What 2439 Reports Taught Me
How I Started
Last year a friend of mine recorded parts of his life in moleskin books. At the end of the year he made an info graphic of some of the data.
The experience made me think about my life and journaling. I was envious of his ability to go back to some of those moments and wanted that for myself, so I waited for ReporterApp to be released.
When it was released, I went all out. It was beautiful and simple to use. I went crazy asking all sorts of questions and couldn't wait to find out all of my life secrets.
At the time of this writing I have 2439 reports. During that time I have learned more about myself, the app, made tools to get more out of my data, and have more ideas.
Highlights About Myself
Out of all the new people I met this year, I saw only 20% of them more than once.
There were 10 people that accounted for more time than the other 224 combined.
Someone I saw almost every day last year appeared only twice this year. We still live within 30 minutes of each other.
I reported 18 different books, but only finished 6 of them. The Dispossessed was my favorite.
I recorded learning over 224 new things in 2014.
I eat out much more than I eat in. Especially for lunch (12% at home) and dinner (14% at home). Favorite destinations were Hruska’s Kolaches (breakfast), Communal (lunch), and WingNutz (dinner).
My top three shows were Archer, It’s Always Sunny in Philadelphia, and Parks and Rec. Out of the 57 shows reported, 20 were comedy.
Next, I want to talk about my experience with ReporterApp itself.
Frequency doesn't mean better data
When I first started, I was asking myself about 10 questions a survey. I wanted to track everything.
For some of the questions, the frequency hurt the integrity of the data. I found myself either skipping surveys because they took too long to fill out, or I would use the previously filled in data to finish faster. This translated into missing or inaccurate data.
It also made the data hard to interpret. A good example of this is a ‘How happy are you?’ (1–10 scale) question. It’s hard to tell how happy you are when asked 6 times a day. If I reported in a scrum meeting at work I would enter a 3. If I reported while visiting friends or family, I would be much higher. Overall, the sadness of the meeting is quite fleeting in the big picture and probably shouldn't be accounted when trying to find my overall happiness.
To fix this, I moved most of my questions to the end of the day. During the day I ask myself 3 questions: ‘Who are you with?’, ‘What are you doing?’, and ‘Where are you?’. Then, when I go to sleep, I ask the happiness, health, and other questions that don’t make sense to ask during the day multiple times. Now my surveys during the day are very fast so I get more surveys and better data.
ReporterApp Doesn't Tell, It Tracks
While the ReporterApp’s visuals answer some pretty basic questions, they don’t tell much of a story. I wanted to answer questions like:
What have I done with person x?
Who am I with when I am doing y?
What is my Low, Average, and High marks for a given question?
Obviously, I needed to get the data myself and do some digging. I wanted context and the ability to find the low, average, and high of my numeric questions.
Being a web developer, I decided to make a personal website that showed all the data. So my site was born. I could finally see my low, average, high and current numeric answers. More importantly, I could see the context of a given answer. My friends loved this feature because they could look themselves up and see some meta data about the section of my life I share with them.
Where I Failed
The project failed in a couple of ways. First, the programming logic was useless to anyone but me. It was hard coded to my questions and didn't support all types. Second, the low, average, and high data was pretty shallow. Almost immediately I wanted to dig deeper. What else was reported on my lowest and highest reports? What factors influenced them? I had no way of easily knowing.
I have addressed one of the failures. First, I made this project so others could download the code to look at their data and, hopefully, contribute!
My next goals are to be able to view reports in a journal style where I can navigate from day to day, and link people straight to specific days. Then link other parts of the app, like lowest happiness, to the report that had it.
After attending Eyeo 2014, I was also inspired by feltron to make a tool that would allow me to easily get my data into a physical book. Then take it a step further and be able to make books using my context logic. That way I could give my friends books of the their context in my life.
I have found great satisfaction in the difficulty of tracking, extracting, and interpreting my data. A part of me thinks that the struggle makes it more memorable and, if it was easier, I probably wouldn't remember as much.
I would also like to plead for help from the community. Let’s get our hands dirty and collectively get more out of our data and, hopefully, more out of life. If you can’t code, design. If you can’t design, use the tools and give feedback.