Blanca, Maria J. , Alarcon, Rafael , Arnau, Jaume , GARCÍA DE CASTRO, FERNANDO JAVIER, Bono, Roser
No
An. Psicol.
Article
Científica
01/10/2024
001284216400012
2-s2.0-85200772878
Adjusted F tests have been proposed as an alternative to the F statistic in repeated measures ANOVA. Despite the existence of previous research, there is a lack of evidence on the behavior of these statistics in case of simultaneous violation of normality and sphericity. The objective of the present work has been to carry out a detailed examination of the type I error and the power of the F statistic and the Greenhouse-Geisser ( F-GG ) and Huynh-Feldt ( F-HF ) adjustments , manipulating the number of repeated measurements. (3-6), the sample size (10-300), the sphericity (Greenhouse-Geisser estimator of epsilon ', from its lower limit to the upper limit), and the shape of the distribution (from slight to extreme deviations from normality ). The results show that the behavior of F-GG and F-HF depends on the degree of violation of normality, sphericity and sample size. In general, it is suggested to use F-GG in case of violation of sphericity and mild or moderate deviations from normality; with serious deviations from both, F-GG can be used with a sample size greater than 10; and with extreme deviations, this statistic can be used with a sample size greater than 30. In case of discrepant results between F-GG and F-HF , the choice depends on the epsilon ' value.
Greenhouse-Geisser adjustment; Huynh-Feldt adjustment; Robustness; Power; Monte Carlo simulation