4 ms·
I recommend that people look for a meta analysis instead of individual trials when searching for medical research. Whilst randomized controlled trials are excel
by pilotInPyjamas 3y ago
I recommend that people look for a meta analysis instead of individual trials when searching for medical research. Whilst randomized controlled trials are excellent, the sample size is often quite small. A meta analysis avoids this problem by combining the results of many trials together.
For example, this analysis: https://pubmed.ncbi.nlm.nih.gov/23440782/ https://pubmed.ncbi.nlm.nih.gov/23440782/ finds that whilst the general population probably finds no benefit to Vit C, people who were "exposed to brief periods of severe physical exercise" did. Note that this meta analysis contains 11,000 participants, far more than the 400 in the original article.
- joshspankit 3y agoPersonally I’m more wary of meta-analysis than I am of individual trials. As a reader of a meta paper it’s almost impossible to separate out bad data and it means to be “good science” the analysis compilers must have even more integrity and scrutiny than the original authors (which feels like it’s hard to come by in the current scientific landscape).