Currently browsing posts filed under "Grading"

Follow this category via RSS

Next Page →

Grade Distribution 2016-2017, 3

A source sent us the official registrar’s report (pdf) on the distribution of grades in 2016-2017. (Relevant background: data for 2008-2009 and 2013-2014, recent Record coverage, and prior EphBlog discussion.) Day 3.

The most embarrassing section of the report:

germ

I suspect that the professors in the German Department (we still have a German department?!) are feeling under some pressure to increase enrollment. One way to respond to that pressure is to give everyone an A!

Which professors in the department are most guilty? Tough to tell.

But not impossible! Here are the German courses offered last year:

germ2

There were 44 students in the beginning and intermediate 100-level language courses, taught by a mixture of professors in the department. Giving half the students A’s and the other half A-‘s is one way to encourage students to take more German classes!

Much worse, however, is the 4.12 (!) average grade for the 20 students who took either German 201 or 202. The only way to get such a high average is for 1/3 of the students to get an A+ (while everyone else got an A). Who is guilty of such absurd grading? Professor Mandt has finished up her one year visiting position. Professor Kone, on the other hand, is on the tenure track. He is teaching Germ 202 again this spring. How many A+ grades do you think he will award?

The best way for serious faculty members to fight the trend toward awarding too many A+ grades is, perhaps counter-intuitively, to celebrate such exceptional achievement. (Recall that Williams gives this grade hundreds of times each year.) The Administration could require that any faculty member who wanted to award an A+ would be required to schedule a 30 minute public presentation at which the student would present her excellent work and the professor would explain why he judged the work to be among the very best he has ever seen from an undergraduate.

There would still be — I hope there will be! — 10 or so A+ grades awarded each year. On occasion, Williams students do produce some amazing work. We should celebrate their achievements! But handing out 200 such grades each year devalues such truly exceptional work.

Trivia Question: Which Williams faculty member in the last decade was most (in)famous for handing out A+ grades?

Answer: Bernard Moore, affirmative action poster hire for the Political Science department — thanks Cathy Johnson! — and convicted felon.

Facebooktwitter

Grade Distribution 2016-2017, 2

A source sent us the official registrar’s report (pdf) on the distribution of grades in 2016-2017. (Relevant background: data for 2008-2009 and 2013-2014, recent Record coverage, and prior EphBlog discussion.) Day 2.

grades1

1) Grade inflation is still a problem. The average grade at Williams in 2008-2009 was 3.39. It was 3.41 in 2013-2014. Last year, it was 3.45. So, in the last 8 years, the average has increased by 0.06. In the last 3 years, by 0.04. If the average grade continues to increase at the rate of about 0.01 per year, then, by 2072-2073, we will reach Nirvana! The average GPA at Williams will be 4.0.

2) Grade inflation may be decreasing in severity, perhaps as a natural result of bumping up against the 4.0 ceiling. The average GPA in 2015-2016 was also 3.45. The faculty seem concerned about the problem. Perhaps things are getting under control?

3) Which faculty are most concerned about the problem of grade inflation? My prior (prejudice?) would be that older (male?) faculty are more concerned. This Record article from 1998 has some interesting background.

Facebooktwitter

Grade Distribution 2016-2017, 1

A source sent us the official registrar’s report (pdf) on the distribution of grades in 2016-2017. (Relevant background: data for 2008-2009 and 2013-2014, recent Record coverage, and prior EphBlog discussion.) Day 1.

1) Someone needs to write a thesis about grade inflation at Williams, an update, 20 years later, to “When A=average : the origins and economic implications of grade inflation at Williams College and other elite institutions,” by Peter Siniawer ’97. (And why isn’t this thesis available on-line?)

2) We need more transparency about grading at Williams. Recall my (unsuccessful) efforts to get the registrar to provide this data. Almost anything that is distributed to hundreds of faculty at Williams ought to be made public. Interested alumni/students/parents should not have to depend on EphBlog’s sources . . .

3) Division 1 should be called out for not holding the line on grades:

grades

The most distressing aspect of the differences across Divisions (and across departments) is the bad signals that it sends to students. If a student gets a B+ in an intro Computer Science class but an A in Theater, she might thing that this means she is “better” at theater than computer science. Isn’t this one way that Williams guides her on choosing a major that matches her abilities? But, of course, the College is lying to her. She is an average student in computer science and in theater. Lax grading by the latter is misleading her.

Of course, if the Theater Department, and Division 1 departments more generally, want more students, then misleading them about their actual talents may be just what the ticket . . .

Facebooktwitter

Grade Distribution for 2013–2014

An anonymous professor provided this Williams grade distribution (pdf) for 2013-2014. Comments:

1) Williams should be more transparent, especially with information like this that is available to hundreds of Ephs. (I believe that all members of the faculty are e-mailed a copy.)

2) Summary:

grades

Key question: Has grade inflation at Williams (almost) stopped? Recall this discussion from 8 years ago which quotes a Record article from 2000:

The most frequently given grade in 1999 was an A- and the mean grade hovered just above a B+ at 3.34.

If the average grade since 2000 has only increased from 3.34 to 3.41 then grade inflation, while still a problem, has at least slowed down significantly. The overall average in 2008–2009 was 3.39. Again, the thing I find most embarrassing (and what we need current data about) is the hundreds of A+ grades handed out.

3) The results by Division are consistent with the standard stereotypes.

division

The most distressing aspect of the differences across Divisions (and across departments) is the bad signals that it sends to students. If a student gets a B+ in an intro Computer Science class but an A in Theatre, she might thing that this means she is “better” at theatre than computer science. Isn’t this one way that Williams guides her on choosing a major that matches her abilities? But, of course, the College is lying to her. She is an average student in computer science and in theatre. Lax grading by the latter is misleading her.

There is much more here. Worth a week to discuss?

Facebooktwitter

Grade Inflation

grad_inflation

Imagine that Professor Kornell wants to do something about this. What advice do you have for him?

Start with transparency. What is the distribution of grades at Williams today? How has it changed over time? How does it vary by department? There is no good reason to keep this a secret, other than shame. Here is the data from 2008–2009. Here is recent data for Middlebury.

Facebooktwitter

Grade Inflation

Grade inflation is a problem at Williams, one we have discussed many times in the past. Start here for a good introduction. The most annoying aspect of the debate is the refusal by Williams to make the data public, or at least available to students and alumni.

Here are the grade distributions at Middlebury.

midd

The average grade at Middlebury has increased from 3.32 to 3.53 in 11 years. How much higher will it go in the future?

Why can’t Williams be as transparent as Middlebury when it comes to this important topic?

Facebooktwitter

Grade Inflation Update for Princeton

Grade inflation is a a scourge at many schools, including Williams. See our previous discussions here. In the last decade, the best hope for those interested in fighting grade inflation has come from Princeton. We discussed there efforts here, almost a decade ago. (EphBlog is old!). Alas, it appears that Princeton may be about to give up the fight.

A number of colleges and universities adopted policies designed to curb grade inflation. But one of the most prominent of those institutions — Princeton University — now appears poised to roll back the most controversial part of its policy: a limit of 35 percent on the A-range grades awarded in each course. A faculty report released Thursday made that recommendation, and it was endorsed by the university’s president.

See their recent report (pdf). I hope that the Princeton faculty stands firm.

Facebooktwitter

Grade Inflation a Decade Ago

From the Record

The Williams faculty voiced its concern over grade inflation at the College as it passed several motions of a proposal by the Committee on Educational Policy (CEP) Subcommittee on Grading, instituting grading targets at the monthly faculty meeting held Feb. 16.

The targets, which range from 3.2 to 3.5 from 100-level to 400-level courses, increasing one tenth of a point per course level, intend to stabilize the mean GPA of the College beginning in Fall 2000 to about 3.3, the mean grade for 1998. The most frequently given grade in 1999 was an A- and the mean grade hovered just above a B+ at 3.34.

“When you take the long view and look from 1960 to 1999, you see overwhelming evidence that grades are moving steadily upward,” Chair of the CEP and James N. Lambert ’39 Professor of Anthropology Michael Brown said.

“The problem is that leaves you no room to move. Grades are so compressed that you start getting into making finer distinctions which are harder and harder to justify at the same time.”
“The so-called ‘Gentleman’s C’ is now the ‘Gentleman’s B+,’” quipped Associate Dean for Student Services and Registrar Charles Toomajian.

Concern over grade inflation is nothing new for the College. Two Williams economics professors issued a report in 1991 indicating that departments that gave out higher grades attracted more students and increased numbers of majors. In 1996 the Steering Committee formed an ad hoc committee chaired by Mark Hopkins Professor of Mathematics and chair of the mathematics department Colin Adams.

And that was ten years ago! Adams, Toomajian and Brown are all still at Williams. The Record ought to do a follow up story next fall. Grade inflation has gotten worse in the last ten years, albeit at a slower pace. My suggestions are unchanged:

Grade inflation is a major problem at Williams. See all the posts in our grading category, especially this one. Although the rate of grade inflation has slowed in the last 10 years, it still exists and the absolute level of grades is too high, leading to all the standard problems.

Fortunately, there is a simple solution. Do what Princeton does.

President Falk ought to appoint a special committee — stuffed with grade inflation opponents like (?) Registrar Charles Toomajian and Professors Michael Brown, George Marcus, Duane Bailey and James McAllister — and charge them with investigating grade inflation at Williams and the experience of peer schools like Princeton. I have little doubt that such a committee would report back with the answer that Falk wants, or at least ought to want: Williams should institute something like the Princeton plan that would force individual departments to start grading both equally and lower.

Want to write great senior thesis or give a memorable MATH/STAT colloquium? Use this approach (pdf) to calculate alternative GPA’s for the Williams class of 2010.

Recall this comment from last summer.

The primary goal of the undergraduate student leadership this year is to repeal the Princeton grading policy because of all of its negative effects on the student body.

As I predicted, that student effort was a failure. The Princeton Plan is working exactly how its proponents wanted it to. See our seminar for details.

Facebooktwitter

Grade Definitions at Pomona

From last fall:

In response to longstanding concerns about grade inflation at Pomona College, a faculty and student committee is developing a set of grade definitions to guide grading across the college.

The Curriculum Committee began reviewing grading practices last semester, after faculty voted in favor of a process to produce college-wide grade definitions. Pomona currently has no official grade definitions to designate what level of student performance corresponds to each letter grade.

The committee plans to submit a proposed set of definitions to a faculty vote in Spring 2011.

“Our goal is to define, as best we can, how the college wants to grade,” said Academic Affairs Commissioner John Thomason PO ’12, who represents the ASPC on the Curriculum Committee. “We want to reflect the wishes of the people that are doing the grading—the faculty—while taking all viewpoints into account.”

A good idea. Williams should do the same. Even better, Williams should copy the Princeton policy for dealing with grade inflation.

Facebooktwitter

Grade Inflation is Making Students Lazy

From the Washington Post:

College students study a lot less now than in the 1960s, yet they get better grades.

For students, these trends must seem like marvelous developments. But they raise questions about both declining rigor and potential grade inflation in higher education.

In a forthcoming study in the journal Economic Inquiry, economist Philip Babcock finds the trends linked. As Babcock related in an e-mail, when the instructor “chooses to grade more strictly, students put in a lot more effort.” And when the professor gives easy A’s, students expend less effort.

Williams should copy the Princeton plan for dealing with grade inflation. What worked there will work here.

In the meantime, looking at the links between grades and student effort at Williams would make for a great senior thesis.

Facebooktwitter

Zero African-American Phi Beta Kappa Members in 2009

I recently observed that, as far as I can tell, none of the students in the 2009 Phi Beta Kappa (pdf) are African American. My methodology is certainly not flawless (based essentially on talking to students in the class of 2009), and it’s possible that a few of these students are African American.

(Previous post about the same results for the class of 2010 here).

I am genuinely interested in why African Americans at Williams appear (at least for two years) to be underrepresented in Phi Beta Kappa. Is it random chance? Academic and extra-curricular choices made by African-American students? Other factors? Does anyone have any thoughts on the relative weights of these factors?

Or is this something that we should never publicly discuss?

Facebooktwitter

Student Request on Group Performance

This week’s seminar focuses on evaluating the academic performance of groups of Williams students.

A February e-mail from a Williams student:

You probably don’t remember me, but I wrote to you a couple of times last year when I was abroad. I’m now doing a project for my senior seminar in Anthropology and need some information about GPA’s at Williams. In short, I’d like to research the extent to which this ‘wonderfully diverse student body’ performs academically.

Though I’ve heard many second hand stories about black and Hispanic GPA’s being below (1) the Williams average, and (2) even further below the white average (a phenomenon that surfaces in graduate rates, too, apparently), I haven’t been able to find any data on the topic. Do you happen to know if the school releases this kind of information broken down by race?

What should I have told this student?

We ended up having a 30 minute phone conversation. Very fun! What approach do you think I recommended?

Facebooktwitter

Apology

I recognize that my past posts on academic performance at Williams have offended people, and I apologize. I realize that I have raised the subject in such a way that does not produce a useful exchange of ideas. Rather, my approach has antagonized readers and has resulted in overly personalized exchanges. I realize that I am responsible for this sad state of affairs.

I am striving for a new beginning. To that end, I am trying to write this post in a way that I hope will generate a civil exchange of ideas.

To begin with, I thought it would be useful if I posed a question and asked readers to offer their own thoughts on this difficult topic. With this in mind, I will do my best to avoid dogmatic assertions and hopefully many readers will be able to provide facts to add to the general knowledge of the subject. So here goes:

Do African-Americans and other minority groups have a disadvantage at Williams because of their background?

Facebooktwitter

Zero African-American Phi Beta Kappa Members in 2010

This week’s seminar focuses on evaluating the academic performance of groups of Williams students.

In a recent thread, Derek (“you then don’t have empirical numbers”), Name-Removed-by-Request (“make up statistics out of thin air”), Rory (“make up numbers”) and others chided me for a lack of data about the racial breakdown of Phi Beta Kappa membership at Williams. Point taken! So, let’s gather some data using the distributed power of the EphBlog readership. There are no African-Americans in Phi Beta Kappa for the class of 2010 at Williams College. See the course catalog (pdf) for the raw data. Start with the Summas:

Bachelor of Arts, Summa Cum Laude
*+Christopher Alan Chudzicki, with highest honors in Physics
*Kristine Grønning Ericson, with highest honors in Art
*Ruth Madeline Ezra, with highest honors in Art
*Cristina M. Florea, with highest honors in History
*Andrew Lawrence Forrest
*Sophie Ariel Glickstein
*Yibai Li
*Zachary Clair Miller, with highest honors in History
*+Ralph Elliott Morrison, with highest honors in Mathematics
*+Kathleen Malone Palmer, with highest honors in Neuroscience

The * indicates membership in Phi Beta Kappa. None of these Ephs are African-American. (Corrections welcome!) See below for the Magnas.

Read more

Facebooktwitter

Athletic Underperformance

This week’s seminar focuses on evaluating the academic performance of groups of Williams students.

Did anyone notice the little trick that I played yesterday? I implied that this was a quote from a Williams faculty committee.

Finally, we recommend that studies of academic performance by under-represented minorities continue to be carried out periodically by the Admissions Committee. We recommend that the interval between such studies be five years. Reports of differences in academic performance tend to contribute to a sense of stigma and to stereotype threat among under-represented minorities, but failure to report on their academic performance is equally undesirable; a 5-year interval is a reasonable compromise.

That is absurd! For decades, Williams has refused to even (publicly) discuss the “academic performance by under-represented minorities.” This is the Administration’s single most closely guarded secret. In fact, the quote is a slightly modified version of one of the conclusions of the 2009 Report by the Athletics Committee (pdf).

Finally, we recommend that studies of academic performance by athletes continue to be carried out periodically by the Athletics Committee. We recommend that the interval between such studies be five years. Reports of differences in academic performance tend to contribute to a sense of stigma and to stereotype threat among varsity athletes, but failure to report on athletes’ academic performance studies is equally undesirable; a 5-year interval is a reasonable compromise.

Williams is the sort of place where studying the academic performance of athletes is perfectly acceptable, even praise-worthy. Studying the academic performance of under-represented minorities is verbotten. Fortunately, EphBlog (or at least Rory and I) have reached agreement on this issue. We both think that Williams ought to study the academic preparation and performance of under-represented minorities.

If Adam Falk were to propose the formation of a faculty committee on that topic, would any EphBlog reader object?

Below the break are excerpts from the MacDonald Report and the 2009 Athletics Report. I can’t imagine Williams ever publishing similar analysis for under-represented minorities. Can you? Political correctness is an ill-defined term, but there is no better example of PC at Williams than the College’s refusal to look as closely and openly at URM admissions/performance as it has at athletic admissions/performance.

Read more

Facebooktwitter

Academic Performance Seminar

This week’s seminar focuses on evaluating the academic performance of groups of Williams students.

Consider this quote from a recent report by a Williams faculty committee.

Finally, we recommend that studies of academic performance by under-represented minorities continue to be carried out periodically by the Admissions Committee. We recommend that the interval between such studies be five years. Reports of differences in academic performance tend to contribute to a sense of stigma and to stereotype threat among under-represented minorities, but failure to report on their academic performance is equally undesirable; a 5-year interval is a reasonable compromise.

1) The Advisory Group on Admission and Financial Aid (AGAFA) is one of the more important but least known faculty committees at Williams. (At one point, it seems to have had student members. Is this no longer true? When did that change? Or maybe it never was true?)

2) “under-represented” is code for African-American and/or Hispanic. Asian-Americans and International students (of all colors) do fine at Williams.

3) Alas, AGAFA does not seem to provide a useful summary of the data. But I bet that what former Williams professor KC Johnson points to for Duke would be similar to what we will find at Williams.

Click for a larger version.

4) When last we were exploring the issue of the academic preparation and performance of different student groups at Williams, Derek recommended that I frame the issue as follows:

Does Williams do enough to serve its minority students, its poor and working class students, its female students, and its international students? It is possible to look selectively at a smattering of evidence — memberships in Phi Beta Kappa and GPA’s come to mind — and conclude that the college could be doing a better job. Williams seems clearly committed to diversity in terms of admissions. Is it doing a good enough job to serve these students after they receive their long-awaited thick envelopes in the spring of their senior year of high school?

I will leave my substantive comments on this point of view for later in the seminar. For now, let me just point out that worry about how Williams serves “its female students” and not mentioning our male students is weird since female Ephs have significantly higher GPAs than male Ephs and are much less likely to be given academic warnings, forced to take time off from the College and so on.

5) What sort of data for Williams would people most like to see?

6) Kudos to the faculty authors of this quote. Although we should always be concerned about any “sense of stigma” and “stereotype threat,” there is no more important policy at Williams than admissions. We reject some applicants and accept others. Are we making the right decisions? (Other schools, like Caltech and the University of Chicago, do things quite differently.) The only way to decide is to have an honest and open discussion about how different groups of students perform at Williams. Without such a discussion, it is impossible to have an informed opinion about admissions policy at Williams.

Would any reader disagree? Would any reader argue that the larger Williams community should not discuss this topic?

UPDATE: As the comment thread documents, the quote at the top is a modified version of one from the Athletics Committee. Annoyingly misleading or an effective rhetorical technique? You make the call!

Facebooktwitter

Phi Beta Kappa ’11

Here is a listing of the 27 Williams seniors inducted into Phi Beta Kappa. These are the students with GPAs in the top 5% of the class. One of them will almost certainly be the valedictorian. Some (potentially offensive) comments:

  • 9 of the 27 are international students. Although this is the largest percentage that I have ever seen, it is typically the case that international students do better, on average, at Williams than US students. (Previous discussions here and here.)
  • 18 of the 27 were male. (See below the break for details.) I have not looked closely at gender numbers in previous years, but my sense is that men are, if anything, over-represented in PBK despite their, on average, lower GPAs. To anyone with a clue about human biodiversity, this would be the expected result. Men have higher standard deviations in most things, so they should be overrepresented in the tails of the GPA distribution. This would also suggest that Williams does not (nor needs to) offer admissions preferences to male applicants.
  • What is the breakdown by race? (Previous discussion here.) I don’t think that there are any African Americans in the list. (Corrections welcome.) If African American students were just as likely to be in Phi Beta Kappa as other students, we would expect to see 2 to 3. In many years of study, I don’t think that I have ever noticed an African American student in PBK. Have you? (I did see an African Eph in PBK one year.) The PC contingent among our readers should be sure to explain this puzzling anomaly in the comments.
  • Celebrities include: Christopher Fox (EphBlog author), William Lee (EphBlog contributor) and Ville Satopaa (Kane Capital summer associate). Want to be in PBK next year? Hang out with me . . .
  • I am trying to figure out what the names Lee, Meng, Li, Shin and Wang have in common. Suggestions welcome.

UPDATE: Thanks to Andy for providing the correct gender breakdown. Post has been updated accordingly.
Read more

Facebooktwitter

Princeton Grading: Departments

Sparked by our conversation about grade inflation at Williams, I will be reviewing different aspects of Princeton’s grading policy at noon each day this week.

From the Princeton FAQ (pdf):

Why have we made the departments responsible for implementing the policy?

We aspire to have students graded the same way in each department, so that there is no advantage or disadvantage to studying in a particular field. But the departments have different mixes of courses and course enrollments and different challenges and opportunities for implementing the grading policy. We leave to each department to determine how to meet the common institutional grading standard, taking into account the range, size, and level of the department’s courses. We’re not asking that every faculty member grade the same way, or that every course have the same grade distribution. Departments are in the best position to know what approach makes sense for their faculty and their courses; the grading policy vests maximum flexibility and room for judgment in each individual department, at the same time that it asks each department to agree to meet a common institutional standard.

Assigning responsibility at the department level was smart. Why? Because every school has professors like our own Derek Catsam ’93:

Oh: and I’m curious what tenured professors at Princeton who oppose this policy are doing. Because grading fits smack dab in the center of academic freedom, and as a tenured professor my answer would be something like this: “That’s a nice idea. Good luck with it. I’m going to grade how I see fit. And here’s my contract, my tenure letter, the number of the American Association of University Professors, the number of the President of the Faculty Senate, and the number of the chair of faculty affairs if you have any questions. Have a nice day.”

The best way to deal with prickly professors like Derek is to make the department the unit of measurement. Derek may be willing to tell off the central administration, but doing the same to his department chair — and his department colleagues — is much harder. Those are the folks that he needs to work closely with for decades. He needs to get along and compromise with them. He needs favors from them, at least occasionally.

Williams should also measure/fight grade inflation primarily at the department level.

Facebooktwitter

Princeton Grading: Student Quality

Sparked by our conversation about grade inflation at Williams, I will be reviewing different aspects of Princeton’s grading policy at noon each day this week.

On student quality:

Aren’t Princeton students better than ever before? Shouldn’t they get more A’s than ever before?

There is a strong temptation to argue that undergraduates today come to college better prepared academically than any previous generations of Princetonians—and, therefore, deserve more A’s. It is certainly tougher than ever before to gain admission to Princeton, but more intense competition does not necessarily mean abler students.
It’s true that the proportion of Academic 1’s and 2’s in the student body has grown over time, but one needs to be cautious about over-interpreting academic ratings. Those ratings are made up of three components: high school grades, rank in class, and SAT scores. As for high school grades, grade inflation is as much a high school phenomenon as it is a college phenomenon. Rank in class is increasingly problematic as a useful measure; in many high schools, there are many students who stand first in the class, for example, and many high schools now decline to provide class rank at all. And SAT scores do not in themselves sustain the argument that current undergraduates are more qualified than previous generations of Princetonians. Moreover, Princeton attracts such excellent students that the difference between a 1 and a 2, or a 2 and a 3, is actually very small.

Suppose, though, that we concede the argument—suppose today’s undergraduates really are more accomplished academically when they matriculate at Princeton. If that’s the case, then the faculty has a responsibility to hold them to higher standards—that is, to expect more of them and stretch them further academically than we have stretched previous generations. And even the best qualified students don’t do their best work on every assignment in every course; the point of the grading policy is that they shouldn’t be getting the same grades for their ordinary work as they get for their best work.

Indeed.

Facebooktwitter

Princeton Grading: Post-Grad Effects

Sparked by our conversation about grade inflation at Williams, I will be reviewing different aspects of Princeton’s grading policy at noon each day this week.

The major concern expressed by students and others against a policy of stopping and reversing grade inflation is that it hurts students in the competition for post-graduation opportunities. Lower grades at Princeton, on average, means less success for Princeton students (even though they are just as smart and work just as hard as, say, Yale/Harvard students) in applying to medical/law/business schools, fellowship awards and private sector companies. Princeton addresses these concerned in detail:

What’s happening to the fortunes of Princeton students in the job market and in admission to graduate and professional schools?

Based on the record thus far, we can report with a high degree of confidence that Princeton students are not being disadvantaged by the new grading policy.

… [Endless details]

We will, of course, continue in the years to come to track carefully the fortunes of Princetonians in the various external marketplaces in which they compete for jobs and graduate and professional school admissions. At present, with all caveats taken into account, and all limits in the data acknowledged, we find that Princeton undergraduates continue to be highly successful, as they were before the grading policy was instituted.

Read more

Facebooktwitter

Princeton Grading: Purpose

Sparked by our conversation about grade inflation at Williams, I will be reviewing different aspects of Princeton’s grading policy at noon each day this week.

The purpose of the grading at Princeton:

Princeton established a University-wide grading policy five years ago for two reasons.

The first is that we wanted grading to be done fairly, so that students in one academic department could expect to be graded according to the same standards as students in any other academic department. Before we adopted the grading policy, there was wide variation in grading standards among departments. In general, students in humanities departments were being graded more leniently than students in social science departments; students in engineering were being graded more leniently than students in the natural sciences. We thought that wasn’t fair, and we set about to fix it. The second reason for the adoption of the grading policy is that we thought students deserved clear signals from their teachers about the difference between their ordinarily good and their very best work. With grade inflation and grade compression, that differentiation wasn’t happening. We thought that we had a responsibility as educators to use grades to give students better information about the quality and effectiveness of their work.

Aren’t both equally true at Williams?

I would add that Williams (and Princeton) have an obligation, not just to their students, but to the outside world, to fairly and accurately represent student accomplishment. A transcript and the grades which it contains should be meaningful. Committees for company hiring, law school admissions, and fellowship awards should be able to look at a transcript and draw accurate conclusions about how well a student has done, relative to her peers, at Williams.

There is no doubt that this makes some students worse off. If you are the laziest, stupidest student at Williams, with a transcript filled with Cs and Ds, then you would be better off with a policy that awarded everyone an A no matter how lousy their work. Similarly, the more accurate and meaningful a transcript is, the better off the very best students are. Less grade inflation/compression (probably) leads to more Harvard Medical School admissions, Rhodes Scholarships and Goldman Sachs jobs offers.

I prefer a policy which benefits the best students at Williams even if it, slightly, hurts the worst. What do you prefer? See below the break for more details from Princeton:
Read more

Facebooktwitter

Princeton Grading: Background

Sparked by our conversation about grade inflation at Williams, I will be reviewing different aspects of Princeton’s grading policy at noon each day this week.

Let’s start with this background from the New York Times.

Type-A-Plus Students Chafe at Grade Deflation

When Princeton University set out six years ago to corral galloping grade inflation by putting a lid on A’s, many in academia lauded it for taking a stand on a national problem and predicted that others would follow.

But the idea never took hold beyond Princeton’s walls, and so its bold vision is now running into fierce resistance from the school’s Type-A-plus student body.

With the job market not what it once was, even for Ivy Leaguers, Princetonians are complaining that the campaign against bulked-up G.P.A.’s may be coming at their expense.

“The nightmare scenario, if you will, is that you apply with a 3.5 from Princeton and someone just as smart as you applies with a 3.8 from Yale,” said Daniel E. Rauch, a senior from Millburn, N.J.

The percentage of Princeton grades in the A range dipped below 40 percent last year, down from nearly 50 percent when the policy was adopted in 2004. The class of 2009 had a mean grade-point average of 3.39, compared with 3.46 for the class of 2003.

Read more

Facebooktwitter

Cornell’s Grading Policy

Continuing our discussion of grading policies, consider how Cornell does things.

WHEREAS, the Committee on Academic Programs and Policies, has determined that it is desirable for Cornell University to provide more information to the reader of a transcript and produce more meaningful letter grades,

THEREFORE, BE IT RESOLVED, that the Faculty senate adopts the following policies on grade reporting:

1. Transcripts and grade reports for undergraduate students shall indicate, along with the grade earned, the median grade given in the course and the course enrollment. Independent study, honors courses, and individual research may be exempted upon recommendation by the department (or program) and appropriate college committee. Courses with enrollments of fewer than ten students will also be exempt from this policy. This policy shall become effective as soon as technically feasible, but will apply only to classes entering after the effective date.
2. The Office of the University Registrar shall publish at the end of each semester, a list of median grades and enrollments of all undergraduate courses with ten or more students. This policy shall become effective in Spring 1997.

Technical Details

The record of a course on an undergraduate’s transcript now reads, for example:

Hyperspace Travel PHYS 999 3.00 B-
This indicates the course title and number, credits received, and grade. Under the new proposal the record would read:

Hyperspace Travel PHYS 999 3.00 B- (110) (B-)
This indicates, in addition, that the course had an enrollment of 110 and that the median grade was B-.

1) Has Williams ever discussed such a policy? Do you think it is a good idea? I am agnostic about adding this information to the Williams transcript but strongly in favor of making class median grades publicly available.

2) Here are the median grades for every Cornell class (with 10+ students) over the last 13 years. Shouldn’t Williams make similar information publicly available? I think so. Williams should be a leader in transparency among elite institutions. If Cornell or Princeton makes Information X public, then Williams ought to as well (unless a compelling case can be made not to). Transparency should be the default.

3) What other elite schools make grading information public? Links welcome.

Facebooktwitter

Grading Background

From the Record in 2008, a quiz on grading:

1) About what percent of all grades at Williams last year fell in the “A” range (-A, A or +A)?
2) What was the college-wide average GPA last year?
3) Are average grades across Divisions 1, 2, and 3 roughly equal?
4) In a consortium of about 26 highly competitive colleges and universities with whom Williams consistently shares data (including college GPAs), where does Williams rank in terms of highest average GPA?
5) Does grade inflation continue to exist at Williams?

Let’s see how you did:
1) About fifty percent.
2) As specifically as can be put in print: what amounts to a low B+.
3) No. At the 100, 200 and 300 levels, grades in Division 1 classes are consistently a tenth of a point higher than the average Division 2 and Division 3 grades at those levels. Grades at the 400 level, however, are roughly equivalent across divisions.
4) Williams is consistently near the top of the list, ranking for the 2006 to 2007 academic year among the top six schools in terms of average GPA.
5) Yes. Though grade inflation has slowed as compared to pre-1999 to 2000 rates, inflation still persists.

Useful background information to our discussion. Since this seems to be a topic of wide interest and divergent views, I am going to provide information on Princeton’s experience in a series of daily noon posts next week. Contain your excitement!

Facebooktwitter

Grade Distribution: 2008-2009

Here is the distribution of undergraduate grades at Williams for the 2008-2009 academic year. There were 15,679 grades awarded with an overall average of 3.39, slightly above a B+.

A+    203
A   3,586
A-  3,951
B+  3,425
B   2,347
B-  1,118
C+    446
C     331
C-    136
D+     36
D      32
D-     20
E      48

More than 49% of the grades given at Williams are some form of A. Recall Peter N. Siniawer’s ’97 thesis: “When A=average : the origins and economic implications of grade inflation at Williams College and other elite institutions.” Things have only gotten worse in the last 13 years. One of Professor David Zimmerman’s students ought to write a senior thesis updating this analysis.

From the course catalog (pdf): Read more

Facebooktwitter

Not Considered Public

In order to have a meaningful discussion about the Gaudino Option, especially with regards to the B- minimum grade required to invoke it, we need some basic data about grade point averages at Williams. So, in my role as EphBlog reporter, I sought out that data. Here is what Registrar Charlie Toomajian wrote back:

Dave:

I am sorry to report that the information you requested is not considered public by the College. By faculty regulation, we do not release information about class rankings and providing the data you requested could jeopardize that ruling.

I hope the discussion proves lively and informative.

Charlie T.

1) Thanks to Charlie for taking the time to answer and for giving me permission to quote him. Note that this ridiculous paranoia is not Charlie’s fault. There are voices at Williams who think that the less transparent that Williams is, the better. Those voices prevent us — or, really, all alumni, students and parents — from having an informed conversation.

2) My original e-mail to Charlie is below the break. But note that I am not asking for class rankings! I just want to know average GPAs and the cut-offs for Latin honors. This is some sort of state secret? Pathetic.

3) If the College were serious about not “releas[ing] information about class rankings,” then it should stop publishing the names of Summa students (or the name of the Valedictorian). (Summas are the top 2.5% of the class by GPAs.) But, obviously, Williams is just using that phrasing as an excuse to keep me (and you!) poorly informed.

4) Given the strong praise that Adam Falk had for transparency at the Boston Alumni meeting, do you think that I should politely follow up with him?

5) Some of this data (like average GPA) is regularly discussed at Williams and reported in the Record. Why keep it a “secret” from us?

Read more

Facebooktwitter

Heresy

Morty and former Williams Economics Professor Mike McPherson on defining college success.

“What college success means depends so much on what [kind of] college you’re talking about and what students you’re talking about,” said McPherson, who is president of the Spencer Foundation and former president of Macalester College. He suggested that the best measures for college success would be specific but tailored to individual institutions. What that means, precisely, “each one can answer that question for themselves,” he said.

The panel was conspicuously divided into two halves: on one side sat McPherson and Schapiro, the president of Williams College; on the other were two representatives of public institutions whose students are much more likely to come from disadvantaged backgrounds and rely on financial aid. Williams, the moderator, didn’t hesitate to point out that the administrators from Miami-Dade and the University of Maryland seemed more willing to embrace strict accountability measures and the data collection that approach requires.

Schapiro, also an economist, suggested that there might be “some appetite” among faculty for more in-house accountability measures, but explained that much of the resistance stems from a fear that increased empiricism could lead to a one-size-fits-all testing regime — like a No Child Left Behind for higher education.

He stressed the need to more rigorously link what colleges do to their students’ professional and other outcomes after they graduate. Otherwise, it’s impossible to tell which teaching methods work and which don’t. Schapiro brought up a hypothetical proposal to compare students’ incoming SAT scores with outgoing GRE scores to determine whether they improved (and presumably correlate those scores to majors and other factors during the college experience).

“I would do that, but then again, I’m an empirical economist,” he said. Professors in the English department, he imagined, would view it as “heresy.”

When colleges experiment with different ways to teach critical thinking skills, as Williams does, Schapiro said, it should be seen as necessary to then empirically test what worked the best. Higher education is “horribly bad at this,” McPherson said — to take one example, colleges tinker with class sizes all the time — but they “never, ever look at the results.”

“Even at Williams, there’s not as much of an appetite as there should be,” Schapiro said.

Well, isn’t it (part of) the president’s job to generate that appetite?

Now, to be fair, Morty is already at the 99th percentile of all college presidents in terms of his willingness to measure Williams performance, so I shouldn’t be too critical. And, to be fair to my English professor friends (Hello Katie Kent ’88!), any measurement plan that uses a tool like the GRE is likely to fail, both because improving standardized test scores is not the purpose of a Williams education and because any such improvement is likely too small to notice.

Instead, my point is that there is an obvious policy change that would a) Allow fair-minded observers to see the causal effect of a Williams education on student achievement and b) Not force Williams professors to do much if anything differently. That change is the public display of student work. Put on the web all the papers that a student writes as a freshman for ENGL 101 and all those she writes as a senior for ENGL 401. If the Williams English Department is doing its job, the latter papers will be much better than the former.

There are, of course, all sorts of difficult issues to consider in any plan which makes student work public (as well as her professor’s comments but not grades). Perhaps freshmen should be exempt. Perhaps students should be allowed to opt-out from the requirement for one class per semester. Applying the requirement to non-paper-writing classes is difficult. And so on.

But the central principal is obvious: Being a part of an academic community requires public participation in the scholarly conversation. Making papers public will increase the quality of work done at Williams. Making the comments (but not the grades) public will have a similar effect. All the good reasons for making senior theses public apply in the context of other classes as well. (See Tim Burke for a related re-imagining of a liberal arts education.)

Assume for a moment that Morty agreed. What should he do? Best next step is to recruit some faculty to try out the experiment. (All the projects done by students in my Winter Study will be posted to the web, along with my comments.) See how it goes. I bet that someone like Joe Cruz would be willing to try it out in philosophy. Perhaps the whole thing will be a disaster. More likely, I think, is that other professors would be impressed with how making academic work public both improved the quality of that work and made it easier for everyone to see the progress that students make.

[Side note: Just noticed that the ENGL department no longer has a 101 (common introductory course for all students) or 401 (common senior capstone course). This is another sign of the Decline of the West, but save that for a separate rant. Just substitute 100-level and 400-level in the above.]

Facebooktwitter

Grade Inflation

What is the current status of grade inflation at Williams? See here for previous discussion. The Record reported 8 years ago that:

The Williams faculty voiced its concern over grade inflation at the College as it passed several motions of a proposal by the Committee on Educational Policy (CEP) Subcommittee on Grading, instituting grading targets at the monthly faculty meeting held Feb. 16.

The targets, which range from 3.2 to 3.5 from 100-level to 400-level courses, increasing one tenth of a point per course level, intend to stabilize the mean GPA of the College beginning in Fall 2000 to about 3.3, the mean grade for 1998. The most frequently given grade in 1999 was an A- and the mean grade hovered just above a B+ at 3.34.

“When you take the long view and look from 1960 to 1999, you see overwhelming evidence that grades are moving steadily upward,” Chair of the CEP and James N. Lambert ’39 Professor of Anthropology Michael Brown said.

“The problem is that leaves you no room to move. Grades are so compressed that you start getting into making finer distinctions which are harder and harder to justify at the same time.”

“The so-called ‘Gentleman’s C’ is now the ‘Gentleman’s B+,’” quipped Associate Dean for Student Services and Registrar Charles Toomajian.

Indeed. But where are things now? A Record editor mentioned to me that they had sought the latest data but been rebuffed by the Registrar. True? Unless something has changed, the faculty should have access to this data.

[Professor Colin Adams said] “We passed one of the motions which said that we would distribute information to all the faculty at the end of every semester which told them where they were in relation to other departments, whether their GPAs were above or below the other departments.”

Is this still going on? If so, surely there is at least one faculty member who thinks that this information should be made public to force Williams to do something. Are you that faculty member? If so, my mailing address is David Kane, 30 Washington Street, Newton, MA 02458. Mail the paper to me (or e-mail an electronic version) and I will post it here.

Every year, the College exchanges information with a group of about 20 peer institutions of liberal arts colleges. Recently, Williams ranked second in the list of highest annual GPA. This fact concerns many faculty members.

“If the faculty doesn’t control grade inflation, in 10 to 15 years everyone is going to have a 4.0 or higher and the transcript will be utterly useless,” Brown said. “There will be no distinctions possible because everyone will get the same grade. That is one of the important issues that as grades become compressed, it is more difficult to present the nuances in a picture of a student’s performance at Williams.”

Perhaps Professor Brown could give us an update. I also suspect that faculty like, say, Sam Crane would agree with me that something needs to be done about this.

And, it is fun to go back even further, to 1998 when our own James McAllister was a new professor.

Assistant Professor of Political Science James McAllister said he made it clear to all of his students at the beginning of the semester that it would not pan out that way for them; they could not expect inflated grades. He said he distributed articles on the problem of grade inflation and vowed that he would not participate in the phenomenon.

McAllister commented that his stance against grade inflation was inspired by the grade reports from the Registrar. He saw that he was eleventh out of 44 on a list of classes with the highest mean grades, and began to fear that his classes were crowded because students thought they were easy.

McAllister said he deflated grades by downgrading borderline grades. Last year, he tended to upgrade grades on the border because he was a new professor and uncertain of the standards.

McAllister said he now knows that Williams does not pressure professors to grade high. “Williams is not a grade inflationary school,” he said.

Really? Even 10 years later? Show us the data.

For those of us who hire from Williams, this is a real problem. If a student tells me that she got an A from Sam Crane or James McAllister, then I really want to believe that she is one of their very best students. My guess, however, is that all this A tells me is that she is in the top 1/3, which doesn’t tell me much. When James MacGregor Burns put an A on a transcript 25 or 50 years ago, you knew that it meant something.

Do the grades on a Williams transcript mean anything today?

Facebooktwitter

Due Date

Grades from professors are “due” to the Registrar today. Cale Weatherly comments, “You are significantly likelier to receive grades on time from professors who do not have tenure than from professors who do.”

Indeed. And this would make for an interesting senior thesis! What is the pattern of timeliness among professors when it comes to getting grades in? You could even run some randomized experiments by sending out reminders to some professors but not others. You could look for department effects. And so on.

I realize that this might seem like a stupid topic. Shouldn’t a senior thesis tackle a bigger problem? No! The more narrow the topic, the better the senior thesis. The more Williams-focussed, the more likely that anyone will ever read it. Thousands of people have read selections from Lindsey Taylor’s ’05 thesis. Moreover, this small topic will provide all the complexities of a bigger issue.

Facebooktwitter

Fighting Grade Inflation

Although I still haven’t read Peter Siniawer’s ’97 thesis, grade inflation is an occasional topic at EphBlog. Here is an idea that would make for a(nother) great senior thesis. How different would class ranks be if adjustments were made for different grading standards and talent levels across classes at Williams?

Facebooktwitter

Next Page →

Currently browsing posts filed under "Grading"

Follow this category via RSS