Name of Test:



Test Review: Phonological Awareness Test-2 (PAT-2)

|Name of Test: Phonological Awareness Test-2nd edition |

|Author(s): Carolyn Robertson and Wanda Salter |

|Publisher/Year: Linguisystems 1997; Linguisystems 2007 |

|Forms: one |

|Age Range: 5 years, 0 months through 9 years, 11 months |

|General comment: |

| |

|The recent second edition (2007) features a new normative sample, division of the subtest into two groupings-Phonological Awareness and Phoneme-Grapheme subtests, and reduced administration |

|time. It contains an optional invented spelling subtest that is not normed. The substitution without manipulatives task has been dropped in order to “shorten the test” (Robertson & Salter, |

|2007a, p. 7). |

| |

|Norming Sample: |

| |

|The standardization study was conducted from January through April, 2007, with a sample the authors describe as reflecting “the national school population demographics from the 2004 National |

|Census for gender, race, education placement, and socioeconomic status” (Robertson & Salter, 2007a, p. 44). |

| |

|Five hundred, ninety master’s level speech-language pathologists participated as test examiners. These examiners were supported with online resources and accuracy and completeness of the data |

|was computer monitored. The students were chosen by the examiners to match the demographic selection criteria. The original PAT did not specify who participated as examiners other than to state |

|that the examiners were speech-language pathologists. |

| |

|Total Number: 1 582 (N.B. an additional 316 at risk students participated in validity studies) |

|Number and Age: The students were between 5 years, 0 months through 9 years, 11 months of age in 10 six-month intervals. |

|Location: 48 states and the District of Columbia. (Original PAT sample consisted of 175 elementary schools in 5 communities in California, 2 in Connecticut, 1 in Florida, 2 in Texas, and 3 in |

|Wisconsin were sampled. (The sizes of these communities are not given). |

|Demographics: 792 male, 790 female, 58% White, 16% Black, 19% Hispanic or Latino, 7% Asian and Others. |

|Rural/Urban: no information |

|SES: Categories divided into: High, Middle, and Low. |

|Comment: The table on page 44 of the examiner’s manual does not define the income ranges for these categories. Such specific information is found in later in the examiner’s manual in the section|

|on race/socioeconomic differences in test performance” ((Robertson & Salter, 2007a, p. 49) where high SES is family income above 60,000, middle SES is 25,000 to 59,000 and low SES is less than |

|24,999. These ranges were referenced as U.S. Census Bureau (2005). This information represents an improvement over the original PAT which contained no such information. |

| |

|Other (Please Specify): The norming sample included children with Individualized Education Plans (IEPs) who attended regular classes. Exclusion criteria included: hearing impairment, children |

|who were nonverbal, children who were non-English speaking, and those living outside of the United States. |

| |

|Comment: The original PAT was missing information commonly provided such as SES, urban/rural, all participants were reported to be normal, and no clinical populations included in the sample. |

|When a sample has no clinical populations included, they will score zero, therefore, severity cannot be assigned. These concerns have been largely addressed in the new edition; however, the |

|information on clinical populations is slim leaving me to wonder what their performance might be. This omission becomes important later when considering aspects of validity when only the term |

|“at-risk” is used to describe the students who participated in the validity studies. |

| |

|Comment: The standardization study includes more students (i.e., 1582 compared to 1235 in the original PAT); however, the sample characteristics are still somewhat vague. A table showing the |

|numbers of students by state could have been easily provided to show distribution. |

|Summary Prepared By: Eleanor Stewart 26 July 2007; revised September 2008 |

|Test Description/Overview: |

| |

|Theory: On the first page of the PAT-2 manual, the authors provide an overview of the key evidence for the relationship between phonemic awareness and reading as presented by the National |

|Reading Panel (2000). Five key findings that form the basis for evidence-based practice are listed. The authors state that the revised PAT incorporates the findings of the panel as well as |

|“expert professional practice” (Robertson & Salter, 2007a, p. 1). The authors state that phonological awareness is linked to early reading development and that Children who lack this skill |

|require intervention. |

| |

| |

|In the Introduction, beginning on page 9, the authors present the rationale for the PAT-2. They trace the growing awareness among professionals about phonological awareness since the original |

|PAT was published ten years ago pointing out how phonological awareness has now been recognized widely as a correlate of children’s reading and spelling abilities. The authors provide a concise |

|overview of what is known about phonological awareness carefully noting that phonological awareness alone is not responsible for reading success. However, they state that without phonological |

|awareness skills, “most of our approaches to reading and spelling are less effective” (Robertson & Salter, 2007a, p. 10). |

| |

|The following section, “Test Description”, provides a list of each subtest identifying the skills being assessed. At the conclusion of the Test Description (Robertson & Salter, 2007a, pp. |

|11-12), the authors offer what they believe are the characteristics of students “whose language disorders are most aptly identified by PAT, namely, difficulty repeating multisyllabic words, |

|over-reliance on visual memory when decoding and encoding, and poor performance on rapid naming tasks” (p. 12). |

| |

|The test kit consists of a sturdy coil bound examiner’s manual, two spiral-bound booklets for the Graphemes and Decoding subtests, plastic blocks for the Substitution subtest, and record forms. |

|A separate statistics manual contains all the standardization, reliability, validity, and related statistics. The examiner’s manual contains dividers for each section as follows: test |

|administration, scoring, discussion of performance, and statistical and normative analyses. A useful glossary of testing terms is included (Robertson & Salter, 2007a, pp. 52-53). |

| |

|The record form contains identifying information and subtest summary scores on the first page. Subsequent pages include age equivalency and standard score profiles, a pronunciation key, |

|well-highlighted (in bold letters) instructions for administering each subtest, and response and scoring items. The layout of the print is easy to read and not too crowded. |

| |

|Comment: Although a pronunciation key is provided, I am likely to use the International Phonetic Alphabet (IPA) to record the child’s responses in the Segmentation of Phonemes subtest. I wonder |

|why the authors did not choose to use IPA, which is familiar and widely-used especially by those likely to use this test (i.e., speech-language pathologists). |

| |

|Purpose of Test: The purpose of this test is to assess phonological processing and phoneme-grapheme correspondence. |

| |

|Areas Tested: The authors state that the subtests are “generally arranged in developmental progression” (Robertson & Salter, 2007a, p.11). The subtests, each briefly described on page 11, are: |

|The Phonological Awareness Subtests: |

|Rhyming: two tasks-discrimination and production, |

|Segmentation: 3 tasks-sentences, syllables, and phonemes*, |

|Isolation: isolate initial, final, and medial* phonemes, |

|Deletion: compounds and syllables, phonemes, |

|Substitution, with manipulatives |

|Blending, syllables and phonemes |

|The Phoneme-Grapheme Subtests: |

|Graphemes, consonants, long and short vowels, consonant blends*, consonant digraphs*, R-controlled vowels*, vowel digraphs*, and diphthongs* |

|Decoding, VC words*, CVC words*, consonant digraphs*, consonant blends*, vowel digraphs*, R-controlled vowels*, CVCe words*, and diphthongs*. |

| |

|*The authors marked these test items as inappropriate for five-year-olds and are to be administered with discretion to that age group. If not administered, these items are marked 0 on the raw |

|score and the subtest score would be 0. |

| |

|Optional subtest: Invented Spelling (no standardized results offered). |

| |

|Areas Tested: |

|Phonological Awareness Segmenting Blending Elision Rhyming |

|Other Phoneme-Grapheme skills (graphemes and decoding) and Invented Spelling. |

| |

|Who Can Administer: |

|The PAT-2 is very specific about the examiner’s qualification. The manual states that, “the test should only be administered by a professional trained in analyzing the phonological structure of |

|speech. For example, a speech-language pathologist, learning disability teacher, reading teacher, or special education consultant” (Robertson & Salter, 2007a, p. 15). The authors further note |

|that these skills are unlikely to be in the skill set of assistants or other support personnel. |

| |

|Administration Time: The administration time is 40 minutes. PAT-2 can be administered over more than one session but individual subtests must be completed during a session. |

|Test Administration (General and Subtests): |

| |

|In the general description of test administration, the authors state that if it is clear that the student is unable to do the task, that task should be discontinued and all subtest items scored |

|as zero. They direct the examiner to use the normative data and scoring procedures with zero scores on subtest, section, and total test results (Robertson & Salter, 2007a, p. 15). |

| |

|Detailed procedures begin on page 19 in the section on Scoring. Administration begins with the Rhyming Discrimination subtest. A demonstration item is presented followed by the test items. All |

|subtests are described in a set format with a brief description of the task followed by details of the procedure, prompts, recording and scoring of responses. |

| |

|Testing begins with the first item for all age groups. There are no basals or ceilings. Tasks not appropriate for the 5 years, 0 months to 5 years, 11 months range are identified on the record |

|form and in the examiner’s manual. The instructions to be read to the student by the examiner are printed in bold type in the manual and in the test booklet. A stimulus phrase is included for |

|each task. Examples of these phrases include: “Tell me a word that rhymes with _____”, “Tell me each sound in _______”, and “What’s the middle sound in the word _____”. Stimulus phrases can be |

|used until it is evident that the student understands the task. In terms of allowable prompts, the examiner is directed to repeat the stimulus. The record form contains a “Pronunciation Key” to |

|guide the examiner in recording the student’s responses. The examiner is encouraged to record the child’s response verbatim. |

| |

|The Substitution subtest introduces the use of blocks to represent each sound in target words. The student is instructed to follow the examiner’s demonstration, and then the test items are |

|introduced. |

| |

|Comment: This is probably an unfamiliar task for many students. I wonder if it introduces other skills/biases. Depending on how the child does, I would be cautious in interpreting performance as|

|I am not clear what else might be tested. Another subtest, Segmentation, introduces clapping with the same concern. It may be that my experience as a preschool clinician limits my view here. |

|Clinicians with more experience with children at the ages targeted by the PAT-2 may find that these tasks are unproblematic for their students. |

| |

|The Graphemes subtest allows for repeating the stimulus phrase “Tell me what sound this makes”. However the authors state that “some clinical judgment may call for further prompting only as |

|noted” in each of the tasks (i.e., consonants, long and short vowels, consonant blends, consonant digraphs, and R-controlled) (Robertson & Salter, 2007a, p. 28). |

| |

|In the Decoding subtest, the examiner instructs the child, “I’m going to show you some made-up words. I want you to read each one to me” (p.30). Only the stimulus phrase can be repeated. No |

|other prompts are allowed. |

| |

|Later, in the Discussion of Performance, the authors caution examiners not to teach students the concepts that are being assessed (Robertson & Salter, 2007a, p. 35). |

| |

|The Invented Spelling subtest is optional that does not include standardized scoring. The procedure involves asking the student to spell some words on a sheet of lined paper which are dictated |

|to the student. The procedure continues until the examiner is satisfied that enough items have been presented to represent the student’s ability. There are no prompts for this subtest. Table 15 |

|of the Statistics Manual is used to compare the student’s performance. This table lists the spelling skill (e.g., initial sounds) and the context (e.g., all words, or examples of individual |

|words with the target underlined as in “shepard”). Stages of spelling are presented in table format as follows: prerepresentational, developmental, representational, and conventional. No |

|reference for this organization of stages is given (Robertson & Salter, 2007b, p. 175). |

| |

| |

|Comment: In general, I wonder what influence accent, dialect, and/or articulation has on both the examiner’s presentation and the student’s responses. However, this concern may be unavoidable |

|unless audio-recoded presentation was introduced. This test is widely used so I doubt that this concern proves to be much of an issue. |

|Test Interpretation: |

| |

|The section, “Discussion of Performance”( Robertson & Salter, 2007a, pp. 35-40), uses a question and answer format to address common questions related to interpreting results. This section of |

|the PAT-2 is unchanged from the original version with the exception that a question regarding socioeconomic effects on performance is addressed. Examples of questions addressed include: |

|How could a student who is at-risk for reading do well on this test? |

|How low does a score need to be before I enroll a student for intervention? |

|My student’s level of performance was below the 25th percentile on the test. What should I do for intervention? |

| |

|In terms of the socioeconomic issue, the authors respond that their standardization data shows that children, ages 7 years, 0 months to 7 years, 5 months, from low SES environments scored lower |

|on 7/8 subtests than children from higher SES levels thus indicating to the authors that “the age range of 7-0 to 7-5 years may be a critical period for the development of phonological awareness|

|for children from the lower socioeconomic (SES) level” (Robertson & Salter, 2007a, p. 37). |

| |

|This Discussion of Performance section also has “Suggestions for Intervention” which includes sample activities at the word, syllable, phoneme, and grapheme levels (Robertson & Salter, 2007a, |

|pp. 37-40). In addition to these brief descriptions of intervention, the authors recommend The Phonological Awareness Kit (Robertson & Salter, 1995a) and the Phonological Awareness |

|Kit-Intermediate Robertson & Salter, 1995b) which provide more comprehensive intervention activities. |

| |

|The next section of the examiner’s manual, “Statistical and Normative Analyses”, contains instructions for compiling the results, information on the standardization study (with sample |

|characteristics), and the interpretation of results with reference to standardized score. With regard to standardized scores, each is defined and described with references to the tables |

|contained in the separate Statistical manual(Robertson & Salter, 2007b). |

| |

|This section of the examiner’s manual also contains the report on reliability and validity studies which will be addressed in this review in the sections which follow. |

| |

|Comment: You could get lost in the Statistical manual. The layout is not very reader/user friendly. Some pages are facing up and some are to be read with the manual turned to the side. There are|

|a total of 175 pages. |

| |

|Comment: The organization of the two sections, “Discussion of Performance”, and “Statistical and Normative Analyses”, is confusing. I would have expected that interpretation of results would be |

|in one section not spread out over two sections. The actual interpretation section is buried in the second of these two sections though the question and answer format of the previous section |

|would lead to the expectation that the first section is where interpretation is found. |

|Standardization: Age equivalent scores Grade equivalent scores Percentiles Standard scores Stanines |

|Other: Profiles of Age Equivalent scores and Standard Scores |

| |

|Reliability: |

| |

|No information is provided on the sample characteristics. Usually test authors provide detailed information about the participants in reliability studies so that the reader is assured that |

|performance was the same across all groups and to allow examiners to compare their local populations. |

| |

|Comment: Interestingly, the authors state that the reliability of the PAT-2 was evidenced by internal consistency and test-retest studies. They separate the inter-rater reliability from these |

|two by grouping the first two in their discussion and then introducing the inter-rater data in the subsequent section. I am not sure what they intended by this separation but it can confuse a |

|novice reader who may be familiar with the grouping together of all three forms of reliability evidence. |

| |

|Internal consistency of items: Item homogeneity was calculated using Kuder-Richardson (KR 20) coefficients. KR20 coefficients are presented by age with averages across all age ranges for each |

|subtest ranging from.58 (Syllables in Blending subtest) to .82 (R-controlled vowels). Subtest totals ranged from .74 (Substitution Total) to .97 (Decoding Total), whereas Phonological Awareness |

|Total was .96 and Phoneme-Grapheme Total was .98. The overall total test consistency score was .99. |

| |

|Test-retest: The test-retest data was collected from 145 students “across ten age ranges” (Robertson & Salter, 2007a, p. 47). The test was re-administered within 14-21 days by the same examiner.|

|Coefficients, presented in Tables 4-1 to 4-5 (Robertson & Salter, 2007b, pp. 51-52), range from an average of .60-.90 for individual subtests with test totals of .90 for Phonological Awareness |

|Total, .92 for Phoneme-Grapheme Total, and .93 for Total Test. |

| |

|Inter-rater: Eight raters, who were speech-language pathologists, rated 11 randomly selected completed test forms. Percentage agreement was reported to range from 95 to 98 with an average of 97.|

| |

|Comment: Missing from the authors’ information is any description of how the raters were recruited or trained or how the eleven test forms were chosen, though presumably they came from the |

|standardization sample, and what age ranges were represented, etc. The use of eleven forms seems like a small pool. Despite this lack of detail, this attempt at inter-rater reliability |

|represents an improvement over the original PAT where no inter-rater data was reported. Although percent agreement makes sense as a way of demonstrating inter-rater reliability, other tests |

|reviewed in this series use reliability coefficients and I am left wondering why the PAT authors did not make such as choice. |

| |

|Other (Please Specify): Standard Error of Measurement (SEM) for test results are described in the examiner’s manual (Robertson & Salter, 2007a, p. 47) and actual age and subtest SEM data are |

|found in table form in the Statistics manual alongside the test-retest reliability coefficient data (Robertson & Salter, 2007b). |

|Sensitivity and specificity data are not given. |

|Validity: |

| |

|Content: Content validity was evidenced qualitatively with the authors’ introduction in which the rationale for developing a test of phonological awareness was presented along with highlight of |

|the research evidence and report of the National Reading Panel. In the section on validity in the Statistical and Normative Analyses, the authors note that all the skills that are necessary for |

|the test domain of phonological awareness are assessed in the PAT-2. Further, they state that the test “was developed following extensive review of available tests and the literature which |

|indicated the particular subtests and skills selected were those reflective of necessary phonological awareness skills of elementary age students” (Robertson & Salter, 2007a, p. 48). |

|Quantitative evidence of content validity such as conventional item analysis is not presented. |

| |

|Comment: Although the authors state that they conducted an extensive review, only a brief summary of the background to phonological awareness containing only two statements with supporting |

|references was presented in the introduction to the PAT-2. The original PAT rationale and literature review spanned only two pages (six paragraphs) and contains only 13 references constituting |

|the shortest rationale I’ve reviewed so far. The new edition contains 35 references. To be fair, what is known about phonological awareness is well established at this point as evidenced by the |

|findings of the National Panel. |

| |

|Comment: Quantitative evidence would strengthen the validity. Item discrimination and item difficulty analysis would have shown how students’ performance could be differentiated. |

| |

|Criterion Prediction Validity: Not reported. |

| |

|Comment: The authors offer contrasted groups validity in their section on criterion-related validity. I find this a bit confusing as the terms criterion-related or criterion prediction validity |

|generally refer to how well a test performs compared to some other established measure of the same phenomenon being tested. Against this definition, the PAT-2 does not offer any evidence for |

|criterion prediction validity. |

| |

|Construct Identification Validity: |

| |

|Group differentiation was addressed in the form of contrasted group validity (Robertson & Salter, 2007a, p. 48) which involved comparing test results of students who participated in the |

|standardization study with a group of students identified as at-risk for reading who were enrolled in special services. Using a matched sample, the authors report that the data, based on t-test |

|results, indicate that the PAT-2 differentiates between the two groups at every level of the test (i.e., subtests, totals, and total test results). The mean test results for subtests and total |

|test scores for at-risk students performed were consistently and significantly lower than those of the normal students. |

| |

|Age differentiation is not addressed in the section on validity. |

| |

|In terms of group difference, identified groups such as minority and gender groups should get similar scores on a test. The PAT-2 examined race and socioeconomic groups in this regard. Using two|

|groups (i.e., race and race/socioeconomic), the authors describe how they analyzed the group differences at the item and subtest levels using z-tests and Chi-square to demonstrate that race was |

|not a factor on the PAT-2. In terms of the race/socioeconomic group, Analysis of Variance F-tests was used to examine mean differences. The authors’ description of the results of these analyses |

|is detailed and I will only summarize their overall findings here by noting that differences in means were statistically significant for race and SES. Refer to pages 49-50 for the authors’ |

|complete discussion of the results. The authors summarize by stating, “Collectively, the results indicate that neither race nor SES has a major impact on The Phonological Awareness Test 2” |

|(Robertson & Salter, 2007a, p. 50). |

| |

|The authors report a trend in scores which indicates that caution should be exercised for the 7 years, 0 months to 7 years, 5 months range where lower scores on 7/8 subtests were found for |

|children at the lower SES level. The authors postulate that this age range may indicate a vulnerable stage in phonological skill development for children living in poverty and that preventative |

|instruction may be valuable. |

| |

|Biserial correlations are presented to evidence subtest, section, and total test intercorrelations (i.e., item validity). These correlations demonstrate that performance on parts of the test are|

|related to performance on the overall test. The results show that 89% of the individual items are statistically significantly correlated with the subtest scores. In this way, the PAT-2 |

|demonstrates that the test items measure separate aspects of phonological awareness but also contribute to the main construct. |

| |

|Comment: Construct identification validity evidence is also measured by exploratory factor analysis and confirmatory factor analysis using data from the normative sample. These analyses allow |

|the authors to demonstrate that the traits which are tested reflect the model of phonological awareness. The PAT-2 does not offer evidence in this way though presumably the raw data could have |

|been analyzed in this way. Instead, the authors have chosen to use biserial correlations, an older method of evidencing validity. |

| |

|Differential Item Functioning: no information |

|Other (Please Specify): Comment: Although the use of t-tests z-tests and ANOVA make intuitive sense to me, I am not sure whether the analysis proves rigorous enough given that most other tests I|

|reviewed used factor analyses. |

|Summary/Conclusions/Observations: |

| |

|I begin with a note on the original PAT. The description of reliability and validity studies in the original PAT was very slim and confusing. So much information was missing that it was |

|difficult to judge what the data presented meant and how the test results could be used. This may reflect some of the inadequacies from the original normative sample. This concern, along with |

|the concerns about the sample and reliability of the original lead me to recommend that the new PAT-2 should be used clinically even though testing standards would allow us to continue using the|

|original PAT for an overlapping period of two years. |

|Clinical/Diagnostic Usefulness: |

| |

|I think that the PAT-2 will be popular with clinicians because the tasks are familiar and make sense given what we now know about phonological awareness. This is an easy test to administer, the |

|scoring is clearly stated, and the record form is well laid out and easy to use. A proficient clinician can administer this test in a short period of time. These attractive features should be |

|balanced with the concerns stated above regarding the sample and psychometric information. |

References

National Reading Panel (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Rockville, MD: National Institute of Child Health and Human Development.

Robertson, C. & Salter, W. (1995a). The Phonological Awareness Kit. East Moline, IL: Linguisystems.

Robertson, C. & Salter, W. (1995b). The Phonological Awareness Kit: Intermediate. East Moline, IL: Linguisystems.

Robertson, C. & Salter, W. (2007a). Examiner’s manual. Phonological awareness test 2. East Moline, IL: Linguisystems.

Robertson, C. & Salter, W. (2007b). Statistics manual. Phonological awareness test 2. East Moline, IL: Linguisystems.

To cite this document:

Hayward, D. V., Stewart, G. E., Phillips, L. M., Norris, S. P., & Lovell, M. A. (2008). Test review: Phonological awareness test-2 (PAT-2). Language, Phonological Awareness, and Reading Test Directory (pp. 1-12). Edmonton, AB: Canadian Centre for Research on Literacy. Retrieved [insert date] from .

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download