Related Experiment Videos
Extracting a maximum of useful information from statistical research data.
Staffan Sohlberg1, Gerhard Andersson
1Department of Psychology, Uppsala University, SE-75142 Uppsala, Sweden. staffan.Sohlberg@psyk.uu.se
Scandinavian Journal of Psychology
|January 22, 2005
Summary
Psychological research is shifting from null hypothesis significance testing (NHST) to include effect sizes and Bayesian methods. This evolution helps researchers better interpret statistical data, reflecting decades of critique against NHST.
Area of Science:
- Psychology
- Statistics
Background:
- Null hypothesis significance testing (NHST) has been the dominant statistical inference method in psychology.
- Decades of criticism have highlighted limitations of relying solely on NHST.
Purpose of the Study:
- To critically review the current practices of statistical inference in psychological research.
- To examine the transition towards more informative statistical approaches.
Main Methods:
- Review of the evolution of statistical practices in psychological literature.
- Analysis of the increasing adoption of effect size estimates and confidence intervals.
- Assessment of the growing interest in Bayesian statistical approaches.
Main Results:
- Psychological research is moving beyond exclusive reliance on NHST.
- Effect size estimates, confidence intervals, and Bayesian methods are increasingly incorporated.
- These changes are viewed as beneficial for data interpretation.
Conclusions:
- The shift in statistical practices enhances psychologists' ability to extract meaningful information from data.
- Criticism of NHST over the past seventy years is influencing current research methodologies.