The mean as a multilevel issue

David Trafimow

doi:10.3389/fpsyg.2014.00180

Abstract

The mean as a multilevel issue

Highlights

Edited by: Jeremy Miles, Research and Development Corporation, USA Reviewed by: Craig Speelman, Edith Cowan University, Australia
Psychol. 4:451. doi: 10.3389/ fpsyg.2013.00451. This General Commentary extends the excellent article by Speelman and McGann (2013) criticizing how psychology researchers often use group means when interpreting psychology research
If we are interested in causation at the group level, obtaining differences between group means is reasonable

Summary

Introduction

Edited by: Jeremy Miles, Research and Development Corporation, USA Reviewed by: Craig Speelman, Edith Cowan University, Australia. This General Commentary extends the excellent article by Speelman and McGann (2013) criticizing how psychology researchers often use group means when interpreting psychology research. The problem is part of a more general multilevel problem in psychology that can be seen if we make salient that researchers compute group means to support a causal hypothesis, as follows.

Results

Conclusion