Suggestions

:speech_balloon:

Find correlations automatically

Instead of needing to search for particular correlations I would love if Bearable would just find them automatically.

So instead of choosing a symptom, I could go to a screen and be shown all correlations sorted by strength. If you’ve ever used Whoop they do a great job of this.

I imagine that periodically I would be presented with the top correlations, then I could choose to ignore them in the future if they weren’t interesting to me.

It seems like a lot of data to process, but I’m hoping the computation could happen in the background in pieces as I use the app normally.

5 votes

Tagged as Suggestion

Suggested 26 October 2022 by user Jason

  • Sign in to comment and vote. Sign in by email
  • 26 October 2022 Jason suggested this task

  • 02 November 2022 Eoghan approved this task

  • avatar

    Hi Jason! Thank you for this suggestion. Are you referring to the monthly and weekly reports in Whoop?

    Or where in the Whoop are you thinking of? OG whoop user here :)

    02 November 2022
  • avatar

    Nice! I’m talking about the monthly Whoop report, where they list all your factors and show you how they impact your recovery (e.g. alcohol reduces recovery by 7%).

    It’s easier for Whoop because they’re only measuring one outcome (recovery). In Bearable’s case I’d like for Bearable to show me all correlations between all factors and all outcomes (ideally between factors and factors or outcomes and outcomes as well).

    Basically I want Bearable to automatically show me anything that is correlated with anything else, without me having to pick specific things. I think this would make it a lot easier to find interesting impacts and motivate me to consider why those factors could be important.

    02 November 2022
  • avatar

    Oh, another important part about the whoop comparison is that whoop will tell you when there are confounding factors. I think that’s an important feature.

    I’d like Bearable to help me figure out which factors are actually important and which are correlated but not important.

    05 November 2022
  • avatar

    I fully support this suggestion. Especially when considering delays in factor - symptom pairings (e.g. 2 days later), there is so much manual searching involved currently.

    17 November 2022
  • avatar

    That’s a great point, I hadn’t even considered the time delay aspect. I imagine a screen with a listing like: * sedentary :: 2 days later :: sleep :: -7% * high activity :: 0 days later :: sleep :: +8% * etc…

    18 November 2022
  • avatar

    Seems like my markdown formatting didn’t come through. Let’s see if dashes work better:

    • sedentary :: 2 days later :: sleep :: -7%
    • high activity :: 0 days later :: sleep :: +8%
    • etc…
    18 November 2022