And yet, there are loads of studies testing things like specific therapy techniques where 2 thirds of the people trying it dropped out, and those who stayed show a tiny marginal improvement. And no mention at all of survivorship bias?! Like if the thing is not working or making my problems worse, I’m more likely to drop out.
But the people doing the study seem to either be oblivious or willfully ignorant of this.
Survivorship bias at its finest
And yet, there are loads of studies testing things like specific therapy techniques where 2 thirds of the people trying it dropped out, and those who stayed show a tiny marginal improvement. And no mention at all of survivorship bias?! Like if the thing is not working or making my problems worse, I’m more likely to drop out.
But the people doing the study seem to either be oblivious or willfully ignorant of this.
TBF, dealing with informative dropout is really difficult. And the best remedy we have is often… more statistics.
Check out survival analysis (quick video on the YouTube link), people dropping from studies is acknowledged and accounted.
https://youtube.com/watch?v=C5BZA-0dloU
In my field it often isn’t.
In a well done RCT of course it is. My field has crazy drop out rates and the majority of reviewers and authors don’t bat an eye.