Check out this article that summarizes a (very long) paper published in Social Psychology Quarterly designed to test the hypothesis that liberals are smarter then conservatives.
Oh wait. That isn't the hypothesis. The hypothesis as described by the author himself, is:
the Savanna-IQ Interaction Hypothesis (hereafter “The Hypothesis” in this blog) suggests that less intelligent individuals have greater difficulty than more intelligent people with comprehending and dealing with evolutionarily novel entities and situations that did not exist in the ancestral environment. "Apparently since liberals are supposedly more willing to give money to support other humans* to whom they are not related, then liberals are more evolutionarily advanced and also more intelligent. (I wonder what Darwin would say about this?)
Unfortunately, the actual hypothesis has been lost in all the media hype, which I can only assume was intended. Who doesn't expect a lot of attention when publishing a paper explaining "Why liberals and atheists are more intelligent"?
I will confess that I have not read the original research article. I am not a psychologist and I am not familiar with the theories and data presented within the paper. However, a person doesn't have to know much about anything to see a bar graph like this one and not get a little bit concerned. I have done enough of my own research and tried to crunch my own data to know that things rarely, if ever, look this neat.
So a little searching led to this article, in which the methods used by the original author are examined and found to be seriously lacking.
As much as this whole topic gets my blood boiling, the idea that someone thinks they can prove that liberals or conservatives are really genetically or intellectually superior is not what upsets me the most. The most glaring issue is that the bar graph which seems to so convincingly support the apparent (if not actual) hypothesis is completely made up! The bar graph shows Mean adolescent intelligence (IQ) versus Adult Political Ideology, but the author never actually measured IQ!
If any of my students ever tried to present their data by claiming it represented some value that it didn't represent, I would give them an F or make them rewrite it. This highlights in the best possible way the importance of presenting your data clearly and being completely transparent about the methods. No matter how much you believe your own hypothesis, you have to represent the data with as little 'spin' as possible. Yes, I know, we all try to make the data support our hypothesis. Of course we do. But we all know that we all do it, and that is why as a graduate student we spend so much time learning how to read other research papers, examining the data and drawing our own conclusions.
|Figure 1: Here is a graph representing data I wish were true. Nom nom nom, I'll go buy stock in Reese's!|
*and that isn't even true, either. Although I'm sure the data in this study is also flawed in some way. Which just proves the point, which is, you cannot take these things too seriously!
PS: If you would like to look at some completely fabricated data, and also get a very good laugh, check out the pie charts presented here.