Bad Science Reporting



Patrick is a mid-40 year old geek with an undergraduate degree in mathematics and a master's degree in Information Systems. Nothing he says here has anything to do with the official position of his employer or any other institution.

Related Post Roulette

43 Responses

  1. Avatar Glyph says:


    [fills corncob pipe with bacon, lights it up]Report

  2. Avatar Oscar Gordon says:

    Ah statistics & context, what kind of BS can’t you spin with them?Report

  3. Avatar Damon says:


    We should be ENCOURAGING people to eat this stuff. They die sooner. That combined with limited medical treatments to folks who have these cancers, ’cause it’s their own fault for not eating healthy, we’ll save MASSIVE amounts of cash which we can then use to cover more people so they can get dental and eye care from the ACA. See how it all works?!Report

  4. Avatar Jaybird says:

    Hitler was a vegetarian who went on to commit suicide.

    If we start extrapolating out from that, there are many conclusions we can reach.Report

  5. Avatar Chris says:

    The title of this piece is redundant.Report

  6. The WHO should be able to conclude that the only cause of death is drug abuse.Report

  7. Avatar North says:

    Yeah I was seriously considering buying some steak in honor of this announcement.Report

  8. Avatar b-psycho says:

    Nuance doesn’t get enough eyeballs. It always has to be “FISH YOUR PLEASURE!! EAT NOTING BUT SPINACH OR YOU WILL DIE MISERABLY!!”

    Eh, I don’t deny the unhealth of the average diet, or of mine. Though when I was doing as close to the right things as I’ve ever done is when I ended up having to fight off the C. I’ve since decided that if DNA is that unpredictable, I’m just going to enjoy what I have left and not care. Death is going to suck anyway, why should life suck too?Report

  9. Avatar Roland Dodds says:

    Thank you Patrick. Like most Americans, I would like to interpret your argument in a beneficial way towards my current behavior and not eat 8 hot dogs.Report

  10. Avatar Richard Hershberger says:

    I take as a baseline assumption that any journalism that involves percentages, and especially comparing percentages, will be a complete botch. Because math is hard, don’t you know? We can’t be expected to keep straight what we are comparing.Report

    • Avatar Glyph says:

      I figured someone around here would be writing about the latest “Hot Hand” news.Report

      • Avatar Chris says:

        We are undoubtedly streaky, but basketball is the worst domain to look at, because there’s so much noise (related, e.g., to the bounciness of the ball). Better to look at golf putts or dart throwing, say. Full geek.Report

      • I had not seen that yet. Thanks for the tip. But big picture here. SABR (the Society for American Baseball Research) is divided into the stats camp and the history camp. I don’t think it is any secret that I am solidly in the history camp (19th century ghetto, over by the latrines). I respect the stats side. They have done impressive work, and my understanding of the game is improved by it. I am happy to eat the sausage, but I am not qualified to make it, or even to discuss how it is made.Report

    • Avatar Morat20 says:

      Or inflation. Try reading some of the economic reports and the ridiculous comparisons they make, because nobody bothers to index to inflation, GDP growth, or population growth.

      Of course, politicians do that too. (“We increased school spending 5%! What are you complaining about? Well, enrollment is up 10% and so we sorta have a problem here….”)Report

  11. Avatar trizzlor says:

    You can never shit enough on science journalism, but this story was covered admirably in The Atlantic by Ed Yong. Strength of evidence vs. degree of risk is probably the most important and misunderstood concept in scientific reporting.Report

  12. Avatar Aika says:

    In this article, you mentioned that smokers are 15 to 30 times more likely to get lung cancer than a non-smoker. But according to the CDC website, statistics show that nonsmokers who are exposed to secondhand smoke have a 20%-30% chance of developing lung cancer. It also shows that secondhand smoke causes more than 7,300 lung cancer deaths among U.S. nonsmokers each year.Report