Hacker News Reader

I agree with you.

For some reason, every now and again studies like this -- ones that are interesting but under-researched -- are posted on HN, and I don't get why. People then treat these studies as if they are conclusive evidence of something. Luckily nothing ever comes of it, but still

View on HN

I regard this as another pilot program, to be scaled up. Gotta publish to get more funding.

maybe the results of their test will spur some doubting Thomases to pick up the mantle and have a go themselves. They'll either find supporting results or disagreeing results. Either way, sounds like good science.

If they are taking the results of the single study to say "This is the final answer", then that would be bad. If they are just saying "These are the results we achieved", then that sounds like what a study is suppose to do.

To be quite frank, repeated hypothesis tests on ad-hoc groupings is _terrible_ science.

Reply on news.ycombinator.com