The person who collects the data has already shaped what can be concluded from it.

Analysis gets scrutinised and collection rarely does, yet the collection settled the question first. Who was asked, what the options were, which cases were excluded as messy, how a category was defined: each of those closed off conclusions before anyone ran a calculation. By the time the numbers are on a page, the interesting decisions are already invisible behind them. Arguing about the analysis alone leaves the real assumptions untouched.

When a finding matters to you, ask how the data came to exist before you interrogate what it says. Who was left out is usually the fastest question to the heart of it.