How to Improve on Statistical Significance: Effect Sizes, CIs, Graphs and Baseline Models
This symposium will introduce and discuss how scholars can improve upon the Null Hypothesis Significance Tests (NHSTs), which are currently constraining the production of knowledge in management science. The extensive use of NHST in quantitative research has led to the accumulation of statistically significant results that are both too small to be practically relevant and so small that they are unlikely to replicate. In a field that aspires to provide useful advice to managers, we need to focus on practically important effects that are robust across a wide variety of settings. The proposed symposium introduces and discusses alternative approaches to address NHST limitations -- such as, effect size measures, confidence intervals, graphs, meta analyses and baseline modeling. A final Question and Answer session will offer additional opportunities for specific discussions, advice and recommendations.
Abrahamson, Eric, S. Holloway, Andreas Schwab, and William H. Starbuck. "How to Improve on Statistical Significance: Effect Sizes, CIs, Graphs and Baseline Models." Academy of Management Proceedings 2015, no. 1 (2015): 13604.
Each author name for a Columbia Business School faculty member is linked to a faculty research page, which lists additional publications by that faculty member.
Each topic is linked to an index of publications on that topic.