Yes, good suggestion :)

This would be awesome!

Would love this

Yes, please!

Yes, thought about this as well!

Similarly, I suspect the symptom insights are messed up bc correlations btwn factors aren’t considered in the model… a more sophisticated model would be so awesome.

I definitely need this!

Would absolutely love this! I keep track of so much in Factors and it would be so helpful to be able to see correlations between different factors. If I could upvote this a million times I would!

This seems to me to be the purpose of an app like this. Again, check out MySymptoms for an example.

A million percent yes. This would be a game changer for me.

Do you all mean stats like:

e.g. When you do “Work” you also seem to do:

• High screen time (100% correlation)
• Moderate productivity (95% correlation)

etc?

That would be most essential to me too. Million time yes!!!

James, yes, that is what I was thinking! Thank you!

This was actually my primary intent in getting this app - to see correlations between factors, and then additionally, to see how individual factors or factor clusters manifest in moods and/or symptoms. I’d like to be able to have the option to manually select several suspect factors at once to compare for strength of correlation with each other, as well as with moods and/or symptoms (displaying as a line graph over time?), as well as have a view where the app analyzes and presents all factors or symptoms or moods (à la bar graph, in order of predominance during a specific window of time). Another idea is an option to automatically cluster factors into tiers by correlation strength, with thresholds beyond which factors fall into the various tiers. Or maybe you could throw together all factors, moods, and symptoms, and note “hotspots” of correlation between the three things where you can zoom in and see the connections shared by them all; alternatively, one could select all or a few of one type (factors, moods, and symptoms) and see what hotspots that cocktail brings up.

Thanks, some very interesting ideas there.

We actually have a design idea for Factor correlations already, just need to consider a few more things to perfect it and hopefully we can have something like this in the next few months

Cool! I’m so excited to have some correlation functionality in the near future. Thanks for the update!

It would also be great to see correlations and reverse correlations- for example, if a certain factor is associated with a certain symptom _% of the time, how often is that symptom associated with the factor (I’ve found that when looking at correlating symptoms- for example headaches might correlate with stress 50% of the time, but if I go to stress and look at correlation with headaches it might be much higher or lower). I think that’s due to the difference in number of data points for each item? But I’m not sure how to calculate the frequency of correlation when there are two very different numbers like that. It would be great to see how when you look at the correlation from both sides and consider number of data points what the overall percentage looks like

Hey all, I am finalising designs for this.

I’m just wondering if you are looking for % correlation for factors you entered simply at any time on the same day as the factor in question, or % correlation for factors you entered in the same time period or before (i.e not showing factors you entered after)? This is assuming you enter factors into time periods of course.

Please let me know :)

For me it would be same day. I get the point of AM to AM correlation, but for my uses (migraines, stress, reflux, psoriasis), many factors could be earlier that day for me. Could also be the day before, but within the day would be biggest concern

I would like to be able to correlate factors from previous events so that I can find the patterns that might be causing the migraine

For me, it would be desirable to identify cause-and-effect relationships between factors here, so I would say - % correlation for factors I entered in the same time period or before (i.e not showing factors I entered after). A simple example is to identify which nutrition factors (foods) lead to a normal stool (or to constipation, or to diarrhea). Another example is to find out which factors most often precede late falling asleep.

Indeed it would be the most flexible if all items could be available to be chosen both as a “factor” and as an “outcome”.

For example, I am interested in how the “time asleep” correlates with my symptoms or custom ratings.

I understand that factors are “true/false” variables and outcomes are “numerical” variables, but for this purpose, I think one can make a “numerical” item become a “true/false” one by applying a threshold (e.g. the average of the period or a user-defined threshold). Then I can see, for example, how “time asleep < average” correlates with mood or with headache etc. Side note, since we are talking about correlation , admirably rightfully so, the word “outcome” is not the best choice, as it suggests causation instead. One could call them “health parameters” or “health descriptors” o “variables”. (Disclaimer: I am not a native English speaker so apologies if I am wrong here).

I love all of the app, function and appearance, so thank you very much! I am happy to be a premium user since the early days. 💙

This is it! I’m really missing something like this… just a list of correlation between factors to find unseen routines and examine possible causations… This is necessary for a creative use of the app, improving routines and discovery of the self

My thought is to label them to emphasize the time occurrence. Maybe “Lead Factors” and “Following Factors.” Take all semblance of cause/effect out of it.

Knowing the timing of the comparison would empower the user to look for possible causation.

I would be very interested in this feature to be able to do exactly what is suggested on your support page “How to Find Correlations”: “Your correlations might not always make sense at first so it’s important to take the time to think about why this is. For example, ‘acts of kindness’ might correlate with lower ‘sleep quality’, which doesn’t make a lot of sense until you take into account the fact that ‘acts of kindness’ are often tracked on days when ‘junk food’, ‘high stress’ and ‘caffeine’ are also being tracked.”

This was moved to Planned in 2021 and then back to Under Consideration 8 months ago. What happened?

Hey James been loving the app and worth every dollar. But myself and seems many others are interested in this improvement. Would you have any update on the status of this by chance?

We had a related suggestion by another member of the community:

“It would be very useful if in the insights section, the user could see the correlations between the things they eat/drink, and their bowel movements. It would give a lot more information regarding how certain food items affect their bowel movements.”

We received a related suggestion from another member of the community:

“It would be good to be able to compare certain factors between one another instead of comparing each one to the lack itself.

For instance if trying to see which solution works best for snoring out of nose strips or mouth guard I’d want to compare both to days with neither nose strips or mouth guard to avoid overlaps.”

Yes, this is exactly what I would love to see. For example, I might be able to see that my “Meditated” factor corresponds to my “Under X hrs of screentime” factor.