A method of analyzing M response measures gathered in a simple analysis of variance design has been suggested. The first step is a factor analysis of the M by M total correlation matrix with rotation to “simple structure” and the extraction of Q factors with uncorrelated factor scores. Secondly, separate analyses of variance are performed on each of the Q sets of factor scores. Some advantages are knowledge of the structure of the response space, economy of description, and orthogonality of response factor measurements.
Get full access to this article
View all access options for this article.
References
1.
BrunswikE.Perception and the representative design of psychological experiments. Berkeley: Univer. of California Press, 1947.
2.
BurtC.Factor analysis and analysis of variance. Brit. J. Psychol., Stat. Sect., 1947, 1, 3–26.
3.
KoganW. S.BoeE. E.GockaE. F.JohnsonM. H.Personally changes in psychiatric residents daring training. J. Psychol., 1966, 62, 229–240.
4.
PearsonK.On the laws of inheritance in man. Biometrika, 1904, 3, 131–190.
5.
Student (GossetW. S.). On the probable error of a mean. Biometrika, 1908, 6, 1–25.
6.
ThurstoneL. L.Multiple factor analysis. Chicago: Univer. of Chicago Press, 1947.
7.
ZinnerL.The consistency of behavior in various situations: A methodological application of functional ecological psychology. Unpublished Ph.D. dissertation, Univer. of Houston, 1962.
