A Defense of Publishing Nonsignificant (ns) Results

Abstract

An argument is advanced for the propositions that a lack of statistically significant findings does not automatically justify rejecting a scientific article for publication or mean that the findings are necessarily uninformative. Systematically declining publication for nonsignificant results leads to negative consequences that include distorting scientific literatures, making hypotheses and theories less falsifiable, depriving meta-analyses of an accurate sample of prior findings, and encouraging questionable research practices. Providing effect sizes and confidence intervals can make findings more informative, regardless of whether the finding is p < .05.

Keywords:

• Falisibility
• Meta-Analysis
• Nonsignificant Results

Additional information

Notes on contributors

Timothy R. Levine

Timothy R. Levine (PhD, Michigan State University, 1992) is a professor in the School of Media and Communication at Korea University, Seoul, Republic of Korea.