Wow, that's terrible. It reminds me of an obesity study that showed no statistical correlation between BMI and greater health problems. The researchers even recognized the lack of correlation, much less causation, in the analysis section. But in the conclusion they start talking about their "conclusive evidence" showing that people with high BMIs are unhealthy. Just cause you think it's true doesn't mean you can ignore your actual data.
The best part of these things is that not only do they conduct bad science, but their underlying premise, even before the bad science is done, is suspect. In this case, conflating political ideology with personality traits (read: treating stereotypes as true a priori), in that, BMI.
According to the government's BMI scale, I'm overweight. BMI is a flawed measurement. Fat and muscle are treated identically - and I'd put good money on that having damaged their sought-after correlation.
Perhaps you should introduce the authors of that article to your fiance... call me when you want me to meet with them. Until then I'll be 409-ing my bust of Obama in the living room. :-P
Comments 4
Reply
According to the government's BMI scale, I'm overweight. BMI is a flawed measurement. Fat and muscle are treated identically - and I'd put good money on that having damaged their sought-after correlation.
Reply
Reply
Reply
Leave a comment