A Correction for the ε̃ Approximate Test in Repeated Measures Designs With Two or More Independent Groups
- 1 December 1991
- journal article
- Published by American Educational Research Association (AERA) in Journal of Educational Statistics
- Vol. 16 (4) , 371-372
- https://doi.org/10.3102/10769986016004371
Abstract
The routine ε̃ approximate test procedure in repeated measures designs when the condition of circularity is not fulfilled uses an erroneous formula in the case of two or more groups. This may lead to a substantial underestimation of the deviation from circularity when the total number of subjects is small.Keywords
This publication has 3 references indexed in Scilit:
- Some Approximate Tests for Repeated Measurement DesignsPsychometrika, 1978
- Estimation of the Box Correction for Degrees of Freedom from Sample Data in Randomized Block and Split-Plot DesignsJournal of Educational Statistics, 1976
- Estimation of the Box Correction for Degrees of Freedom from Sample Data in Randomized Block and Split-Plot DesignsJournal of Educational Statistics, 1976