greetings

I want to compare data for two independent groups, at different follow-up periods.

Example:

Control Group:

(X) - The first follow-up period

(Y) - second follow-up period

Group 2:

(Z) - The first follow-up period

(W) second follow-up period

The comparison will be as follows:

X Vs. Z

Y Vs. W

--------

my questions:

1- is the normal distribution (shapiro wilk) test performed on each group in each follow-up period alone or the whole data at once?

2- If one side of the comparison groups does not follow the normal distribution and the other does, do I perform the paramitric test or the non prarmetric?

3- Is there a statistical impediment to perform a parametric test on one of the control periods (e.g. X Vs. Z) and a non-parametric test on another period (e.g. Y Vs. W).