PhD Comics, via Missives from Marx and Dmitriy T.M.
Common Mistakes in Polling and Poll Results
11 Comments
Comments
-
Genevieve
I’m not sure “mistakes” is the right title for this. When I’ve seen news media doing 1-3, my first thought is usually not that they need to hire a new statistician. Most of the statistical “mistakes” I spot on the news seem to be done intentionally, assuming (or just hoping) that the average viewer won’t be “someone who took a basic statistics course”, so that they can spin a more sensational story.
-
mercurianferret
It’s hilarious because all these things are basic points that anyone should know at the end of their first basic statistics course. Further, I don’t understand why such literacy isn’t a mandatory part of general societal education. (You should minimally be able to determine if someone is yanking your chain about the meaning of the presented statistics.)
Of course, if “everyone” were able to tell the above 1-3 as chain-yanking, then obfuscation will just [have to] become more canny.
-
Jared
As mercurianferret pointed out it’s dangerous to laugh and assume that these are unintentional mistakes. When a station like FOX presents “Climate Change” as an issue to be decided by an opinion poll of their Southern US viewers, it’s not a cute little mistake. It’s extremely malicious.
-
Austin C
I knew Larry King was old, but sheesh.
-
Kat
What I find really really maddening is the reporting of studies produced by universities/research teams etc, which do not include basic facts such as the number of people who took part in the study, how they were recruited, what exactly they were asked, who financed the study and the background of the researchers and the research institute. There have been so many BS studies, whose BS findings were then trumpeted in very sensationalist fashion by news outlets.
-
Bagelsan
I also hate when the article just flat-out lies: “Bitches are CRAZY!” the title will trumpet! Then the rest of the article will be like “relationships! Ladies! Amirite? They want commitment!” and then buried somewhere 2/3 down the page there will be an actual research/researcher quote to the effect of “men and women pretty much responded the same, also it was just college kids surveyed, also we didn’t adjust for socialization.”
It makes me die a little inside.
-
Leslie
Yeah, I stopped reading the Science section of the New York Times for that reason. I like the NYT in general but their reporting of scientific studies sucks.
A little while ago they had this big headline like “Study Shows Anti-Depressants Don’t Work!!!!” (without the exclamation points, but they were implied). Then a week or so later they published another article that was entitled something like “Read This Before You Quit Antidepressants,” which explained the study in more detail and put it into the context of many other studies which have shown that antidepressants do work. Both articles made it into the “Top Ten Most Emailed” list on the paper’s website. Very clever, NYT, but I see what you’re doing there – two sensations are better for site traffic than one!
-
-
Jamie
Haha, yeah, because people using television statistics to write news stories *totally don’t understand statistics.* I agree with most posters here that calling it “mistakes” bit naive, and a bit dangerous.
~~~Doubly maddening is the fact that people didn’t believe Earth was flat in the time of Galileo, because this was resolved in classical Grecian times. The debate being referenced was actually between geo- and heliocentric theories, but whatevs.
Signed,
-someone who took a basic history course. -
Rachel Kelly
Jamie, I love you
Signed,
-Someone that is a history geek. -
What Was in the 1941 College Woman’s Closet? » Sociological Images
[...] find it rather charming that a popular magazine would even bother to clarify anything about their polling methods. So…earnest! Leave a Comment Tags: clothes/fashion, consumption, gender, gender: [...]
Sitemeter
© 2010 Sociological Images
