Related Experiment Videos
Testing discrepancy effects: a critique, a suggestion, and an illustration
Miron Zuckerman1, Marylène Gagné, Iris Nafshi
1Department of Clinical and Social Sciences in Psychology, University of Rochester, Rochester, NY 14627, USA. miron@psych.rochester.edu
Behavior Research Methods, Instruments, & Computers : a Journal of the Psychonomic Society, Inc
|October 25, 2002
Summary
Difference scores can be misleading in research. Interaction analysis is a better method for assessing the discrepancy between independent variables, offering clearer insights into their effects on a criterion variable.
Area of Science:
- Social Psychology
- Quantitative Research Methods
Background:
- Researchers often use difference scores to analyze discrepancies between two independent variables.
- Correlating difference scores with a criterion variable is a common, yet potentially flawed, analytical approach.
Purpose of the Study:
- To evaluate the limitations of using difference scores in psychological research.
- To highlight the advantages of interaction analysis for assessing variable discrepancies.
Main Methods:
- The study critically examines the statistical underpinnings of difference score correlations.
- It contrasts difference score analysis with interaction testing using social psychology research examples.
Main Results:
- Correlation of difference scores with a criterion is explained by the constituents' correlations and variances.
- Difference score analysis can be misleading when predictions extend beyond main effects.
Conclusions:
- Interaction analysis provides a more accurate method for assessing the effects of discrepancies between independent variables.
- Researchers should consider interaction testing over difference scores for nuanced predictions in social psychology.