A/B test
Because the frames with the black theme were based on an existing page design, I made some new ones light-themed so I could do an A/B test to see which one performed better.
My presupposition was that an easy-on-the-eye design (a lighter theme) would have better conversion rates. To verify that, I did an A/B test using a 548,000 sample size for each variation.
For the whole group of users, the sample size was statistically significant. In this test, I used the number and amount of donations to measure the conversions.