This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by McQuitty, L. L. Search for Related Content Social Bookmarking                         What's this?

Comprehensive Analysis of Test Item Variance

University of Miami Coral Gables, Florida

This paper develops and illustrates a comprehensive method for analyzing the endorsements to test items (called CATIV). It (1) accounts for whatever kind of variance is present; (2) is analogous to analysis of variance, but is applied to test-item responses; (3) can analyze many more variables than analysis of variance; (4) shows which test items are additive and which are configural; (5) includes a criterion for identifying test-item answers which belong to a universe of answers other than the one dominant in the study; (6) selects patterns of answers which yield substantive types; (7) assesses the relative dependability of the types which are isolated; (8) isolates all substantive types present in the data; and (9) can be applied by pencil and paper to a column matrix which shows the number of endorsements of every answer to every item of a test, provided these are arranged in order from the most popular to the least popular answer in terms of number of endorsements.

Educational and Psychological Measurement, Vol. 36, No. 1, 51-84 (1976)
DOI: 10.1177/001316447603600106


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?