You are not currently logged in.
Access JSTOR through your library or other institution:
Methodological Issues in Estimating Main and Interactive Effects: Examples from Coping/Social Support and Stress Field
John W. Finney, Roger E. Mitchell, Ruth C. Cronkite and Rudolph H. Moos
Journal of Health and Social Behavior
Vol. 25, No. 1 (Mar., 1984), pp. 85-98
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2136706
Page Count: 14
You can always find the topics here!Topics: Regression analysis, Analytical estimating, Psychological stress, Stress functions, Modeling, Social behavior, Depressive disorders, Moderator variables, Statistical models, Covariance
Were these topics helpful?See somethings inaccurate? Let us know!
Select the topics that are inaccurate.
Preview not available
Although used increasingly frequently, statistical analyses to explore moderating as well as main effects have not always been well-understood by researchers in the stress and coping/support field. We attempt to clarify three areas of potential confusion. First, three approaches for estimating main effects in multiple regressions with significant interaction effects and two methods for estimating main effects when a test for interaction has not reached statistical significance are evaluated. Second, multiple within-groups analysis is explored as an alternative to product-term regression analysis for probing interaction effects. Third, stress-attenuation analysis is distinguished from analyses of stress-moderating effects by locating both within the general framework of the elaboration paradigm.
Journal of Health and Social Behavior © 1984 American Sociological Association