The purpose of this study was to conduct a critical statistical analysis of the Iowa Silent Reading Advanced Tests, Form Cm. This battery of nine subtests has been designed for the diagnosis of the reading ability of students from Grade 9 to junior college.
During the months of November, December and January of the academic year 1947-48, the test was administered to a total of 433 students in ten sections of English 205 at the University of British Columbia. The data from sixteen students who were unable to complete the test were rejected.
(1) The mean difficulty, expressed as the [formula omitted] x 100, of the
subtests ranged from 47 to 79%. That of four tests, 4, Word Meaning, 5, Sentence Meaning, 6, Paragraph Comprehension, and 7B, Selection of Key Words, ranged from 47 to 79 percent. The distribution of raw scores on subtests 5,6, and 7B, was determined to be markedly negatively skewed.
(2) Subtest standard score equivalents for the subtest raw scores have been published by the test authors. With the present group, these scores for subtests 4, 5, 6, and 7A (Use of Index) were not directly comparable with those of the remaining subtests.
(3) Difficulty of items in all subtests were ranged from approximately 10 percent to 99 percent passing. In three subtests, 4, 5, and 6, over 40 percent of theitems were passed by 90 percent of the group. With the exception of Part A of subtest 1C (Comprehension), items were arranged in order of difficulty for the group.
(4) As an expression of item validity, phi coefficients were determined for each item, with the subtest scares as criteria. In spite of the lack of difficulty of many items for the group, most items, with the exception of the first half of subtest 5, correlated significantly with the subtest scores.
(5) Reliabilities of the subtest and median scores, estimated by the Kuder-Richardson formula No.20, ranged from .725 to .916 for the subtest scores; while that of the test median was estimated as .955. Only subtests 2 (Directed Reading), 4, 5 and 7B might be considered sufficiently reliable for individual diagnosis.
(6) Factor analysis by Thurstone's Centroid Method revealed three common factors, accounting for 34.3, 6.7 and 4.1 percent of the variance of the subtests respectively. In subtests 1C, 5 and 6 the variance due to the first factor exceeded 40 percent. Variance due to specific factors in subtests 1R, 2, 3, 7A and 7B exceeded that due to common factors.
(7) To study the validity and predictive value of the Iowa tests, coefficients of correlation were determined between both subtest and median scores and final grades in five second year subjects, English, Economics, Geography, Mathematics and Accounting. These coefficients ranged from -.03 to + .45. With average final grades in second year pharmacy (N=47) coefficients of correlation of test scores ranged from .28 to .61. The subtests tended to correlate more highly with grades in those courses which required more reading.
(8) Coefficients of correlation between both subtest and median scores and the Otis S.A.Test of Mental Ability, Higher Form A, administered in the fall of 1946, for a sample of 105, ranged from -.07 to + .71. There is some support for the hypothesis that the relationship between the scores on the two tests may be due to the nature of the common factor of the Iowa tests as revealed by factor analysis. / Arts, Faculty of / Psychology, Department of / Graduate
Identifer | oai:union.ndltd.org:UBC/oai:circle.library.ubc.ca:2429/41250 |
Date | January 1949 |
Creators | Butler, Alfred James |
Publisher | University of British Columbia |
Source Sets | University of British Columbia |
Language | English |
Detected Language | English |
Type | Text, Thesis/Dissertation |
Rights | For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use. |
Page generated in 0.0022 seconds