from Hacker News

Being a Night Owl May Be Bad for Your Health

by yogi123 on 4/9/15, 4:33 AM with 2 comments

  • by smt88 on 4/9/15, 4:45 AM

    Like most mass-media writeups of scientific studies, there's no science here. The sample size is absolutely tiny, and it's poorly designed on top of that.

    Here are some problems with it:

    - The sample size is absolutely insignificant. For studies that we want to generalize to all humans, a sample of at least 100,000 is ideal.

    - This is not longitudinal or experimental data. We have no idea if we're just seeing a random correlation or not. There are simply too many variables to control for, especially with such a small population.

    - Only people 47 to 59 were studied, and the questionnaire was administered in the present. We don't know what their sleep patterns have been in the past.

    - The groups were self-reported, which likely caused problems. What if you're a morning person on weekdays, but not on weekends?

    - Correlated health issues, like obesity or diabetes, might be the cause rather than the effect.

    - They're presumably all Korean. I don't know how this might influence the study, but it very well might.

  • by anigbrowl on 4/9/15, 4:59 AM

    Not as bad as obsessing over wellness porn in the popular press.