how to spot unrepresentative sampling flaw.

[gplaya123]

I was wondering when to say "oh that's unrepresentative sampling flaw."

Because the other day, I ran into this type of question:

Gemma3 is the best flu medicine ever.
Because the symptoms of those who have flu yet take Gemma3 are as half severe as those who don't take it.

I was like hmm....

Could it be the case that those who have flu yet take Gemma3 is "unrepresentative" since it could be the case that they belong to a group where they have light symptoms of flu?

I was wrong...

Is there clear ways to spot the unrepresentative sampling flaw?

Thank you in advance!

[tommywallach]

Hey G,

There needs to be something about the group that is distinct. If it's just a group of people, with no further description, there's no reason to believe it's unrepresentative. It could be too small, but it wouldn't be unrepresentative.

-t