« Why are GMAT scores going up? | Main | Reader's guide to the power pose controversy 1 »


Feed You can follow this conversation by subscribing to the comment feed for this post.


While I understand your main points, I don't think you can use Independence Day and Labor Day to refute the points about Thanksgiving, Christmas, and Easter.

Yes, this is anecdotal too, and I realize it's secondary to your overall point, but there are huge differences in the way the holidays are often celebrated.

The former often involve cookouts, and many people undoubtedly eat a lot of food for a portion of a day, often while also consuming large volumes of alcohol. Foods most often involve things like burgers, hotdogs, chicken, and various salads.

But the latter holidays often involve travel, going back home, bring family together, spending several days baking pies, cookies, wide arrays of side dishes, with main courses often involving roasted meats, smothered in gravies, stuffings and starches, with leftovers that last for days, etc.

There is a fair likelihood that there is a significant difference to be found in people's weights as a result of these holidays, whether or not this study found it, or was properly conducted.

Equating one type of celebration with the other is not really one of the criteria that can be used to criticize this analysis, in my opinion.



jlbriggs: The title of the research is "weight gain over the holidays." The concluding paragraph starts with "In three prosperous countries, weight gain occurs during national holidays." The authors recommend "advising a patient to have better self-control over the holidays." So, the researchers claim to have learned something about all national holidays. As for different eating habits during different holidays, we can have a debate if they actually have data on food intake but they don't!


Weight gain is a surrogate for food intake. All they need to do is change the hypothesis to holidays with food traditions cause people to put on weight and it works. One of teh problems with the analysis is that people seem to put on weight during winter, which is not unsurprising, as they spend more time indoors. THis especially makes the effect of thanksgiving higher than it should.


Kaiser - like I said, I fully understand your points.

I am just pointing out that one of your arguments was, IMO at least, rather flawed.


Ken/jlbriggs: I don't agree the argument is flawed. I don't think it is possible to explain a weight loss around July 4th, when there should clearly be an uptick in food intake. I think for this study to be credible, it needs to explain what is going on around all major holidays not just the ones that fit the hypothesis. In addition, tailoring one's hypothesis to exclude data that do not conform to it is not to be encouraged.

There are many other problems with this analysis. As Ken pointed out, they use one year of data so there is no accounting for seasonal effects. Plus, are people who weigh themselves frequently the ones more likely to have weight gains? etc.

jlBriggs: thanks for starting this debate. At least, we both agree on the main point of this post: there are many porblems with the blueprint for analyzing Big Data/found data.


Well, again, I think it's easy enough to explain - again, purely anecdotally - by the simple fact that the "uptick" in food intake is not as universally true on the 4th of July (many people will not eat any more than usual at all, they simply eat a meal in the context of a party), and that even when people do eat more, they're eating more at a single meal on a single day.

I am not sure what your experience with Thanksgiving, Christmas, and Easter is like, but for many Americans, it is a prolonged and significant increase in food consumed.

Obviously, to capture the effect of eating on weight, we would need to be able to capture a great deal more information than just people's weights, and I would certainly want to be able to explain, with data, the difference between such holidays if I were reporting on such a study. Ignoring the holidays that don't fit, rather than explaining why they don't fit, is a poor approach.

But if we're speaking anecdotally, there are explanations to be found :)

The comments to this entry are closed.

Get new posts by email:
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.

See my Youtube and Flickr.


  • only in Big Data
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here

Principal Analytics Prep

Link to Principal Analytics Prep