Comparing Dependent Correlations: Statistical Tests and Practical Advice
Author Information
Author(s): García‐Pérez Miguel A.
Primary Institution: Universidad Complutense Madrid
Hypothesis
How do different statistical tests compare when assessing dependent correlations with overlapping variables?
Conclusion
The study identifies five reliable tests for comparing dependent correlations, emphasizing the need for proper statistical methods over isolated significance tests.
Supporting Evidence
- Five tests were found to be reliable for comparing dependent correlations.
- Statistical tests should replace isolated significance tests for better accuracy.
- Non-normal distributions can affect the accuracy of correlation tests.
Takeaway
This study helps researchers choose the best way to compare how two things relate to a third thing, making sure they use the right tools.
Methodology
The study used simulation methods to assess the accuracy, power, and robustness of 10 statistical tests for dependent correlations.
Potential Biases
Potential biases may arise from the choice of statistical tests and the assumptions of normality.
Limitations
The study's findings may not generalize to all forms of non-normal distributions.
Participant Demographics
The study analyzed data from various psychological research papers, but specific demographics were not detailed.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website