3 ms·
If the natural variability is high enough, then any effect will be hidden by it. Sure, we can say with some certainty that the effect of microdosing on most sca
by zosima 4y ago
If the natural variability is high enough, then any effect will be hidden by it. Sure, we can say with some certainty that the effect of microdosing on most scales are not that huge. But we wouldn't have expected very large effect sizes anyway, because they are almost unheard off. (Unless the subjects become really severely impaired)
And there may still be meaningful treatment effects, at least judging by current standard of care for many psychiatric ailments and how those have performed in studies.
Again, I'm quite confident that the effect of many current psychiatric standard of care treatments would never have been picked up by this study. Not because they don't work (at least somewhat), but simply because there is too much noise and natural variability.
- jamiek88 4y agoQuite confident but utterly wrong. As has been pointed out to you several times. Science does not and cannot work by cherry picking.
- pfisherman 4y agoIf the observed effect of an intervention does not rise above the background variance it means that the intervention does not do what you want it to do. Either (1) it has no effect or (2) the strength and direction of the effect are random. Either of those qualities renders the intervention ineffective. You can’t justify giving a patient something that will have no effect, or will make them worse off half of the time when there are more effective options are on the table. Of course this is comes with the caveat that the study is adequately powered to detect the strength of the effect you are looking for. That being said, there is plenty historical data for the placebo, so I don’t think that being underpowered because of misestimation of the background variance would be an issue.