• @[email protected]
    link
    fedilink
    292 months ago

    n=40 isn’t actually bad for generalized conclusions, given a reasonable spread in the results. Your second point is a much stronger argument. The sample is entirely non-representative.