[parfe]

>Why do comments like this get upvoted so much?

They're dressed up versions of "correlation != causation". Low thought comments appealing to people with not-even stats 101 knowledge.

There is no such thing as a perfect study, and a generic version of the OPs comment could be copy and pasted on any study ever. Unless you study the entire population of the planet, you're going to miss subgroups. Unless you study the entire population of the planet, you're going to need some sort of selection criteria. Unless you have infinite funding and time, you're going to need to make trade offs and sacrifices in your experiment design.

A study will disclose these shortcomings for readers to balance the significance of results against.

Take this complaint from OP:

>* The sample size was tiny. (20 students)

What sample size would satisfy him? Why is 20 too small? 40? 80? 1037? Is he basing his opinion of a proper sample size on his gut? 20 just doesn't feel right?