I’ve seen some users suggest the ability to mute or ignore some correlations. It seems that basically changing the wording to when x happens x is more likely to happen can also reduce this frustration. Because it is true, if unhelpful, that when you feel bad you are more likely to use a heated blanket. That doesn’t mean heated blankets make you feel bad. The trouble is this is kind of the deep problem with this app and all similar apps. Correlation is not causation. And I think there is already another suggestion about making a much bigger deal about what that means and how to use and interpret the data. I’d love to hear what the developers think about this problem. It seems to make up a decent chunk of complaints and it kind of a big deal but it’s a tricky problem– it’s a tricky problem in stats in general not just in this app.