« Starting Early | Blog Main | Scientiae Carnival is up and better than ever. »

So Is This Research Design Republican Or Democratic?

I was planning to write a post on "transcending the debate," the theme of this month's Scientiae Carnivale but I did write about that earlier this year and if I don't write the following, I may explode!

I have on my wall a newspaper clipping from a 1998 USA Today headlined "Republicans, Statisticians Disagree on Sampling" . The headline made me laugh and the article made me want to cry. Partisan politics was trumping statistical good sense.

First it was sampling, now it's research designs that are being colored blue or red. The American Educational Research Association’s August/September Research Policy Notes tells us that Democratic senators Sherrod Brown (Ohio), Barack Obama (yes that Barack Obama) and John Kerry (MA) introduced an amendment to the FY 2008 appropriations bill that includes a provision to prohibit the Secretary of Education from conducting an evaluation of Upward Bound using a randomized controlled trial. This is the opposite side of the coin (or the aisle) from the Republican sponsored requirements for randomly controlled trials.

Come on folks—whatever happened to choosing a research design based on whether the results will answer the research question rather than choosing one based on your party affiliation? There are many times when randomized controlled trials are the best choice, there are lots of other times where they aren’t feasible and other times when they aren’t appropriate.

Here is a radical idea, how about leaving the choice of the research design to researchers rather than legislators?

PS The Research Policy Notes aren't on line so if you want a copy of the paper version, let me know