In the below table if we want to pool the correlation of neck circumference with waist circumference, what will be the p-value?
Thank you in advance
Hello Sagar,
Do you mean, pool the correlation (of neck, waist circumference) across sexes? Or, did you mean pool the correlation across similar tabled data sets that may exist?
The key ingredient for determining p-value for a correlation (to address the question, is it different from zero?) is sample size, or more accurately, df. If sample sizes are unknown from your source, that will make the task more challenging, to be sure.
For example, a correlation as strong as .78 has different significance values depending on N, but they tend to stabilize pretty quickly (as being very small) as N increases:
N = 5, p (non-directional) = .1197
N = 10, p = .0078
N = 20, p = .00005
N = 30, p = .00000037
On the other hand, a weak correlation (e.g., r = .12) may take bigger N to stabilize:
N = 5, p = .8476
N = 10, p = .7412
N = 20, p = .6143
N = 30, p = .5276
N = 50, p = .4065
N = 100, p = .2344
N = 200, p = .0905
Good luck with your work!
Hello David, thank you.
I am confused what will be the exact p-value, For example, for the correlation coefficient between neck circumference and waist circumference, the p -value is given as P value
Hello Sagar,
Depending on sample size and the number of digits used in the computation, the answer is, perhaps so!
Here's a web site that will do the p-value computation: https://www.danielsoper.com/statcalc/calculator.aspx?id=44
Good luck with your work.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis