4 ms·
This study performs a randomized, blinded, placebo-controlled (all good) trial to determine the effect of probiotics using a self-report survey. However, the tr
by thanatosmin 12y ago
This study performs a randomized, blinded, placebo-controlled (all good) trial to determine the effect of probiotics using a self-report survey. However, the trial is horribly underpowered [1] (N = 20 per group). There's little reason to believe the differences are true effects, rather than statistical noise.
[1] http://en.wikipedia.org/wiki/Statistical_power http://en.wikipedia.org/wiki/Statistical_power
- kszx 12y agoIt may be true that this particular sample is too small; I haven't looked at the experiment in detail. However, just pointing out that n<=20 per group isn't enough of an argument to justify this statement. Significance is about more than just the sample size. The idea that it's impossible to get significant statistical result with small samples is a common misconception.
- thanatosmin 12y agoTrue. The study provides the requisite observation variances and effect sizes for this analysis as well.