Another recent hobby of mine is looking at data about my life. I think we can all agree that sometimes we just don’t remember things the way they happened. More often than not, this is observable in disagreements I have with my fiancee over various sequences of events. Given that problem, we become unreliable narrators of our own lives. In order to get at and understand some of the minutia of daily living it becomes important to have instruments, sensors, and data collection methods that fulfill a few criteria:
- Accurate – that these measurements are taken in parallel to qualitative information inputed by the user (me).
- Automated – that data collection takes place with little need for my direct involvement.
- Analyzable – that data collected present some meaning or capture something worth understanding.
On a quest to understand, how do you get all that data?
Undertaking this quest, I’ve fallen into the rabbit hole and found a web of services, solutions, and self-guided practices that have allowed me to begin tracking more.
I started early on with google calendar, excel, and a notebook. This required too much effort from me at times and was often forgotten in light of more important or immediate issues. Slowly I moved to using smart phones apps. Many of them I’m still using, but in more complex ways than before:
Saga: Probably the first thing I needed to get myself off of google calendars. It tracks the where, when, and how long while integrating information from other connected applications. It is both an end point and an entry point for much of the data I’m now collecting.
Sleep-bot: Sleep bot was a newer find, and was something that replaced a similar app I was using on my iPhone earlier last year. When I got an android phone I went without any sort of smart alarm/sleep analysis. Now that sleepbot has about three months of data, I’m starting to see trends and am slowly improving both the quality and quantity of sleep I get.
MyFitnessPal (MFP): I’m a bit off and on with this service, it integrates well with other apps and programs, but it requires attention and manual entry of food. There isn’t yet a good solution for calorie and food tracking, so this is about as close as I can expect anyone to get. I had originally used Lose It! but lost interest after some changes to the application and discovering MFP.
RunKeeper: The first app I started using (and began using again recently), RunKeeper has come such a long way since I first used it. It logs more than just running and now has support for a heart rate monitor. Easy to use, nice customization options on the audio cues, and definitely a main stay on my phone.
Have data, want additional data.
Recently I’ve also done some research into getting a fitness tracking device like fitbit flex, jawbone up, or basis. I’m currently waiting on an invite to purchase basis due to the additional sensors offered there.
The major problem with “quantified self” is the difficulty in pooling and analyzing data across systems, programs, and devices. Currently no simple solution exists, but Saga, MFP, and RunKeeper are all integrated via API/integration features. Basis is still fairly closed, but offers what I feel is the most robust device. An API would do much and more to help me pool my data and for developers to connect the services.
Where are we going here?
I’m really hoping that in the next few years I’ll be able to pool my data, and run regressions like crazy on it (the one really useful mathematical skill I learned in the PhD Program I recently left). In addition, I’d love to be able to take this information and show it to my doctor. As you’ll likely come to know, I’ve got a bit of a problem with weight (really its not that bad), and being able to show my physician whats going on, what working, and what is not would be nice.
I currently don’t have anything in place to help me understand how stress relates to my psoriasis, but that’s something I’m exploring for the future. I know stress and certain diets are affecting it, but I don’t know how. We don’t know as much as we could about psoriasis, and this might be one way of starting toward the big-data approach to understanding commonalities among people with psoriasis instead of the voodoo-like-advice I see on internet forums about it.
And another thing..
Two things I think are worth mentioning: the spatial dimension and the lack of qualitative data.
Spatiality is something I’ve talked about quite often in my (former) PhD program. Space, whether we like it or not, has interesting and unique impacts on the way we live. Even something as simple as bid rent theory suggests a complex relationship with economic thinking, sociological concepts that operate and motivate breaking or imposing social barriers, and psychological patterns of how we understand the world. Tracking some things without understanding the where will prove fruitless.
Qualitative data is also something I feel is worth logging alongside quantitative data. Some of my most insightful exploratory questions have come from looking at personal, group, and internet-consensus qualitative trends. Even something as simple as looking at a thread on Reddit can prove useful in finding new avenues of exploration and preliminary insight into how the question might be answered.
Every app and service I mentioned above has rudimentary elements of spatial and qualitative data collection, but it is not as robust nor as integrated as I would like. Hopefully that’s also in the future.