A New Look at Treatment Differences

A New Look at Treatment Differences

The analysis of dichotomous item response data in a repeated measures format is considered. Earlier results which indicate that, with regard to Type I error, these test procedures have much practical merit are cited. Results in this paper indicate that power computations based on the non-central F d...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:American educational research journal 1972-01, Vol.9 (2), p.311-321
1. Verfasser: Mandeville, Garrett K.
Format: Artikel
Sprache:eng
Schlagworte:
Analysis of variance
Binary data
Covariance matrices
Curricula
Data models
Educational research
Mathematical independent variables
Modeling
Ratios
Statistical variance
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The analysis of dichotomous item response data in a repeated measures format is considered. Earlier results which indicate that, with regard to Type I error, these test procedures have much practical merit are cited. Results in this paper indicate that power computations based on the non-central F distribution are also in good agreement with empirical power estimates based on Monte Carlo simulation.
ISSN:0002-8312
1935-1011
DOI:10.3102/00028312009002311