Friday, January 13, 2012

Inside Duke: hurting the ones we love?

This very interesting study had access to comprehensive data ranging from Duke admissions office evaluations of applicants, to students' intended majors and subsequent shifts, to grades awarded and student composition (including abilities!) for each course offered at Duke. Interesting factoid: 40% of fathers of White students at Duke have doctorates.

For similar studies (although not emphasizing ethnicity) using U Oregon data, see Data mining the University , Psychometric thresholds for physics and mathematics.

What Happens After Enrollment? An Analysis of the Time Path of Racial Differences in GPA and Major Choice

Peter Arcidiacono, Esteban M. Aucejo, Ken Spenner

May 24, 2011

Abstract
If affirmative action results in minority students at elite schools having much potential but weak preparation, then we may expect minority students to start off behind their majority counterparts and then catch up over time. Indeed, at the private university we analyze, the gap between white and black grade point averages falls by half between the students' freshmen and senior year. However, this convergence masks two effects. First, the variance of grades given falls across time. Hence, shrinkage in the level of the gap may not imply shrinkage in the class rank gap. Second, grading standards differ across courses in different majors. We show that controlling for these two features virtually eliminates any convergence of black/white grades. In fact, black/white gpa convergence is symptomatic of dramatic shifts by blacks from initial interest in the natural sciences, engineering, and economics to majors in the humanities and social sciences. We show that natural science, engineering, and economics courses are more difficult, associated with higher study times, and have harsher grading standards; all of which translate into students with weaker academic backgrounds being less likely to choose these majors. Indeed, we show that accounting for academic background can fully account for differences in switching behaviors across blacks and whites.

For a review of Richard Sander's analysis of affirmative action in law school admissions, see here.

The results of all of these studies can be summarized as: to first approximation, psychometric predictors work, and in an unbiased way across ethnicities.

12 comments:

  1. JustinLoe10:39 PM

    Note the significant grade inflation at Duke:
    White students:Year:1 3.38 GPA2. 3.45 GPA3. 3.59 GPA4. 3.64 GPA

    ReplyDelete
  2. See also...

    http://isteve.blogspot.com/2009/06/numbers-on-affirmative-action-in.html

    "A recent paper, "Does Affirmative Action Lead to Mismatch: A New Test and Evidence"
    by four economists (Arcidiacomo, Aucejo, Fang, and Spenner) has some
    inside info from the Duke U. admissions department on Duke students,
    including SAT scores and freshmen grades."

    Asian: SAT 1464, freshman year GPA 3.40
    White: SAT 1417 (s.d. 100), freshman year GPA 3.33 (s.d. 0.46)
    Latino: SAT 1349, freshman year GPA 3.13
    Black: SAT 1281, freshman year GPA 2.90

    ReplyDelete
  3. JustinLoe10:48 PM

    "
    For example, a UC Berkeley professor noticed that African American students were struggling in science classes while their Asian classmates thrived. The professor decided to take a closer look at students’ study habits. He noticed that Asian students studied in groups. If a student struggled with a problem, his classmates would help him think it through. African American students studied on their own, without the benefit of peer support. The professor then required all students in his class to study in groups. The performance of African American students improved dramatically as a result."
    http://ucsdnews.ucsd.edu/thisweek/2011/05/02_HigherEducation.asp 

    ReplyDelete
  4. LondonYoung10:23 AM

    I object to the economist authors of this study grouping "economics" with STEM rather than with H&SS.

    ReplyDelete
  5. Guy_Brodude4:43 PM

    I concur. Why not include the data for each specific major?

    ReplyDelete
  6. LondonYoung7:22 PM

    All I saw was them saying that economics majors self-reported studying a lot in one of the footnotes.  

    ReplyDelete
  7. Richard Seiter9:26 PM

    I wonder how the analysis results would change if economics switched categories.  Does anyone here know anything about the economics department at Duke?  I know economics is much more STEM-like at some universities than at others.

    ReplyDelete
  8. Richard Seiter10:24 PM

    For more about Sander's analysis see his web page with additional information here: 
    http://www2.law.ucla.edu/sander/Systemic/SA.htm

    ReplyDelete
  9. Richard Seiter10:30 PM

    Thanks for the anecdote.  It's good to hear about someone doing something that helped.  Sounds like someone needs to do a study of studying approaches by groups ;-)

    Sad data about number of hours spent on homework.  The rest of that paragraph was illuminating as well:
    "About one-third of undergraduates have never taken a course with more than 40 pages of reading per week; and half haven’t taken any courses requiring them to write more than 20 pages over the course of the semester."

    ReplyDelete
  10. Guy_Brodude10:51 PM

    Yeah, I think that's about it. So lame. Prof. Hsu's scatter plots were much more informative. I still wonder about the men (and women) behind some of those dots ;)

    ReplyDelete
  11. All of I actually spotted appeared to be these folks announcing this economics majors self-reported reviewing considerably around among the list of footnotes.

    ReplyDelete