|
Since 2002, 4 million visitors plus:
| |||||||
|
|
ORAL AND WRITTEN LANGUAGE SCALES John O. Willis The Oral and Written Language Scales (OWLS; Carrow-Woolfolk, 1996) includes a Written Language Scale designed to assess writing skills. It is has norms for ages 5-0 through 21-11 and fall and spring norms from the spring of kindergarten through the spring of grade 12 and takes about 20 minutes (10 to 30) to administer. In addition to the total score, the OWLS Written Expression provides percentile ranks for nine special Skills Areas categories in the domains of Conventions, Linguistics, and Content. The child is administered one of four overlapping blocks of items (1-20, 12-24, 21-32, or 25-39) depending on age. However, the use of Rasch-type Ability Scores for the blocks allows out-of-level testing for children with unusually weak or strong writing skills. A maximum of 70 to 82 raw score points is available for each block, permitting considerable bottom and top for each examinee. There is a variety of item types, including copying printed words and sentences; writing letters, words, and sentences from dictation; and writing sentences and paragraphs according to specific, oral instructions. The OWLS Written Expression scale may be administered in small groups "with examinees 8 years and older who are being assessed for reasons other than placement decisions" (Carrow-Woolfolk, 1996, p. 33), as was done in some cases during the standardization. Scoring Categories The OWLS Written Expression Scale does not have subtests, but does provide reproducible Descriptive Analysis Worksheets permitting calculation of percentile ranks and determination of strengths and weaknesses for nine of the 15 Skills Areas at each year of age. Conventions
Scores Each item is scored on the basis of one to nine very explicit, completely independent scoring rules in the various Skills Areas so a particular response might, for example, receive points for meaningful content, details, and supporting ideas, but not for spelling and capitalization/ punctuation. Not all scoring criteria require perfection (e.g., "No more than one incorrect word."). Each of the several criteria for each item has a maximum possible score of one to four points. Scoring criteria for each item are listed on the record form and explained, with many examples, in the Manual. Further helpful information is given in the Glossary. The raw score for an item set is converted to a Rasch-type Ability Score with 68%, 90%, or 95% confidence limits, which can be converted to standard scores (M = 100, SD =15), percentile ranks, normal curve equivalents, stanines, and grade- and age-equivalent scores. Age-based norms are provided at three-month intervals for ages 5-0 through 8-11, at four-month intervals for ages 9-0 through 13-11, six-month intervals for ages 14-0 through 18-11, and 12-month intervals for ages 19-0 through 21-11. Grade-based norms are provided at half-year intervals from the spring of kindergarten through the spring of twelfth grade. Standardization The standardization group contained a representative national sample of 1,373 students stratified to match the U.S. census data for 1991 on the basis of age, sex, and four categories each of mother's education, race/ethnicity, and geographic region. An additional 185 students with language impairments, mental handicaps, learning disabilities, hearing losses, and reading delays were tested for clinical validity studies. There were 115 to 124 children at each year of age from 5 through 11, and average of 65 per year from 12 to 15, and an average of 42 per year from 16 through 21. There were 111 to 159 children per grade from kindergarten through grade 6 and an average of 51 children per grade for grades 7 through 12. Reliability Internal consistency reliabilities range from .77 to .94 (Mdn rxx = .87). Test-retest reliabilities over 18- to 165-day intervals for a sample of 84 examinees were .66 with a mean gain of 0.0 standard score points for ages 8 through 10 and .83 with a mean loss of 1.3 points for ages 16 to 18. [Corrected for the variability of the norm group (SD = 15), the reliabilities were .88 and .87, respectively.] Interrater reliability for four raters scoring 60 protocols ranged from .91 to .98 at various ages (Mdn rxx = .94). Validity Construct validity of the OWLS Written Expression Scale is based on extensive development efforts to match the content and format of the test to language theory (e.g., Carrow-Woolfolk, 1988, 1996; Carrow-Woolfolk and Lynch, 1981). The correlations with other writing tests reported in the Manual were .67 with the Kaufman Test of Educational Achievement (K-TEA) Comprehensive Form Spelling (n = 31), .78 with the Peabody Individual Achievement Test-Revised (PIAT-R) Spelling (n = 31), and .77 with the PIAT-R Written Language Composite (n = 31). [Corrected for the variability of the norm group, these were .82, .73, and .71, respectively.] Correlations with total reading scores were .75 with the K-TEA, .84 with the PIAT-R, and .80 with the WRMT-R (n = 29). [Corrected for the variability of the norm group, these were .86, .80, and .87, respectively.] Correlations with oral language tests were .57 with the PPVT-R (n = 100), .74 with the Clinical Evaluation of Language Fundamentals-Revised (CELF-R), [Corrected for the variability of the norm group, these were .62 and .79, respectively.] .57 with the OWLS Listening Comprehension Scale (n=1,364), .66 with the OWLS Oral Expression, and .67 with the OWLS Oral Composite. Correlations with intelligence tests included .61 with the Wechsler Intelligence Scale for Children, 3rd ed. (WISC-III) Verbal IQ, .51 with the Performance (nonverbal) IQ, and .59 with the Full Scale IQ (n = 34). [Corrected for the variability of the norm group, these were .72, .64 and .70, respectively.]. Correlations with the Kaufman Brief Intelligence Test (K-BIT) were .62 with Vocabulary, .32 with Matrices, and .52 with the Composite (n = 62). [Corrected for the variability of the norm group, these were .67, . 41, and .58, respectively.] Students in the clinical samples with language impairments, mental handicaps, and learning disabilities (reading) and learning disabilities (undifferentiated) groups all scored significantly lower than students in matched control groups. Students in a Chapter One Reading Program scored lower (M = 90.7, SD = 11.7) than the norming sample. Comment on the OWLS Written Expression Scale The OWLS Written Expression Scale offers a brief, but comprehensive assessment of writing skills. The use of multiple, direct and indirect items rather than a single writing sample does not penalize students who write only a few words when given fifteen or twenty minutes to write a story or a letter to a friend. For students who are willing and able to write a longer sample, you can supplement the OWLS with one of the various story- or letter-writing subtests or with an informal writing sample, e.g., "You have been asked to give a graduation speech at a teachers' college. Your speech will be the last advice these college graduates will hear before they go out to teach students like you for the next 40 years. Please give them the most important advice you think they need. You may use this paper to write an outline, web, or notes before you begin." The 241-page Manual is clear, explicit, and helpful. After a little practice, scoring and interpretation quickly become efficient. The OWLS Written Expression Scale can be purchased and used alone, but it can also be used with the OWLS Listening Comprehension and Oral Expression Scales, which are brief, but very useful instruments a psychologist can use to assess the impact of oral language difficulties on the student's functioning and to decide whether to make a referral for an in-depth language assessment. The OWLS Manuals include data on statistical significance and base rates of differences among the three scales.
|