Working Paper Series,
Lund University, Institute of Economic Research

Abstract: This paper argues that there is a norm in the publication of quantitative management research that accepted hypotheses are conceived of as good hypotheses. This has far-reaching implications for quantitative research in that, in the worst case scenario, this norm encourages commonsensical research questions and manipulation of data in order for an article to be published. Naturally, such behavior is not otherwise encouraged by the research community and, therefore, there is not even transparency in the decisions that are made. Instead, the choices that are made in the course of a research process (both in terms of how variables are operationalized and in terms of how results can be interpreted) are down-played unless in cases where hypotheses are rejected. Sample- and instrument-issues are then the most frequently used excuses for non-expected findings. Rather than trying to live up to demands set outside the social sciences (and failing), this paper calls for more reflexivity in quantitative management studies. The world is a messy place and so is our data.

Keywords: hypothesis testing; significance bias; publication bias

15 pages, August 25, 2004

Download statistics

• Lund University, Institute of Economic Research
• Ordering Working Papers
• Home page for this series

Questions (including download problems) about the papers in this series should be directed to Elsbeth Andersson ()
Report other problems with accessing this service to Sune Karlsson ().

This page generated on 2024-02-05 17:10:08.