Hi, I am looking for some help here.

I work for a non profit organization and we sent out a direct mail package to donors - a standard control and a test language in the copy. Our test copy showed an $8 increase in average gift size but a lower response rate in overall donations. Is it possible to run a confidence percentage if this same test was run again that the average gift size would still show the increase in the test treatment? I think that just one large gift can create a skew in the average gift that confidence would be low but wanted to toss this out to the intelligent folks here.

Pieces mailed control: 29,042
Pieces mailed test: 29,078

Response control: 251
Response test: 228

Income control: $5,118
Income test: $6,432

Average Gift Control: $20
Average Gift Test: $28

Thanks in advance for the help!