Tag Archives: T Test

N of 1 Testing: Butter Redeemed

I had a self-created story about butter. It was such a strong story that it was changing my behavior around what I ate. I had come to believe butter “gummed me up” and after eating it my body would react by feeling lousy.

bloay

Two months ago I transitioned to a low carb, medium protein, ketogenic diet. As I described at the time, the experience of weaning myself off carbohydrates involved feeling poorly for a number of the early days. As butter was a good source of fats I was eating a lot of it, so I started associated feeling lousy with butter. And I developed the story, “This butter is throwing me a beating.”

I didn’t like the story because it cut down on my food options. And I knew from my early QS studies that the majority of my stories were self-induced fictions.  Was this story about butter a fiction or did it have some basis in fact?

My Question

Did my body feel lousy after eating butter?

What I Did

Each morning I would take butter or an alternative source of fat with my coffee. The alternative was usually flaxseed or coconut oil. I randomly determined which I would take each day. Four hours after my coffee I would capture how I was feeling in a Google Spreadsheet via a Google Form tool I call my DIY Tracker.

How I Did It

I used a Google Spreadsheet to generate a random “one” or “zero” for each day in the period. This gave me instructions on what to put in my coffee each morning, butter or the alternative. This was important because I needed to not choose what I took based on some bias, or more importantly, the story about butter. If I felt rough in the morning and arbitrarily chose the alternative fat source because of my belief  butter  would make me feel worse I would skew the data. So I stuck to my random schedule when I took my coffee after a workout at around 7am.

I captured how I felt each morning at 11am using my DIY Tracker. This provided my data for how I was reacting to butter and the alternative.

IMG_0112

I threw out data that did not meet my control criteria, which is a fancy way to say if something out of the norm was happening I did not use that data. If I did not follow the randomly generated instruction or the amount of fat I consumed was not within a set range, I considered the reading invalid. I also only included readings on days I had exercised for 20 to 30 minutes prior to the coffee.

After thirty-seven days I had twenty-seven good data points. I separated the list into a “butter” and “other” list and ran a T-Test for butter and a value for feeling lousy that was embedded in the way the DIY Tracker poll solicited information.

What I Learned

Butter was redeemed. I compared it both to an average score that would indicate I was consistently feeling lousy and to the oil based source of fat. The way I gathered information in the DIY tracker a consistent score of “2” (Yawny, tired) would have shown I was feeling lousy.

The data showed I wasn’t consistently feeling lousy. There was a statistically significant difference between the feeling score after eating butter (avg of 3.6) and the expected average score of feeling lousy (avg of 2.5). The T-Test returned a value of .004. Within the data I only felt lousy on two of the 27 days.

When I consumed the oil based sources of fat my average feeling score was higher than with butter, but the results were not statistically significant when I compared oil to butter (T-Test returned .17). So butter remains in the morning coffee mixture rotation.
butter

Behold the redeemed!

Try Your Own Test

Do you have a food story? A cherished belief that you can’t eat a certain food because it makes you feel bloated, or bad? N of 1 Testing is quite easy to set up when you have a specific question you are asking based on a story you have. Get free instructions on how to create you own DIY Tracker or we can help you out at QuantXLaFont.

Sign up for the QuantXLaFont Newsletter
Get our lifestyle tips and studies delivered to your inbox.
Thank you! We don't spam :)