Related Experiment Videos
Effect sizes for experimenting psychologists
1Temple University, USA. rosnow@temple.edu
Summary
This study details effect size estimators for psychologists, covering between- and within-group designs. It introduces the counternull statistic to prevent interpretation errors in null hypothesis significance testing.
Area of Science:
- Psychological research methodology
- Statistical analysis in psychology
Background:
- Effect size estimation is crucial for interpreting research findings.
- Null hypothesis significance testing (NHST) is widely used but prone to interpretation errors.
Purpose of the Study:
- To describe three families of effect size estimators.
- To illustrate their application in general and specific psychological research contexts.
- To introduce the counternull statistic for improved NHST interpretation.
Main Methods:
- Discussion of correlation (r-type) effect size indicators.
- Description of difference-type and ratio-type effect size estimators.
- Explanation of the counternull statistic's utility.
Main Results:
- Provides a comprehensive overview of various effect size estimators.
- Demonstrates applicability across between-group and within-group (repeated measures) designs.
- Highlights the role of the counternull statistic in mitigating interpretation errors.
Conclusions:
- Effect size estimation enhances the practical significance of research findings.
- The counternull statistic offers a valuable tool for rigorous interpretation of statistical results.
- This work provides psychologists with practical guidance on effect size estimation and interpretation.