Hello Kanwer,

The short answer is, 100*(post - pre) / pre (for each group/condition).

The better answer is, don't rely on percentage of change as an ES metric, as that is conditioned by whatever the pre-treatment score happens to be. Instead, opt for a standardized effect, such as Cohen's d (post mean - pre mean) / preSD (or use Hedges' adjustment to same). That deconnects the metric from the numeric value of the level of the pre-treatment score.

Good luck with your work.