In my reviewing articles in many good journals, I observe that sometimes very low correlation coefficient like 0.14 or 0.18 is significant because the sample size is big (more than 300). In this case is it accepted correlation coefficient??
Isn't statistical significance sometimes misleading ???
https://www.researchgate.net/post/What_is_the_minimum_recommended_correlation_coefficient_among_items_for_factor_analysis
' Isn't statistical significance sometimes misleading ? '
Misunderstanding what statistical significance tells you is misleading. You need to check you understand your null and hypothesis before you interpret. For correlation the null is that the correlation is zero, the alternative is that its magnitude is not zero. I see many articles that only worry about significance but don't worry about the magnitude.
A statistically significant result for correlation means it is unlikely that a correlation that size was due to random chance in an uncorrelated relationship. It tells you nothing about whether the correlation is of a useful magnitude. That is what the correlation coefficient itself does. When reviewing a journal article I don't worry too much about statistical significance for correlations (it become meaningless above 100 samples), the confidence intervals are more useful for quantifying the certainty of the magnitude.
Bear in mind that a low correlation does not mean a factor is not interesting, but it may need support from other independent factors to make a useful contribution. Complex phenomena that are affected by many factors may have low individual correlations with the phenomenon being studied, but all the factors must be considered together to form a useful understanding of the phenomenon, e.g. weather forecasting
Dear Dr. Suhad Mohammed
Thank you very very much for your reply.
Best wishes
Dear Dr. Ruaa SH Redeiny
Thank you very much for your contribution.
'Best regards
Dear Dr. James Renwick Beattie
Thank you very much for your contribution.
'Best regards
https://statistics.laerd.com/statistical-guides/pearson-correlation-coefficient-statistical-guide.php
https://www.statisticssolutions.com/pearsons-correlation-coefficient/
https://journals.lww.com/anesthesia-analgesia/fulltext/2018/05000/correlation_coefficients__appropriate_use_and.50.aspx
Article Statistics Corner: A guide to appropriate use of Correlation...
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size