Riddle me this, stats/medicine people.
I know about Justify Your Alpha.
Has anyone in *medicine* (or otherwise, but particularly interested in US academic medicine) actually proposed a study where they said they'd use an alpha threshold above 0.05? How was it received? (cont)
I know about Justify Your Alpha.
Has anyone in *medicine* (or otherwise, but particularly interested in US academic medicine) actually proposed a study where they said they'd use an alpha threshold above 0.05? How was it received? (cont)
(Also, please do me a favor, spare me the arguments about NHST being a flawed paradigm on this particular thread)
Clearly not all studies have the same tradeoffs of a false-positive vs a false-negative finding, and in some cases a higher alpha threshold seems like it should be warranted...
...my understanding is that in some contexts (pharma companies in early stage research, for example) this is more accepted since they just want to identify things that are promising enough to pursue further...
...but I'm curious, in a grant-reviewed academic-medicine setting, has anyone actually proposed this? How was it received? Happy to hear stories publicly or privately, if anyone willing to share.
Of course, I ask because I've specifically been asked to work up a study design where the sample size is simply not going to feasibly allow good power (by conventional reckoning) at an alpha=0.05, but I think there's a semi-legitimate case to be made...
...that in the specific context of this study, a higher alpha level should be permitted, and I'm curious to hear from folks that have attempted this if they were received sympathetically or got the "it's unusual to use alpha more than 0.05, everyone knows that is significance"
• • •
Missing some Tweet in this thread? You can try to
force a refresh
