The relationship between variables in experimental research is often investigated through the use of extreme group designs. Alf and Abrahams (1975) have described three statistical strategies and their merits for estimating the relationships in such designs. This Monte Carlo study is designed to investigate empirically the efficacy of these procedures in terms of their power and mean square error. Alf and Abrahams' "covariance information statistic" (Strategy 2) proved to be the best in terms of estimation of correlation, power and mean square error for the correlation.
Get full access to this article
View all access options for this article.
References
1.
Abrahams, N. M. and Alf, E. F., Jr.Relative costs and statistical power in the extreme groups approach. Psychometrika, 1978, 43, 11-17.
2.
Alf, E. F., Jr. and Abrahams, N. M.The use of extreme groups in assessing relationships. Psychometrika, 1975, 40, 563-572.
3.
Feldt, L. S.The use of extreme groups to test for the presence of a relationship . Psychometrika, 1961, 26, 307-316.
4.
McNemar, Q.At random: Sense and nonsense . American Psychologist, 1960, 15, 295-300.
