For those interested in Expereal's data export feature, I wanted to offer some initial impressions of its value. I think there's great potential, but I feel there are a bunch of things required to "make it work" and be more useful and interesting.
1. Quantity of Data: I enter in multiple ratings every day, often at various times, with a range of tags. This is quite important if you'd like to know when your high and low points (more likely, quintiles) are occurring. I suspect most users aren't entering in data as frequently or as much; in other words, they might enter a rating once a day, but no other information like descriptive tags, location, etc. So I need to determine how to encourage more frequent rating and make it easier to do so, including tag entry. Without question, there's friction in its usability.
2. Quality and Consistency of Data: As I've tried to determine correlations, I've found that it's been somewhat difficult - if not impossible - for a range of reasons. There is little normalization of tags: sometimes, I'll call work "work" and at other times the name of the company. Then, sometimes I'll type in the name of the company; other times, I'll use Facebook places picker, which is a very messy data source. As I describe my state of mind, I don't consistently use the same words, but whatever comes to mind. The consequent challenge is: I don't use the same terms often enough to be able to draw any interesting conclusions from them in their analysis. For example, if rather than simply using "content" and "ecstatic" as tags, I use twenty other very similar words, it's difficult to draw any correlations between a particular location or tag and a number "8". There are simply too few instances and too many word variations. The same challenge goes for people tags: if I call my wife "wife", "Britt" and "B", I don't benefit from the fact that they all refer to the same person.
So, what's my solve?
I hope to figure out a way to suggest and save tags within the App, so that when you enter a given rating, you can select from a table of previously used tags depending on whether it's a description, location or person. I'm hoping that this might cut down the number of unique tags and actually give people some data that can be used to generate useful insights.
Expereal is both a work-in-progress and an experiment. I've written that I personally find it useful to "take stock" of one's life and review it. It's unclear, as of yet, whether such data will show any correlations and, even further out, inform actionable, "life changing" insights - that is, will be predictive. It's just too soon to tell.
I recently corresponded with a journalist who doubted the possibility. He might be right, but I think Expereal needs to evolve and likely integrate with other measurement tools (physical self-trackers, etc.) to be declared useful, useless or somewhere in between. Regardless, it's sure fun to think about and be part of a community interested in such matters.