Effect-size estimates: Issues and problems in interpretation

Edward F. Fern, Kent Bourdon Monroe

Research output: Contribution to journalArticlepeer-review

Abstract

In recent years, researchers have recognized the importance of the concept of effect size for planning research, determining the significance of research results, and accumulating results across studies. However, the uncritical use of effect-size indicators may lead to different interpretations of similar research findings because of differences in assumptions underlying the nature of the research, aspects of the phenomenon being investigated, or the methodological characteristics of the research. This article reviews the substantive, measurement, and methodological issues that influence the relative magnitude of an empirical effect size. The relationships and transformations between different types of effect-size indicators are presented. It is the thesis of this article that the meaningfulness of an estimated effect size should be interpreted with consideration of the type of research (relational vs. experimental), the anticipated application of the results obtained (effects application vs. theory testing), and the research history in the domain of inquiry. Researchers must be cognizant of the many different causal factors that influence effect size before using the magnitude of an effect for assessing the importance of research results, calculating the statistical power of a test, or synthesizing findings across different studies.

Original languageEnglish (US)
Pages (from-to)89-105
Number of pages17
JournalJournal of Consumer Research
Volume23
Issue number2
DOIs
StatePublished - Sep 1 1996

ASJC Scopus subject areas

  • Business and International Management
  • Anthropology
  • Arts and Humanities (miscellaneous)
  • Economics and Econometrics
  • Marketing

Fingerprint Dive into the research topics of 'Effect-size estimates: Issues and problems in interpretation'. Together they form a unique fingerprint.

Cite this