Hacker News Reader

Correct me if wrong, but this doesn't appear to have a non-Covid-19 control group?

This study on Long Covid found there was no statistical difference between teenagers that had Covid compared to those that didn't. Suggesting that these symptoms may be getting entangled and misattributed.

https://www.medrxiv.org/content/10.1101/2021.05.11.21257037v...

View on HN

If you read through the limitations section of the report, you will find they covered this along with several other limitations of the study.

When a study has unsurmountable limitations, which make it impossible to usefully reason about the "results" of the study, is being frank about those limitations more useful to society than not publishing the study?

To me the obvious answer is "no, you toss it in the bin, or better yet you think for awhile and don't perform the useless study", but opinions can differ here.

Completely agree. The term "limitations" is being used here to mean faults. When your study has faults, its conclusions are of little to no use and only serves as fodder for the media. It should be binned.

Reply on news.ycombinator.com