A somewhat useful data logging of a years' worth of meals

One woman documented her meals for a year and turned them into a series of data visualizations. What did I learn? Well, for one thing, how important it is to gather and group data into logical and easily comparable categories. For another, this woman ate more french fries than fruit. (Via Liz Landau)


  1. Come on Maggie….

    The 1st rule of making a good graph is labeling your axis and conveying that information in a proper way. From looking through all her graphs I’m guessing they represent the number of times she had a particular thing.

    So in that sense she does eat a lot of chicken, almost everyday.

    But in a different respect all these graphs do not tell us how are calories are consumed. And isn’t a french fry really a veggie? (Sure it has been deep fried, but what about veggies I saute/stir fry/grill?)

    It’s interesting, but I’m not sure why we have some things labeled as groups (veggies, seafood) and others as specific items (sushi, french fries). My questions abound…

    1. Perhaps I should have been more clear.

      “Well, for one thing, how important it is to gather and group data into logical and easily comparable categories.” = a critique.

      I think the choices she made about what data she collected are confusing and some of the visualizations don’t tell me anything at all. But it’s an interesting project.

  2. You made me think about it, and amazingly enough I now eat more fruit than french fries! It hasn’t always been that way, because what some people claim is definitely true: it’s expensive to eat right. I’m able to grab a bag of apples, oranges, mandarins, bananas, whatever I want when I’m at the store. I love having fresh fruit around now. French fries are cheaper. (Per calorie.)

  3. Also, the color wheel thing is cute, I guess, but it makes it really hard to see any lines whatsoever between many neighboring sections…

  4. “…this woman ate more french fries than fruit.”

    I think you’d be hard-pressed to find an American for whom the above statement is not true. Unless you count diet cherry coke as a fruit (which is reasonable, since it has “cherry” in the name).

    1. :raises hand: Um, that’d be me. Fruit at least every day, fries….only once or twice a month.

      1. I don’t even like french fries, which might be why I have room for seven to tens portions of fruits and vegetables. I can’t even understand why anyone would find fried starch more appealing than roasted asparagus or a ripe apricot. Although I wouldn’t be surprised, particularly for women, if it’s not about getting more salt.

    2. I haven’t eaten fries in years, although they are delicious.

      And I haven’t drunk soda in decades. I prefer beer.

      Don’t assume all Americans are like the ones on TV.

      1. We don’t need to assume anything. We have statistics and our damn eyes to corroborate this empirycally.

    3. “I think you’d be hard-pressed to find an American for whom the above statement is not true.”

      That can’t possibly be true, can it? Does everybody have their own deep frier or something?

      1. Well I have Fry Daddy up in the cabinet…

        But most of the time I just use a deep pot and a candy thermometer…

        About the only thing I fry at home is chicken, and that might be once every couple of months. It taste great, but that grease smell lingers in the house for a couple of days.

        -I’m not counting bacon as a deep fried food, mostly because I don’t add oil. Even though it’s cooking style is more like deep frying that anything else, assuming you are doing it in a pan on the stove.

        And all the non-Americans wonder how/why we eat some much fried food (or just french fries)? It’s the de facto standard that every restaurant serves as a side. Unless you are talking higher dollar places, or specific regional things (mexican/asian), every place else serves fries. (Of course YMMV, here in NC fried things are just as common as humidity in the summer.) Now that my wife and I have been watching what we eat it pains me that sometimes I just want a salad or fruit to eat. I don’t need 300+ calories in a small thing of fries.

        Of course I do like fries, but that might be because of this:
        When I was a kid, like 10 or so, my mom would work weekends 2nd shift. So dad and I were stuck home fending for ourselves. Which meant serious crap food. And on multiple occasions we would french fry a whole 5 pound bag of potatoes for dinner. That and polish off a 2 liter of Mountain Dew… No wonder I am/was overweight.

  5. Of all of her graphs, these are the only ones that I can actually Read and be able to glean any information from.

    Food dots, masked food photos, vector food percentages, pies monthly, 2009 vs 2010, bubbles, tree map, pie labeled, single bar graph, single pie.

    Special note for Fries By Month. I liked that one & understood it fine.

    The others are all just colorful blobs and lines and crap that I’d have a tough time coming to any conclusion from if I had to. How you present your data I’m sure greatly influences how it is reacted to.

  6. Information is Beautiful, but anyone trying to emulate Tufte or Felton needs to consider that the informational content needs to come before making it look pretty. If your graphic doesn’t enhance the data, or worse, it takes longer to parse the graphic than it would have taken to just parse a table of data, you’ve failed. No matter how many bold colors you slap on pristine white backgrounds.

  7. Why exactly does she separate ham out from pork?

    Reminds me of this Simpsons quote:

    “(Lisa) “I’m going to become a vegetarian” (Homer) “Does that mean you’re not going to eat any pork?” “Yes” “Bacon?” “Yes Dad” Ham?” “Dad all those meats come from the same animal” “Right Lisa, some wonderful, magical animal!””

  8. Yep, definitely an InfoViz fail.
    40+ different presentations of a badly conceived dataset.
    Her categories are illogical; some of them overlap, many of them represent different levels in a super/subclass hierarchy. No units information is given. One of the graphs suggests the unit is “number of meals that contained this item”, which makes most of the visualizations inappropriate, because they suggest comparisons in quantities that the data does not support.

    Oh, and they’re not really different visualizations either, as she makes no attempt to offer different aggregations/categorizations.

  9. I know she mainly did it for fun, but it bugs me a bit that she did not include beverages at all.

    She’d fit right in in Québec: It’s like every second restaurant is a chicken rotisserie of some sort. My family ate chicken so often that I swore it off for a few years when I moved out.

    And count me in as another fruit enthusiast: I need them every single day or I feel sluggish. It’s true that fresh fruit is expensive but I always saw them as an investment.

  10. I am thinking of a diet tracking website where you upload a picture of every single meal you eat and software roughly calculates the serving size and a bunch of other nutritional metrics. You also upload info after every physical and maybe your weight on a regular basis. Sure, with an N of 1, there could be a lot of slop, but you do this a million times over and I bet you would get some useful correlations.

    I just thought of the next piece of hardware to build into a smart phone – a scale.

  11. I hate to criticize a stranger as much as I do a friend, but… wow! This project seems to have a large percentage of failures in it.

    Anyone who claims to have “hoped to learn the art of data visualization” and wanted to find “the most effective way to visualize information” and then still presents the data with gray-on-white legends in tiny type has wasted an awful lot of her time.

    Some of the graphs are totally meaningless. I am left with a thesis of my own, and that is that the project was more about effective data obfuscation than data visualization. What the heck are we to make of “BAR CHARTS MONTHLY SIMPLE”?

    Even the methodology described for the “Experience Cards” she used to get feedback seems designed to avoid discovering which visualizations are bad and why. Hint: The ones I looked at the longest weren’t necessarily the most interesting or informative.

    Hoping to end on a positive note, I can say that Ms. Manning’s study has shown that chicken is probably not “brain food”.

Comments are closed.