For personal experimentation of the "enabling/disabling [an Eight Sleep mattress] over a few phases before analyzing sleep data collected by my Oura ring" variety, I suggest looking into the iOS app Reflect https://ntl.ai/reflect/
Besides switchback-style experiments, it supports day-level randomization, which is necessary if you're running more than one such experiment at a time. And it's tolerant of less-than-full compliance (it will tell you whether today should be T or C for each intervention, but allows you to log the treatment status even when you don't comply). I'm afraid that all such experiments are hopelessly underpowered and so have given up on the approach myself. But the tool is decently well-designed for it.
Incredible tip, I'll have to play with this as I set out along that path. When I last thought along these lines ~2 years ago, there were apps meant to hunt for correlations, but nothing that supported real experiments.
The Eight Sleep experiment might be more real than hypothetical as summer is ramping up here... I share your concerns about power but find variance in my Oura Sleep Score to be encouragingly low. Only one way to find out
For personal experimentation of the "enabling/disabling [an Eight Sleep mattress] over a few phases before analyzing sleep data collected by my Oura ring" variety, I suggest looking into the iOS app Reflect https://ntl.ai/reflect/
Besides switchback-style experiments, it supports day-level randomization, which is necessary if you're running more than one such experiment at a time. And it's tolerant of less-than-full compliance (it will tell you whether today should be T or C for each intervention, but allows you to log the treatment status even when you don't comply). I'm afraid that all such experiments are hopelessly underpowered and so have given up on the approach myself. But the tool is decently well-designed for it.
Incredible tip, I'll have to play with this as I set out along that path. When I last thought along these lines ~2 years ago, there were apps meant to hunt for correlations, but nothing that supported real experiments.
The Eight Sleep experiment might be more real than hypothetical as summer is ramping up here... I share your concerns about power but find variance in my Oura Sleep Score to be encouragingly low. Only one way to find out