Suggestions

:speech_balloon:

Using variance to make data more useful

Maybe as advanced insights for people with months or more worth if data. Or to give people option tick box to use predictive data. Use regression, ANOVA, Bayesian or so to give predictive results instead of just correlations. Correlations in the app are often superfluous and require very critical eye to reach conclusions. E.g. app says meditation causes anxiety next day, even though person probably meditates more due anxiety period. Or eating bananas causes depression, but person just eats many bananas in general. Of course, you can do this with exported data, but in case the selling point is meaningful analysis these are some logical next steps.

Related: www.tylervigen.com/spurious-correlations

5 votes

Tagged as Suggestion

Suggested 05 September 2021 by user W

  • Sign in to comment and vote. Sign in by email
  • 05 September 2021 W suggested this task

  • 01 October 2021 Eoghan approved this task