Alternate Form Reliability Coefficients Research Articles

Technical Characteristics of Curriculum-Based Measurement With Students Who Are Deaf.

This study compared the reliability and validity of student scores from paper-pencil and e-based assessments using the "maze" and "silent reading fluency" (SRF) tasks. Forty students who were deaf and hard of hearing and reading between the second and fifth grade reading levels and their teachers (n = 21) participated. For maze, alternate form reliability coefficients obtained from correct scores and correct scores adjusted for guessing ranged from r = .61 to .84 (ps<.01); criterion-related validity coefficients ranged from r = .33 to .67 (most ps < .01). For SRF, reliability coefficients obtained from correct scores ranged from r = .50 to .75 (ps<.01); validity ranged from r = .25 to .72. Differences between student performance on paper-pencil and e-based conditions were generally non-significant for maze; significant differences between conditions for SRF favored the paper-pencil condition. Findings suggest that maze holds promise, with inconclusive results for SRF.

CBM maze-scores as indicators of reading level and growth for seventh-grade students

The technical adequacy of CBM maze-scores as indicators of reading level and growth for seventh-grade secondary-school students was examined. Participants were 452 Dutch students who completed weekly maze measures over a period of 23 weeks. Criterion measures were school level, dyslexia status, scores and growth on a standardized reading test. Results supported the technical adequacy of maze scores as indicators of reading level and growth. Alternate-form reliability coefficients were significant and intermediate to high. Mean maze scores showed significant increase over time, students’ growth trajectories differed, and students’ initial performance levels (intercepts) and growth rates (slopes) were not correlated. Maze reading level and growth were related to reading level and/or growth on criterion measures. A nonlinear model provided a better fit for the data than a linear model. Implications for use of CBM maze-scores for data-based decision-making are discussed.

Curriculum-Based Measurement in Science Learning

The technical adequacy of curriculum-based measures in the form of short and simple vocabulary-matching probes to predict students’ performance and progress in science at the secondary level was investigated. Participants were 198 seventh-grade students from 10 science classrooms. Curriculum-based measurements (CBM) were 5-min vocabulary-matching probes administered once weekly over a period of 14 weeks. Criterion measures were knowledge pre- and posttests, the science subtest of the Iowa Test of Basic Skills (ITBS), and course grades. Alternate-form reliability coefficients ranged from r = .64 to .84. Coefficients increased over time and by combining scores across probes. Correlations between scores on the vocabulary-matching and criterion measures ranged from r = .55 to .76. The estimated mean group growth rate on the vocabulary-matching measure was .63 correct matches per week, which was significantly different from zero. The measures produced significant interindividual differences in growth rates, and growth on the measures was related to performance on the ITBS, course grades, and pre–post gains on the knowledge test. Results provide initial support for the technical adequacy of vocabulary-matching as an indicator of performance and progress in science.

Silent reading fluency using underlining: Evidence for an alternative method of assessment

AbstractAssessing silent reading fluency in classroom environments is challenging. This article reports on a method of assessing silent reading using underlining, an approach that solves many problems other silent reading fluency assessment measures face. This method computationally monitors readers' silent reading fluency by the speed they underline words in a text. Traditional silent reading fluency measures were compared with the new underlining methodology. Fourth‐ and sixth‐grade students completed silent reading fluency measures (i.e., moving windows, underlining, and paper‐and‐pencil reading tasks), along with measures of their oral reading fluency, reading comprehension, and vocabulary knowledge. Strong alternate‐form reliability coefficients were found for underlining, which significantly correlated with other measures of silent reading fluency and reading comprehension. Underlining methodology also correlated with common language factors, such as word length and word frequency. Together, these results provide support for the psychometric properties of underlining and suggest that it is a promising alternative method of assessing silent reading fluency. © 2012 Wiley Periodicals, Inc.

Curriculum-based measures of writing across grade levels

This study compared the technical adequacy of curriculum-based measures in written expression across diverse grade levels. Students ( n = 484) completed two narrative writing samples in response to story starters. Samples were scored for the number of words written (TW), the number of correct word sequences (CWS), and the number of correct minus incorrect word sequences (CWS-ICWS). Alternate-form reliability and criterion-related validity coefficients of the curriculum-based measures were examined to determine differences between the coefficients according to grade level. Results indicated moderate to strong alternate-form reliability coefficients and variable criterion-related validity coefficients. The data from this study support the need for curriculum-based measures of writing to become more complex as students age and become more skilled.

Consistency of interform content for the Gordon Test of Visual Imagery Control.

A total of 116 undergraduates (49 men, 67 women) completed the Gordon Test of Visual Imagery Control followed immediately by another form. Scores were higher for women than for men and on the original than on the new version. Alternate-form reliability coefficients were .517 for the complete sample, .712 for men, and .210 for women. It was concluded that consistency across content was lower than desirable.

CONSISTENCY OF INTERFORM CONTENT FOR THE GORDON TEST OF VISUAL IMAGERY CONTROL

A total of 116 undergraduates (49 men, 67 women) completed the Gordon Test of Visual Imagery Control followed immediately by another form. Scores were higher for women than for men and on the original than on the new version. Alternate-form reliability coefficients were .517 for the complete sample, .712 for men, and .210 for women. It was concluded that consistency across content was lower than desirable.

Alternate form of the california verbal learning test: Development and reliability

Abstract An alternate form of the California Verbal Learning Test (CVLT) was developed and compared with the original form. The two CVLT forms yielded equivalent mean scores for all 19 learning and memory variables analyzed. No significant practice effects were found regardless of the order of administration of the two forms. The two forms yielded almost identical relationships between overall memory performance and age and education. Sixteen of the CVLT variables resulted in significant alternate form reliability coefficients. The coefficients for the traditional recall measures of the CVLT were particularly robust and higher than those reported for other, commonly used clinical memory tests.

Alternate forms of the AVLT: A procedure and test of form equivalency

The Rey Auditory Verbal Learning Test is frequently used in neuropsychological evaluation and research. However, its utility in the measurement of progressive change is limited by the availability of alternate and equivalent forms. Criteria were developed for word selection to generate new lists. Two alternate forms generated according to these criteria, as well as the original Key AVLT and alternate form, were administered to elderly Medical Center patients and undergraduate volunteers. All four AVLT forms yielded comparable mean recall scores, and alternate form reliability coefficients for each trial varied from.67 to.90. Conclusions from this study, though, must be made with caution as the sample size used was small.

Alternate form reliability and equivalency of the Rey Auditory Verbal Learning Test.

Assessed alternate form reliability and equivalency for the Rey Auditory Verbal Learning Test (AVLT) in a clinical sample. A test-retest, counterbalanced design was utilized with a diagnostically heterogenous group of 85 VA Medical Center patients. The mean test-retest interval was 140 min. Alternate form reliability coefficients were highly significant, all p less than .001, and ranged from .60 to .77. The forms yielded comparable means with differences of less than 1 point on each of the five learning trails and the postinterference (VI) and recognition trials. Total words recalled on trials I through V differed by less than 3 points across forms. From a statistical point of view, when the alternate form was administered second it was slightly more difficult than the original. In the reverse order, the two measures were comparable. Overall, differences between forms lacked clinical significance and the tests were judged to be equivalent measures.

A correlational analysis between the PPVT-R and Woodcock-Johnson achievement cluster scores for nonreferred regular education and learning disabled students

Both forms of the PPVT-R, along with the reading, mathematics, and written-language subtests of the Woodcock-Johnson Psycho-Educational Test Battery were administered concurrently and in counterbalanced order to 28 nonreferred regular education and learning disabled students. With the exception of correlations between form M and the reading and written-language subtests for the LD group, values did not reach significance. Alternate form reliability coefficients for the RE and LD groups were .70 and .65, respectively. Mean standard scores for forms L and M were comparable for the RE group, but form L yielded a significantly higher score than M for the LD population.

Reliability, homogeneity and the construct validity of Cattell's 16PF

This paper reviews criticisms of the psychometric properties of Cattell's Sixteen Personality Factor Questionnaire and examines Cattell's claims that 16PF scale reliabilities and factor validities are seriously underestimated by conventional internal consistency approaches to reliability estimation. On large samples of British adults and undergraduates no evidence was found to support the claim that the 16PF is deliberately heterogeneous in item content. No presence of suppressor effects or balancing of specifics were found in the 16PF scales, most of which were reasonably homogenous by conventional standards. Conventional internal-consistency reliability estimates in fact give very similar results to alternate form reliability coefficients and Cattell's factorial validities. 16PF scales which were particularly poor in terms of scale consistency were L, M and N with C, O and Q4 emerging as measures of essentially the same trait of anxiety.

Alternate Form Reliability of the Peabody Picture Vocabulary Test

Dunn ( 1 ) stated that coefficients of equivalance for Forms A and B of the Peabody Picture Vocabulary Test appeared to be satisfactory for both average and mentally retarded children. The present study sought to determine alternate form reliability coefficients for the Peabody. Data were collected from 317 Negro males, 241 white males, 161 Negro females, and 151 white females ( N .= 870), ages 6 through 15 yr. and Grades 1 through 10. The children were performing poorly in school, and, as a result, were being considered for placement in classes for the educable mentally retarded. The children were from 37 integrated public schools in South Central and Southeast Alabama. Forms A and B of the Peabody were administered individually in counterbalanced order to all subjects by the same white male examiner. Mean IQs and standard deviations on the Peabody Form B were as follows: for Negro subjects, M = 67.92; SD = 11.67; for white subjects, M = 77.33; SD = 15.27; for male subjects, M = 73.60; SD = 13.83; for female subjects, M = 69.66; SD = 14.83; and for all subjects, M = 71.93; SD = 14.69. On Form A of the Peabody, mean IQs and standard deviations were as follows: for Negro subjects, M = 68.63; SD = 11.28; for white subjects, M = 77.21; SD = 14.30; for male subjects, M = 74.23; SD = 13.95; for female subjects, M = 71.71; SD = 13.69; and for the total sample, M = 73.16; SD = 14.32. Mean IQs on Form B were significantly lower ( p 6 .05) than Form A mean IQs for Negroes ( t = -2.12), whites ( t = -3.54), males ( t = -2.57), females ( t = -3.35), and for the total sample ( t = -4.21). Pearson correlation coefficients berween Form A and Form B of the Peabody were as follows: for Negroes, .85; for whites, .75; for males, 3 8 ; for females, .72; and for the total sample, .83. All rs were significant ( p L .001). For this sample, Form A of the Peabody yielded significantly higher mean IQs than did Form B. Coefficients of equivalence were comparable to others': .72 to .84 ( I ) , .84 ( 3 ) , and .86 ( 4 ) , but higher than .61 ( 2 ) .