Addressing Flaws in My Quantified Self Data Collection Approaches

I started this post before QS15, the quantified self conference I attended last week, but I thought I’d finish it now. I did a show and tell presentation ) that focused on what I did, how I did it, and what I learned from adopting tracking approaches and technologies for weight loss and maintenance.  You can see a pdf with notes here:  A Picture is Worth 1000 Calories notes 6-16-15.  Over the past year, I’ve blogged about various projects in my Fitness and Food series on the JMP blog, shared my personal motivation for adopting long-term weight tracking, learned about my fluctuations and patterns, and found flaws in my data that made me reconsider the tools and approaches I was using to collect it. The presentation was a chance to share some of what I learned from visualizing what I’d collected.

I could argue that the most important outcome of my projects has been that they have given me the motivation and knowledge I need to successfully maintain in a healthy weight range.  Although I have been at a healthy weight many times before in life, I’ve struggled with both motivation and methods and not been able to achieve long-term maintenance.  But another very important outcome has been that along the way, I’ve learned that I can make improvements to my data collection approaches but this requires looking at my data in a more comprehensive way.  Some of my data issues were non-obvious until I was able to look beyond the shorter time frames of the data summaries provided by apps and view my data over a several year period.

One example I used in my talk was how I swapped out my BodyMedia FIT for a Fitbit Charge HR because I found patterns in my data that showed I was using the band much less in the summer (short sleeve weather) than in the winter (sweater weather).  In other words, my step numbers and hours of activity were tracking exactly with my hours of device usage.

Seasonal compliance

Immediately upon seeing this graph, which I created for my JMP Discovery poster last year, I knew it was my vanity that had cost me, since I was doing this intentionally to avoid tan lines on my upper arm.  I just didn’t realize how much this was impacting the quality of my data when I was only looking at daily or weekly reports.  Once spring hit this year, I was in the market for a wrist-worn device with less obvious tan lines and so far, I have loved my Fitbit Charge HR!

The four years of food log data I got out of BodyMedia’s software  was very hard to extract, as it was buried in monthly PDF files.  I experienced some data loss in the conversion, with portion sizes and quantities less than 1 being truncated to 0.  Plus the app had a few read/write quirks that were irritating, and no way to copy multiple food items at once.  I have since switched my food logging over to MyFitnessPal (MFP), which lacks these quirks and has a much more comprehensive food database than BodyMedia.  Although I am somewhat wary of MFP’s user-contributed items, MFP does mark selected items as verified.  With my many years of food logging experience, I can usually assess quickly whether calorie estimates for a given item seem reasonable.

With food logging, it’s less about precision to the calorie or even the absolute value of the numbers and more about habits and accountability.  I’ve heard negative opinions on logging that range from “nailing jello to a wall” to “too time-consuming” but personally, committing to long term logging has been the best decision I’ve ever made.  Food logging works best for me when paired with a food scale or measurement tools like cups and spoons.  When you use both together, though, you start to get a sense that the portion sizes listed on packages are actually rounded relative to the grams of food each contains.  The worst discrepancy I have ever seen is peanut butter, which lists a serving size by amount as 2 TB and grams as 28.  I have seen 2 TB of peanut butter weigh as much as twice as what the label lists the serving size to be, which would mean instead of getting the 180 calories you think you’re getting, you might be getting twice that!  Not worth the calories in my opinion.

There is a widespread perception out there that calories don’t matter as long as you’re eating “healthy” foods.  But there are plenty of people out there who have gotten fat while eating too much of “healthy foods.”  And there are plenty of people-myself included-who lost weight and have maintained that loss while eating some proportion of calories from “junk foods,” which in my case included a considerable amount of dark chocolate.  I like that in MFP, I can log specific brand names for items in a way that I couldn’t do “out of the box” with BodyMedia’s app.

I did enter quite a few recipes and custom items into my personal BodyMedia database over the years, but I lacked the patience to enter, for example, detailed nutritional info for every variety of dark chocolate I ate during that time period. As a result, I can’t trace back how many pounds of Lindt sea salt dark chocolate bars I ate over the past few years in my BodyMedia data, but I can do this for the past 80 days I logged in MFP.  Knowing data at this level of detail is interesting and perhaps even helpful for food budget planning.

While I do try to get my whole foods, veggies and lean proteins in every day’s meals, I am pretty philosophical when it comes to how I view various food religions.  And make no mistake, that is what specific styles of eating have become to most people.  The concept of orthorexia was a new one to me several years ago, and I learned more about it via the Venus Factor podcasts.  But there is a rich history of psychological research that shows people who are orthorexic, who subscribe to food ideologies and lists of “good” and “evil” food have a lot harder time recovering from dietary “indiscretions.”  While I know there are some foods that just don’t work well for me, I try very hard to allow myself to have all foods in moderation and avoid orthorexic thinking. If my goal is to keep total calories in a certain range, there is more flexibility about food choice.

A friend relayed a comment recently on Facebook when she mentioned calorie counting that someone made to her,”I’m glad you found something that works for you!”  I had to laugh because calorie counting, when done at a sufficient level of detail, can pretty much work for anyone who is willing to collect some data and prove it to themselves that eating in a deficit causes weight loss.  The harder part is to admit some things that may be unpleasant to consider:

  • I ate too much to get to this overweight state
  • I need to eat less than I burn to lose weight
  • I’m probably going to need to eat a lot less than I want to in order to lose weight

Calorie counting just one tool that can prove the truth of these statements to you.  I’ll be blogging in the coming weeks about some of the tools and technologies I learned about at the conference that compliment food logging in the weight loss and maintenance process, like tools that can help estimate metabolic rate and provide high resolution information about body fat percentage.  I’m excited to incorporate some of these into my own maintenance strategies to ease this long-term process!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s