Anecdotes are still anecdotes regardless of how many you have. For data you need to actually measure a response in a controlled environment. Making sure you re

Anecdotes are still anecdotes regardless of how many you have. For data you need to actually measure a response in a controlled environment. Making sure you record both the interesting, and non-interesting, responses so you can compare them. Humans have a very strong natural tendency to only remember and evaluate interesting responses. Causing us to vastly overestimate the frequency of interesting outcomes. That paired with our brains strong desire to pattern match, means we’re terrible at making accurate observations on a day-to-day basis. Hence scientists do experiments like this to discover what aspects of our intuition are accurate, and what parts are inaccurate.