LIES, DAMNED LIES, AND YOU KNOW FUCK ALL ABOUT STATISTICS

Ah, college. I remember those brain cells fondly...

I certainly have my doubts about whether AnonymousPoster is really Virginia_1776, but one thing I have no doubts at all about is that AP is a fucking moron. In this post AP had this to bleat about the article in Science about attitudes toward evolution:

I too read JC's article and have to wonder how he based his frazzled opinion on one chart he found posted on someone else's liberal blog. There is no statistical control data listed as to who was polled or how many were polled. There is no data at all about any educational level of those polled or their political affiliations.

Fascinating. And by fascinating, I mean I am fascinated by how lazy and ignorant some people can be. It took me all of 5 minutes looking through the links from the newspaper article, to the magazine article, to find this link (PDF) to the actual Materials, Methods and Results of the survey. You'd expect that if AP wanted to try and slice at me over not having my data ducks in a row, that you'd, I don't know, make sure the data isn't right fucking there, for all to see?

Right away, AP seems to be heaving around some weighty terms, in an effort to make the reader think she has some expertise. It happens a lot when someone knows just the tiniest sliver about a topic, and wants to bluff that they have more. Anyone who spends time on Paltalk will be more than familiar with the "PT Professor." AP, my little friend, control groups are used in experimental designs, not in correlational surveys. Experiments depend on differences between groups. Correlations depend on the interactions of different variables within the same sample.

In the Supplemental Materials you'll see a description of all the surveys, including the number of participants for each survey. Further on, you can see that the surveys included data on age, gender, educational level, religious belief, pro-life beliefs, political ideology, knowledge of genetics, as well as attitudes toward evolution and science & technology. They used a simple regression model to find the degree of effect of each independent variable on the dependent (attitudes toward evolution), and lo and behold found a significant negative correlation for religious belief, pro-life beliefs, and political ideology. It also comes as no surprise that having knowledge of genetics is positively correlated with acceptance of evolution.

Of course, it goes without saying that correlation does not equal causation, and variables can have an interactive effect, but it seems like the data supports the idea that the more religious and conservative you are, the less likely you are to accept evolution. You see how easy this is when you actually bother to do your homework?

JC