Wednesday, May 18, 2011

Partisan Grading: Democratic Professors Are More Likely to Redistribute Grades Than Republicans

From a very interesting forthcoming paper in the American Economic Journal titled "Partisan Grading" by economists Talia Bar (Cornell) and Asaf Zussman (Hebrew University):

Abstract: We study grading outcomes associated with professors in an elite U.S. university who were identified using voter registration records as either Republicans or Democrats. The evidence suggests that student grades are linked to the political orientation of professors: relative to their Democratic colleagues, Republican professors are associated with a less egalitarian distribution of grades and with lower grades awarded to Black students relative to Whites.

Dataset:  59,874 grade observations from 17,062 students taking 3,277 undergraduate level courses with 417 Republican and Democratic professors.

Results: The variance of grades is higher in courses taught by Republicans than in courses taught by Democrats. Moreover, in additional analysis we find that relative to their Democratic colleagues, Republican professors tend to assign more very low and very high grades: the share of the lowest grades (F, D-, D, D+, and C-) out of the total is 6.2 percent in courses taught by Republican professors and only 4.0 percent in courses taught by Democratic professors; the share of the highest grade (A+) out the total is 8.0 percent in courses taught by Republican professors and only 3.5 percent in courses taught by Democratic professors. Both differences are highly statistically significant. These suggestive results are consistent with our grading egalitarianism hypothesis.

A different illustration of the relationship between political identification and grading egalitarianism is contained in the chart above. The figure displays mean grades by student SAT score ranges in courses taught by Republican and Democratic professors. The observed pattern is consistent with the hypothesis that Republican professors are associated with a steeper slope of the grade-ability profile, i.e. with higher returns to student ability.

Conclusion: We found that relative to their Democratic colleagues, Republican professors are associated with a less egalitarian distribution of grades and with lower grades awarded to Black students relative to Whites.

Professors control the allocation of grades which serve as the primary currency of academia. Our results suggest that the allocation of grades is associated with the worldview or ideology of professors. This finding may inform the public debate on potential reforms to university grading practices. To the extent that the application of objective standards is an important university goal, policy makers should consider limiting the discretion professors enjoy when it comes to grading and making it more difficult for them to use student characteristics as factors in the grading process.

MP: One conclusion here might be that highly motivated, high-achieving students should prefer classes from Republican professors because it's more likely they'll be rewarded with a really high grade (A or A+), and less motivated, lower-achieving students should prefer classes from Democratic professors, because it's less likely that they'll receive a really low grade. 

Thanks to Norman Berger. 

22 Comments:

At 5/18/2011 10:26 PM, Blogger BBL Jr said...

So the question becomes Do we value achievement or do we value lack of achievement? After all do we want students to earn academic degrees or certificates of self esteem?

 
At 5/18/2011 11:06 PM, Blogger Che is dead said...

Petition to Redistribute GPA Scores

 
At 5/19/2011 4:27 AM, Blogger George Phillies said...

There does not appear to have been a control for academic discipline. It is fairly well known that some fields have had more grade inflation than others, by a lot. Also, some fields really tend to draw only very good students. In my 30+ years experience, students tend not to become physics majors unless they are really motivated to do the work, or unless they really really have little connection with what is happening around them.

It might also be of interest to ask if the students objectively we showing similar levels of performance from the Republican and Democratic professors, or whether they earned their grades, professor by professor.

As I hear discussion from here and there, there seems to be a marked increase in the extent to which grade distributions are bimodal, with more really good, motivated students and more students who are helping their classmates by padding the bottom of the grade distribution.

 
At 5/19/2011 5:36 AM, Blogger geoih said...

Of course the recommendation is for more centralized control and 'standardization'. Any guesses on what that would actually produce?

 
At 5/19/2011 6:11 AM, Blogger Ecclesiastes said...

Perhaps it's because I'm used to reading reports and letters in Science and Nature, but this paper looks like an op-ed column.

 
At 5/19/2011 6:47 AM, Blogger Rand said...

Are grades awarded (as in a prize) or earned (as in a paycheck)?

 
At 5/19/2011 7:07 AM, Anonymous Anonymous said...

I thought of the same thing as che.


http://www.overcomingbias.com/2011/04/natural-hypocrisy.html

 
At 5/19/2011 7:17 AM, Blogger VangelV said...

Perhaps it's because I'm used to reading reports and letters in Science and Nature, but this paper looks like an op-ed column.

Papers in the soft sciences tend to be like op-ed pieces. Sadly, Nature and Science tend to publish many such papers. All you have to do is to look at their 'climate science' literature and you see op-ed pieces sexed up by adding some dubious statistics.

 
At 5/19/2011 8:12 AM, Blogger MikeRussellMcK said...

@George_Phillies: Actually they do try to control for department, see table 4 and the discussion around it.

What I'm can't find is any control for population. Wouldn't we expect a wider variation in average scores when looking at 27 Republicans versus 370 Democrats?

The race findings are problematic, but as the authors point out Democrats are also grading black students lower their their white peers, suggesting it is more complicated.

 
At 5/19/2011 8:32 AM, Blogger Michael Hoff said...

"Republican professors are associated with ... lower grades awarded to Black students relative to Whites."

Oh, so it's a hit piece! See, I thought they were maybe trying to demonstrate a difference in philosophy between democrats and Republicans, but no. No, they just wanted to prove that Republicans are racists. And, hey lookie here! Here's the proof!

Someone needs to be swatted right in the teeth for this.

 
At 5/19/2011 8:40 AM, Anonymous Anonymous said...

Rand, grades should be earned, and students should always know where they stand. I allow discussion about points that lead up to a percentage that is ultimately translated into a grade at the end of the semester, but I don't discuss the final grade itself unless there is a proven mistake. If you allow it, too much class time is burned on attempted grade negotiation, so I clearly spell out my policy in the syllabus and stick to it. The word quickly gets around the student population on who bends on grades and who doesn’t.

Other than requiring a clearly-written grading system in the syllabus, college administrators should stay out of the grading business because they are too far removed to make that call. If I had read the last sentence of the paper first, I don't think I would have spent my time reading it.

 
At 5/19/2011 9:07 AM, Blogger Methinks said...

There are Republican professors? I learn something new from this blog every day!

 
At 5/19/2011 9:09 AM, Blogger Sean said...

Objective standards would be associated with something like standardized tests, which don't exist in many serious disciplines. And that doesn't even factor in the skill of the professor teaching. So grading tends to be done on a curve, with the expectation that some students will always do well, others poorly, etc. That works better than you might expect (although still not really well) because of its ability to adjust to so many more variables than objective standards.

What they're really arguing over here is simply the shape of the curve.

 
At 5/19/2011 9:16 AM, Blogger George Phillies said...

With 27 Republicans in the sample, the fluctuations are going to be a bit large.

However, with respect to the racial issue, I am reminded of two universities with which I have in some sense been associated, both fairly large. Both had what were said to be support programs of very different sorts for African-American students. At university A, performance of African-American students on the whole -- a few sterling exceptions -- with was somewhat dismal, including in fields and exam methods (A solution of 6 grams per liter sodium chloride in water has a molarity of (a) (b) (c) (d) ) that appear challenging to bias.

At university B, emphasis was at creating a culture among students that failure was morally unacceptable and hard work and helping each other over difficult points was the solution. Student performance, almost entirely in science and engineering, was higher among African-American students than among white students.

 
At 5/19/2011 9:20 AM, Blogger morganovich said...

i find the downturn in grading from democratic professors for the students of the very highest SAT scores VERY interesting.

could it be that the professors can't handle students who are smarter than they are and therefore penalize them?

such a result should be unaffected by the biases from subject latter (neuroscience vs sociology) as it's the same classes. (and i agree that much of the divergence simply shows that science professors are republicans and humanities are democrats and that it's very difficult to get anything less than a b in humanities)

so why do those with 1600 SAT's fare worse than those with 1500's when taught by a democrat but do much better when taught by a republican?

that's a helluva divergence.

 
At 5/19/2011 10:24 AM, Blogger Ironman said...

MP writes:

One conclusion here might be that highly motivated, high-achieving students should prefer classes from Republican professors because it's more likely they'll be rewarded with a really high grade (A or A+), and less motivated, lower-achieving students should prefer classes from Democratic professors, because it's less likely that they'll receive a really low grade.

Assuming the observed patterns are valid, which kind of professor would university administrators prefer to hire? (Or rather, which kind of professor might cause fewer problems for university administrators because of their grading practices?)

 
At 5/19/2011 11:10 AM, Blogger Buddy R Pacifico said...

Ironman writes:

"Assuming the observed patterns are valid, which kind of professor would university administrators prefer to hire? (Or rather, which kind of professor might cause fewer problems for university administrators because of their grading practices?)"

Yes, the administrators want to hire professors that demand less of the students. This then grows the adminstration; with more programs, administrators and funding for higher administrative salaries.

 
At 5/19/2011 11:44 AM, Blogger juandos said...

"The evidence suggests that student grades are linked to the political orientation of professors: relative to their Democratic colleagues, Republican professors are associated with a less egalitarian distribution of grades and with lower grades awarded to Black students relative to Whites"...

Hmmm, could that explain Obama's alledged college degree and supposed law degree?

 
At 5/19/2011 4:00 PM, Blogger Bobby Caygeon said...

I am reminded of the time when after working very hard to EARN the highest grade in a senior level Finance course the soon to be retiring Professor came in and told everyone that as his "gift" the top quartile were all getting A's, the bottom quartile C's, and everyone in between B's.

Needless to say it made a lot of underachievers very happy and me a Republican.

 
At 5/19/2011 4:33 PM, Blogger andyweintraub said...

When I was teaching economics in a university, the economics department, the finance department, the engineering departments, the accounting department, and other "quantitative" departments were dominated by politically conservative, probably Republican, professors. The difficulty of the subject matter alone would probably produce a wider distribution of grades than one might find in the English or history departments.
Did these authors take that into consderation?

 
At 5/19/2011 6:48 PM, Blogger bix1951 said...

ever hear of a "blind audition"?

 
At 5/20/2011 9:20 AM, Blogger Hayrack said...

If admission is biased to improve diversity within a student body by accepting lower academic credentials of minority students, wouldn't that necessarily be reflected in a non-biased grading system?

 

Post a Comment

<< Home