It is ANOVA if you compare four groups on one interval-scale
variable. If the hypotheses are separate enough, it would be
two ANOVAs in order to test two variables.
If they are not separate and distinct hypotheses -
you could test a composite score by ANOVA, or you could use
discriminant function for two variables (which is the convenient
form of MANOVA for the simple case).
If the two variables have some particular relationship, like
Pre and Post, some other analysis might be better.
--
Rich Ulrich