I appreciate the detailed notes/ feedback on the research process. I think the points you make are very reasonable and definitely helpful.
I expect that I will come back to this stress article, and the sleep one before that, to improve the quality of recommendations through a better research process. I can see a number of ways to do that including points that you bring up. - I’d like the evidence tables to cover 90+% of interventions that appear valuable from the outside/ have been suggested elsewhere, which I agree is not a bar I think this article reaches. - Leaving out the ‘% of studies that were significant’ makes good sense, especially given it’s quite inaccurate as I’ve put it here in its treatment of meta-analyses. - I’d like to keep some form of prioritisation by effect size. I think there’s probably a better way of doing this than what I’ve done here but prioritising interventions seems particularly valuable to me in terms of increasing the practicality of the recommendations and the likelihood a reader starts doing something useful. Perhaps it would be better to include the effect sizes as a range and highlight the different measures used (e.g. subjective stress vs. cortisol) to make this issue clearer to the reader. More work to do though on figuring this out.
I appreciate the detailed notes/ feedback on the research process. I think the points you make are very reasonable and definitely helpful.
I expect that I will come back to this stress article, and the sleep one before that, to improve the quality of recommendations through a better research process. I can see a number of ways to do that including points that you bring up.
- I’d like the evidence tables to cover 90+% of interventions that appear valuable from the outside/ have been suggested elsewhere, which I agree is not a bar I think this article reaches.
- Leaving out the ‘% of studies that were significant’ makes good sense, especially given it’s quite inaccurate as I’ve put it here in its treatment of meta-analyses.
- I’d like to keep some form of prioritisation by effect size. I think there’s probably a better way of doing this than what I’ve done here but prioritising interventions seems particularly valuable to me in terms of increasing the practicality of the recommendations and the likelihood a reader starts doing something useful. Perhaps it would be better to include the effect sizes as a range and highlight the different measures used (e.g. subjective stress vs. cortisol) to make this issue clearer to the reader. More work to do though on figuring this out.