Hypothesis testing: differences between groups expressed by raw data

If you have objects grouped by a single dichotomous variable, a Hotelling's T-test may be sufficient.

If you have objects grouped by one or more nominal explanatory variables, consider MANOVA.