But you'd need other data to prove (or at least strongly suggest) that the data isn't representative. You can't just say "the data COULD be unrepresentative",

But you'd need other data to prove (or at least strongly suggest) that the data isn't representative. You can't just say "the data COULD be unrepresentative", then just use some random guys anecdotes on the internet as more valid than national statistics.