My raw-data (n=25) does not meet the assumption of normal distribution. Therefore, I want to do correlation matrix analysis with non-parametric Kendall tau correlation.
But there is a linear relationship between my X and Y in my data (all quantitative). Later, I want to do simple linear regression model with one X and Y variable, with the same data set (n=25). For the simple linear regression model, I use lm function in R. Doing simple regression I want to know the p-value and the R2 to interpret the association between Y and X. (not for the prediction purpose). I thought I would do simple linear regression because my data somewhat rely on linear patterns.

I am worried about this. In my case,

  1. Is it okay to generate a correlation matrix with non-parametric methods and doing simple linear regression using the same data set? In other words, in my case, is that okay doing a non-parametric correlation matrix and the simple linear regression? or is there any contradiction in doing these two with the same data set? or Does Pearson correlations fits with my data set? so any help would be much appreciated. Thanks in adavnce!
  • $\begingroup$ you can do all at the same time, Kendall, lm, Pearson, non-gaussian etc. no contradiction $\endgroup$
    – Aksakal
    Oct 12 at 21:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.