Hi Super stars,
I have 6 datasets, and each dataset has two groups of values (~100 values in each group). I want to measure if the mean of values in the two groups are significantly different. First, I tested whether each dataset follows a normal distribution. If yes, t-test was performed, otherwise, wilcoxon rank sum test was performed. After that, I want to perform a meta-analysis across these six studies, and my question is can I do meta-analysis if the p values are derived from different tests? If yes, how should I do this meta-analysis?