We have a "test" and a "control" group. We calculated a market share percentage for both groups, the test group was in a market that received advertising from our client, the control group was in a market that did not see advertising. After the advertising ran, we then went back and recalculated market share for both groups. Market share for the "test" group went from 10.8% pre adversiting to 12.09% post advertising. For the "control" group, market share went from 9.48% to 9.12%. So the test group saw a .46% change in market share and the control group saw a -.36% change, how do I determine if the difference b/w .46% and -.36% is statistically significant?