I logged my snack intake for a month and made charts out of it.
Awhile back, I visited the Big Bang Data exhibition at the ArtScience Musuem in Singapore. One of my favorite exhibits there was a collection of postcards sent from one data scientist to another, titled ‘Dear Data’. You can see some of Giorgia Lupi and Stefanic Posavec’s postcards in their soon to be published book. They would collect personal data ranging from times they were annoyed in a day to the number of goodbyes said and hand draw charts on postcards to be mailed to each other.
I really liked the idea of collecting data on something seemingly mundane on a day-to-day basis, and decided to start with this: snacks.
I organized the types of snacks into three categories, high-guilt, neutral and low guilt (healthy!) snacks. They’re relative to how I feel about them, so what might be a low guilt snack may not apply to you. Examples of each follow:
- high guilt: chocolates, canned drinks, cakes, ice-creams, instant noodles, Krispy Kreme donuts, potato chips, Hello Panda chocolate-filled cookies (omg).
- neutral: wheat crackers, bread, sugar-free drinks, coffee, Milo.
- low guilt: yoghurt, granola bars, fruit juice, milk, dried fruits and nuts
Every time I had a snack, I’ll mark it as ‘1’ intake. If I ate two servings of fruit in a day, it will get marked down as ‘2’. The idea is to track frequency rather than actual amounts of snacks. This went on for a total of 4 weeks.
Disclaimer: This wasn’t done very scientifically by any stretch, but still fun.
Here’s what I found out:
- I snack an average of 3.5 times a day
- The highest number of times I had a snack in a day was 8 (it was a rough day)
- I tend to snack twice as much on Mondays than I do on any other day (you can tell from the regular spikes)
- I am twice as likely to snack on high guilt snacks than neutral or low guilt ones.
- My snack intake went down as time went on with this project. (I also started eating more low guilt snacks)
For the curious: link.
I have a month’s worth of financial spending data, which is slightly more complex and fascinating (to myself, at least). More charts may follow!