ERIC Number: ED367697
Record Type: RIE
Publication Date: 1993-Dec
Reference Count: N/A
A Multivariate Generalizability Analysis of the 1989 and 1990 AAP Mathematics Test Forms with Respect to the Table of Specifications.
Colton, Dean A.
Tables of specifications are used to guide test developers in sampling items and maintaining consistency from form to form. This paper is a generalizability study of the American College Testing Program (ACT) Achievement Program Mathematics Test (AAP), with the content areas of the table of specifications representing multiple dependent variables. The results are presented with respect to variance and covariance components and to estimates of error variance. Also discussed are alternative weightings of the content categories. The pattern of mean difficulty of the four content areas is the same for 1989 and 1990 data. It appears that the specifications for the mathematics examination are resulting in fairly consistent levels of difficulty at both the test and content strata levels, although the form-level standard deviations for trigonometry are generally smaller than the standard deviations of the other content areas. In general, examination of the variance components leads to the conclusion that universe scores, item effects, and residuals make fairly consistent form-to-form contributions to the variability of scores in the content areas with relatively large proportions of the test's items. Thirteen tables and three figures are included. (Contains 42 references.) (Author/SLD)
Descriptors: Achievement Tests, Difficulty Level, Error of Measurement, Generalizability Theory, High Schools, Item Analysis, Mathematics Achievement, Mathematics Tests, Multivariate Analysis, Test Construction, Test Format, Test Items
ACT Research Report Series, P.O. Box 168, Iowa City, IA 52243.
Publication Type: Information Analyses; Reports - Evaluative
Education Level: N/A
Authoring Institution: American Coll. Testing Program, Iowa City, IA.
Identifiers: ACT Assessment; Table of Specifications; Variance (Statistical)