I frequently read the claim that India reduced their COVID-19 cases a lot via Ivermectin but it's very unclear to me whether the relevant data to make that claim is cheripicked.
India has 34 region and each of those region has their own policy on Ivermectin. While we don't have exact data about Ivermectin usage, Google Trend data provides a proxy for usage of Ivermectin and does provide data that's fine-grained enough to separate individual region. John Hopkins provides data for COVID-19 cases by Indian region as well. 34 regions sounds to me like it's plausible to find statistically significant results if Ivermectin had an effect that's as great as proponent of Ivermectin claim
Given p-hacking concerns, it makes sense to first decide on how to analyse the data and then run the statistics. Does anyone have suggestion for how the data should be analysed?