6 ms·
I find the design and results somewhat underwhelming (weak significance in a N=800 sample with >50% dropout). The authors also mention that they lack power, but
by uniqueuid 2y ago
I find the design and results somewhat underwhelming (weak significance in a N=800 sample with >50% dropout). The authors also mention that they lack power, but the only correct answer then is not to do the study, include more people, more data points, or reduce measurement variance.
That said, I think it's a refreshing perspective, and it seems like a good idea to look at effect heterogeneity in psychedelics use. Thinks like bayesian trees (BART) can help identify subgroups that react differently to stimuli along specific characteristics.
- WesternWind 2y agoA lower power study can still lend weight to a good meta analysis.
- uniqueuid 2y agoThat is true, but there are also conditions where low-powered studies degrade meta-analyses. For example, if the panel attrition is caused by self-selection on unobserved variables, then the study contains bias that is not (and cannot be) accounted for, which skews the meta analysis. But yeah, those concerns are mostly about measurement not sample size.
- coldtea 2y agoMeta-analyses are even less rigorous than a low power study