Correlations

Many emails come in, asking if Symple will ever show correlations between symptoms and factors:

  • Is acupuncture effective?
  • Does exercise alleviate depression?
  • Does gluten trigger migraines?
  • Are the meds working? 

This is a tricky one, but I think there is an answer in sight.

You see, I have been fearful of Symple interpreting your data for several reasons. First, what if our tool, our magical algorithm, is wrong? Second, can we do correlations and not run afoul of medical device regulations or something like that? And finally, how well will people understand the difference between a correlation (yep, those numbers seem to move together) and causation (and therefore, let's bet our life savings on RED!!).

We often confuse a correlation (a relationship) with a causation (a reason for something happening).

Here's an example: if you take your medications every morning at the same time each morning, with a bowl of cereal and a shot of vodka, are the meds to blame for your woozy head? Probably not. But if you only track "medications" against the symptom "wooziness," you'll definitely see a correlation and incorrectly identify your meds as the cause for your wooziness. 

Maybe the medications do cause wooziness. But until you ditch that morning shot, you won't know for sure. It's up to us to observe all our potential culprits and heroes. And the app helps us do that, graphs them for us, and make visible what hadn't been visible before. But let's also get a doctor, pharmacist, dietary specialist or therapist to weigh in and help interpret our data with his/her own trunk full of knowledge.

But an app? Well, Symple hasn't gone to medical school.