They didn't assume anything. They asked. First you get the data, then you can think about correlation.

Because humans are imperfect and prone to bias.