ERIC Number: ED439145
Record Type: Non-Journal
Publication Date: 1999-Apr-22
Reference Count: N/A
Measurement Models for a Testlet-Based Test.
Lee, Guemin; Dunbar, Stephen B.; Frisbie, David A.
It has been shown that the fundamental assumptions associated with conventional one-factor measurement models are frequently violated in analyses of scores from a test composed of testlets. Eight different measurement models were conceptualized for this kind of situation, and the goodness of fit of each model was examined. Measurement models incorporating correlated errors appear to be more appropriate than conventional measurement models with uncorrelated error specifications when testlets are involved. Also, the congeneric assumption about part tests or items appears to be more plausible than the essentially tau-equivalent assumption for a test composed of testlets. The one-factor congeneric model with correlated error specifications would be the best measurement model for a test composed of testlets if dichotomously-scored items are used as unit of analysis. However, a congeneric model using passage (testlet) scores can be considered as an alternative for a test composed of testlets when passage (testlet) scores are used as the unit of analysis. (Contains 6 tables and 19 references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A