Increasing the Power of Your Study by Increasing the Effect Size

17 Pages Posted: 7 Nov 2017 Last revised: 19 Apr 2018

See all articles by Tom Meyvis

Tom Meyvis

New York University (NYU) - Department of Marketing

Stijn M. J. van Osselaer

Cornell University - Samuel Curtis Johnson Graduate School of Management

Date Written: 2018

Abstract

As in other social sciences, published findings in consumer research tend to overestimate the size of the effect being investigated, due to both file drawer effects and abuse of researcher degrees of freedom, including opportunistic analysis decisions. Given that most effect sizes are substantially smaller than would be apparent from published research, there has been a widespread call to increase power by increasing sample size. We propose that, aside from increasing sample size, power can also be increased by boosting the effect size. If done correctly, removing participants, using covariates, and optimizing experimental designs, stimuli, and measures can boost effect size without inflating researcher degrees of freedom. In fact, careful planning of studies and analyses to maximize effect size is essential to be able to study many psychologically interesting phenomena when massive sample sizes are not feasible.

Keywords: Experiments, Research Methods, Effect Size, Statistical Power, Replication Crisis

Suggested Citation

Meyvis, Tom and van Osselaer, Stijn M. J., Increasing the Power of Your Study by Increasing the Effect Size (2018). Journal of Consumer Research, 44, 1157-1173. Available at SSRN: https://ssrn.com/abstract=3065720

Tom Meyvis (Contact Author)

New York University (NYU) - Department of Marketing ( email )

Henry Kaufman Ctr
44 W 4 St.
New York, NY
United States

Stijn M. J. Van Osselaer

Cornell University - Samuel Curtis Johnson Graduate School of Management ( email )

Ithaca, NY 14853
United States

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
187
Abstract Views
1,380
rank
174,988
PlumX Metrics