Poster 02-04 Sponsored byPearson and Prevue HR Systems
Expanding the social goal taxonomy: Construction and validation of the Social Obligation Scale
Ronnel B. King (The University of Hong Kong, Hong Kong SAR, China)
David A. Watkins (The University of Hong Kong, Hong Kong SAR, China / Melbourne University, Australia)
Research in cross-cultural psychology has highlighted the importance of social goals (social reasons for studying) when studying the academic motivation of students from collectivist countries. Dowson and McInerney (2004) recently proposed a 5-factor social goal model which includes: social affiliation, social approval, social concern, social responsibility, and social status goals. A preliminary qualitative study conducted in the Philippines indicated that this five factor model might not be adequate. A sixth type of social goal, which we call social obligation goal was identified. Therefore, the aim of this study was to construct and validate a Social Obligation Scale which can be used to measure the aforementioned construct. Filipino secondary school students were recruited as participants. Within-network and between-network approaches to construct validation were used. Results of the confirmatory factor analysis (CFA) indicated that social obligation was distinct from the five social goals previously identified by Dowson and McInerney. Multigroup CFA showed that it was invariant across different kinds of students. Social obligation was also found to be related to other theoretically-relevant constructs such as academic engagement, achievement, and well-being. Implications for cross-cultural research are discussed.
Poster 02-05 Sponsored byNOA and Schuhfried
Cross-cultural differences between standardization samples of the Test of Motivational Orientation
Rakhubovska Khrystyna (OS Ukraine, Ukraine)
The aim of this study is to compare cultural features of different TOM (Test of Motivational Orientation, Borgogni, Petitta and Barbaranelli, 2004) standardization samples. Psychometric characteristics of Italian, Bulgarian, Romanian, Hungarian and Ukrainian TOM versions were compared. The following results were obtained: (1). Standard deviations of Italian, Hungarian and Romanian TOM scales are higher than standard deviations of Bulgarian and Ukrainian scales. It could be due to high personal significance of the research for Bulgarian and Ukrainian respondents. Standard deviation of Leadership Orientation scale of Romanian TOM is less than standard deviations of the same scale of other versions. It may be related to high importance of leadership in Romanian culture. (2). Bulgarian Achievement and Innovation Orientation factors explain less variance than Leadership and Relationship Orientation factors. It could be related to great importance of these factors in Bulgarian culture. (3). In Romanian and Italian TOM versions the biggest correlation is between Achievement and Innovation Orientation scales – according to McClelland (1987), the need for innovation is an aspect of the need for achievement. In conclusion, differences were revealed between TOM versions related to various factors including cultural features of samples, their composition, question formulating etc.
Poster 02-06Sponsored by Boom test uitgeversand Zeeuws Bureau voor Toegepaste Psychologie
Reading tests: A Decade of sources of validity evidence
Betty-Jean Usher-Tate (Buros Institute for Assessment Consultation and Outreach, USA)
Validity evidence is a critical component to establish credence and for the defensibility of a test being challenged in a court of law (USA), yet within the testing industry the concept of validating a test may take on different meaning, foci, and utility for test publishers, consumers, policy makers, and litigators. Sireci (2009) ascribes that validity is not an inherent characteristic of a test; validity is based on interpretations or inferences from the scores, and should be evaluated within the context of purpose and use of the test. This study uses manuals and reviews of commercially available tests published in the Mental Measurement Yearbooks to examine the validity evidence presented for Reading tests in the last decade.
Poster 02-07 Sponsored byCubiks and National Board of Medical Examiners
History and Current Developments of Clinical Assessment in Bulgaria
Zornitsa Kalibatseva (Michigan State University, USA)
This paper examines the history and the current state of psychological assessment in the field of clinical psychology in Bulgaria. First, the role of historical, sociocultural, and economic factors is discussed to present an integrated description of the contemporary state of mental health and clinical assessment. Then, the various domains that utilize clinical psychological assessment are analyzed. In particular, clinical psychological assessments have been primarily used in schools, private practices, medical offices, and hospital settings as well as research settings. However, there is little information on standard procedures for adaptation and use of clinical psychological assessments. The paper discusses the use of diagnostic manuals, such as the DSM-IV-TR and ICD-10, and the use of questionnaires for depression, anxiety, and other types of psychopathology. Lastly, recommendations are provided for standardizing the process of adaptation and use of such assessments and advancing the field of clinical assessment in Bulgaria.
Poster 02-08
Population Heterogeneity in the Effect of Consuming Value on Compulsive Consumption
Jieting Zhang (South China Normal University, China)
Minqiang Zhang (South China Normal University, China)
Jingxian Li (South China Normal University, China)
Yuexin Xiong (South China Normal University, China)
Yi Wang (South China Normal University, China)
Jing Liu (South China Normal University, China)
The current study aims to investigate the diversity in the relationship between consuming value (namely the value and attitude of consumption) and compulsive consumption (consumption by enthusiasm, with less ration or plan), among college students with different consuming types. 740 students from nine universities in Guangzhou (China) completed the measures of consuming behavior, consuming value and consuming decision-making style. Latent class model was built referring to the consuming behavior, and there were three types of consumption. Generally, students of Type A (51.5%) have an intermediate level of consumption; students of Type B (34.1%) have low expense for all consumptions; students of Type C (14.4%) have highest expense for all consumptions. The result demonstrated that the regression of impulsive consumption to various dimensions of consuming value is different across the three types of consumption. For Type A students, their impulsive consuming was moderately affected by experiential value (the goal of experiencing in consumption), and slightly by symbolic value (the goal of symbolizing), the emotional value (goal of meeting emotional needs) and excessive consumption. For Type B, impulsive consuming was moderately affected by symbolic value and experiential value. For Type C, impulsive consuming was strongly influenced by emotional value and moderately by symbolic value. Conclusion could be made that students with different levels of consumption have different reasons for compulsive consumption. Reason for the heterogeneity was further discussed in respect of socioeconomic status and its related latent consequence.
Poster Session 2b July 3, 13.30-15.30 Room: Rotonde
Poster 02-13
Assessing teacher students’ knowledge about the educational system in Germany
Friederike Hohenstein(University of Kiel, Germany)
Friederike Zimmermann (University of Kiel, Germany)
Olaf Köller (University of Kiel, Germany)
Jens Möller (University of Kiel, Germany)
Thilo Kleickmann (University of Kiel, Germany)
Previous research on professional knowledge of teachers has shown that it is crucial for the quality of education and the achievement of students. The development of professional knowledge therefore is to be considered highly important for teachers. Shulman (1986) classified three categories of professional knowledge: pedagogical knowledge, pedagogical content knowledge and content knowledge. The standards for teacher education in Germany presume knowledge about the educational system and the organization of a school to the category of pedagogical knowledge. There is still a lack of tests to assess the knowledge about the educational system. The aim of this study is to examine an assessment tool to test the knowledge about the educational system. The test consists of 20 items (multiple-choice) and was tested in a sample of more than 1000 Bachelor- and Master students. The psychometric criteria show adequate values for reliability. There exists a correlation to cognitive abilities, which were measured by the “Cognitive Ability Test”, and correlations to scales, which measure educational scientific variables like learning and school development. Differences in the level of studies are shown between beginners and advanced students. Further results from a pilot study and additional aspects of validity will be presented in this study.
Poster 02-14
A review on the reform of high school certificate examination in China
Yang Wu(South China Normal University, China)
Lingbiao Gao (South China Normal University, China)
High school certificate examinations (HSCE) were in the charge of provincial examination boards in China. This makes the situation very complicated due to the vast population and great differences among the provinces. With the new curriculum reform started at 2001, all provinces renewed their school evaluation policies and reconstructed the HSCE. It is very important to have a clear understanding of the policies, real situation and the orientations of the HSCE since it has always a very strong wash-back effect on school teaching, learning, educating, and even the society. It is also important to find out the commonalities and differences among these provinces given their great differences. This project reviewed the HSCE policies, strategies and techniques of 29 provinces in Mainland China. We aimed at answering a series of questions related to the direction, focuses, strategies, format, and techniques of HSCE, as well as the criterion and marking rules, the interpretation and application of the exam results, etc.. Understand the direction of change in HSCE policy and practice in China since 2001. Having reviewed if HSCE in China really progresses on the right track, and the problems on their ways, we present our suggestions to the innovation of HSCE in China.
Poster 02-15
The effect of assessment systems reform on schools and classrooms
Masoomeh Khosravi (Ministry of Education, Iran)
Hamid Reza Badiei (Ministry of Education, Iran)
In our country (Iran), from five years ago there was a reform in assessment systems of primary schools which is still under development. In this reform we implement qualitative assessment instead of quantitative assessment with focus on portfolios, self-assessment, and peer assessment. In this research, we select Markazi province and we are studying the implementation and impact of assessment reform on school and classroom practices, focusing on the subjects of writing and mathematics. Recently, surveys were sent to a representative sample of about 50 elementary school principals and a representative sample of about 150 teachers, who are teaching in first, second and third grades. These are grades in which the qualitative assessment system is implemented in our province (Markazi Province). The surveys asked about the respondents familiarity with the reform and their opinion about it. Principals were also asked about implementation of the reform and changes in assessment system.
Poster 02-16
Assessment of general education goals: Challenges and solutions
Ramezan P. Dowlati (Northern Virginia Community College, USA)
New technology and advances in measurement have raised our hope for more objective assessment in many areas of interest. However, major challenges still exist in the assessment of general goals that most nations and institutions have set for their educational systems. Schools at all levels, from kindergarten to college, need to prove their positive role in helping students achieve general education goals like critical thinking, communication, quantitative reasoning, scientific reasoning, information literacy, cultural and social understanding, personal development, civic responsibility. The assessment of general education goals has become more important as many educational institutions have chosen to focus on training specialized workforce for the job market and allocated only limited resources to courses and activities that help the general goals of education. The current trend has caused concerns among the public who expect schools and colleges play a significant role in achieving the general education goals. This makes the assessment of such goals a sensitive and important task that needs the collaboration of educators and testing experts. This presentation discusses some of the challenges that the assessment of general education goals is currently facing. Using examples from community colleges and universities in the United States, practices of assessment are reviewed, some major challenges are highlighted, and a number of practical solutions are proposed.
Poster 02-17
Higher education evaluation in Brazil
Claisy Maria Marinho-Araujo (Universidade de Brasília, Brazil)
Mauro Luiz Rabelo (Universidade de Brasília, Brazil)
Higher education has experienced a great growth around the world which has brought challenges and demands to the related institutions in order to promote the students success. The Brazilian experience in large-scale assessment gained over the years enabled the deepening and expansion of the techniques applied in this process and its use in other systemic reviews, especially in the Brazilian Exam of Students Performance - ENADE - in the context of the Brazilian System of Higher Education Evaluation - SINAES. In this context, this paper intends to do reflections concerning the importance of the development of competences as one of the privileged strategies in higher education evaluation. The focus of the article, besides including the underlying methodological theoretical conception on evaluation reference matrix, is to discuss the construction of the subjectivity and the professional formation mediated by evaluative processes, in the contexts of teaching and learning. In spite of the complexity of development competences process, it is evidenced that the use of the reference matrix is an excellent alternative as guide of the evaluation process, especially for its systemization and as an opportunity of doing inferences about the achievement of the expected profile of the professionals.
Poster 02-18
Mapping and evaluation of competencies in higher education
Cynthia Bisinoto (Universidade de Brasília, Brazil)
Claisy Marinho (Universidade de Brasília, Brazil)
Higher Education is regarded as a fundamental element of social inclusion and economic development based upon the argument that the development of a nation and its competitive ability are closely related to the quality of its Higher Education System. However, a complex set of factors is involved in the preparation of higher education and in the development of competent professional profiles, that are socially committed and sustained upon the development of technical, ethical, political, social and other competencies.Developing and evaluating these competenciesthroughout the course of training have been a challenge taken up by researchers and teachers committed to the quality of Higher Education. In this paper, we discuss the process of mapping and evaluating the profile of students in the Natural Sciences course from a university in Brazil with particular emphasis on the construction processes of a “reference matrix”that guides and systematizes the construction, validation and application of an evaluation questionnaire. The results from this evaluation aim to provide teachers and institutional administrators with a broad overview of student performance and, moreover, assist in promoting the necessary changes in the teaching and learning process.
Poster 02-19
Psychometric equivalence of PISA2009 across languages in Spain
Josu Mujika (University of the Basque Country, Spain)
Paula Elosua (University of the Basque Country, Spain)
Paola Bully (University of the Basque Country, Spain)
Spain has participated in PISA since its first edition in 2003. In the 2009 edition 15 of Spain’s autonomous regions also took part, providing local samples. In some of them the questionnaires were administered in the co-official language of each region: Galician, Catalan, Valencian and Euskera. Given the demonstrated influence that the adaptation of questionnaires can have on validity, the objective of this paper is to analyze the psychometric equivalence between language versions of the reading test. Taken as a reference group that performed the test in Spanish, we compare the average of the evaluated dimensions, factor structures and differential item functioning of the versions in Euskera, Galician, Valencian and Catalan. In addition, a quantitative analysis seeks to explain the discrepancies found.
……………………………………………………………………………………………………………………………………………………....
Poster Session 3 July 3, 15.30– 17.15 Room: Rotonde
Poster 03-01
The Chinese revision of the SON-R 6-40 non-verbal intelligence test
Jianping Xu (Beijing Normal University, China)
Zhi Guo (Beijing Normal University, China)
Qian Wang (Beijing Normal University, China)
Houcan Zhang (Beijing Normal University, China)
The new version of the SON-R Test for age 6-40 has been revised into Chinese. With the help of Dr. Peter Tellegen from the University of Groningen, this project started in 2009 and was accomplished at the end of 2011. The norm group was collected in different provinces around the whole country. In total 1721 individuals were selected according to the results of national census, with reference to sex, social-economic status, educational level etc. Before application in China, a cross-cultural adaptation study was achieved. By comparing the Chinese norm and the Dutch/German norm, some positive results were obtained. First, both the reliability and validity indexes of the Chinese version of the SON-R 6-40 are qualified psychometrically and close to those of the original, which means cross-cultural consistence exists. Second, the study of cross-cultural equivalence showed some minor cultural differences between Chinese and Western subjects, such as thinking style and speed; however, the construct equivalence and measurement unit level of cross-cultural equivalence can be clearly observed. In conclusion, the Chinese version of SON-R 6-40 is suitable for use in China and also suitable for cross-cultural study.
Poster 03-02
Interpretation of the Performance Scale of the SON-R 2½-7[a] a nonverbal test of intelligence
Felipe Valentini (University of Brasília, Brazil)
Jacob Arie Laros (University of Brasília, Brazil)
This study aimed to interpret the Performance Scale items of the SON-R 2½-7[a] based on IRT scores. The SON-R 2½-7[a] is a non-verbal test of intelligence with Brazilian norms. The Performance Scale (PS) consists of two spatial tests: Mosaics and Patterns. In Mosaics the children are required to copy several mosaic patterns using colored squares. In Patterns the children need to copy an example using a pencil. The data of the Brazilian normatization study were used (N = 1,200). The minimum IRT score was determined at which 65% of the children gave the correct answer. The items were grouped in three sets: The first consists of easy items that can be resolved observing only a limited amount of stimuli. The second set represents items with some degree of asymmetry. The items of the third set are more asymmetric, and the number of lines in Patterns increases while in Mosaics the borders of the colors do no longer represent the borders of the squares. These results are consistent with the difficulty theory developed for the SON-test, which relates item difficulty to asymmetry. The item difficulty seems to increase along with the requirement to deal with visual detail and spatial orientation.
Poster 03-03
Verification of differential item functioning in the Brazilian version of the SON-R 2½-7[a]
Jacob Arie Laros (Universidade de Brasília, Brazil)
Camila Akemi Karino (Universidade de Brasília, Brazil)
The presence of Differential Item Functioning (DIF) in measuring instruments can affect the validity of their scores. In this context, the occurrence of DIF in the Brazilian version of the SON-R 2½-7[a] was investigated. This version is the abridged edition of the SON-R 2½-7, a non-verbal test of intelligence of Dutch origin for young children. The data of the Brazilian normatization sample of 1,200 children with ages between 3 and 7 years were used in this study. The occurrence of DIF was verified in relation to gender and in relation to the five Brazilian regions (North, North-East, South, South-East, and Centre-West) using the IRT method. Of a total of 60 items, 5 items were indicated as having DIF for gender and 13 items as having DIF between regions. An important observation in this respect is that much more items with DIF were detected in the Performance Scale than in the Reasoning Scale. This result seems to indicate that previous attempts to reduce the occurrence of DIF in the Reasoning Scale have been successful. Considering that the vast majority of the items of the SON-R 2½-7[a] did not show DIF, it might be concluded that the test can be adequately used in Brazil.
Poster 03-04
The adaptation of the Brazilian Wechsler Abbreviated Scale of Intelligence
Denise Balem Yates(Universidade Federal do Rio Grande do Sul, Brazil)
Carlos Henrique Sancineto da Silva Nunes(Universidade Federal de Santa Catarina, Brazil)
Clarissa Marceli Trentini(Universidade Federal do Rio Grande do Sul, Brazil)
The Wechsler Abbreviated Scale of Intelligence (WASI) is an American instrument developed to be a brief and reliable measure of intelligence. The aim of this paper is to present the adaptation of WASI’s verbal subtests. The adaptation of WASI’s Vocabulary and Similarities subtests followed the International Testing Commission Guidelines on Adapting Tests and, therefore, consisted not only of the translation of the original instrument in English for Brazilian Portuguese, as well as a semantic equivalence of the words was sought and psychometric evidences were obtained with Rasch analysis from Item Response Theory (IRT). Analyses were performed in Winsteps Version 3.72.0. The steps of the verbal adaptation were: simple translation, experts’ evaluation of the translation, pilot study, generation of extra items, assessment of adaptation sample, scoring, interscorer agreement, Rasch analysis, exclusion of items with misfit, selection of items with different levels of difficulty and higher item-theta correlation and ordering of items in terms of item difficulty. It was concluded that verbal adaptation requires semantic equivalence with the original test as well as good psychometric properties. In WASI’s adaptation study, Rasch analysis provided a reliable tool for item selection and consequent construction of the adapted version.
Poster 03-05
Assessment of intelligence in Greece using the General Ability Measure for Adults: Concurrent validity with the WASI
Elena C. Dieti (American College of Greece, Greece)
Achilles N. Bardos (University of Northern Colorado, USA)
The measurement of intellectual ability is an important component of psychological assessments conducted around the world. The General Ability Measure for Adults (GAMA; Naglieri & Bardos, 1997) utilizes abstract designs to assess general cognitive ability and was designed and normed in the USA for individuals ages 18 to 96 years old. It provides a measure of intellectual ability that is not confounded by a person’s linguistic skills (receptive & expressive). Having these features the GAMA has the ability to be used in various cultures. Given that very few tests are available for Greek psychologists that do not involve linguistic components, this study examined the concurrent validity of the GAMA with another similar subtest (Matrices) of brief intelligence, from the Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999). Fifty adults were administered the tests in a counterbalanced order. Minimum mean score differences were observed with the sample performing in the average range. Significant correlations were also observed (r = .48, p< .001) although this value is affected by the sample’s restricted distribution of scores (GAMA SD = 10.7). Implications of these concurrent validity findings for use of the GAMA in Greece will be discussed.
Poster 03-06
Construct validity of the WISC–IVUK with a large Irish sample of referred children
Gary L. Canivez (Eastern Illinois University, USA)
Marley W. Watkins (Baylor University, USA)
Trevor James (Éirim: The National Assessment Agency, Ltd., Ireland)
Kate James (Éirim: The National Assessment Agency, Ltd., Ireland)
Rebecca Good (Éirim: The National Assessment Agency, Ltd., Ireland)
Irish educational psychologists often use the Wechsler Intelligence Scale for Children–Fourth UK Edition (WISC–IVUK; Wechsler, 2004) in clinical assessments of children with learning difficulties. Construct validity studies of the WISC–IVUK with Irish samples have not yet been reported. WISC–IVUK core subtest data obtained from evaluations to assess learning difficulties in 794 Irish children (494 male, 300 female) were examined. One through four, first-order factor models and indirect versus direct hierarchical models were examined and compared using CFA. The four–factor (VC, PR, WM, PS) model provided the best fit to these data, but no meaningful differences in fit statistics (c2, CFI, RMSEA, AIC) were observed between the four, first-order factor; the indirect hierarchical; and direct hierarchical models. For statistical and theoretical reasons, the direct hierarchical model provided the best explanation of the WISC–IVUK factor structure. The general factor accounted for 63.7% of the common variance while VC, PR, WM, and PS factors accounted for 9.4%, 8.6%, 8.7%, and 9.6% of the common variance, respectively. Thus, the results with referred Irish children were similar to those from other investigations (Gignac, 2005, 2006; Watkins, 2010) further demonstrating the stability of the WISC-IV factor structure across cultures.
Poster 03-07
The WISC Arithmetic and Digit Span subtests: What do they measure?
Christina M. Ravert (Manheim Township School District, USA)
Marley W. Watkins (Baylor University, USA)
Confirmatory factor analyses were conducted to investigate hypotheses regarding the Arithmetic and Digit Span subtests that formed the third factor of the Wechsler Intelligence Scale for Children. Previous research suggested that these subtests may be measures of working memory, quantitative ability/reasoning, or some other ability. When the Wechsler Intelligence Scale for Children-Third Edition was administered to a referral sample of adolescents in conjunction with marker tests for memory and quantitative reasoning, a five factor solution was optimal. The resulting factors were verbal comprehension, perceptual organization, processing speed, quantitative reasoning, and memory. Based on these results, the WISC Arithmetic and Digit Span subtests appear to be measures of different abilities with Arithmetic being a measure of quantitative reasoning and Digit Span a measure of memory. Given these results, interpretation of the Arithmetic subtest as a measure of memory may not be accurate.
Poster 03-08
Analysis of French WISC-IV index score scatter and level of performance with gifted and nongifted children
Thierry Lecerf (University of Geneva, Switzerland)
Isabelle Reverte (University of Geneva, Switzerland)
Sotta Kieng (University of Geneva, Switzerland) (presenting author)
Nicolas Favez (University of Geneva, Switzerland)
Jérôme Rossier (University of Lausanne, Switzerland)
Nathalie Scherrer (University of Geneva, Switzerland)
The French version of the Wechsler Intelligence Scale for Children- Fourth Edition (WISC-IV) was administered to 20 “non consultant” gifted children (FSIQ = 126.6) and 20 nongifted children (FSIQ = 96.9), aged from 8 to 12. Samples were matched on chronological age, sex, and grade level. In addition to the four factor scores of the WISC-IV (VCI, PRI, WMI, and PSI), five Cattell-Horn-Carroll (CHC) composite clusters were computed: Fluid reasoning (Gf), Comprehension-Knowledge (Gc), Visual processing (Gv), Processing speed (Gs), and Short-term-memory (Gsm). The first objective of this study was to determine the extent of the difference between gifted and nongifted in subtests scores, factor scores and CHC composites scores. The second objective was to determine whether factor-score or CHC-composite score scatter was larger for gifted than for nongifted children, and whether scatter using significant deviation from the mean index score was larger for gifted than for nongifted children. To deal with these goals, T-tests were conducted with Bonferroni correction. Although sample size was small, it was sufficient for detecting differences for all comparisons. Results were relatively consistent with previous studies and indicated that gifted children scored significantly higher on all indices, except on Visual Processing score (Gv). The results of the profile analysis indicated that gifted and nongifted children display similar scatter. Thus, composite-score scatter was not larger for gifted than for nongifted children.
Poster 03-09
Long-term stability of the French WISC-IV: An exploratory study
Sotta Kieng (University of Geneva, Switzerland)
Isabelle Reverte (University of Geneva, Switzerland)
Nathalie Scherrer (University of Geneva, Switzerland / Distance Learning University, Switzerland)
Nicolas Favez (University of Geneva, Switzerland)
Jérôme Rossier (University Lausanne, Switzerland)
Thierry Lecerf (University of Geneva, Switzerland / Distance Learning University, Switzerland)
Intelligence is presumed to be an enduring trait. However, few studies have investigated the long-term stability of the WISC-IV scores. The goal of this study was to explore the long-term stability of the French WISC-IV scores with an average test-retest interval of 2.64 years. Test-retest reliability coefficients of the four Index scores (Verbal Comprehension: VCI; Perceptual Reasoning: PRI; Processing speed: PSI; and Working memory: WMI) and the Full Scale IQ (FSIQ) were assessed. The sample consisted of 50 nonclinical French-speaking Swiss children aged between 8 and 12 years (mean age = 9.94; SD = 1.43). Mean difference from first assessment to second assessment was not statistically significant for FSIQ (M1 = 105.5; M2 = 104.9), VCI, PRI, and PSI. Mean difference from first assessment to second assessment was statistically significant for WMI (M1 = 96.88 and M2 = 93.76; t(49) = 2.23, p = .03). Finally, test-retest reliability correlations between first and second assessment were calculated, and were very high for VCI (r = .82), and FSIQ (r = .83), and only moderate for WMI (r = .58), PSI (r = .67) and PRI (r = .64). In sum, these results supported only the long-term stability of the French FSIQ and VCI.
Poster 03-10
The relationship between results on the WISC-IV-HR intelligence indices and academic achievement
Krunoslav Matesic (University of Zagreb, Croatia)
The adapted and standardized Wechsler Intelligence Scale for Children (WISC-IV-HR) was published in the Republic of Croatia in 2009. The norms were obtained on the basis of a stratified sample of 1200 children aged 6.0 to 16.11 years. This paper presents data on the relationship between the WISC-IV-HR intelligence indices, IQ and general half-term academic achievement in higher primary school grades (grades 5 to 8) which were included in the standardization. The obtained correlation coefficients for samples ranging from N = 105 to N = 113 are from 0.281 for processing speed and general school success in the 5th grade to 0.586 for IQ and general academic achievement in the 8th grade.
Poster 03-11
The correlation between general intelligence and a general factor of personality
Anders Sjöberg (Stockholm University, Sweden)
Sara Henrysson Eidvall (Sandahl Partners, Sweden)
Maria Åkerlund (Sandahl Partners, Sweden)
Spearmans general intelligence construct (g) is well known. Recently research have proposed that there is also a General Factor of Personality (GFP) which accounts for the relationships found between the big five personality dimensions. Rushton (1985) suggested that a positive correlation should exist between personality and intelligence based on the r-K life history theory. The theory claims that hierarchically organized traits can be explained by a single super-factor. The r-K life theory highlights that traits need to be harmonized, not work independently of each other. Based on r-K life history theory this study test the hypothesis of a positive correlation between general g and GFP in a sample of manager candidates (N = 286). The particpants had completed a big five personality inventory (NEO-PI-R) and a measure of general intelligence (BasIQ). Contrary to the hypothesis, the correlations between the GFPs and g were found to be non-significant, suggesting independence between GFP and g. Theoretical and methodological implications of the results are discussed.
Poster 03-12
Concurrent validity of the General Ability Measure for Adults (GAMA) with the K-BIT2 in Greece
Demetrios Katsikis Aristotle (University of Thessaloniki, Greece)
Stiliani Kasapi (Provate practice, Thessaloniki, Greece)
Achilles N. Bardos (University of Northen Colorado, USA)
There are limited psychological tests developed in Greece and those adapted lack the proper documentation that ITC guidelines have been followed in their development and adaptation process. This study utilized two nonverbal measures of intelligence and examined their concurrent validity as part of the validation of the General Ability Measure for Adults (GAMA; Naglieri & Bardos, 1997) in Greece. The GAMA uses abstract designs to assess general cognitive ability. It was developed in the USA for adults (18 and older) and assesses intelligence without the confounding effects of expressive language skills. The concurrent validity criterion was the Matrices subtest, a similar nonverbal measure adapted from the Kaufman Brief Intelligence Test (K-BIT-2; Kaufman & Kaufman, 2004). Fifty one adults (33 males and 18 females) with an average age of 32 yrs (SD = 7.5) were administered the tests in a counterbalanced order. The sample performed in the average range, with mean scores of 99 and 101 for the GAMA and K-BIT2 respectively while the differences where small and nonsignificant (USA-GAMA = 103 ) with a matched sample from the US standardization sample. Significant correlations were also observed (r = .60, p< .001). This data provide evidence for the use of the GAMA in Greece.
Poster 03-13
Subtests Matrix Reasoning and Coding of BETA-III in Brazilian population
Ivan Rabelo (Casa do Psicólogo Publishing, Brazil)
Silvia Pacanaro (Casa do Psicólogo Publishing, Brazil)
Gisele Alves (Casa do Psicólogo Publishing, Brazil)
Irene Leme (Casa do Psicólogo Publishing, Brazil)
This study aimed at investigating the influence of variables on performance in two subtests (Matrix Reasoning and Coding) of BETA III. 1,044 people were participants, 532 men (51%) and 510 women (48.8%), aging from 14 to 83 years old (M = 29.22, SD = 11.6). Their schooling ranged from elementary school (19%), High (34.1%) to College (45.7%). The result of the Matrix Reasoning subtest was significantly associated with age, with a trend for lower scores as ages increase. About schooling, an increase in scores was observed as the schooling also increased. As for the subtest coding in relation to age, there were significant differences in the performance of people between 20 and 36 years old, and those participants who were 37 years old or more had significantly lower scores. Participants with higher education had significantly higher means, indicating that the higher level of schooling led to changes on performance improvement in processing speed. Thus, it can be said that, in this sample, there was an influence of age and schooling in the performance on both subtests, indicating the use of separate normative tables for each specified variable.
Poster 03-14
Test Information Function and DIF of a 12-item-version of the Raven's Matrices
Silvia Galli (University of Florence, Italy)
Marianna Donati (University of Florence, Italy)
Caterina Primi (University of Florence, Italy)
Francesca Chiesi (University of Florence, Italy)
Due to the wide use of the Raven Progressive Matrices to assess fluid ability, the present work aimed to investigate the psychometric proprieties of Arthur and Day (1994) Advanced Progressive Matrices-Short Forms (APM-SF) applying Item Response Theory (IRT) in order to provide evidence that it might be employed for a sound assessment of general fluid ability with a short time frame. The APM-SF was administered to 1,956 students (Mean age = 19.30, SD = 4.08 years). Once attested the unidimensionality, the increasing order of items difficulty level, Test Information Function and Differential Item Functioning across age and gender were estimated through the 3PL model. Results revealed that the APM-SF is adequately accurate in measuring general intelligence and it is metrically equivalent across age and gender, nevertheless findings did not completely confirm the progressive order of the items difficulty level. Therefore, following the ranging empirically determined, we administered the reordered scale to a sample of 335 students (Mean age = 17.75, SD = 1.05 years), in order to examine the effect of the reordering on the test performance. These findings replicated evidences of the scale’s accuracy in the assessment of the latent trait, confirming the adequacy of the Arthur and Day’s short form as a measure of general fluid ability.
Poster 03-15
A set of methodological strategies to Identify Developmental Stages on Cognitive Tests
Hudson F. Golino (Universidade Federal de Minas Gerais, Brazil)
Cristiano Mauro (Universidade Federal de Minas Gerais, Brazil)
Michael Lamport Commons (Harvard Medical School / Dare Institute, USA)
Patrice Marie Miller (Salem State College, Salem MA, USA)
This study will show a set of methodological strategies to identify developmental stages, having the Model of Hierarchical Complexity (MHC) as reference. The MHC is a general measurement theory, part of the normal Mathematical Theory of Measurement applied to the phenomenon of difficulty of tasks and items. It was demonstrated, in axiomatic terms, that task difficulty or complexity, beyond other sources, increases in two ways: horizontally and vertically. The first refers to the accumulation of informational bits necessary to successfully complete a task, while the second refers to the organization of information in the form of action in two or more subtasks, in a coordinated way. Tests designed to identify developmental stages should control horizontal complexity and employ items with different vertical complexity. In order to check this assumption, an inductive reasoning test (IRTD), based on the MHC, was constructed. It was administered to a convenience sample composed by 188 Brazilian people (42.3% men, 57.7% women) aged between 6 to 65 years (M = 21.45, SD = 14.31). In the first part of the data analysis the dichotomous Rasch Model was used. In the second part of the analysis, the spacing of Rasch scores between items of adjacent OHC where verified using the formula [pic]. To verify if the differences between difficulties of items from stage k and the mean difficulty of the stage k+1 are statistically significant, the One-Sample t-test is used, with a 95% confidence interval. The effect size is calculated using the Cohen’s d. The results can be summarized as follow: 1) The reliability for the full scale was .99, and its Infit mean was .94 (SD = .22; Max = 1.46; Min = .56); 2) The person reliability was .95; 3) The one-sample t-test showed that the comparisons between all adjacent OHC were significant, with large effect sizes. Thus, our evidences suggest that developmental stages can be empirically verified using the methodological strategies used.
Poster 03-16
Bridging intelligence and developmental testing - The Intelligence and Development Scales
Priska Hagmann-von Arx (University of Basel, Switzerland)
Alexander Grob (University of Basel, Switzerland)
The poster presents the Intelligence and Development Scales (IDS; Grob, Meyer & Hagmann-von Arx, 2009) which were developed to provide a combined assessment of intellectual and developmental functioning in children aged 5 to 10 years. The IDS include a measure of general intelligence and additionally provide a developmental profile for the functional areas of cognition, psycho-motor skills, social-emotional competences, mathematics, language, and achievement motivation during the testing session. The IDS were standardized in Germany, Austria, and Switzerland with 1330 children. The internal consistency of the functional areas range from α = .78 for psycho-motor skills to α = .92 for cognition; construct validity with the WISC-IV is given (r = .83). Comparisons of normal and gifted children, learning-disabled children, children of foreign-language, children with ADHD disorder, children with Asperger’s syndrome, and children with aggressive behavior offer additional proof of the test’s differentiation validity. We conclude that the IDS are well suited to measure children’s strengths and weaknesses in six functional developmental domains and their interplay. Therefore we strive to extend the IDS for preschool children (IDS 3-5; Grob, Reimann, Gut, Frischknecht, in prep.) and for the second decade of life (IDS 10-20; Grob & Hagmann-von Arx, in prep.).
Poster 03-17
Predicting intelligence and academic achievement in primary school: The predictive validity of the Viennese Developmental Test
Ursula Kastner-Koller (University of Vienna, Austria)
Pia Deimann (University of Vienna, Austria)
The Viennese Developmental Test (WET; Kastner-Koller & Deimann, 2002) is a developmental test for three- to six-year olds, which allows for a comprehensive developmental assessment of the entire range of important areas of functioning such as motor, visual, memory, cognitive, language, and psychosocial development. As the WET was designed for assessing preschool development with respect to the child’s future adaptation to academic challenges, analyzing the test’s predictive validity is crucial. Two studies are presented: Study 1 deals with the follow-up of 27 children who were assessed at preschool age using the WET and a second time two years later at school age using the German version of the WISC IV. Additionally, reading and writing competences were examined as well as socio-emotional adaptation to school. Study 2 analyzes the predictive validity of the WET using the AID 2. The sample of Study 2 comprises 22 children having been assessed at the age of 5 and reassessed one year later. Both studies yield high intercorrelation coefficients between the developmental scores at preschool age and IQ scores at school age with more than 50 % shared variance. Moreover, Study 1 points to some WET-subscales predicting reading and writing competences, others predicting socio-emotional adaptation to school.
Poster 03-18
The influence of lexical knowledge on the proportions correct of verbal intelligence test items
Jan van Bebber (Meurs HRM, The Netherlands)
Liesbeth E. van Leeuwen (Meurs HRM, The Netherlands)
Investigating the construct validity of measures, assuring that actually has been measured what one purposed to measure, is an ongoing process that accumulates information in the nomological network surrounding the instrument. With respect to verbal intelligence items, an important facet of this network is the degree to which the items draw on lexical knowledge. In this study, we first validated a new approach for estimating word frequency indices based on text-analysis of Dutch subtitles called SUBTLEX-NL (Keulers & Brysbaert, 2010) by correlating four different word frequency indices with the item-difficulties of vocabulary items (Q1000 Vocabulary). Additionally, we used these indices to explain variations in item-difficulties on two other verbal scales that are part of the Q1000 test battery, verbal analogies and diagrams. With respect to verbal analogies, Roccas et al (2003) found that the familiarity index of the most unusual of the four words of the itemstem and the keyed response correlates highest with item difficulty (rxx=.22). Surprisingly, we found that word frequencies only accounted for approximately 20% of the variance in item difficulties of the vocabulary items. . For Q1000 Analogies we found negative correlations of-.17 (stem + keyed response) and -.36 (distractors) of the word frequencies of the words being used and the corresponding item difficulties. The correlation for Q1000 Diagrams was weak, but positive (.08) as was expected. Information regarding the relationship between word frequency measures and item-difficulties of various types of verbal intelligence items could be used as a prior information in item calibration studies.
Poster 03-19
Reliability and validity of the Syllable-Word Working Memory Test in Russian language
Tatjana Turilova-Mišcenko (University of Latvia, Latvia)
Malgožata Rašcevska (University of Latvia, Latvia)
The purpose of the study was to determine the psychometric properties of the original created Syllable-Word Working Memory Test Russian version (SWWMT-R). This test was developed on the base of the working memory model by Baddeley and Hitch (1974; 2000). The test contains series of word syllables in a wrong order and each respondent’s task is to construct words from the syllables. The SWWMT-R consists of 11 items. The sample included 110 participants aged from 13 to 16 years old and was roughly gender balanced (51% female and 49% male). All items of the SWWMT-R show good difficulty and discrimination indices. Results show that the average item difficulty index of the test is .40 and average item discrimination index is .40. Spearman-Brown split-half reliability is .75. The concurrent and convergent validity was established for SWWMT-R using WISC-III Digit span subtest, r = .63, p< .001.
……………………………………………………………………………………………………………………………………………………....
Poster Session 4 July 4, 08.45-10.45 Room: Rotonde
Poster 04-01
Assessment of creativity in the educational setting: Brazilian instruments
Eunice M. L. Soriano Alencar (Catholic University of Brasilia, Brazil)
Denise de Souza Fleith (University of Brasilia, Brazil)
Creativity is a complex and multi-faceted phenomenon, manifested in multiple forms and domains. Its importance in diverse contexts has been increasingly recognized, given the benefit not only for the individual but for the society as well. An increasing number of instruments were developed in recent decades with the aim of collecting data on the multiple dimensions of the construct and of investigating variables that influence creativity in diverse environments. However, there is scarcity of instruments designed to assess different factors that impact the expression of creativity in educational settings. This presentation will provide a brief description of instruments designed to assess factors associated with the promotion or inhibition of creativity in educational settings, such as climate for creativity in elementary school classrooms, educational practices for creativity in higher education, and personal barriers to creativity. The rationale for the construction of the instruments, their psychometric properties, and studies conducted with samples of teachers and students will be described, as well as suggestions for future investigations to strengthen their psychometric properties will be pointed out. The presentation will finally address major issues in creativity assessment in educational settings
Poster 04-02
Thinking and Creating Styles: Psychometric studies with a Portuguese sample
Sara Ibérico Nogueira (Universidade Lusófona de Humanidades e Tecnologias, Portugal)
Leonor Almeida (ISG-Business School, Portugal) (presenter)
Solange Wechsler (PUC-Campinas, Brasil)
We present some results concerning the psychometric characteristics of the Thinking and Creating Styles Scale (Wechsler, 2006) for a Portuguese adult sample with 475 participants, 254 males and 221 females, aged between 18 and 64 years old, composed mostly of college graduates (84.2%). The results are based on exploratory factorial analysis and on the analysis of the reliability with Cronbach’ Alpha method. The results reveal a good level of internal consistency (Cronbach Alpha = .73). From the factorial analysis emerge five dimensions that explain 24.4% of the total variance. We discuss the implications of the results for the evaluation of the thinking and creating styles for the Portuguese population.
Poster 04-03
Thinking and Creating Styles Scale: Portuguese adaptation
Margarida Pocinho (University of Madeira, Portugal)
Soraia Garcês (University of Madeira, Portugal)
Solange Wechsler (Pontifical Catholic University of Campinas, Brazil)
Saul Neves de Jesus (University of Algarve, Portugal)
This objective of this study was to validate the Thinking and Creating Styles Scale to an adult Portuguese population. The first step was a pre-test in order to obtain the final version of the instrument to be used in the research. The administration was on-line and also by paper survey at the University of Madeira, with a total of 331 participants. The data was then analyzed using SPSS version 17.0.
First, by analyzing the “α if item deleted” and the “item total-correlation” the need to eliminate some items from the original 100 was evident. Consequently, the final version contains 49 items with a cronbach’s alpha of 0.91. An exploratory factorial analysis showed the presence of 5 factors, named: non-conforming/transformer; emotional/intuitive; relational/divergent; independent judgment and logical/objective. It is concluded that this reduced version of the scale shows great validity. However, new studies to confirm the validity are needed. We recommend extending the sample to a broader population.
Poster 04-04
The Creative Functioning Test (CFT): Validity
Emils Kalis (Daugavpils University, Latvia)
Liga Roke (Riga Teacher Training and Educational Management Academy, Latvia)
Studies on creativity mostly are based on measures of divergent thinking and self-report. Traditional assessment of divergent thinking has many psychometric drawbacks but self-reports seem to be dependent on people’s ability of self-reflection and both could be considered only as parts of minor importance evolving on the more general basis of creative personality. An instrument claiming to account for this is The Creative Functioning Test (CFT) which measures a creative personality with a process-oriented approach employing perceptgenetic (PG) technique (Smith & Carlsson, 2001). Measure of CFT has two main dimensions where the first represents a person’s inclination for active interpretation of the world but the second reveals a person’s tendency to refuse already assigned meanings and readiness for reinterpretation of the perceived world. To test convergent and divergent validity of CFT, the Task of Unusual Uses, the Test for Creative Thinking and the Big Five Inventory were administrated along with CFT in a sample of 50 respondents. Investigation of correlations suggests that CFT has adequate convergent and divergent validity thus it could be regarded as a measure of creativity. Also some weak points of CFT are discussed concerning difficulties in testing procedure and in conducting retest studies.
Poster 04-05
Do you feel cautions as an ant or nonconformist as the grasshopper?
Leonor Almeida (ISG-Business School, Portugal)
Sara Ibérico Nogueira (Universidade Lusófona de Humanidades e Tecnologias, Portugal)
Sara Goulão (ISG-Business School, Portugal)
Considering the thinking and creating styles as behaviour trends, resulting from certain cognitive skills and personality characteristics, as well the life values as the matrixthat guidesthe behavior, we expect both of them be related and mutually influential. The aim of this study is to analyze the correlation between the thinking and creating styles and the life values of a Portuguese adult workers sample. The thinking and creating styles are assessed by the Thinking and Creating Styles Scale (Wechsler, 2006) and the life values of life by the LVI-Life Values Inventory (Brown & Grace, 1996). The sample has 180 participants, 72 males (40%) and 108 females (60%), with age range between 18 and 64 years old (M = 35.7; SD = 9.1) and with different educational levels: 22 primary level (6.1%), 125 secondary level (35.6%) and 32 college graduates (58.3%). In spite of the fact that we found some expected correlations, we shall reflect about the lowest correlations between the Emotional/Intuitive style and the Relational/Divergent style, on one hand, and the creativity value, on the other, which raises the question about the gap between our values and the trends from which we behave. It also raises the question of the social desirabilityof the answers. Furthermore, we must think about the correlation between these thinking and creating styles and creativity values and the real creative performance.
Poster 04-06
Constructing guidelines for psychological assessment teaching for undergraduate psychology courses in Brazil
Maiana Farias Oliveira Nunes (Faculdade Avantis, Brazil)
Monalisa Muniz Nascimento (Fundação do Ensino Superior do Vale do Sapucaí, Brazil)
Over the last decade, Psychological assessment in Brazil has improved noticeably, specially regarding the psychometric properties of tests and their respective user manuals. Despite these improvements, psychologists haven’t being sufficiently prepared to conduct psychological assessments, because teaching of this subject has been overlooked. There are no laws or national guidelines on what would be the minimum content to teach on undergraduate psychology courses regarding the topic. For instance, some psychology courses offer only one class for this subject during five years of study. Given that, this paper aimed at proposing guidelines for teaching psychological assessment in Brazil. For this purpose, we revised national and international publications as well as we interviewed psychological assessment teachers. A set of 23 skills that psychologists should develop when taking psychology undergraduate courses was listed, such as knowledge of the history of psychological assessment and psychometrics. We also proposed the period of the course these skills should be taught and listed references for theoretical and practical support. We found some difficulties in attempting to adapt international standards to the Brazilian context, since the knowledge and dissemination of information considered as basic topics in other countries has not been sufficiently explored in ours.
Poster 04-07
Psychological testing in Argentina: Some indicators about research developments (2001-2011)
Mercedes Fernández Liporace (Universidad de Buenos Aires, Argentina)
María Josefina Scheinsohn (Universidad de Buenos Aires, Argentina)
Alejandro Castro Solano (Universidad de Buenos Aires, Argentina)
In order to diagnose current status in local research on psychological testing, some developmental indicators corresponding to the last decade are presented. Results allow rational long-term planning, oriented by weak points’ recognition and can stimulate actions and improvement with respect to vacant areas. By taking into account some descriptors of Argentinean scientific productivity in this area this study also considers additional academic and professional points: 1. Local specialized peer-reviewed journals (number, indexation, impact, scope). 2. Local psychological peer-reviewed journals, referred to general topics but with special interest in testing (number, indexation, impact, scope). 3. Main research groups (number, productions in terms of test constructions or test adaptations, visibility). 4. Specialized professional associations (number, scope, main purposes and activities). 5. Specialized papers in peer-reviewed journals (number and citations obtained). 6. Specialized books and chapters (number, scope, scientific committee reviews). 7. Specialization Careers and Masters (number and type). 8. PhD theses analyzing psychological tests as main subject (number). 9. h indexes for main authors. 10. Collaborations between research groups. Data collection: main searching engines and databases; public universities. Conclusions are analyzed regarding concrete implications linked to quality standards for psychological testing and instrumental developments, concerning challenges involved in design and construction processes.
Poster 04-08
The development of Guidelines and a Primer Book for psychological testing in Brazil
Lucas Dannilo Aragão Guimarães (University of the State of Piauí, Brazil / Court of Justice of Maranhão, Brazil)
Ana Valeria Lopes Lemos(University of the State of Piauí, Brazil)
Andrea Thaís Xavier Rodríguez Hurtado(University of the State of Piauí, Brazil)
Ana Rayonara de Sousa Albuquerqu (University of the State of Piauí, Brazil)
Fernanda Tamyris de Oliveira Lopes(University of the State of Piauí, Brazil)
Larissa de Carvalho Muniz(University of the State of Piauí, Brazil)
Brazilian surveys point to the negative social impact that psychological testing has in psychological evaluation in general, and more specifically in the clinical and forensic area. The negative consequences of these stigmas are great, the more because this negative attitude is associated with issues such as the incorrect use of tests by psychologists along with the illegal access to their content, which undermines the credibility. The aim of the paper is to discuss the development of a Primer book, designed for the layperson public in order to demystify the use of psychological tests, and the development of a Handbook of Guidelines about psychological testing to be used by graduated psychologists. A literature review combined with the data of an Experience Report study made up the basis for these books. The educational and professional textual productions were made in the Research Group of Psychological Assessment and Health Psychology of University of the State of Piauí, Brazil. Respondents were clients recruited in human resource departments, traffic clinics and job-preparatory courses. They got a semi-structured interview. These data were used for the development of the Primer Book. As a result of these interviews 10 myths and truths of psychological tests in Brazil were noticed regarding its application, use, definition, and psychometric qualities as well as its social and legal consequences. With respect to the Handbook of Guidelines about psychological testing, we addressed ethical issues, the educational training for psychologists and the use of the psychological tests in the perspective of Brazilian reality.
Poster 04-09
Can tests define human knowledge?!
Ahmad Shahidov (Parliament of Republic of Azerbaijan, Azerbaijan)
Specialists who search on the human psychology for long years try to use from different methods for defining knowledges and skills of individuals. Test exams were tested as written, oral, face to face, interview from time to time. But these methods which were not perfect aroused. At last, mankind dwelled on test method at the end of the XX century. Although majority of people approve a test method, but persons who emphasize negative sides of this test method are found. So the young who dont have enough information firstly about deeplly of test method, but the who have deep knowledge in their field excite serious stress during test exam, cannot fill correct corresponding columns because of excitement and according result, they demonstrate very low indicators. Another moment, it is possible to meet with the students who copy of from correct answers, the young who use help of near comrade abusing from inattention of the observers during test exams, these facts show that test method is not always effective. The test method is considered wide spread exam method in my country – in Azerbaijan. Checking of entrant for admitting to universities, checking of personnel for getting fixed up in a job of state organs with test method is applied in large form in our country. For now although any serious seditions dont happen, insufficiencies of test method crop up from time to time that, I would like to share those seditions with conference participants in my speech.
Poster 04-10
Assessment of quality of life in individuals with chronic low back pain
Sergio Fernando Zavarize (Pontifícia Universidade Catíloca de Campinas, Brazil)
Solange Muglia Wechsler (PUC-Campinas, Brazil)
Low back pain is considered an important public health problem affecting 70% of the adult population in Brazil. When chronic, it may compromise the quality of life at various levels. This study evaluated the quality of life in patients with chronic low back pain. The sample consisted of 158 participants, 105 women and 53 men aged 30 to 88 years old, diagnosed with lumbar osteoarthritis, enrolled in five clinics in the state of São Paulo, Brazil. Instruments used were the WHOQOL-BREF of Quality of Life, the Visual Analogue Scale (VAS) for self-assessment of the perception of pain and a supplemental questionnaire which measures coping strategies. Results showed that there were no significant differences in quality of life between sexes, although women have sought treatment in greater numbers, displayed higher levels of pain and showed a higher frequency of coping strategies as socialization, reading, handicrafts and sports. By Multivariate Analysis of Variance (MANOVA) it was concluded that the quality of life was significantly influenced by age in the physical domain and by pain in all areas at the people with chronic back pain. These results contribute to a better understanding of low back pain, favoring their evaluation and treatment.
Poster 04-11
Assessing the dimensionality of the Revised Life Orientation Test in a Brazilian sample
Micheline R.Bastianello (Federal University of Rio Grande do Sul, Brazil)
Claudio S. Hutz (Federal University of Rio Grande do Sul, Brazil)
Optimism is defined as generalized outcome expectancies and it plays an important role in maintaining goal-directed behavior. It was assessed with the Revised Life Orientation Test (LOT-R), a six-item self-report measure (plus four filler items). The number of underlying factors of the LOT-R is controversial and different solutionswere proposed in diverse cultures. The aim of this study was to evaluate the LOT-R structure in a Brazilian undergraduate sample (N = 519). Initially, a one-factor solution with the six items loading on optimism was tested with CFA (Maximum Likelihood), and next a two-factor solution with three items loading on optimism and three items loading on pessimism was tested with the same method. The two-factor solution presented a high correlation (.84) between the factors, suggesting that they are not orthogonal for the sample studied. Nevertheless, the one- and two-factor solutions presented a poor fit. Therefore, a new one-factor solution, suppressing one item that presented similar semantic content to the other items, was tested and presented a good fit. In conclusion, no evidence was found to support the assumption that optimism and pessimism are independent constructs, and the one-factor solution with five items presented the best model to interpret Brazilian students’ scores.
Poster 04-12
Testing for measurement invariance in the Satisfaction With Life Scale: A comparison between USA and Brazil
Cristian Zanon (Federal University of Rio Grande do Sul, Brazil)
Claudio S. Hutz (Federal University of Rio Grande do Sul, Brazil)
The Satisfaction With Life Scale (SWLS) is often used to establish cross-cultural comparisons. However, the comparability of the SWLS scores across groups has been more assumed than empirically demonstrated. Scalar invariance is required to conduct mean comparisons. The present investigation examined the measurement invariance of SWLS scores across U.S. and Brazilian student samples. Multigroup confirmatory factor analyses were performed. Three main questions guided model evaluations: Is the factor structure of the SWLS similar in both samples (configural invariance)? Are the items perceived and interpreted in the same way across the groups (metric invariance)? Are the item intercepts invariant in both samples (scalar invariance)? A configural model was set at the baseline of comparisons to the other nested and increasingly restrictive models. The configural model presented a reasonable fit. The metric model presented an acceptable, but slightly worse fit. However, the chi-square difference test between the models was significant and imply that metric invariance is not supported. Inspection of the data showed that item 4 was noninvariant. This finding suggests that Americans might think that they are more succesfull in life than Brazilians. In any case, the results recomend care when making comparisons between U.S. and Brazil using the SWLS.
Poster 04-13
Quality of life of adults with intellectual disability: Individual, proxy and caregiver inter-rater agreement
Alessandra Coscarelli (University of Torino, Italy / University of Valle d'Aosta, Italy)
Giulia Balboni (University of Pisa, Italy)
The current approach for the evaluation of Quality of Life of adults with Intellectual Disability is to interview the individual himself/herself for self-report assessment and his/her caregiver for direct observation. Nevertheless, if an individual cannot be interviewed on the self-report assessment, a proxy is usually asked to respond on the individual’s behalf. To provide evidence supporting the reliability and validity of proxy responses, inter-rater agreement between individuals with ID, proxies and caregivers was investigated using the Personal Outcomes Scale. The POS_Self Report was filled out independently by individuals with ID and by his/her proxy for 47 verbal adults with ID, and by two independent proxies for 44 non-verbal adults with ID. In both cases, the POS_Direct Observation was completed by his/her caregiver. Responses of individuals and proxies on Self Report were found to be different only for the emotional and physical well-being domains. Moreover, the individual/proxy agreement was found to be higher than the corresponding individual/caregiver agreement on Self Report/Direct Observation. Finally, no differences were found between two independent proxies on Self-Report. Present findings demonstrate that reports provided by proxies are reliable and closer to those of the individuals with ID than those of the caregiver.
Poster 04-14
Personality, habits, and the quality of life
Katharina Lochner (cut-e group, Germany)
Achim Preuss (cut-e group, Germany)
Maike Wehrmaker (cut-e group, Germany)
During the growth of the Positive Psychology movement, factors that contribute to subjective well-being and happiness have been subject to extensive research. Such factors are positive emotions (e.g., Fredrickson, 1998, 2001, 2009) and social relationships (e.g., Diener & Seligman, 2002). In a study with N = 3,131 participants from various educational backgrounds (from primary education to Ph.D. degree) and different ages (from 14 to 81), we replicated these findings: The positive emotions joy and balance as well as participants’ self-rated happiness with their social contacts contributed significantly to their ratings of their own life satisfaction. Furthermore, sleep and recreation were significantly related to happiness, as well as happiness with the own financial situation. Finally, exercising regularly was related to overall life satisfaction. The study is part of an online programme intending to help its participants improve their well-being and mental fitness. Implications of the study results for the programme are discussed.
Poster 04-15
Precondition guidelines and Test translation/adaption guidelines. A practical guide.
Paola Bully (University of the Basque Country, Spain)
Paula Elosua (University of the Basque Country, Spain)
Josu Mujika (University of the Basque Country, Spain)
Leandro Almeida (University of the Basque Country, Spain)
In the framework of the second edition of the Guidelines for test adaptation elaborated by the International Test Commission, the aim of this work was to propose a practical guide for the two first categories: Precondition and Test translation/adaptation. Using as example the Spanish adaptation of the Bateria de Provas de Raciocínio (BPR; Reasoning Test Battery) the paper shows the processes were carried out in order to get: a) legal permissions, b) approval from the Ethics Committee, c) cultural and linguistic equivalence, d) equivalence between stimulus materials and administration procedures, e) equivalence between speediness and item difficulty. The analyses were made in such a way that culture- and language-related problems that can affect the validity of the inferences drawn from the scores were minimized. We try to show the problems involved in those first steps and some procedures to deal with them. Although these issues are relevant in any adaption process, they are not usually reported in the technical manuals of tests.
……………………………………………………………………………………………………………………………………………………....
Poster Session 5 July 4, 10.45-12.45 Room: Rotonde
Poster 05-01
A comparison of methods for univariate and multivariate time series analyis
Tanja Krone (University of Groningen, The Netherlands)
C.J. Albers (University of Groningen, The Netherlands)
Marieke E. Timmerman (University of Groningen, The Netherlands)
Time series analysis gives the possibility to analyse data over an extended time and forecast following scores. However, little is known about the different possible methods of time series analyses. At the moment, the most used method is ARIMA, which needs a static time series with a stable mean and variance. This paper shows two univariate methods, ARIMA and the Baysian DLM, and three multivariate methods, VARIMA and the Bayesian MVDLM and LMDM, used with two continuous variables. The methods are demonstrated with a dataset of 12 Parkinson patients who filled in the PANAS questionnaire every day for 10 weeks. The twenty questions were combined into two factors measuring extraversion and emotional instability (Krone et al.). The ARIMA and VARIMA models fit less well than the DLM and MVDLM. The LMDM did not fit the data well, which is explained by the fact that the data was orthogonal. For static data, (V)ARIMA can be used, for more dynamic datasets the (MV)DLM is very useful, and for multivariate dataset with dependent or correlating variables, the LMDM can be used.
Poster 05-02
The use of Automated Test Assembly in continuous improvement of fixed length tests
Thomas Proctor (The College Board, USA)
Rosemary Reshetar (The College Board, USA)
Gerald Melican (The College Board, USA)
Testing programs use a variety of methods to measure performance – forced choice, free-response, and simulations as examples. Each method has strengths and weaknesses and the most effective combination to measure the appropriate knowledge and skills reliably needs to be evaluated regularly. When the mix of item types needs to change, the change must be weighed against other competing psychometric qualities such as the ability to equate future forms. Automatic test assembly allows the possibility of assembling test forms that reduce the number of forced-choice items, considered less authentic, and increasing the number of other item types while maintaining same testing time, and psychometric qualities. The purpose of this study was to investigate the feasibility of altering the mix of item types while still meeting content and psychometric properties. Several reasonable, from a logical and “face validity” perspective, combinations of item types were posited. Each of these combinations including the current test specifications then was used to assemble multiple forms of the examination. Simulations were performed to generate results to compare to current performance. Criteria include meeting content specifications, and psychometric requirements. The test under evaluation is used to determine five performance levels so misclassification is a major criterion.
Poster 05-03
A comparison of two approaches of dimensionality assessment in Rasch measurement
Rudolf Debelak (Schuhfried GmbH, Austria)
This study compares two approaches of dimensionality assessment which have been suggested in the field of Rasch measurement by the means of a simulation study. The first approach is based on a principal component analysis and parallel analysis of smoothed tetrachoric correlation matrices. The second approach is based on a cluster analytical approach that uses item fit statistics for the Rasch model as a distance measure. To evaluate both approaches, they are applied to different types of simulated datasets. In each dataset, two item sets, each fitting a unidimensional Rasch model, were combined with each other. The simulations differed in the size of the simulated item sets, the size of the person samples, the distribution of the difficulty and discrimination parameters and the correlation between the person parameters. The results of this study suggest that under the conditions of our simulations, the first approach leads to a correct assessment of dimensionality more often than the second approach. Both approaches are further demonstrated by applying them to the data of 222 test takers who worked on the intelligence test battery IBF.
Poster 05-04
Exploring the relationship between item difficulty distribution and passing rate
Brad Ching-Chao Wu (Pearson VUE, USA)
The importance of statistical parameters such as item difficulty and test information on assembling parallel forms and stabilizing passing rate in a criterion-reference test has been widely discussed and emphasized in modern testing practice. However, the knowledge of item difficulty and test information do not always guarantee parallel forms and consistent passing rate. Difficulty distribution of the item bank could place significant constraint on item selection and form parallelism. This study presents the challenges in creating parallel forms when 1) banked item difficulty distribution does not match target, and 2) banked item difficulty distributions across content domains vary significantly. The first part of the study examines the relationship between difficulty distribution and passing rate using simulated data with difficulty distribution held constant across content domains. The second part of the study examines the effect of varied difficulty distributions across content domains on passing rate with real data from a certified test. The study asks to what degree unparallel forms can affect passing rate and what level of discrepancy in difficulty distribution can test developers tolerate in order to achieve stable passing rate, which are both important because exact match of difficulty distribution and test information across forms is usually unattainable.
Poster 05-05
Current best practices in item selection
Tzu-Yun Chin (Buros Center for Testing, USA)
Anja Römhild (Buros Center for Testing, USA)
Yongwei Yang (Gallup Inc., USA)
This study will provide an up-to-date survey and critique of the practices regarding psychometrics-based item selection during test development. First, we will review and synthesize the literature on statistical indicators and criteria used to guide item selection for educational and licensure tests. Specifically, we will present the types of statistics a test developer should consider along with their strengths and weaknesses as well as the cut-off values recommended by the research community (if available). This literature synthesis will be couched in terms of test score usage and interpretations (e.g., criterion-referenced tests and norm-referenced tests) and test administrations (e.g., fixed-form tests and computerized adaptive tests). The second component of this presentation will focus on the empirical usage of item selection statistics in current practice. We will survey prominent international testing programs and compare their item selection criteria against best practice synthesized from the literature. The results of this study will provide clearer guidance and benefit international test practitioners and test developers.
Poster 05-06
Psychometric analysis of a 4-point unequal scale
William Fernando Durán (TEA Ediciones, Spain)
Pablo Santamaría (TEA Ediciones, Spain) (presenting author)
The Personality Assessment Inventory (Morey, 1991, 2007) uses items with a 4-point unequal scale (False, Slightly True, Mainly True and Very True). This scale does not assume equal distance between each point. The first point indicates disagreement, the second mild agreement, the third moderate agreement and the fourth full agreement. Currently, there are no studies that support any methodological advantage of this scale. The purpose of this paper is to analyze the effect of this four unequal scale against 4-point equal scale in the criterion validity of the test. The sample considered for this research was composed of 4715 participants, 3655 non-clinical and 1650 psychiatric outpatients, who were gathered from major regions of Spain. IRT and classical models are used as well as simulation data. The results confirm the psychometric properties of 4-point unequal scale and improved in correct classification. The main conclusions of this study will be discussed in the poster as well as the simulations data.
Poster 05-07
An alternative to Cronbach's alpha coefficient: The ordinal index of reliability
Josu Mujika (University of the Basque Country, Spain)
Paula Elosua (University of the Basque Country, Spain)
Bruno Zumbo (University of British Columbia, USA)
Cronbach’s alpha reliability coefficient is the most widely used in the social sciences to estimate internal consistency. Among the underlying assumptions is the continuous nature of the variables. However, the items rarely satisfy this condition. This paper presents an alternative to Cronbach's alpha coefficient: the ordinal index of reliability. It was first introduced by Zumbo, Gadermann, and Zeisser (2007) to quantify measurement reliability for scales and measures comprised of rating scale (i.e., Likert-type or ordinal) item response formats. We will (a) describe the statistical and conceptual assumptions of the ordinal coefficient alpha, (b) compare how ordinal alpha performs relative to the conventional coefficient alpha with both simulated and real data, and (c) provide examples from real data of how to compute the ordinal coefficient.
Poster 05-08
Coping with Heterogeneity thru Heterogeneity
Hadassah Weitzman Bennett (David Yellin College, Israel)
Rachel Haller (University of Amsterdam, The Netherlands / David Yellin College, Israel)
The objective of this study was to assess the value of an indigenous multicultural professional research team in adapting a test battery for use with the multi-culturally complex population in the city of Jerusalem, Israel. The adaptation process utilized a mixed methods approach. In addition to traditional psychometric methods of item building, qualitative sessions including field feedback were added, providing in-depth considerations of cultural sensitivities such as relevance, friendliness and respectfulness. The large pool of resulting relevant items was administered to a large number of subjects in the various sectors. Process and resultant responses were qualitatively analyzed. Findings included the Identification of a large number of items that were culturally appropriate and did not discriminate between the groups. However some conflicts were noted between what is “friendly” and “relevant” to the different sectors. In addition some unintentional ignoring of the majority hegemonic group was noted. In conclusion we propose a model emphasizing the utility and value of a multicultural team in all stages of test adaptation. Sensitivity to balance the needs of the majority along with the needs of other groups and continual validation of culturally sensitive methods for scoring and interpreting results are urged.
Poster 05-09
Applying principles of universal design to the development of psychological tests in Brazil
Carlos Henrique Sancineto da Silva Nunes (Federal University of Santa Catarina, Brazil)
Cassandra Melo Oliveira (Federal University of Santa Catarina, Brazil)
Universal design is the process of creating devices, environments, systems, and processes which are usable by people with the widest possible range of characteristics. The application of principles of Universal Design to the development of psychological tests is one way to make them accessible for groups of people with different characteristics and needs. The objectives of this research are: enumerate the applications of Universal Design principles in psychological tests; present the Universal Design applied to psychological tests in Brazil; and summarize the advances proposed by the work in the field of psychological assessment that aim full accessibility. This research consists in a systematic literature review regarding papers published in national and international electronic databases. Results indicate that there are no tests authorized for professional use by the Brazilian Federal Council of Psychology that adopt the concept of universal design with appropriate studies of reliability and validity. In Brazil, even the attempts to adapt tests to make them suitable for people with disabilities are reduced, unlike the international situation, in which many studies use new technologies as a way to make testing more accessible. The conclusions indicate the need for improvement of this area in Psychological Assessment, especially in Brazil.
Poster 05-10
Construction and validity studies of a Computerized Adaptive system for personality assessment in Brazil
Carlos Henrique Sancineto da Silva Nunes (Federal University of Santa Catarina, Brazil)
Maiana Farias Oliveira Nunes (Federal University of Rio Grande do Sul, Brazil)
Cássia Roettgers (Federal University of Santa Catarina, Brazil)
Luciane Guisso (Federal University of Santa Catarina, Brazil)
Leonardo Pereira Lima (Federal University of Santa Catarina, Brazil)
Research on computerized adaptive testing in different contexts has received attention from many researchers who work with Psychological Assessment. This study aimed at developing a Brazilian Computerized Adaptive Test for personality assessment of the Five-factor model (FFM). This study started with the selection of anchor-items from a paper and pencil personality test with validity studies in Brazil. New items were created and pretested together with anchor-items so they could be equalized. Three forms of pretests were created, with 140 items each (the same 40 anchor-items were used in all test forms). The samples used for pretest studies were approximately 400 people from different Brazilian regions, for each test form. Psychometric analyses for the new and anchor-items involved Factor Analysis, estimation of psychometric parameters using Rasch model, Principal Component Analysis of Misfit, and DIF related to gender and educational level. At the moment, we have an item pool composed by 635 items that assess personality in the FFM with good psychometric properties. The assessment software for adaptive testing is already functional but it is under revision to accommodate some key concepts of Universal Design.
Poster 05-11
Trust behavior in a computerized investment game
Pei-Chun Shih (Universidad Autonoma de Madrid, Spain)
Alejandra Montoro (Universidad Autonoma de Madrid, Spain)
Agustin Martinez (Universidad Autonoma de Madrid, Spain)
Eduardo Estrada (Universidad Autonoma de Madrid, Spain)
The aim of this study is to develop and validate a computerized version of the Investment Game (Berg et al., 1995), the most common behavioral measure of trust, in a web setting. Three different performance conditions were carried out to explore the psychometric properties of the measure and to compare the effect that positive and negative reciprocity has on trust behavior. The sample comprises 112 participants (82 females, 73.2%), whose age ranged from 17 to 52 years (Mean = 20.02, SD = 5.516). Internal consistency was good (alpha = .85) and the instrument showed convergent validity with other trust measures (r = .24, p < .05 with the Yamagishi General Trust Scale and r = .331, p < .01 with NEO-PI-FFI Agreeableness). A significant difference between positive (mean = 51.59, SD = 21.36) and negative (mean = 37.44, SD = 17.24) reciprocity conditions was found (F = 15.06, p< .001; partial eta-squared = .12). Results show that this instrument has good reliability and validity. Moreover, it allows conducting experimental research on trust behavior with a large number of subjects while overcoming time and space restrictions.
Poster 05-12
INSIGHT: Development of a new computer adaptive test measuring abstract reasoning
Matteo Ciancaleoni (Giunti O.S. - Organizzazioni Speciali, Italy)
Luisa Fossati (Giunti O.S. - Organizzazioni Speciali, Italy)
Abstract reasoning is well-known as a significant predictor of working performance in a wide range of jobs. Traditional instruments require too much administration time for personnel selection. To reduce that time a Computer Adaptive Test (CAT), called INSIGHT, was developed applying Item Response Theory (IRT). Two different booklets composed by 75 items, including 15 anchor items, were administered to two samples of Italian fresh-graduated and manager (n1 = 278; n2 = 224). Because of IRT local independency assumption was not met, the 75 items were collapsed into 25 testlets composed by 3 items each. Explorative Factor Analyses indicated that IRT unidimensionality assumption was met. Testlets showed acceptable fit indices under the Graded Response Model (GRM). To link GRM testlet parameters between the two booklets, anchor testlets were used. Testlets covered a wide range of difficulty and had medium to high discrimination power: therefore, information functions were adequate. Testlets exposure was controlled using the a-stratified (Chang & Ying, 1999) and “randomesque” (Kingsbury & Zara, 1989) strategies. Stopping rule concerning standard error lower than .50 was met administering 4 to 8 testlets. IRT analyses show that testlets have good psychometric properties. Compared to traditional tests, INSIGHT measures abstract reasoning in a shorter time frame with higher precision.
Poster 05-13
Influences of item selection methods on item exposure control in multidimensional computerized adaptive testing
Po-Hsi Chen (National Taiwan Normal University, Taiwan)
Jar-Wen Kuo (National Taiwan Normal University, Taiwan)
Yao-Ting Sung (National Taiwan Normal University, Taiwan)
Kuo-En Chang (National Taiwan Normal University, Taiwan)
Hsien-Sheng Hsiao (National Taiwan Normal University, Taiwan)
Hak-Ping Tam (National Taiwan Normal University, Taiwan)
Tzu-Chien Liu (National Central University, Taiwan)
Precision of ability estimation and item exposure control are always in a dilemma in computerized adaptive testing. Methods developed to reduce the item exposure rate would also decrease the precision of ability estimation more or less in unidimensional computerized adaptive testing. In this research, the two item selection methods were used in multidimensional computerized adaptive testing (MCAT) in order to know the effect of item exposure control and ability estimation. The Fisher’s maximum information method and progressive method (Revuelta & Ponsoda, 1998) were used and compared with their effect on MCAT. Two independent variables, dimensionalities (two, six and ten) and correlation between dimensions (high or low), were manipulated in this research. Root mean square of error (RMSE) and item exposure rate were computed to compare the effect of item selection methods. Results indicated that the RMSE of MCAT was slightly reduced in the progressive method, independent of the dimensionalities and correlation between dimensions. However, the item exposure rates were controlled in a reasonable range when using the progressive method. Suggestions and applications of using these two item selection methods were discussed in this research.
Poster 05-14
Equivalence between paper and pencil and computer based tests to assess personality traits
David Arribas-Aguila (TEA Ediciones, Spain)
The objective of this study was to investigate the equivalence of paper-and-pencil (PPT) and computer based tests (CBT) in two widely used tools for personality assessment. There are contradictory results regarding PPT and CBT equivalence. This lack of homogeneity could be explained by differences in computer systems for administering tests used by researchers and perceptive or motor demands of tasks. Assessment purpose or speed level of the test could be important to understand this heterogeneity as well. Using a common system of Internet based administration (e-), equivalences between PPT and CBT forms of compeTEA (Arribas and Pereña, 2009) and TPT (Corral and col., 2002) are analyzed. Samples were composed of more than 25,000 participants per test. Differences between raw scores on both conditions were significant, although small effect sizes were observed As ITC Guidelines on CBT points out, before the implementation of a PPT in a CBT format, psychometric characteristics under both administration conditions should be studied, as well as the effect size of the difference in order to develop specific norms.
Poster 05-15
Real time item collusion detection in Computerized Adaptive Testing
Ji Eun Lee (University of Minnesota, USA)
Shungwon Ro (Kenexa, USA)
A recent trend in the current testing industries (educational assessment, certification/licensure testing and employment testing) is a web-based test delivery. It is a very efficient and effective way to maximize a testing program’s capacity in test production and delivery, breaking the barrier of limited accessibility. However, no testing program (even with an appropriate proctoring system in place) is free from the threats of test security such as item collusion or breach. Item collusion studies are done, employing item response theory (IRT)-based person fit statistics, for delivery modes such as linear computer-based testing and computerized adaptive testing (CAT). However, they are post-hoc analyses and thus not practical in coping with collusion. This study aims to develop a real time solution in item collusion detection under CAT. Through realistic simulations of CAT administrations given the predefined testing conditions such as item location and percentage of breached items, etc., groups of people (and testing sites) committing item collusion can be detected real time. The results show the utility of person fit statistics in immediate item collusion detection in CAT. This study provides a tool to overcome the utmost threat of test security.
Poster 05-16
Supervised online testing: A way to reduce faking?
Matthias Stadler (Psytech International, Germany)
This study examined for the first time the effect of supervised online testing on faking in psychological testing under high stakes conditions. In two independent studies, both Chinese and European job applicants completed the Psytech 15FQ+ Personality Questionnaire within either a supervised (N>1000) or an unsupervised condition (N>14000). As hypothesized, applicants showed significantly higher scores on the 15FQ+ Big-Five Emotional Stability and Conscientiousness scales in the unsupervised condition, reporting they were more emotionally stable and more conscientious than those responding in the supervised condition. No significant differences were found between the applicant groups on the social desirability scale of the 15FQ+; however, the unsupervised applicants scored higher on the “Fake Good” scale derived from specific item responses within the test as a whole. The same pattern of results was found in both the Chinese and European sample data. This is interpreted as evidence for elevated impression management within the unsupervised condition. Examinations of effect sizes showed mean differences equal in magnitude to effect sizes found in other faking studies on unsupervised internet testing. The results may be an indication that supervised online testing is an effective administration procedure which might help reduce faking on job relevant scales of personality questionnaires.
Poster 05-17
Computer-assisted testing of stability vs. susceptibility to failure stress
Iryna Arshava (Oles Honchar Dnipropetrovsk National University, Russia)
Eleonora Nosenko (Oles Honchar Dnipropetrovsk National University, Russia)
The study was aimed at designing a computer-assisted information-processing technique, simulating the exchange of information between a virtual pilot and a flight controller, which was patented by the authors as a means of testing the individual's stability vs. susceptibility to failure stress. By exposing the subjects to conveying visually displayed information to a virtual interlocutor under the conditions, when its scope exceeds at one of the stages of the simulated activity human short-term memory capacity, an unavoidable failure is elicited. The results of assessing the prognostic validity of the technique on a sample of 110 subjects showed that those of them, who demonstrated statistically lower efficacy of information processing after failure than prior to it, appeared to be diagnosed as susceptible to stress by other methods. Their susceptibility to stress was most vividly manifested by the inability to simultaneously process two interrelated activities: the identification of the perceived objects and reporting the “call names” assigned to them and their virtual interlocutor. The results allow to conclude that the designed technique can be used for predicting emotional stability vs. susceptibility to stress and can find practical application in the area of personnel selection.
Poster 05-18
Openness in assessment for the stakeholders of a Computerized Adaptive Test
Eduardo Doval (Universitat Autònoma de Barcelona, Spain)
Carme Viladrich (Universitat Autònoma de Barcelona, Spain)
Rebeca Garcia-Rueda (Universitat Autònoma de Barcelona, Spain)
Laura Riera (Universitat Autònoma de Barcelona, Spain)
Mick Sumbling (Universitat Autònoma de Barcelona, Spain)
Pablo Sanz (Universitat Autònoma de Barcelona, Spain)
Since its foundation in 1990 the language service of the Universitat Autònoma de Barcelona (UAB) has been responsible for the teaching and certification of English as a foreign language (EFL). In 2001, as a part of a plan to align its courses with the Common European Framework of Reference (CEFR), changes were introduced in the evaluation criteria for productive language skills, while computer-based and adaptive (CAT) tests were also first used to test receptive skills. This new context meant a dislocation between the teaching and evaluation processes, with consequences for all stakeholders: language testers, teachers and test-takers. A decade’s evidence regarding the design, implementation, development and monitoring of the evaluation system is presented. The optimization achieved in item development, psychometric quality, the empirical validity of results and saving in time is underlined. Positive empirical evidence is contrasted with the subjective sensation of a lack of face validity in the tests and their results, as expressed by some test-takers and teachers, largely due to changes in the testing tradition. With the goal of improving this situation, a network of communication has been developed by which all stakeholders can become more fully informed and aware of the quality of evaluation in question.
Poster 05-19
From paper to digital testing
Andress P. Kooij (Pearson Assessment and Information B.V., The Netherlands)
Karinke Jonkman (Pearson Assessment and Information B.V., The Netherlands)
The testing world is rapidly changing from paper to digital assessment. Pearson developed a new assessment system that makes widely used assessment instruments (e.g. WAIS-IV, CELF-4) available for testing on a tablet. The system is also intended to make new ways of assessment possible. The system is intended to improve user decision making, treatment planning and interventions per individual client. It helps experts to increase accuracy, productivity and provide increased time efficiency in their assessments. Furthermore it helps the examiner to focus more on the examinee by simplifying and streamlining the test administration procedures. The tablet system provides the possibility to use electronic record keeping and simplify the analyses of the instruments taken. In this poster presentation we give a brief overview of the system, its purposes and benefits for the professional. We give insight in the way the transfer from paper to digital assessment was made and the obstacles that had to be overcome. It gives the benefits of digital testing above paper-testing. We conclude the poster with a demonstration of the system, its features and future opportunities such as new ways of (digital) assessment.
Poster 05-20
Interaction of item bank, scoring method and termination criteria in Computerized Adaptive Testing
Lisabet M. Hortensius (University of Minnesota, USA)
David J. Weiss (University of Minnesota, USA)
Computerized Adaptive Testing (CAT) provides good measurement precision with fewer items than conventional tests, by only administering items that are relevant to the examinee in difficulty. When designing a CAT, decisions need to be made on test termination and θ (score) estimation. This study compares different termination methods (minimum standard error (SE), a combination of SE and minimum information in remaining items, convergence of the θ estimate, and four fixed-length conditions (average number of items for each of the three other methods, and full bank) and different θ estimation methods (maximum likelihood estimation, weighted likelihood estimation, expected a priori and maximum a priori). To date, no studies have addressed this combination of conditions. Each combination is compared across four hypothetical item pools: a large and a small dichotomous and polytomous item bank. For each item bank response vectors for 10,000 simulees will be generated. Keeping the response vectors constant, post-hoc analysis will be run using CATsim, varying only θ estimation methods and termination methods. The tests are compared on test length, correlation of θ estimate and true θ, bias, root mean square error, and standard error. An interaction between termination criterion, item bank and θ estimation method is expected.
Poster 05-21
Comparison of empirical methods in the study of discriminant validity
Paola Bully (University of the Basque Country, Spain)
Paula Elosua (University of the Basque Country, Spain)
Clinical and diagnostic categorizations derived from test scores have important practical consequences. The aim of this work was to assess the efficiency and type I error of several classification procedures. Using Monte Carlo simulation we compare traditional (M±2SD, Percentiles), statistical (logistic regression, quadratic discriminant analysis and latent class analysis) and artificial intelligence (neural networks and support vector machines) methods. Different factors were manipulated: Sample size (200, 800 and 2000), group size ratio (50:50, 75:25), mean differences effect size (0.2, 0.5 and 0.8), and normal and non-normal data distributions. All the manipulated variables were fully crossed for a total of 36 simulation conditions. For each combination of conditions, 100 simulations were run in order to average out the possible bias in any single sample run. The results of the study can aid in the design of classification systems in which several classification methods can be employed to increase the reliability and consistency of the classification in psychology. The implications of the results for discriminant validity are discussed.
……………………………………………………………………………………………………………………………………………………....
Poster Session 6 July 4, 13.30-15.30 Room: Rotonde
Poster 06-01
Latent structure of Posttraumatic Stress Disorder symptoms among Chinese adolescents after an earthquake
Richu Wang (Chinese Academy of Sciences, China)
Li Wang (Chinese Academy of Sciences, China)
Zhanbiao Shi (Chinese Academy of Sciences, China)
Jianxin Zhang (Chinese Academy of Sciences, China)
Posttraumatic stress disorder (PTSD) is defined by the Diagnosis and Statistical Manual of Mental Disorder - fourth edition (DSM-IV) as an anxiety disorder characterized by re-experiencing, avoidance, and hyperarousal symptoms. However, the DSM-IV’s three-factor model of PTSD symptoms was vastly disproved by empirical researches. In the past decade, two four-factor models got most empirical supports, namely King’s numbing model (King et al., 1998) and Simms’s dysphoria model (Simms et al., 2002). More recently, a five-factor model comprised of intrusion, avoidance, numbing, dysphoric arousal, and anxious arousal, was proposed by Elhkai et al. (2011). Our previous studies among Chinese samples supported this model. In the current study, we investigated the latent structure of PTSD symptom among an adolescent sample who had recently experienced an earthquake in South China. Using confirmatory factor analysis, the five-factor model showed the best-fit among alternative models. We further examined the depression and anxiety levels, and tested their correlation with PTSD clusters of symptoms. The result showed symptoms of intrusion, avoidance and anxious arousal correlated with anxiety more than depression, numbing correlated with depression more than anxiety, whereas dysphoric arousal correlated equally with anxiety and depression. These findings are pertinent to the upcoming DSM-5.
Poster 06-02
Psychometric properties of the Millontm Clinical Multiaxial Inventory – III personality disorders scales in Russian Language
Jelena Kolesnikova (University of Latvia, Latvia)
The aim of the study is to determine the psychometric properties of the Millontm Clinical Multiaxial Inventory – III (MCMI-III, Millon, Millon, Davis, & Grossman, 2006) for personality disorders scales in Russian language. The MCMI-III includes 14 scales of personality disorders and 10 scales of clinical syndromes with 175 true-false items. In the present research only the scales of personality disorders was adapted. The sample consists of 111 drug addicts patients from psychiatric centers, aged from 21 till 35 (female 22% and male 78%). Results of the study indicated that many the items of the personality disorders scales show good reaction and discrimination indexes, but the average 2.42 items with low discrimination index were excluded from each scales. The internal consistencies (Chronbach`s alphas) for the scales of personality disorders respectively were of .60 to .80: Schizoid, Narcissistic and Obsessive-compulsive personality disorders scales were .68, 0.60 and .66, but Avoidant, Depressive, Dependent, Histrionic, Antisocial, Sadistic (aggressive), Masochistic (self-defeating), Negativistic (passive-aggressive), Schizotypal, Borderline paranoid personality disorders scales were above .70. The test-retest reliability coefficient of each scales were above .70 and only masochistic (self-defeating) personality disorders scale was .43. In further research it is needed to exclude the existed items and replace them with the new ones.
Poster 06-03
A comparison of persons with PTSP and the normative sample with the Trauma Symptom Inventory
Krunoslav Matešic, jr. (Naklada Slap, Croatia)
The Trauma Symptom Inventory (TSI, Briere, 1995) has been translated and adapted for use in the Republic of Croatia (Briere, 2010). The TSI consists of 100 items and serves for the evaluation of acute and chronic trauma symptoms which include, but are not limited to, the consequences of rape, spousal abuse, physical assaults, serious accidents and natural disasters, permanent consequences of abuse in childhood as well as other early traumatic experiences. This paper deals with the existence of differences in TSI scales between the normative sample of males (N = 235) and war veterans (N = 51) who suffered from PTSD. The normative sample consisted of non-traumatized males aged between 18 and 54. The patient sample consisted of males aged between 38 and 61. The statistical significance of the differences was determined using a simple analysis of variance for independent samples. Statistical significance was determined in all cases, namely the values of all scales of the PTSD sample were higher in comparison to the control sample. The results show that the Croatian version of the TSI has good diagnostic validity in the assessment of persons with PTSP.
Poster 06-04
Rasch analysis of a preliminary self-report measure of psychopathy
Nelson Hauck-Filho (Universidade Federal do Rio Grande do Sul, Brazil)
Marco Antonio Pereira Teixeira (Universidade Federal do Rio Grande do Sul, Brazil)
Psychopathy is a personality condition characterized by impulsivity, delay discount, manipulation, narcissism, and lack of remorse and empathy. Despite there are several psychometric measures of psychopathy available, the majority of them were developed using Classical Test Theory analyses, which produce largely sample-dependent results. The aim of the present study was to develop a self-report measure of psychopathy with unidimensional Rasch modeling, an Item Response Theory model optimal for building sample-free, truly intervalar scales. Participants were 224 undergraduate students (mean age = 23.76; SD = 8.02; 58.9% females). The instrument was composed of 63 items, distributed in four dimensions theoretically derived from neuropsychological models of psychopathy and from a review of 34 instruments: Disinhibition, Interpersonal Exploitation, Emotional Deficits and Low Response Modulation. Analyses showed four items with local dependence or differential item functioning for sex. Dropping out those items resulted in good fit to the Rasch model, with item reliability for the four measures ranging from .95 to .99. Partial Credit Model also showed good fit for the Likert scale used (1 = Definitely no; 4 = Definitely yes). Accordingly, the preliminary version of the measure showed excellent psychometric properties. Further studies should refine items and address convergent and criterion validity issues.
Poster 06-05
Necker cube copying and cognitive dysfunction in longterm social welfare recipients
Tove K. Marberger (Center for Work and Qualification, Municipality Sagene, Norway)
Øyvind Sundseth (Oslo University Hospital, Norway)
We examined a group of long-term social assistance recipients in Oslo, Norway (N = 300). The participants had diverse problems related to alcohol/drugabuse, psychiatric and/or somatic disease, and they had various ethnical and educational background. The study goal was to identify individuals who struggled with a marked cognitive dysfunction. The initial cognitive status was assessed through a Necker cube drawing/copy. The drawing was scored as 1 (near perfect), 2 (sufficient), 3 (loss of 3D impression) or 4 (Unrecognizable). 103 clients completed WAIS-III and were assigned to four groups depending on their initial Necker score. The scaled scores of the WAIS subtests Information, Similarities, Block design and Matrices were transformed to an estimated IQ. A one-way ANOVA indicated significant group differences, F (3, 9) = 32.5, p < .002. Subsequent pairwise comparisons revealed that group 1 scored significantly better than group 2, 3 and 4. The mean difference between group 1 and 4 was estIQ = 28.6; 95% CI = 20.4, 36.3; p< .000). The differences between group 1 and group 2 and 3 were significant and large (Mean difference estIQ = 16.9 ; 95% CI = 7.2, 26.5; p < .000, and 16.1 95% CI= 4.2, 27.7; p = .002 respectively). There were no significant differences between group 2 and 3. It is concluded that problems with Neckers cube drawing was associated with lowered estIQ, indicating a need for further evaluation of cognitive function.
Poster 06-06
Development and preliminary validity of the new scales of Halstead Category Test
Jorge Costa(University of Aveiro, Portugal)
Paulo Rodrigues (Beira Interior University, Portugal)
Isabel Santos(University of Aveiro, Portugal)
Mónica Ferreira(University of Aveiro, Portugal)
Hugo Almeida (University of Aveiro, Portugal)
Nick DeFillipis(Argosy University/Georgia School of Professional Psychology, USA)
The Halstead Category Test (HCT) is a neuropsychological measure that taps into multiple domains of complex non-verbal reasoning and cognitive flexibility, but only yields a single error score, limiting the clinical use of the test. Currently, several authors have proposed additional subscales in order to increase the clinical utility of the HCT. The purpose of this study is to present the new scales and scoring criteria of HCT, as well as to analyze the construct validity of the new scales recently developed to assess specifics domains of executive function: cognitive flexibility, failure to maintain cognitive set, perseveration and memory. Using a sample of Portuguese undergraduate students aged 19-50 years, we will relate the values obtained in the new scales of the HCT with well-established neuropsychological measures (Trail making A and B, Test of Non-Verbal Intelligence, California Learning Verbal Test and Symbol Digit). After data analysis, and based in previous results, we hope to find that the new scales of HCT (perseveration, set of loss, categorization and memory), show clearly a significant associations with the neuropsychological measures used in this study. These results will be compared with international data, with the important starting point for the clinical use of the new scales.
Poster 06-07
Benton Visual Retention Test: An Item Response Theory analysis
Joice Dickel Segabinazi (Federal University of Rio Grande do Sul, Brazil)
Denise Balem Yates (Federal University of Rio Grande do Sul, Brazil)
Stela Maris de Jezus Castro (Federal University of Rio Grande do Sul, Brazil)
Jerusa Fumagalli de Salles (Federal University of Rio Grande do Sul, Brazil)
Denise Ruschel Bandeira (Federal University of Rio Grande do Sul, Brazil)
The psychometric development and validation of neuropsycological tests and tasks have been limited, for the most part, to methods based on Classical Test Theory. In this study, we conducted analyses based on Item Response Theory (IRT) to evaluate the Benton Visual Retention Test (BVRT) items for measuring visuospatial memory in students. We conducted IRT analyses using ‘R’ free software to the 10-item Administration A (Form C) for 107 male and 133 female Brazilian secondary school students. Item fit statistics adjustment to one, two and three parameters model were compared with the likelihood ratio test. The two parameters model was best suited, as expected for a neuropsychological paper-and-pencil task. The test covered a full range of the latent trait with good reliability (test information) despite most items tap average ability. Difficulty was related to the drawings’ complexity. A new ordering with increasing difficulty of the items, different from the original test, was proposed.
Poster 06-08
Identification of symptoms in childhood by CBCL
Roselaine Berenice Ferreira da Silva (Santa Cruz do Sul University, Brazil)
Marcio Luis Paveglio da Silva (Santa Cruz do Sul University, Brazil)
This study assessed frequency of internalizing and externalizing disorders by CBCL in 413 children between five to 12 years old. The objective was to determine the association with sex and age. Internalizing disorders include anxiety and depression, depression and withdrawn, somatic and thought problems. Externalizing disorders are related with impulsive and aggressive behaviors. The prevalent age was between six and eight years (73.8%), prevailing girls (55.2%) and internalizing symptoms (38.8%). Related to the measures association (chi-square), significant statistical associations were: (1) Gender and internalizing disorder, predominantly boys; (2) Age and internalizing and externalizing disorders; (3) Anxiety and depression at seven years old, depression and withdrawn at 11 years old; somatic problems at six years old; social problems at seven years old; thought problems between six and eight years old; attention problems at seven years old; impulsivity at six e aggressive behavior at seven years old. By ANOVA, higher symptoms incidence has been identified between six and 11 years old – school age. It is concluded that developmental stage is crucial to onset symptoms, indicating that psychopathology should be related with developmental view. Moreover, it is important for school, family and social environments to be able to identify symptoms associated with each child developmental stage, promoting the early intervention.
Poster 06-09
Human Figure Draw in the Rain: Evidence of validity in Brazil
Roselaine Berenice Ferreira da Silva (Universidade de Santa Cruz do Sul, Brazil)
Marcio Luís Paveglio da Silva (Universidade de Santa Cruz do Sul, Brazil)
This study investigates the validity of HFD-Rain in a 440 children’s sample. The children were 5-12 years old (69.3% in the age range 5-9), 44.5% boys and 55.5% girls. Three tests were administered individually: HFD-III (Wechsler, 2003), HFD-Rain and CBCL (Ackerman, 2001). Related to the cognitive development, HFD-III identified 68.2% of children in the medium score, similar to the normative sample. According to the information collected by CBCL, 23.8% of children showed internalizing disorders, whereas 16.4% showed externalizing disorders. A significant association between performance in HFD-Rain, age and sex of the children was found (p< 0.05). Results indicate that younger children (5 to 6 years) showed less detail in the drawings than older ones. Related to gender, boys showed less detail. Beyond this, HFD-Rain has a correction item that investigates emotional maturity. We found an association of this item with age and it was found that children with lower intelligence tend to skip this item. This illustrates cognitive and personality aspects, as indicated by the literature. This research is ongoing, being necessary to apply the instrument in a clinical sample to compare with the normative sample of this study.
Poster 06-10
The assessment of delusional-like experiences in college students
Eduardo Fonseca-Pedrero (Universidad de La Rioja, Spain)
Mercedes Paino (Universidad de Oviedo, Spain)
Marta Santarén-Rosell (Universidad de La Rioja, Spain)
Serafín Lemos Giráldez (Universidad de Oviedo, Spain)
José Muñiz (Universidad de Oviedo, Spain)
Delusions show high prevalence in the general population, and can be considered a risk marker for psychotic disorders. Although the assessment of these experiences has made considerable progress in recent years, there is still room for improvement in the measurement quality of the self-reports available for such assessment. The goal of the present work was to analyze the measurement quality of the Peters et al. Delusions Inventory-21 (PDI-21) in Spanish college students. The final sample was made up of 660 participants (29.5% men) with a mean age of 20.3 years (SD = 2.6). The results revealed that a high percentage of the sample reported some symptom of paranoia. Analysis of the internal structure of the PDI-21 by means of exploratory factor analysis based on the tetrachoric correlation matrix yielded an essentially unidimensional solution. Cronbach’s alpha for the total score was 0.91. Scores on the PDI-21 correlated statistically significant with Trait and State Anxiety and Negative Affect. These results provide new evidence of the validity of the PDI-21 and endorse its use a measurement instrument for assessing the extended psychosis phenotype in non-clinical population.
Poster 06-11
Development and psychometrics of an adult self-report on emotion regulation strategies within negative emotions
Dina Horowitz (University of Basel, Switzerland)
Alexander Grob (University of Basel, Switzerland) (presenting author)
We show the conceptual framework and the psychometric properties of a newly developed German instrument to measure emotion regulation strategies (ERS) in adults, the Questionnaire of Emotion Regulation for Adults (‘Fragebogen zur Erhebung der Emotionsregulation bei Erwachsenen’; FEEL-E). The FEEL-E has been conceptualized on the basis of the existing version for children and adolescents (FEEL-KJ; Grob & Smolenski, 2009). The adult self-report measure assesses ERS for the discrete negative emotions of fear, anger, and sadness. The instrument is able to differentiate 7 adaptive ERS (problem-oriented action, distraction, seeking social support, acceptance, cognitive problem-solving, reappraisal, forgetting) and 8 maladaptive ERS (aggressive behavior, suppression of expression, withdrawal, self-blame, resignation, rumination, catastrophizing, other-blame). We present preliminary results of the standardization studies (N = 858; age range 18-88 years, M = 44.8 years; 52% females), i.e., item characteristics, scale construction, and reliability information (consistency and stability). Furthermore we provide information on the FEEL-E’s construct (adaptive vs. maladaptive, and comparability), criterion (e.g., personality, well-being, satisfaction, depression), and clinical validity (71 patients with ICD/DSM-IVTR diagnoses). The preliminary results provide evidence for satisfactory to excellent psychometric properties of the FEEL-E.
Poster 06-12
The Hand Preference Test HAPT 4-6 – a comprehensive procedure to assess handedness and consistency of hand preference
Pia Deimann (University of Vienna, Austria)
Ursula Kastner-Koller (University of Vienna, Austria)
Johanna Bruckner (University of Vienna, Austria)
The Hand Preference Test for 4- to 6-year-olds (HAPT 4-6; Bruckner, Deimann, & Kastner-Koller, 2011) is a new observational method to assess handedness and hand preference. Compared to existing tests for handedness, its added value is an explicit and detailed assessment of consistency of hand preference at a very young age. The HAPT 4-6 is an age-appropriate test with especially appealing and motivating testing conditions by integrating the items into the context of a treasure hunt. The test consists of 14 items which include one-handed grasping of an attractive object or picking up an object, one-handed or two-handed manipulation, movements of the arm, shoulder, or body axis, and movements of the fingers and hand. Two qualities of execution, namely, precise tasks as well as rapid, automated tasks were chosen. To ensure objectivity, instructions are given on an audio-tape. Each of the 14 tasks are observed three times to assess overall hand preference and hand preference within the tasks (defined as consistency of hand preference). Test results yield a laterality quotient and a consistency-of-hand-preference score. Reliability of the HAPT 4-6 is high (Cronbach’s alpha = .95). Concurrent validity was estimated by correlating HAPT-scores with parents’ ratings of hand preference (.76) and observation of the preferred drawing hand (.75). Consistency of hand preference is less distinctive in children with drawing problems.
Poster 06-13
A systematic evaluation of ten early childhood social-emotional development measures
Rebecca Gokiert (University of Alberta, Canada)
Rebecca Georgis (University of Alberta, Canada) (presenting author)
Clara Lee (University of Alberta, Canada)
Melissa Daniels (University of Alberta, Canada)
Karen Edwards (University of Alberta, Canada)
The early detection of social-emotional (SE) problems is critical in the promotion of healthy development, however, selecting measures with appropriate psychometric and clinical properties is challenging. The purpose of this study was to systematically evaluate 10 early childhood SE screening and assessment measures. A 16-item protocol was developed to examine important properties such as validity, reliability, standardization, and usability. Each of the 16 items possessed criteria (e.g., internal consistency estimates >.80) to which each measure was evaluated, resulting in a 4-point scale score (0 = no evidence to 3 = criterion met). Findings show there is great variability in the type, quantity, and quality of reported evidence, particularly validity evidence, with large differences between screening and assessment tools. Although all measures provide psychometric evidence, when evaluated against stringent criteria, they may fall short (e.g., 90% of measures met criterion for test-retest and internal consistency but only 10% met criterion for inter-rater reliability; only 70% had adequate sample sizes for the preschool years). We will highlight implications for test users (e.g., how to weigh in on different and sometimes competing technical evidence) and recommendations for sharing knowledge with early childhood professionals about selection, and test developers on what constitutes sufficient evidence.
Poster 06-14
Mood scale: Adaptation and validation in Spanish children using CFA
V. Prado-Gasco (University of Valencia, Spain)
A. Ordóñez (University of Valencia, Spain)
I. Montoya (University of Valencia, Spain)
M.C. Mateu (University of Valencia, Spain)
R. González (University of Valencia, Spain)
The development of emotional competence is necessary to identify, understand, regulate and communicate our emotions (Garner, 2010; Saarni, 2000). When there are difficulties in this area, negative emotions may not be adequately addressed, and lead to negative moods (Rieffe et al., 2009; Rieffe et al., 2010). In the evaluation, our research team uses the “Mood Questionnaire” (MQ), self-report directed at children and adolescents of Dutch origin. This study presents a first approach to the validation of the Mood Questionnaire in the Spanish context using CFA. The sample was made up of 1,166 children from 11 schools between 8 and 14 years olds (girls 52.3%). The Mood Questionnaire (Rieffe et al., 2004, 2006), consisted of 4 mood scales: happiness, anger, sadness and fear. Four items represented each mood scale. Children were instructed to score each item on a three-point response scale. The result was a 16 item scale with adequate psychometric properties supporting the underlying construct (CFA: the whole index between .912 – .960; SRMR = .036; Cronbach's alpha = .76). Future research should increase sample scope in terms of representativeness. The Mood Questionnaire is seen as an appropriate instrument for assessing moods in the Spanish child population.
Poster 06-15
Cross-cultural validation of the Peer Aggressive and Reactive Behavior Questionnaire in the Brazilian and Italian context
Juliane Callegaro Borsa (Federal University of Rio Grande do Sul, Brazil)
Bruno Figueiredo Damásio (Bologna University, Italy)
Paola Gremigni (Bologna University, Italy)
Denise Ruschel Bandeira (Federal University of Rio Grande do Sul, Brazil)
The Peer Aggressive and Reactive Behaviors Questionnaire (PARB-Q) is a newly self-report instrument that consists of two independent scales (Peer Aggression, PA and Reactions to Peer Aggression, RPA) aiming to assess child aggressive behaviors and reactions to peer aggression, respectively. The objective of this study is to present the psychometric properties of the PARB–Q in the Brazilian context and the results of a cross-validation study of the PARB-Q in both Brazil and Italy. Participants were 587 Italian and 727 Brazilian students of elementary public schools. Firstly, exploratory and confirmatory factor analyses were performed in order to evaluate the factor structure of the scales in the Brazilian context. Consistently with the original Italian version, the results supported a one-factor solution for the first scale and a three-factor solution for the second scale. Convergent validity analyses and multivariate analysis of variance (MANOVA) by gender and age were also performed. The results indicated that the PARB-Q has satisfactory validity and reliability indexes as well as excellent fit indexes, configured as a useful tool for evaluation of aggressive behavior and the child's response to peers aggression in the Brazilian context. To evaluate the cross-cultural validity of the PARB-Q, two multigroup confirmatory factor analyses were conducted on the previously obtained factor structures of the scales for both Brazilian and Italian groups. Full configural, metric, scalar and structural invariance were achieved for both PARB-Q scales, possibiliting cross-cultural comparisons. A MANCOVA was conducted in order to evaluate differences between boys and girls and in Brazilian and Italian children regarding the scores on each scale of the PARB-Q, using age as a co-variate. Boys presented higher levels than girls in the Peer Aggression Scale (PA) and in the reactive aggression factor (RA). Girls presented higher levels than boys in the seeking teacher support factor (STS) and in the internalized reactions factor (IR). Compared to Brazilian children, Italian children presented higher levels in PA and Brazilian childrenhigher levels in IR. These results are discussed based on the literature.
Poster 06-16
Ipsative and nonipsative versions of the Picture Based Value Survey for Children
Jan Cieciuch (University of Finance and Management in Warsaw, Poland)
Justyna Harasimczuk (University of Warsaw, Poland)
Anna, K. Doering (University of Muenster, Germany)
The Picture Based Values Survey for Children developed by Doering and colleagues (2010) turned out to be a very useful measure of children’s values in the framework of Schwartz’s theory. Results obtained in many countries (Germany, Poland, Bulgaria, Italy, France, New Zealand, and the USA) revealed a circular structure of values with four higher order values and also some possibilities of differentiation between more types of values. To date however the analyses were performed with multidimensional scaling, while confirmatory factor analysis was impossible because the data were ipsative (a kind of Q-sort data). To test the structure of children’s values with confirmatory factor analysis, we developed a second version of the PBVS-C. In the original version, children place the pictorial items according to the importance they ascribe to them on five levels ranging from “very important” to “not at all important”. In the second version, children assess the importance on a 6-point rating scale as in Schwartz’s Portrait Values Questionnaire. The research was conducted in Poland with children aged 7-11 years. Results of confirmatory factor analysis and exploratory structural equation modelling will be presented and discussed.
Poster 06-17
Second version of the Picture Based Personality Survey for Children
Jan Cieciuch (University of Finance and Management in Warsaw, Poland)
Marta Mackiewicz (University of Cardinal Stefan Wyszynski in Warsaw, Poland)
The prior years have witnessed a growing interest in measuring children’s personality within the framework of the Five Factor Model by McCrae and Costa (1997). However, children as respondents can have difficulties with completing word-based questionnaires. For that reason measures for children should consider level of their cognitive development. Such a measure was recently proposed by Mackiewicz and Cieciuch (2011) – the Picture Based Personality Survey (PBPS) for Children. After analysis, the instrument was improved and the second version of PBPS-C was developed. The second version of the inventory consists of 25 behaviour-oriented items. Each scale incorporates 5 items. Every item is comprised of 2 pictures. Each of the five personality factors is described by the two polar ends of each dimension. Therefore, in the pictures, the main character is in the same situations but behaves completely different. The first behavior indicates a high intensity of a factor whereas the second one indicates a low intensity of the same factor. Children have to choose the picture which demonstrates their behaviour most adequately. Psychometric properties of the measure are satisfactory. Reliability was assessed by Cronbach’s alpha and index of quality of Saris and Gallhofer (2007). Validity was assessed in confirmatory factor analysis and multitrait-multimethod matrix with data obtained from parents.
Poster 06-18
Different applications of the Scale of Self-Perception for Preschool Children
Zlatka Cugmas (University of Maribor, Slovenia)
The author presents the Scale of Self-Perception for Preschool Children (SPSC) which she developed on the basis of her findings about children’s self-perception and on the basis of the experience of other authors developing psychological instruments of this type. She discusses about four types of SPPC implementation (verbally, with puppets, with drawings and digital). The purpose of the study was researching the psychometric characteristics of different types of the scale implementation. The scale was tested in a study which involved 214 children attending kindergarten. They were 4 to 5 years old. The poster describes the sub-scales of SPPC and presents psychometric characteristics and descriptive data of the children’s self-estimates, their motivation for solving the scale and their understanding of the items, as well as the correspondence between the self-evaluation of the children and the evaluation of the children’s self-perception conveyed by the mothers and teachers. The results are presented separately with respect to the different types of SPSC implementation. It was established that the psychometric characteristics of the scale are adequate, however, in further research the implementation with drawings or puppets is recommended. Process of digital application and its strengths and weaknesses are discussed.
Poster 06-19
Psychologycal assessment in hearing handcapped in Brazil
Eliezer Fernandes Gums (UNASP – SP, Brazil)
Tatiana Cassia Nakano (PUC Campinas, Brazil)
Solange Muglia Wechsler (PUC Campinas, Brazil)
Today has been the importance of validated instruments for deaf people in Brazil, has been emphasized by federal Council of Psychology. The purpose this study was to investigate the state of art of psychological instruments. The Brazilian electronic data basis for theses and dissertations (CAPES) and articles (Scielo, Pepsic) was analyzed from 2000 - 2010. Key words selected were Deaf, evaluation and psychological assessment. Results, eight articles were located, 6 with foreign tests and 2 with Brazilian instrument. Among the dissertation 9 utilized Brazilian measures and 3 foreign tests. 5 thesis were located, 3 with foreign tests and 2 Brazilian. In conclusion, these is a predominance of validated foreign tests, but Brazilian constructed tests are also increasing, this indication the need of national measures to assess and understand deaf handicap people.
Poster 06-20
Phonological awareness: State of art in Brazilian scientific production
Yung Sun Lee Damasceno (Pontifícia Universidade Católica de Campinas, Brazil)
Cristina Maria D´Antona Bachert (Pontifícia Universidade Católica de Campinas, Brazil)
Eliezer Fernandes Gums (Pontifícia Universidade Católica de Campinas, Brazil) (presenting author)
Solange Muglia Wechsler (Pontifícia Universidade Católica de Campinas, Brazil)
Considering the importance of phonological awareness (PA) as the fundamental factor of the reading and learning process, the aim of this research was to analyze and investigate the scientific production about PA from abstracts of theses and dissertations on phonological awareness in children, available in Coordenadoria de Aperfeiçoamento de Ensino Superior (CAPES) database, without limitation of time. Ninety-three abstracts of theses and dissertations were selected using the keywords: evaluation, children, phonological awareness. The results showed the increase of research in this area from 2005 (13%), concentrated in public institutions (74%). The empirical studies (93%) predominated, and the sample is composed of students from first grade of elementary school (50%) and final stages of early childhood education (23%). The issue that occurred with the highest incidence was the relation between PA / reading and writing (22%), PA / reading (16%). Regarding of the assessment tool, 52% of the researchs did not specify the assessment tool used and 13% applied the Phonological Awareness Test Capovilla & Capovilla (1998). We conclude that despite the increase in research related to PA and the assessment tool used have received increased attention from researchers in recent years, although it is rather scarce.
Poster 06-21
A validation study of phonological awareness assessment for Chinese kindergarteners who learn English as a second language
Susanna S. Yeung (The Hong Kong Institute of Education, Hong Kong SAR, China)
Ronnel B. King (University of Hong Kong, Hong Kong SAR, China)
The purpose of this study was to determine the psychometric characteristics of a phonological awareness assessment for Hong Kong Chinese kindergarten children who start to learn Chinese and English simultaneously at a very young age. One hundred and eighty five kindergarteners were given syllable deletion, rhyme detection and phoneme identification tasks that were extensively used in previous L2 reading research. They were also assessed on letter name knowledge and English word reading which have been shown to be closely related to phonological awareness. Both within-network and between-network approaches to establish a construct validity approach were adopted. The phonological awareness assessments have good internal consistency reliabilities and evidence is shown for its construct validity in terms of its factorial structure and correlations with letter name knowledge and English word reading. The psychometric properties of the assessment justify its application to Chinese ESL children. Implications for future reading research are discussed.
……………………………………………………………………………………………………………………………………………………....
Poster Session 7 July 4, 15.30-17.30 Room: Rotonde
Poster 07-01
Development of an instrument for the assessment of the quality of mother-child interactions
Nicole Hirschmann (University of Vienna, Austria)
Ursula Kastner-Koller (University of Vienna, Austria)
Pia Deimann (University of Vienna, Austria)
INTAKT is a new observation-based instrument, designed to help assessing the quality of mother-child interactions. So far it has mainly been used in the research context, but by now its use should be extended to practical applications as well. Especially, in the context of educational counseling a thorough assessment of the interaction-quality between a child and his/her caregiver can be helpful for the development of appropriate interventions. The implementation of INTAKT into practical use was accompanied by an evaluation of the validity of the behavioral observational categories. Mother-child dyads who had consulted a psychologist for educational counseling were compared with unscreened mother-child dyads. In counseling-dyads interaction-quality, diagnostic evaluations of the psychologist, and developmental measures of the child were related to each other. Overall, INTAKT-scales proved to be very useful in identifying problems in mother-child interactions as well as during the counseling-process. Special problems arose and therefore further consideration will be necessary concerning the usability of the instrument. Especially, economic aspects will have to be considered since coding behavioral categories proved to be quite time-consuming.
Poster 07-02
Construction of a scale to assess parent-child relations in the child custody dispute context
Vivian de Medeiros Lago (Universidade Federal do Rio Grande do Sul, Brazil)
Denise Ruschel Bandeira (Universidade Federal do Rio Grande do Sul, Brazil)
The Evaluation System of the Parental Relationship (ESPR) is a set of techniques designed to assess parent-child relations in the context of child custody dispute. These techniques comprise an interview with the parents, an assessment protocol for children (5 to 12 years old) and a scale, scored by the own examiner. The score is based on the information collected with parents and children. To construct the scale several procedures were used, such as: a systematic review about parent-child relations, divorce and child custody; an empirical study with parents, children, law officers and psychologists about parent-child relations; analysis of dimensions in existing instruments associated with parent-child relations. This resulted in 26 items spread over eight dimensions. The dimensions include parental capacities and children’s needs, such as education, leisure, affection, orientation and basic care. To demonstrate the external validity of the instrument, one study has been carried out, using the Inventory of Parental Styles with the parents, and the Family Apperception Test with the children. The results have shown appropriate correlations between those instruments and the ESPR. More validation studies of the ESPR are to be launched. This instrument aims to improve the quality of assessments carried out by forensic psychologists and social workers.
Poster 07-03
Needs and challenges for the psychologists using forensic assessment instruments in Brazil
Lucas Dannilo Aragão Guimarães (University of the State of Piauí, Brazil / Court of Justice of Maranhão, Brazil)
Ana Rayonara de Sousa Albuquerque (University of the State of Piauí, Brazil)
Ana Valeria Lopes Lemos(University of the State of Piauí, Brazil)
Andrea Thaís Xavier Rodríguez Hurtado(University of the State of Piauí, Brazil)
Fernanda Tamyris de Oliveira Lopes(University of the State of Piauí, Brazil)
Erica Diana Rodrigues do Nascimento(University of the State of Piauí, Brazil)
Indira Rachel Arrais Oliveira(University of the State of Piauí, Brazil)
Larissa de Carvalho Muniz(University of the State of Piauí, Brazil)
The paper aims to analyze the importance of psychological instruments for forensic assessment in Brazil and to discuss psychological strategies used for forensic assessment in criminal and civil Court. Researchers have demonstrated that the amount of required psychological expertise in order to assess forensic demand has increased and, consequently, some problems have been found in the psychologist’s practice, which may be harmful. It must be emphasized that there is no standard for these procedures in Brazil, nor specific instruments to apply, and that these issues are not regulated by Brazilian Law. In addition, other problems were found, such as overusing unstructured and semi-structured interviews and devaluation of family attachment observation (father-son or mother-son) in psychological assessment practiced in family Court. Literature Revision methodology was used for a search on Brazilian publications about forensic psychology, scientifically based on Scielo, PePSIC, MedLine and Lilacs. 32 articles were selected, published in the last 10 years, through the following descriptors: forensic assessment, psychological expertise, Forensic Assessment Instruments (FAI), psychologist’s practice in Criminal and Civil Court. This research showed the non-existence of psychological instruments related to forensic assessment. It is concluded that this has both negative consequences for the assessed subjects as well as ethical and technical implications for forensic psychological assessment.
Poster 07-04
A self-report version of the Scale of Effectiveness of Adaptation: Development and validation
Elisa M. P. Yoshida (PUC-Campinas, Brazil)
Mileny Cia (PUC-Campinas, Brazil)
Adriana R. Freitas (PUC-Campinas, Brazil)
Rodhys R. Sigrist (PUC-Campinas, Brazil)
The study is an ongoing research program whose objective is the development and validation of a self-report version of the Scale of Effectiveness of Adaptation (SEA). The effectiveness of adaptation is a measure of the quality of the individual’s responses to the vicissitudes of life. The quality of adaptation of the individual’s responses is assessed according with three criteria: the possibility of solution of the problem; the degree of satisfaction and the degree of conflict that they generate. Based on data obtained on a clinic interview, the SEA was designed to measure effectiveness of adaptation according to four dimensions: affective-relational (AR), productivity (Pr), socio-cultural (SC) and organic (Or). The SEA is a Brazilian instrument, largely used in mental health context. A self-report version of this instrument should be useful in situations in which a clinic interview is not possible or desirable of being carried out. For the self-report version (SEA-SR), items were developed for two dimensions : AR (n = 28) , Pr (n = 19). 201 university students, 18 to 35 years (M = 21.6, SD = 4.56) filled in SEA, SEA-SR and the Symptom Assessment Scale-40 (SAS-40). Internal consistency - Total score: α> 0.80 , AR : α> 0.70; Pr: α> 0.60; test-retest reliability (one month): 0.68 to 0.60; 58 participants were interviewed and evaluated according to the SEA and responded to the SEA-SR; significant correlations (p< 0.01) between SEA and SEA-SR, between Pr and total scores of SAS-40; between AR and two dimensions of SAS-40: anxiety and somatization. The results suggest the need for adaptation of some items, especially those of Pr. The factorial structure of the scale must be investigated in the next steps of the research.
Poster 07-05
The process of constructing the Bandung Family Relation Test
Kai Welzen (Radboud University of Nijmegen, The Netherlands)
Efi Fitriana (Padjadjaran University, Indonesia)
Peter Nelwan (Padjadjaran University, Indonesia)
Toni Toharudin (Padjadjaran University, Indonesia)
Leo Polhaupessy (Padjadjaran University, Indonesia)
Marisa Moeliono (Padjadjaran University, Indonesia)
Lucas Pinxten (Radboud University of Nijmegen, The Netherlands)
The Bandung Family Relation Test (BFRT) is adapted from the Dutch Nijmegen Family Relation Test (NGT, Oud & Welzen, 1989). It is an operationalization of concepts in the work of Boszormenyi-Nagy and Stierlin, to diagnose and evaluate the family process. In the 90’s Indonesian researchers and the NGT's authors started with the BFRT, resulting in a first publication of validation research (Fitriana, 2011). The research began with the translation and reformulation of items within the six dimensions of the NGT, in cooperation with experts in indigeneous ways of raising children. The reliability analysis was done before the main study in which 529 children (9-12) and their parents were interviewed. The sample design was a stratified cluster. The child data came from individual interviews while the parents were asked to complete a bio-graphic questionnaire at home. Both situations were guided by trained psychology students. Validity of the BFRT was determined by confirmatory factor analysis of a hypothetical model. Reliability coefficients ranged from .58-.78. The goodness of fit statistics for the factor model gave RMSEA = .07 and CFI = .98. From the results we can conclude that the BFRT is reliable and valid.
Poster 07-06
Desire for children: Development and validation of a scale with a Brazilian sample
Jean Carlos Natividade (Federal University of Rio Grande do Sul, Brazil)
Heitor Barcellos Ferreira Fernandes (Federal University of Rio Grande do Sul, Brazil)
Claudio Simon Hutz (Federal University of Rio Grande do Sul, Brazil)
The aim of this study was to develop and validate a measure of the desire to have children. An on-line questionnaire containing items designed to assess the intensity of present and future desire for children and romantic relationship-related questions was responded by 419 individuals from all regions of Brazil. The participants’ mean age was 27.1 years and 64% of them were women; most of the participants were undergraduate students. About 86% of the participants declared they wanted to have children someday, and the average of desired children was of 1.71. A Principal Axis Factoring analysis was conducted and the three factors extracted explained 58.7% of the variance of the 19 items of the scale. The factors were related to the intensity of the desire to have children, to the intention of having children at the moment, and to the intention of having children in the future; the three subscales were moderately correlated and presented satisfactory reliability. The scale discriminated people who used birth control methods regularly from those who did not or who forgot to use them occasionally, and presented a correlation with the intended number of children.
Poster 07-07
A cross-cultural validated scale of emotional and sexual jealousy
Jean Carlos Natividade (Federal University of Rio Grande do Sul, Brazil)
Heitor Barcellos Ferreira Fernandes (Federal University of Rio Grande do Sul, Brazil)
Claudio Simon Hutz (Federal University of Rio Grande do Sul, Brazil)
This study aimed to test the cross-cultural validity of a scale designed to measure emotional and sexual jealousy. Two studies were conducted using on-line questionnaires. In the first study the scale items were developed and their validity and reliability were tested with a Brazilian sample. In study 2, the items were translated into English and responded by native speakers. Participants of study 1 were 461 individuals, mean age of 26.8 years, 59.2% of whom were women. Participants of the second study were 167 individuals from English speaking countries (85% of them from the USA), mean age of 22.8 years, 66.5% of whom were women. The Principal Axis Factoring analyses showed an identical structure for the scale (two factors) in both studies. The first factor related to distress elicited by emotional infidelity cues, and the second factor to sexual infidelity cues. Besides presenting satisfactory reliability indices, the two versions of the scale discriminated people who said they have started an argument due to jealousy in the last month from those who said they did not. These results are in agreement with theories about jealousy and provide evidence that the scale is valid to assess jealousy in different cultural contexts.
Poster 07-08
CUIDA, a self-report questionnaire for psychological assessment of adoption applicants
Pablo Santamaría (TEA Ediciones, Spain)
The aim of this paper is to present the main characteristics as well as the psychometric properties of CUIDA, a self-report questionnaire created for the psychological assessment of adoption applicants. It is scored under response item theory for polytomous items using Samejima´s Graded Response model. Adoption is a way of protection for those children whose families are not able to meet all suitable needs for their development. Responsible institutions must guarantee that parents who adopt these minors will be able to perform the appropriate parental function. For this reason, a strict assessment of psychological and social characteristics must be conducted. Interviews have regularly been used as the principal instrument. Psychological tests have found some difficulties due to their lack of specificity to this context. CUIDA has been created with the objective of becoming a specific instrument for the assessment of adoption applicants. It assesses applicants characteristics which are considered essential in adoption in order to avoid risks for the minor and to guarantee a favourable development. CUIDA it is composed of 189 items, fourteen personality scales, four global factors and three validity scales. CUIDA´s scales include basic aspects collected in technical criterion about adoption established in the different countries where children come from, as well as different theoretical knowledge of the professional experts. A careful work in the development, standardization and validation of CUIDA has been carried out for three years before its publication. It is scored under response item theory for polytomous items using Samejima´s Graded Response model. Different reliability and validity studies has been conducted with satisfactory results: internal consistency, criterion related validity (NEO-PI-R and 16 PF-5) and construct validity. CUIDA could be considered a reliable and valid instrument to assess adoption applicants.
Poster 07-09
The Romanian validation of the High-K Strategy Scale: Preliminary results.
Ioana R. Cocia (Babes-Bolyai University, Romania)
Lavinia C. Uscătescu (Babes-Bolyai University, Romania)
Andreea Butucescu (Ovidius University, Romania) (presenting author)
Alina S. Rusu (Babes-Bolyai University, Romania)
Cezar Giosan (Berkeley College, New York, USA)
The concept of fitness is central to evolutionary psychology. One fitness enhancing strategy in humans is the high-k (i.e. fewer and fitter offsprings), measured with the High-K Strategy Scale (HKSS; Giosan, 2006). HKSS has 25 items, the last three addressing people involved in a relationship. We aim to adapt the High-K Strategy Scale on the Romanian population, beginning with its linguistic validation. We recruited 12 bilingual English-Romanian speakers (8 women and 4 men; M age = 26.90; SD = 6.23), of which half were involved in a relationship. We followed the translation-back translation method (Brislin, 1970), and further applied both versions using Butcher and Gur’s 1974 method. There was no difference between the two language versions for both 22 (Z = -.41, p = .67) and 25 item versions (Z = -.13, p = .89). In addition, there is an association between the Romanian and English HKSS, Spearman`s coefficient (rs) varying between .81 and .89, α = .01. HKSS has an acceptable internal consistency for the English 22/25 (Cronbach α= .86/.71), and Romanian 22/25 item version (Cronbach α= .88/.79). Results showed a suitable translation and a good reliability of the instrument. Further on, we will assess the instrument’s reliability and validity on the Romanian population.
Poster 07-10
The linear and nonlinear relationship between personality and friends on social network sites
Zheng Zhang (Tsinghua University, China)
Yinfei Li (Tsinghua University, China)
Studies have shown a connection between the individual personality of the SNS (Social network sites) user and the way he or she behaves online including the social network use. Cyberspace gives the individual a possibility to show behavior which is opposite to the real world. Therefore, we supposed that perhaps there were some nonlinear relationships between personality and social network use. The present research aimed to investigate how the Five-Factor Model of personality relates to the friends on social network sites, as well as test the linear versus nonlinear hypothesis. 313 Chinese university students from both Sciences and Liberal Arts took part in the survey. The results showed that Openness to experience had a positive linear relationship with individual friend number on social network sites, Conscientiousness and Agreeableness had no significant relationship with the dependent variable. These results confirmed previous studies. Meanwhile, we also found that there was a letter U curved relationship between two personality traits (Extraversion, Neuroticism) and the dependent variable, implying that both individuals who have higher or lower scores on Extraversion and Neuroticism tend to have more friends on the social network sites. This study extends our insight on personality and social network use.
Poster 07-11
Uncovering systematic sources of variability: The case of self-perceived communication competence
Flaviu A. Hodis (Victoria University of Wellington, New Zealand)
Georgeta M. Hodis (Victoria University of Wellington, New Zealand)
This research aims to uncover the systematic sources of variability underlying data collected using the self-perceived communication competence (SPCC) instrument. Although the 12 items comprising the scale are created by crossing four potentially distinct traits with three different (pseudo-) methods of gauging the traits, the factorial structure of SPCC has yet to be thoroughly investigated. Instead, past and current studies have assumed a unidimensional configuration of SPCC. A structural equation modelling multitrait-multimethod investigation was conducted. Informed by the correlated trait correlated method (CTCM) and correlated trait correlated method minus one (CTC(M-1)) frameworks, a comprehensive taxonomy of models was created and evaluated with respect to fit to empirical data and substantive interpretability. The hypothesis of unidimensional configuration of SPCC received no empirical support, whereas several well-fitting multidimensional alternatives have been identified. The substantive interpretation of these multidimensional structures provides interesting insights regarding the systematic sources of variability underlying the covariances among observed scores. To raise researchers’ awareness of the practical differences associated with employing competing factorial structures of SPCC, as well as to encourage an appropriate usage of this scale, the study compares and contrasts correlation matrices of SPCC and theoretically-driven covariates under uni- and multidimensional scenarios.
Poster 07-12
Construction assessment tool Brazilian Communication Styles
Maria Angela Lourençoni (Faculdade Max Planck, Brazil)
Solange Muglia Wechsler (PUC-Campinas, Brazil)
The construction of the Brazilian instrument of psychological assessment of communication styles is based on the temperament types (with four dichotomous dimensions: thinking-feeling, practical-imaginative, organized-flexible, extrovert-introvert). The purpose of this study is to build a range of communication styles with four dimensions, namely: argumentative-kind, pragmatic-reflective, careful-trusted, expressive-discreet. We developed 100 forced choice items with items A and B (25 for each dimension). The judges chosen for analysis were three professionals with experience in psychological assessment for clinical purposeses, two doctoral students in the field of psychological assessment and one doctor in the area of communication and psychological evaluation. The pilot sample was composed of 19 people who answered 200 options from the dichotomous items. They complained of being tired because of this item format, thus suggesting changing to Likert-type scales. Besides the delay of completion, changing from dichotomous to Likert-type scale tends to be more accepted in Brazilian culture. Further studies are being carreid on to investigate the criterion and construct validity of this instrument comparing with participants' communication styles with peer observations. The aim is to obtain a well validated tool that can help to understand Brazilian communicating styles, thus bringing improvements to their verbal expression in the workplace.
Poster 07-13
Social Skills Scale for the Greek population: Psychometic properties and factor structure
Kostas Mylonas (The University of Athens, Greece)
Aikaterini Gari (The University of Athens, Greece)
Social Skills are considered an ability with significant effects on our everyday social interaction as a social skill is the capability to respond effectively to various demanding social situations. The socially effective person can achieve all his/her goals uninhibited by consideration of others. A Social Skills Scale for the Greek population has been devised and is presented in respect to its psychometric characteristics. The initial item pool consisted of 91 items, addressing 11 theoretical social skills factors. Our sample consisted of Greek students of both sexes (N = 307) with the initial analysis reveraling 10 factors, with internal consistency for each factor at acceptable levels. Several other issues were considered such as item bias due to response styles examined through independent samples. Most of the initial theoretically expected dimensions were present in the data being evident in validity studies which followed, as well.
Poster 07-14
Measuring interpersonal sensitivity: Introducing the Geneva Emotion Recognition Test
Katja Schlegel (University of Geneva, Switzerland)
Didier Grandjean (University of Geneva, Switzerland)
Klaus R. Scherer (University of Geneva, Switzerland)
The ability to recognize others’ emotional states (emotion recognition ability, ERA) is a central component in emotional intelligence and interpersonal sensitivity. However, to date no psychometrically sound test to measure individual differences in ERA exists. Previous tests have been criticized for focusing on a single modality (mostly the face) and for including only a small number of emotions. The goal of the present research was to develop a new ERA test (Geneva Emotion Recognition Test, GERT) that features dynamic and multimodal actor portrayals (videos with sound) and a large number of positive and negative emotions, and that meets high psychometric standards. Here, we present the results of two studies: In study 1, we asked 295 participants to watch 195 actor portrayals and to choose, for each portrayal, which of 14 emotions had been expressed by the actor. We then applied Item Response Theory to select 83 items for the first GERT version. In study 2, we investigated the construct validity of the GERT in a sample of 130 students. The GERT was moderately correlated with three other ERA/ interpersonal sensitivity tests, with ability emotional intelligence, and with general cognitive and verbal abilities. Implications for future validation studies are discussed.
Poster 07-15
Validating the Filipino version of the Classroom Friendship Questionnaire
Fraide A. Ganotice, Jr.(Palawan State University, Philippines)
Jonalyn B. Villarosa (Palawan State University, Philippines)
It is assumed that friendly students are those who have a deep concern for and understanding of others. Numerous studies have shown that friendliness is positively associated with a lot of adaptive outcomes. Therefore, educational psychologists have been increasingly interested in measuring this construct. A promising instrument for tapping into this construct is the Classroom Friendship Questionnaire (CFQ, Miscenko & Rascevska, 2008). However, its cross-cultural validity still needs to be tested given that it has only been studied among Latvian participants. In this study we developed a Filipino version of the CFQ using a translation process that combined a committee approach with back translation. The primary objective of this investigation is to describe the translation process done on the original English version of Classroom Friendship Questionnaire (Miscenko & Rascevska, 2008) and determine the psychometric acceptability of the newly developed Filipino version (CFQ-F) of the scale. The CFQ-F with four dimensions: social contacts, trust, support and cooperation, and lack of hostility, and trust, was completed by a sample of 466 high school students. Both within-network and between-network approaches to construct validation were adopted in the study. Results of statistical analyses performed suggest that the instrument has a good internal consistency and support is provided for its construct validity in terms of its factorial structure and correlations with other external variables. Using confirmatory factor analysis, we found strong evidence for the four-factor structure of CFQ-F. The psychometric characteristics of this scale justify its usefulness in future research involving Filipino participants. Implications for cross-cultural research are discussed.
Poster 07-16
Examining the cultural validity of the Rosenberg Self-Esteem Scale across social contexts: Implications for a revised measure
Pedro Portes (University of Georgia, USA)
The present study explores the cultural dimensions self-esteem across cultures and a method for testing the cultural validity of psychological measures in general. For example, pride-oriented self-esteem has been traditionally referred to as ‘self-esteem,’ which is based on the individualistic values or independent construal of self. Another factor related to self-esteem works as an inhibitor against extremely high self-esteem, which in its extreme form is arrogance or unrealistic self-image. American culture encourages enhancement of pride-oriented self-esteem, while Japanese culture encourages that of humility-oriented self-esteem. A proposed bipolar model addresses issues of discordant findings regarding the relationship between self-esteem and other variables. The item contents of the Rosenberg Self-Esteem Scale (1965) were investigated for identical meanings for both female European American (n = 234) and female Japanese (n = 311) university students, and a survey questionnaire was developed. Multivariate analyses, including factor analyses, revealed that the structure of self-esteem as measured by the Rosenberg Self-Esteem Scale differed for European Americans and Japanese. The methodological approach in the present study may be applicable for other research studies where cultural validity needs to be established. Data from other populations will be also highlighted in this session.
Poster 07-17
Using the SOGS-RA to measure adolescent problem gambling: IRT-based analyses
Maria Anna Donati (University of Florence, Italy)
Silvia Galli (University of Florence, Italy) (presenting author)
Francesca Chiesi (University of Florence, Italy)
Caterina Primi (University of Florence, Italy) (presenting author)
The South Oaks Gambling Screen – Revised for Adolescents (SOGS-RA, Winters, Stinchfield, & Fulkerson, 1993) is the most widely used measure for adolescent gambling. It was specifically developed in order to classify gambling severity into the three categories comprising no problem, at-risk and problem gambling. Since the knowledge of the SOGS-RA’s psychometric proprieties is limited, the aim of this study was to investigate them applying Item Response Theory (IRT), in order to examine the instrument’s accuracy along gambling severity levels. The test was administered to 401 regular adolescent gamblers (Male = 78%, Mean age = 16.54, SD = 1.65). Once the prerequisite of unidimensionality was confirmed, item parameters and the Test Information Function were estimated. Results showed that items have different levels of severity ranging from medium to high values and that the test accurately measures medium-high gambling severity levels. These findings provide evidence of the SOGS-RA’s adequacy for assessment of adolescent gambling indicating that it satisfactorily identifies at-risk and problem gamblers.
……………………………………………………………………………………………………………………………………………………....
Poster Session 8 July 5, 08.15-10.15 Room: Rotonde
Poster 08-01
Validation of Saville Consulting Wave Professional Styles and Swift Analysis Aptitude in Mauritius
Rainer Kurz (Saville Consulting, UK)
Nailah Moussa (University of Mauritius, Mauritius)
This poster looks at the validity of the Professional Styles personality questionnaire and Swift Analysis Aptitude across 14 leading companies in Mauritius. In a concurrent validation study 221 Managers and Professionals completed the Saville Consulting Wave® Professional Styles questionnaire and the Performance 360 Self-assessment. Their performance was also evaluated by at least one self-nominated external rater. External Performance Rating (EPR) values were calculated by averaging across Boss, Peer and/or Report/Sub-ordinate rater categories. Point-to-point validation of a-priori hypothesised relationships of Great 8 Competency Potential composites (methodology based on MacIver et al., 2009) correlated with their criterion EPR counterpart at .16 on average (uncorrected). The same uncorrected value as OPQ in the meta-analysis of Bartram (2005, table 13). The average uncorrected value is .22 for Professional Styles in the Project Epsom co-validation study of different personality instruments (Saville et al., 2012, table 20.56). For a sub-sample (N=58) Swift Analysis Aptitude (SAA) total score correlated .38 with overall performance EPR.
The results demonstrate empirically that the Saville Consulting personality and ability instruments predict work performance in Mauritius thus providing strong support for the cross-cultural application of the psychometric instruments.
Poster 08-02
Facebook vs CVs: The impact of social networking sites on recruitment
Paula Cruise (OPP Ltd, UK)
This paper considers the use of social networking sites (SNS) in recruitment and selection. It examines best practise and equality issues and presents recommendations for psychologists and HR Policy. 1,212 UK & Irish working age participants completed the 16PF personality questionnaire, alongside questions assessing attitudes towards the use of SNS in recruitment, and online privacy. The results revealed that 56% of people said that during a recruitment process they would be likely to search SNS for a job applicant. However, 27% indicated that, if they were the applicant, they would not want an interviewer to search for them. The research suggests that those spending more time on SNS tend to be more extraverted, younger, and have greater access to the Internet and new technologies. 37% of the sample stated they would change their character when online, calling into question the accuracy of information presented online. Best practice concerns for SNS include: the validity, objectivity, and fairness of SNS as a selection tool; inequality of opportunity for applicants to use new technology; and potential for direct and indirect discrimination.
Poster 08-03
The Five Factor Personality Inventory as predictor in police selection processess
Erdogan Gultekin (Turkish National Police Academy, Turkey)
Pre-employment psychological evaluation utilizing personality inventories is common for police selection in law enforcement settings. The Five-Factor Personality Inventory (FFPI), developed from the five-factor theory of normal personality functioning, has revealed to be useful in Turkish police selection. The present study empirically examines the structure of personality traits in self-reports of 5579 police candidates suggesting five major underlying dimensions, that is, ‘Neuroticism’, ‘Extraversion’, ‘Openness’, ‘Agreeableness’ and ‘conscientiousness’. This study also examines the predictive validity of FFPI facet scales for predicting job descriptions elements of police and relationships between the competence dimensions observed in the selection interviews and their personality trait building blocks. The use of the FFPI as a selection instrument for police officers is discussed.
Poster 08-04
Personality traits and intelligence as predictors of police officers' performance in Estonian Police College
Helle Pullmann (University of Tartu, Estonia)
Margus Tõnissaar (University of Tartu, Estonia)
Helle Niit (Police and Border Guard Board, Estonia)
Maria Veltmann (Assessment Centre Tripod, Estonia)
The main goal of the study was to examine the predictive validity of general intelligence and Big Five personality traits for predicting graduation and performance of police officers in Police College in Estonia. The sample consisted of 309 police cadets (41% females, 59% males) who were enrolled in the Police College of the Academy of Security Sciences in Estonia in 2004–2007. Cadets completed two measures during admission procedure: the Personality Inventory (NEC-S/V4) developed by Tripod to measure five personality domains (Emotional Stability, Extraversion, Openness, Agreeableness, and Conscientiousness) and general intelligence was measured by the Mental Abilities Scale (VVS/3) developed by Tripod, consisting of three subscales (verbal, numerical, and visuo-spatial). Academy performance was measured as an overall grade point average from six proficiency exams. Similarly to previous studies, conscientiousness and intelligence had a significant independent contribution to predict academic achievement. The results demonstrated that successfully graduating cadets (n = 222) of competency based learning in Estonian Police College had significantly higher levels of cognitive abilities and conscientiousness trait compared to nongraduating cadets (n = 87). Admitted police cadets who score above average on the intelligence test or Conscientiousness domain can upwards of double the rate of successful police college performance.
Poster 08-05
Criterion validation study of an integrity questionnaire using an experimental design
Louisa Tate (Cubiks, The Netherlands)
Dimmy van Zanten (Cubiks, The Netherlands)
Marina Goldmann (Cubiks, The Netherlands)
Integrity measures have a good reputation as predictors of work performance (Smith & Hunter, 1998). However, due to current debate around ethical assessment, the need for transparency and openness puts pressure on the usage of such instruments, with designing a criterion validation study proving especially challenging. We have developed an online multimedia tool based on Kohlberg’s theory of moral development (Kohlberg, 1981; 1984; 1991). For each moral dilemma, participants are asked to rate different arguments they might consider in making a decision as to how they would act, with the arguments being classified in relation to different stages of moral development. In order to research its criterion validity, an experimental design was used, where participants were offered the opportunity to cheat on an intelligence test. Afterwards they were asked if they did cheat, enabling the tool to be assessed against an integrity criterion. The data are in the process of being analysed, but early indications suggest that the internal structure of the tool is good, e.g. internal consistency in the pilot trial ranged between 0.68 and 0.84 (n = 137), and there was a significant difference in Honesty (HEXACO-PI-R) for participants operating at different levels of moral development (F = 6.705, p = 0.002, n = 92). Further research is needed to provide robust evidence of criterion validation. The generalisability of the findings from an experimental design to high-stakes real-life situations is considered.
Poster 08-06
Mental abilities vary by managerial level: A case of Hansabank Estonia
Maria Veltmann(Assessment Centre Tripod, Estonia)
Terje Veide(Tallinn University of Technology, Estonia)
Helle Pullmann(University of Tartu, Estonia)
Research has confirmed that the validity of cognitive ability increases as the complexity of jobs increases. Based on data collected in Hansabank Estonia in 2006–2007, the variation of general mental abilities across managerial levels was examined in Estonia. About 949 employees completed the The Mental Abilities Scale for Hansabank (VVS/HP; VVS-SB) developed by Tripod, consisting of four subscales (Verbal and Numerical Ability, Logical and Spatial Thinking). The 89 managers from the total sample were divided into three groups (executives, middle-level, and first-level managers) according to their managerial level. Similarly to previous studies (e.g., Veltmann & Pullmann, 2011), all groups of managers had significantly (p < .001) higher levels of mental abilities compared to the averaged mean level of non-managerial occupational groups. The analysis of variance revealed that the levels of mental abilities varied significantly across managerial level groups. Spearman rank order correlation between the managers’ test scores and managerial level was .34 (p < .001). The results indicated that executives and middle-level managers scored significantly higher on the ability indexes compared to first-level managers, attesting to the utility of mental abilities scales used within the occupational community, for selection and assessment of suitability for promotion to senior managerial roles.
Poster 08-07
Personality profiles of occupational groups: A case of Hansabank Estonia
Maria Veltmann(Assessment Centre Tripod, Estonia)
Terje Veide(Tallinn University of Technology, Estonia)
Helle Pullmann(University of Tartu, Estonia)
Personality traits are enduring dispositions of individuals to feel and behave in certain ways across situations and settings. Knowledge of the effects of homogeneity of personality in occupations may have numerous practical implications, particularly in the area of selection and development. Based on data collected in Hansabank Estonia, which has been ranked as the most competitive company in the Estonian financial sector, personality profile patterns for five occupationalgroups were examined. From the total of 3241 employees of the organisation, 947 incumbents completed the Personality Questionnaire (NEC-L/NEO-TRI) developed by Tripod to measure five personality domains (Emotional Stability, Extraversion, Agreeableness, Conscientiousness, and Openness to Experience) in 2006–2007. The employees were divided into five occupational groups: managers (n = 126), analysts (n = 76), specialists (n = 425), assistants (n = 93), and service employees (n = 227). The levels of personality scores varied significantly across occupational groups for Emotional Stability, Extraversion, Agreeableness,and Conscientiousness. The analyses revealed distinct personality profiles for employees with different job specifications. Individuals working in the management were highly more emotionally balanced, active and sociable, relationship-oriented, tending to work towards goals in an industrious and dependable fashion compared to all other job families. The results demonstrated, for instance, characteristically contrasting mirror-image personality profiles for managers and analysts.
Poster 08-08
Exploratory study of patterns in the personality traits of employees at an ICT company
Haniza Yon (MIMOS Berhad, Malaysia)
Kok Mun Yee (MIMOS Berhad, Malaysia) (presenting author)
Tala Mirzaei (MIMOS Berhad, Malaysia)
The present study explored patterns in the personality traits of ICT company employees. A Malaysia-focused personality inventory based on the five-factor model was administered to 645 employees (46% technologists, 26% researchers, 20% managerial staff, 7% apprentices, 1% unknown). A one-way ANOVA analysis demonstrated significant differences between managers, researchers, and technologists in terms of altruism, emotional control, deliberation, sociability, and openness in action. Managers were found to be more sociable, more susceptible to stress, more impulsive, and more prone to depression. Technologists tended to be more deliberate in their work as compared to other groups, whereas researchers were more prone to altruism and more open to change. This study implies differences between employees according to whether their work is technological, managerial or research-oriented. The results carry implications for predicting employee job satisfaction and turnover.
Poster 08-09
Effects of culture, motivation, and test platform on online personality test results: Implications from the SOSIE
Margreet Kerkmeer (Pearson, The Netherlands)
Xiaobin Zhou (Pearson, The Netherlands)
Digital assessment platforms make it possible to deliver fair and standardized assessments and interpretations in an efficient manner to meet the increasing need of international companies to obtain comparable test results across platforms/languages. However, previous studies have shown that foreign adaptation of tests may unintentionally introduce bias in test construct, assessment method, and item interpretation. This could occur both between and within countries (different ethnical subgroups). The present research investigates the impact of culture, language, motivation, and test platform on the results of SOSIE, an online personality questionnaire used for personnel selection and assessment. The standardization data collected for the SOSIE-II international and the Dutch derivative were used for the analyses. All the data were collected during 2011. The sample was composed of Dutch examinees tested on international platform (N = 122), Dutch examinees tested on Dutch platform (N = 994), and Flemish examinees tested on Dutch platform (N = 106). On the Dutch platform, examinees were recruited in both selection setting (N = 127) and general testing setting (N = 867), thus different motivations for test taking. The distributions of scores on the 20 SOSIE scales will be compared across the groups. Implications of the differences and similarities on the bases of platform, culture, and motivation will be discussed.
Poster 08-10
Need for cognition and affective-priming on accuracy of faking detection in employment interviews
Chenxiao Tang (Beijing Normal University, China)
As shown in previous research on affect and social cognition, higher need for cognition (NFC) leads to better discrimination of truthful from deceptive messages, while negative mood increases judges’ skepticism and improves their accuracy in detecting deceptive communications. Is this still true given the situation of an employment selection interview? In the present study, after a mood induction using positive, or negative films, each participant was asked to act as a job expert and make employment selection decisions based on their impressions on two pieces of interview recordings. Applicants in both recordings had sent some untruthful messages about themselves to gain more desirable results. 38 subjects were included in the experiment, the NFC scale (Cacioppo,J.T.,1984) was used to define NFC groups (hign/low). Results show that both NFC and affect-priming have a significant influence on interviewer’s accuracy of interview faking detection. Consistent with part of the prediction, interviewer with higher NFC indeed showed more accurate faking detection. But contradictory to the other part of hypothesis, positive mood verses negative mood lead to more accurate faking detection. This result shows a significant difference between the daily social interaction and the employment selection scenario. The author argues that in employment selection, a rater naturally adopts a skeptical perspective, negative mood would only make them suffer more from the great deal of cognitive load they already have. The true reasons behind this distinguished result need future research.
Poster 08-11
Establishing measurement invariance of the Openness Scale across job specialisations
Haniza Yon (MIMOS Berhad, Malaysia)
Nur Ayu Johar (MIMOS Berhad, Malaysia) (presenting author)
Kok Mun Yee (MIMOS Berhad, Malaysia)
Nur Salihah Raston (MIMOS Berhad, Malaysia)
Items in a psychometric instrument must measure the same construct even when the instrument is administered to two or more distinct groups. Furthermore, the instrument should display measurement invariance, meaning that the trait scores for the distinct groups should fall on the same measurement scale. In this study, measurement invariance of a scale called Openness in a personality inventory was investigated by comparing factorial invariance for three different job tracks: technologist, manager and researcher. The Openness scale comprises such traits as ideas, values and aesthetics. The study sample was comprised of 315 employees at an ICT company, ranging in age from 23 to 55 years. Confirmatory Factor Analysis (CFA) was performed across the three job tracks (technologist: 138, specialist: 85 and researcher: 92). Scalar invariance across the job tracks was demonstrated based on the Comparative Fit Index (CFI) among the models for the three groups, as well as additional indices of fit including the Tucker-Lewis Index (TLI), the Root Mean Squared Error of Approximation (RMSEA) and Chi²/df values. The results (?²/ df =1367.58, CFI = .773, TLI = .762, RMSEA = .048) supported an equivalent factor model of the construct under study across the three different tracks. These findings show that measurement of Openness was invariant with respect to job track; therefore, inter-track comparisons in terms of Openness may be deemed appropriate based on current evidence. Future studies may explore the measurement invariance of the scale using the Item Response Theory (IRT) approach.
Poster 08-12
The impact of organizational citizenship behavior, emotional intelligence and job satisfaction on the employees in Haft Almas Company
Morteza Taheri (Statist Data, Iran)
To explore the impact of organizational citizenship behavior, emotional intelligence and job satisfaction on the employees, a descriptive case study was conducted about the sort of interdependence in Haft Almas Company. The sample consisted of 227 employees of the above mentioned company selected by stratified random sampling. To gather the required data, questionnaires of organizational citizenship behavior, emotional intelligence, and job satisfaction were used. The Alpha coefficient factors of the above questionnaires were respectively, 0.87, 0.73, and 0.72. Moreover, executive analysis of the study was made by interdependence and structural equation methods. The results of the study showed that the relation between organizational citizenship behavior, emotional intelligence and job satisfaction is statistically significant. Furthermore, the results suggest that organizational citizenship behavior accounts for 48% of variation in job satisfaction, and emotional intelligence accounts for 76% of variation in job satisfaction. Other findings of the study showed that there is a statistically significant relation between job satisfaction and the level of education; according to the present study, however, there was no statistically significant with respect to gender in terms of job satisfaction.
Poster 08-13
Verbal ability, personality, and emotional intelligence as predictors of sales proficiency
Janneke K. Oostrom (Erasmus University Rotterdam, The Netherlands)
Lonneke A.L. de Meijer (Erasmus University Rotterdam, The Netherlands)
Nicole Soetiin (Ixly, The Netherlands)
Marise Ph. Born (Erasmus University Rotterdam, The Netherlands)
The aim of this study was to examine the relationships of verbal ability, personality and emotional intelligence with sales proficiency. Data were collected among 199 students, who completed a verbal ability test, a Big Five personality questionnaire, a self-report emotional intelligence scale, and a sales proficiency test. To test sales proficiency, a constructed response multimedia test was used. In line with our hypotheses, results showed that the two dimensions of emotional intelligence (regulation of emotion and the use of emotion to facilitate performance) and three Big Five traits (extraversion, openness to experience, and emotional stability) were related to sales proficiency. Verbal ability was not related to sales proficiency. Yet, verbal ability moderated the effects of emotional intelligence on sales proficiency. Specifically, the correlation between the emotional intelligence dimensions and sales proficiency becomes stronger as verbal ability decreases. Thus, students with low verbal ability were able to show effective sales proficiency as measured with the constructed response multimedia test if they are emotionally intelligent. In other words, emotional intelligence and verbal ability are compensatory with respect to sales proficiency.
Poster 08-14
The meaning of unidentified multidimensionality in the assessment of emotional intelligence
Elke Veirman (Ghent University, Belgium)
Johnny R. J. Fontaine (Ghent University, Belgium)
Research on the Mayor-Salovey-Caruso Emotional Intelligence Scale, an assessment instrument of emotional intelligence from the ability perspective, is frequently challenged with issues such as instability of the nomological network, difficulties to validate the internal structure, and high residual item variances. A first study is presented in which it was investigated to which extent these recurrent problems may be attributed to unidentified multidimensionality. In a heterogeneous child and adolescent sample emotional intelligence was measured by the Mayor-Salovey-Caruso Emotional Intelligence Scale (Youth version). In a first random split-half results showed that only a small portion of the variance could be attributed to systematic inter-individual differences. Further exploratory factor analysis revealed the existence of two higher-order factors, namely the ability to identify and process signs for the absence of particular emotions and the ability to identify and process signs for the presence of particular emotions. These results were confirmed in a second split-half. Moreover, both factors showed a differential nomological network with pathology and personality traits, which suggested motivationally driven differences. In a second study, the relationships with Higgins’ promotion and prevention focus were examined in depth.
Poster 08-15
The adaptation and validation of the EAQ to the Spanish context using SEM
V. Prado-Gasco (University of Valencia, Spain)
A. Ordóñez (University of Valencia, Spain)
I. Montoya (University of Valencia, Spain)
L. Villanueva (University of Valencia, Spain)
R. González (University of Valencia, Spain)
The “Emotional competence” is the set of knowledge, skills, abilities, and attitudes necessary to understand and use emotional vocabulary. It also helps children to perceive, understand, be aware of, control and regulate emotions perceiving (Bisquerra, 2009; Garner, 2010; Lahaye, 2011; Saarni, 2000). In the evaluation, we use the “Emotional Awareness Questionnaire” (EAQ), self-reports directed at children and adolescents of Dutch origin (Rieffe et al., 2007; 2008). The aim of the study is to present a first validation of the EAQ in the Spanish context using SEM. The sample was made up of 1,166 children from 11 schools (8-14 years olds; girls 52.3%). The EAQ (Rieffe et al., 2008) consisted of 30 items and 6 factors: Differentiating emotions, Verbal sharing of emotions, Not hiding emotions, Bodily awareness of emotions, Attending to others’ emotions, and Analyses of emotions. The result was a 30 item scale with adequate psychometric properties supporting the underlying construct (χ²/DF = 996.491 (389); AGFI= .932; GFI= 943; SRMR = .048; RMSEA=0.37; α= .67). Future research should increase sample scope in terms of representativeness. The EAQ is appropriate, useful and necessary, since the instruments that assess emotional competences in the Spanish child population are scarce.
Poster 08-16
Motor performance tests in talent promotion: The MT2-B Leichtathletik for Sport Schools
Cornelia Moll (Karlsruhe Institute of Technology, Institute of Sports and Sports Science, Germany)
Ilka Seidel (Institute of Applied Science of Sports Training Leipzig, Germany)
Klaus Bös (Karlsruhe Institute of Technology, Institute of Sports and Sports Science, Germany)
An important aspect of high-performance sport system is the promotion of talented adolescents. The NRW sport schools are focused on this mission. Due to limited resources, fundamental principles of decisions must be detected to choose the most talented. To ensure a highly accurate selection, empirically evaluated motor performance tests have to be used. So far such diagnostic tools, which are appropriate to athletes at sport schools and sufficient to scientifically proved criteria, are lacking. Based on a test developed by the Institute of Applied Science of Training (Rost & Schön, 2002), the Motorischer Test 2-B Leichtathletik (Seidel, Grüneberg, Moll, & Bös, in prep.) was designed. It comprises a 50-m-race, ball-throwing, triple-hop and a 1000-m-race. The retest-reliability and the criterion-related validity (criterion: corresponding competition results) are to be measured. Data of 29 male and female athletes for the reliability study (13.47 ± 0.50 years) und 22 male athletes for the validity study (14.09 ± 0.32 years) are yet available. First results show good reliability of all tasks (.88 1). In this sense, the present study aimed to investigate the dimensionality of the Brazilian PSS-10 in a sample of school teachers (N = 517) using robust exploratory techniques and more parsimonious retention methods. A minimum rank factor analysis (MRFA), on a polychoric correlation matrix, with a sample permutation parallel analysis (PA), as the factor retention method indicated that the single-factor solution was the most adequate to data. All items presented adequate factor loadings (> .37) and Cronbach´s Alpha reliability was .80. The retained factor explained 67% of the scores variance. This study provides validity evidence of the PSS-10 in the Brazilian context. The authors also suggest the use of adequate methods of factor analyses in future studies with the PSS-10, in order to provide reliable results and diminish the incongruence on its factor structure due to inappropriate analysis.
Poster 09-11
Mean-level change in FFM traits: A meta-analysis of longitudinal studies with the NEO-PI
Sofia Sjöberg (Stockholm University, Pearson Sweden, Sweden))
Anders Sjöberg (Stockholm University, Sweden)
The goal of this meta-analysis was to estimate mean level change of self-reported personality traits by analyzing longitudinal data of self-report character. Moderating effects of age and time interval between assessments was investigated concerning the patterns, direction and magnitude of mean level change. The five factor model provided the theoretical framework for personality and eight studies (N = 2677) using the NEO-PI were included. The results indicate that mean level changes do take place, but only for some traits and in certain ages. The significance and meaning of these results will be discussed both from a biologically based trait perspective and from a measurement perspective concerning self-ratings of personality traits.
Poster 09-12
Testing the psychometric properties of the Adaptive Behavior Assessment System - II in Latvia
Malgozata Rascevska (University of Latvia, Latvia)
Pauls Legzdiins (University of Latvia, Latvia)
Sandra Sebre (University of Latvia, Latvia)
Aleksandra Andrejeva (University of Latvia, Latvia)
The purpose of this study was to evaluate the psychometric properties of the Adaptive Behavior Assessment System – Second Edition (ABAS-II, Harrison & Oakland, 2003) in Latvia. Ten skill areas and three adaptive domains were measured using the Parent, Teacher and Adult Form of ABAS-II. The participants were selected accordingly with regional proportion of Latvian students/adults from 5 to 21 years old for the Teacher and Parents Form (N = 720), and from 16 to 21 years old (N = 70) for the Adult Form, 50% female and 50% male. The internal reliability of ABAS-II for 3 forms of ABAS-II were above α = .96 for total scale, above α = .84 for skill areas, and above .92 for 3 domains, retest reliability of scales were above .70. The results of item analyses showed that 98% of the items showed a psychometrically acceptable discrimination index. The correlations of skill areas varied from .45 to .77 in the Parent Form and .63 to .89 in the Teacher Form (closer than in the USA sample). A one factor model for the ten skill areas were confirmed (and not the three factor models). The ABAS-II Latvian version showed high reliability indices, but the structure of three domains was not confirmed. The implications and comparison of the results with other adaptations of the ABAS-II are discussed.
Poster 09-13
Divergent validity in measuring approach and avoidance goals: It is only goal-engagement?
Scott R. Ross (DePauw University, USA)
Heidi N. Keiser (University of Minnesota, USA)
Since Kurt Lewin (1935) first identified approach and avoidance conflict, researchers have been interested in the measurement of not only conflict, but goal-orientation motivated by an appetitive or aversive consequence. Elliot and Sheldon (1997) attempted to measure approach and avoidance goals, separately. In the current study, student participants (N = 330) rated the importance of 51 goals—including both approach and avoidant—as well as other measures of achievement-related constructs. Contrary to their original findings, we found a high correlation of .88 between approach and achievement dimensions, respectively. In addition, these scales—when summed according to common psychometric practice—resulted in a highly similar pattern of relationships—using Steiger’s (1980) test for dependent correlation—with scales measuring competition, cooperation, facilitating and debilitating anxiety, as well as all five factors of the five-factor model of personality, and indices of Gray's Behavioral Inhibition and Activation systems. These findings call into question the use of these scales to measure separate aspects of approach and avoidance goal-orientation. Instead, the pattern of correlations indicates that they measure the same construct, goal-engagement. Future directions for scale development are proposed and considered in the context of Elliot and Sheldon's measurement approach as well as established psychometric theory.
Poster 09-14
Psychometric analysis of the Enright Forgiveness Scale: What do subscales add?
Scott R. Ross (DePauw University, USA)
Heidi N. Keiser (University of Minnesota, USA)
Amanda Brinkman (DePauw University, USA)
Joanna Will (University of Virginia, USA)
Jarrod A. Johnson (Purdue University, USA)
The Enright Forgiveness Scale (EFS; 1995) was the first major scale developed to measure forgiveness. Notable for its 65-item length, the EFS is composed of 6 subscales. A psychometric examination of the EFS in 293 college students indicates that analysis at the subscale level lacks incremental validity when predicting other forgiveness criteria (e.g., Transgression Narrative Forgiveness Scale, Heartland Other Forgiveness, Mauger Other Forgiveness, Forgiveness Likelihood, Trait Forgiveness Scale), with the EFS-Negative Affect subscale accounting for 9-17% of the variance, and additional subscales collectively accounting for 1-2% of incremental variance. Intercorrelations among subscales ranged from .72 to .87 (Median r = .81). Factor analysis revealed that when subscales were included with other forgiveness criteria, all EFS subscales loaded on the same factor in a two-factor solution. When the EFS total score was included in lieu of subscales, forgiveness measures comprised a single factor. Finally, using Steiger’s (1980) test for dependent correlations, analysis of subscales with major Five-Factor domains indicated no significant differences in the correlations of any subscale with any domain. In sum, these findings suggest that subscales add little to the measurement of forgiveness and that a substantially shortened version of the EFS is warranted.
Poster 09-15
The structure and diagnostics of action control: The role of the additional subscale reflective orientation
Olga Mitina (Lomonosov Moscow State University, Russia)
Elena Rasskazova (Moscow City University of Education and Psychology, Russia)
Two types of action control (action versus state orientation) suggested by J.Kuhl are widely used in self-regulation research. However, these alternatives are not opposite towards each other. Also, state orientation might reflect either self-regulation failure or an intentional choice of the person to take the time considering the whole situation. Thus, we suggest a third component of action control called “reflective orientation”. The aim of this study was to test this hypothesis. 287 students filled in the modified Russian version of the Action Control Scale and a number of self-regulation questionnaires. In order to reveal the scale ambivalence in the ACS, we substituted the ipsative scale by a Likert scale. Also items correcponding to reflective orientation for each situation were added. The state and action orientations were found to be moderately negative correlated to each other. The new reflective orientation scale was shown to be reliable and distinct from other scales. Also it was the best predictor of planning, programming, focusing on the problem and positive reappraisal. It is concluded that the reflective orientation could be considered as a third component of the action control and that the eflective orientation scale was shown to be reliable and valid.
Poster 09-16
Measuring Mindfulness: Insights from Tibetan, Theravada, and Zen Clergy
Michael Christopher (Pacific University, USA)
Varinthorn Christopher (Portland State University, USA)
Sukjai Charoensuk (Boromarajonani College of Nursing, Thailand)
Phenpan Kajohnsilp (Boromarajonani College of Nursing, Thailand
Lisa Woodrich (Pacific University, USA)
Kristin Tiernan (Pacific University, USA) (presenting author)
There are important discrepancies in how mindfulness has been operationalized, assessed, and practiced. Therefore, in this research we examined mindfulness from the perspective of experts – Theravada, Tibetan, and Zen Buddhist clergy. A sequential exploratory mixed methods design was used (the qualitative results are reported here – the quantitative phase will begin in Summer 2012). We developed a coding schema through a series of steps to identify core themes. Participants were 42 Buddhist clergy (14 Zen, 21 Theravada, and 7 Tibetan) who completed a qualitative interview and several existing mindfulness measures. Although data analysis is ongoing, we identified several categories that were most salient among our participants. These are: present focus, nonjudgmental awareness, sensory perception, and impermanence. Although these categories overlap with Western conceptualizations of mindfulness, several areas of divergence also emerged, including, an emphasis on ethics and the discrepancies between “basic” and “advanced” mindfulness. Valid assessment of mindfulness is essential, and although there are areas of convergence between Western and traditional Buddhist conceptualizations, questions regarding the validity of these measures remain. Our progress toward developing a measure of mindfulness in which we integrate elements of existing measures and information gleaned from these interviews will also be discussed.
Poster 09-17
Exploring measurement invariance of a teligiosity scale across two religion groups
Haniza Yon (MIMOS Berhad, Malaysia)
Jie Xin Lim (MIMOS Berhad, Malaysia)
Rohani Pa (MIMOS Berhad, Malaysia)
Mun Yee Kok (MIMOS Berhad, Malaysia)
The quality of a measurement tool to be invariance across groups is important as a property for latent mean comparisons (Kim & Yoon, 2011). As a part of the scale development process, measurement invariance or measurement equivalence (MI/E) of a religiosity scale across two groups of different religion (Muslim and Buddhist) was tested using multi-group confirmatory factor analysis (CFA). 151 female and 60 male students, aged between 19 and 33 (M = 22.59, SD = 2.45) were recruited from three local universities. Of the 211 students, 178 were Muslims while the others were Buddhists. All participants completed a set of questionnaire which consisted of a six-item religiosity scale among other scales in the Malaysia national language. CFA with maximum likelihood (ML) estimation was conducted. A series of increasing-restrictive model testing showed evidence of configural invariance with one-factor model with five items across the groups. However, the data did not reveal an acceptable fit with a more stringent model of metric invariance as interpreted using fit indices. Partial measurement invariance was not explored due to the exploratory nature of the technique. The unequal group sizes had to be taken into consideration when interpreting the result (Brown, 2006).
……………………………………………………………………………………………………………………………………………………....
................
................
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 searches
- federal minimum wage poster pdf
- poster contest rules
- poster contest rules sample
- poster contest guidelines
- scientific poster presentation guidelines
- ash poster guidelines
- scientific poster template downloads
- scientific poster design
- new scientific poster format
- scientific poster design template
- free poster templates
- movie poster template download