There's a surprising source for research on controversial topics in higher education: economists. In their journals and at their scholarly meetings, they are spending a lot of time analyzing issues that are important to many academics.
The American Economic Association met this weekend in Philadelphia. Here are some of the findings of interest to academics who aren't economists:
Female professors in some disciplines have real impact as role models. Conventional wisdom has it that female students are encouraged by seeing female professors, but many critics of affirmative action have questioned whether empirical research backs that up. Eric P. Bettinger, an assistant professor of economics at Case Western Reserve University, and Bridget Terry Long, an associate professor of education at Harvard University, analyzed the records of 54,000 students. They found that female students whose first course in a field is taught by a woman are significantly more likely to take more courses in the field in certain disciplines, including such traditionally male dominated fields as mathematics, statistics, and geology. In other fields, there was no difference or even a negative influence. And before anyone cries for equal time for men, you should know that the study found that male students whose first course in education -- a female-dominated field -- was taught by a man took many more credit hours than those whose first course was taught by a woman.
More money, not more fellowships, will get more scientists. For years, experts have worried about the growing share of science degrees in the United States that are awarded to foreigners. And with some regularity, politicians say that some new or expanded fellowship program would help. Well, maybe. Richard Freeman,  a professor of economics at Harvard, analyzed data on more than 200,000 individuals who applied to the National Science Foundation Graduate Research Program between 1952 and 2004. He looked at the quality of the applicants, and noted changes in the program's size and stipend value. His conclusion: Increasing the total number of awards would result in a modest decline in the quality of recipients. But increasing the value of the stipends would lead to increases in the number of applicants and their quality.
"Open access" journals can be high quality. The high costs of journals has angered many librarians and scholars, some of whom back a new model for publishing. "Open access" journals are free to everyone; they make their money with fees to authors (or authors' institutions). One criticism of this approach is that journal editors could sacrifice quality because of their need to make money from submissions. But Mark J. McCabe, an assistant professor of economics at the Georgia Institute of Technology, and Christopher M. Snyder, a professor of economics at George Washington University, have done research that suggests otherwise. Analyzing the decisions and pricing policies of different journals, they say that the quality issue arises only when all fees are charged at the point an article is accepted. If journals split the fees between submission and acceptance, there is evidence that quality does not have to suffer.
The latter two papers may be found on the economics association's Web site. 
If you haven't read economics research on education issues, it's a little different from the studies released by advocacy groups. There is much more of an emphasis on numbers (and yes, regression analyses lurk everywhere) and less emphasis on why given policies or approaches may be morally right. But I wonder if some of the economics research -- precisely because of its emphasis on numbers -- may have the potential to reach policy makers who normally ignore research about education.