Tuesday, November 4, 2014

Support for Science by Party ID

I wrote yesterday about a survey question on how often respondents believe government should rely on science to solve problems. You can click back to see the question wording, etc. I thought I'd toss this graphic in below looking at Democrats versus Republicans. While the differences are statistically significant (X2 = 115.2, df=4, p<.001 for you nerds out there), as you can tell from the graphic they're not all that substantively different. Still, Republicans clearly are less supportive of this idea, especially in the "some of the time" and "half of the time" answers as compared to Democrats on "most of the time" or "always."

Percent Answering How Often Science Used to Solve Problems?

Looking at simple relationships, you're more likely to want the government to use science if you are liberal, a Democrat, younger, more educated, and of greater income. A regression model in which various factors statistically control for one another finds that age (younger, beta = -.06, p<.001), education (more so, beta = .21, p<.001), gender (male, beta = -.11, p<.001), not being a Republican (beta = -.08, p<.001) and being liberal (beta = .14, p<.001), are all associated with a preference for government using science, even as they statistically control for one another. 

Data Source

No comments: