Pearsons correlation, confidence intervals, bootstrapping... Statistical nightmare

#1
I am currently running a project where I have 4 variables and a questionannaire with 4 subscales. I have performed a Pearsons correlation on this (due to high frequency of extreme scores in the questionairre and non-normal distribution of the questionnaire) and have found 2 significant results.

However I wanted to assess these using confidence interval as the p values are .033 and .037 so are significant but the correlation coefficients are small and so I wanted to check using a method other than null hypothesis significance testing. However the internet has recomended a bootstrapped confidence interval due to the non-normal distribution of my data. Is this correct?

If the CI's straddle the zero point but only slightly (one is -.432 and .025) can it still be assumed that there is no correlation? Is there any way of formally assessing the CI's?


Cheers for any help anyone can offer!