HELP PLEASE :) ANOVAs repeated measures

More I read more I get confused, so please help me !

1. In 2 (2 conditions)x2 (2 time intervals) within subjects Anova if normality assumption is violated- what is the best option to go about? Ignore?
2. In 2x2 within subjects Anova if I have an interaction effect (all data ok)? Do I explore it with paired t-tests and Bonferroni correction applies?
3.In one -way anova within subjects Anova (time vs. 3 categories) do I just request post-hocs (paired-t-tests) when doing my analysis from SPSS?
4. When 2x2 Anova shows one of the main effects significant, 2nd main effect is marginally/not (p= .059) but there is an interaction what should I do?
Thasnk you in advance.