Depend about your sample number ,95% population. Si your study is comparison method i recommend use Bland Altman plot, because good correlation but not good concordance.
I mean that if I have four samples of dependent variable and four samples of independent variable, can I use correlation coefficient to see the relationship between them or not. I mean four samples is an enough for analysis or I need more samples.
Leave alpha at 0.05 and beta at 0.2 (so that power is set at 80%)
Put in the value for r ie what you expect the correlation to be or what correlation you consider to be important to your study.
You will get out the sample size that you will need to ensure you have enough power to detect the correlation that you think is important enough to detect.