Darrow and Kahl's (1982) paper on detecting moderator variables is misleading because it contains both logical and methodological problems. The present paper was written to respond to two of those problems: their distinction between pure and mixed moderators and their choice of model for testing for moderator effects. In this paper we show that the distinction between pure and mixed models is uninformative and that the standard hierarchical regression model represents the appropriate method for identifying moderator variables.
Get full access to this article
View all access options for this article.
References
1.
Blood, M.R., & Mullet, G.M. (1977). Where have all the moderators gone? The perils of Type II error. (Tech. Rep. No. 1). Atlanta: Georgia Institute of Technology, College of Industrial Management.
2.
Cohen, J. (1978). Partialed products are interactions; Partialed powers are curve components. Psychological Bulletin, 85, 858-866.
3.
Cohen, J., & Cohen, P. (1975). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.
4.
Darrow, A.L., & Kahl, D.R. (1982). A comparison of moderated regression techniques considering strength of effects. Journal of Management, 8(2), 35-47.
5.
Kerlinger, F.N., & Pedhazur, E.J. (1973). Multiple regression in behavioral research. New York: Holt, Rinehart, & Winston.
6.
McNeil, K.A., Kelly, F.J., & McNeil, J.T. (1975). Testing research hypotheses using multiple linear regression. Carbondale, IL: Southern Illinois University Press.
7.
Peters, L.H., O'Connor, E.J., & Wise, S.L. (in press). The specification and testing of useful moderator variable hypotheses. In G.D. Ferris & T. Bateman (Eds.), Method and analysis in organizational research. Reston, VA: Reston Press.
8.
Saunders, D.R. (1956). Moderator variables in prediction. Educational and Psychological Measurement, 16, 209-222.
9.
Winer, B.J. (1971). Statistical principles in experimental design. New York: McGraw-Hill.
