Or on a Harvard or Yale undergraduate transcript? As to Harvard or Yale, I think they often don’t. It’s all about the brand name instead. (Contrast Cornell, where students actually can, and do, flunk out.)

Given that tenure decisions are about 50% based on student course evaluations, there is indeed pressure to keep students happy and this has the potential, at times, to translate into higher grades. Now higher grades do NOT automatically mean higher evaluations, but the reality is a student who is getting B+’s on papers is likely to be in a better mood come eval time than one who has been getting B-’s. My sense is that faculty do their best to be completely fair about grading and give low grades when they are deserved, but there is no denying that aspects of the incentive system work against this.

Still, the grades certain do mean something. Students with A- averages genuinely perform better than those with B averages. Straight A averages are quite rare at Williams and those students stick out (as do those with C averages).

Like it or not, Williams students are evaluated against students from other colleges by GPA far, far more frequently than they are evaluated against their fellow ephs. Consequently, our concern with grades at Williams should tend towards being concerned about how accurate a measure of performance are eph grades as compared to grades at colleges nationally (not just Harvard, Yale, and Middlebury, but U-Mass, Emory, Penn State…etc).

I think based on probably any measure you would want to compare, Williams students face grade deflation when compared to colleges nationally. I would be surprised if incoming student gpa vs outgoing student gpa dips far more at Williams than schools nationally. Factor in SATs or “other factors” of selectivity in there and the dip only increases.

“Ok,” you might say, “but let’s look at how hard these students work for these grades.” Perhaps Williams students’ see a dip from their hs to their college grades because they are slackers nationally. However, I have no doubt that if we factor in hours-per-day spent on classwork/studying (or any other measure of “hard work” that you would want to pull out), the relative “dip” in the grades of Williams students would increase dramatically; my experience is that students work significantly more at Williams on average than any but a few other colleges.

The final factor I can think of is classroom expectations. Perhaps Williams profs require less work/studying on average of students to achieve any particular grade. However, even on this count, I would wager that expectations for classwork (both in terms of quantity of work and quality of work) are significantly higher at Williams than they are in any but a few other colleges in the country.

Now, you would be correct in pointing out that my arguments above are based almost entirely on guesses; I don’t have any statistics showing that the average Williams student works harder than the average Ohio State or Boston-U student. However, I doubt few would argue with this statement.

Even if you were to consider each of these factors individually, it’s notable how low Williams students’ grades are. Considering all of the factors in conjunction, and the difference becomes staggering.

In conclusion, while it is true that Williams students are often given a “bump” over students from average private or state schools, I doubt that the “bump” accurately measures the actual difference in earned grade at the two schools. Furthermore, in many cases of employment (probably more than anyone here would want to admit), no bump at all is given, either because the employers haven’t heard of Williams or because they don’t give “bumps” to anyone.

I understand David’s reasoning for railing against grade inflation, but I would urge David and others considering this topic to look at the complete picture. Besides all of that, an A at Williams represents incredible excellence, whether it reflects that students’ falling in the top 10% or top quartile of grades in that class. The fact is that Williams students are largely doing very very high quality work across the board.

Incidentally I happen to know that for the 2007 class, the cum laude cut-off fell below a 3.6 (although I don’t know how far below–it could have been a 3.59 or a 3.4). In other words, the top 1/3 of Williams students graduated from Williams with below an A- average.

Doesn’t over half of the Harvard student body graduate every year with above an A- average?

I also happen to know that the phi beta cappa cut-off fell below a 3.8 (again, i do not know how far below it was).

Grade inflation is indeed a serious problem here at Yale (and certainly at Harvard, too). I will try to dig up some numbers when I am less busy with midterms. I didn’t know you lived in Newton, very cool! I’m 02459.

By the way, kudos on the smooth transition to WP : )

Grade inflation is one of my favorite subjects. It can’t realistically be stopped without upsetting everyone. Grades will no longer be useful in identifying the best students. Hirers will have to take up the burden of sorting prospective candidates by adopting more rigorous screening procedures.

I know it sounds crazy, but I like the new system. It’s more democratic. The GPA-based system places too much power in the hands of professors, and is prone to abuse by cheating, copying, purposely picking easy classes and other chicanery. But as GPAs inflate, class performance will matter less. What you retain from the class will matter more. Independent work will matter more. You can’t point to a 4.0 GPA as the reason a company should hire you — you have to prove your value to them. The task of sorting graduating seniors is best left to the company doing the hiring.

I don’t think this problem cannot be solved without addressing the inter departmental differences. In departments such as economics, the workload is markedly easier than in, say, physics. This leads to an automatic “self-selection” process — for instance department “ease” is likely inversely correlated with the likelihood of its graduates seriously considering academic careers.

It is this self-selection, rather than the intrinsic difficulty of the classes, that makes the situation somewhat unfair. If all classes had students randomly drawn from the college population, class difficulty would ostensibly not matter much in terms of effort/grade. (Assuming they would be curved to the same mean and that grade variance isn’t significantly affected by difficulty, to be precise)

But if some departments have students who are on average willing to expend more effort in their classes, more effort will likely be needed to obtain an average grade. As a result, the difficulty of getting 300 level A’s varies tremendously accross departments.

My guess is some departments attempt to balance this by giving slightly higher average grades — but this, of course, gives everyone else carte blanche to raise their grades, as well.

Current eph:
Your analysis falls right in line with the experience of my frosh.

Based on high school workload/performance/result, vs. Williams workload/performance/results, (and keeping SATs and strengths/weaknesses in mind) I see no evidence of ‘easy’ grades to be had at Williams.

This is also weighing the Williams experience against the experience of other freshmen I know, at other schools, and their past performance/results, SAT scores, and present performance/results.

Granted, my frosh is a public school grad…but if SATs are any indication of the quality of that public education, then I’d say it was on par with my frosh’ private school friends.

The biggest adjustment has been a much heavier reading load and general ‘organization’ of time. Also the commitment to study groups was initially resisted until the advantages were realized.

Frosh mom:

I dont know. Last sem I took a great class in which I was a few notches below perfect. A good friend of mine was in the same class but was flawlessly perfect. We both got As. What differentiates his performance from mine? He couldnt get an A+ as he wasnt THAT stellar, but I couldnt get an A- as that would push a whole lot of people below too… Essentially, we have a 4 tier system. There is no separation. Perhaps professors should be allowed to give number grades and not letter grades, so that I could have got a 3.8 and him a 3.9 in the class. I see that as the only solution.

Make 1/4 of everyone’s grade in every class based on a bell curve from a series of multiple choice tests given for part of the midterm and final… and the average will go down to around 3.0. This will not leave any room for complaints because that portion is based on competition with your piers. If you do not want it to be all subjective and inflated, it is the only way. It forces everyone to be hard and honest. What do you want to bet that the students will study harder, and learn more data if they know that 10% will fail, 20% will get a D, 40% will get a C, 20% will get a B, and 10% will get an A twice during every class, every semester.

Such a system would also force Profs to look at their students in a more honest way. The Profs will have a great tool that evaluates peformace amongst peers, and identifies areas in which their students, and perhaps their classes, fail.

They may even be shocked by who performs, and face their own biases. It is the only way.

to 10 nerd:
Perhaps your professor saw something in you that you don’t yet see in yourself. Based on your comment (and it’s inherent honesty and humility) , I venture to say, that might be the case.

Look… I don’t profess to know all the politics and idiosyncrasies behind the grading system at Williams. But, I do know that Williams has the wherewithal to hire the best of the best. With that in mind, I say, you should give your professor the benefit of the doubt, be damned what you hear otherwise, and accept your grade with the confidence that you earned it.

P.S. to 10 nerd:

My frosh had a couple of surprises as well…one grade lower than expected…and one higher than expected.

And so, there you have it…life in a nutshell.

PTC:
It might be interesting to hear what others think of multiple choice testing.

I personally think they are the bane for anyone that begins to think outside of the box. They require that you backpedal and second-guess…particularly in certain subjects.

“Make 1/4 of everyone’s grade in every class based on a bell curve from a series of multiple choice tests given for part of the midterm and final…”

What a truly awful idea. It’s so bad that at first I assumed you were PTC was kidding, but I now think he/she may have been serious. Some of the most important things that Williams can teach its students are how to write, how to construct arguments, and how to think creatively about problems. Multiple choice questions have essentially no role in developing any of these skills. They may have a limited place as one component of evaluation in a small number of classes, but to use them across the board would a massive blunder (and fortunately, one that Williams profs would never make.

Outside of enormous intro lectures like PSYC 101, multiple choice tests are almost unheard of at Williams. The idea that students should “study harder and learn more data” is exactly what’s wrong with standardized testing / No Child Left Behind / etc; a good education is at least as much about skills as about raw facts and I think most Williams classes recognize that.

Disturbed- It does not have to be 1/4 of the grade… it could be 1/8th, or 1/10th, or whatever. My idea still leaves a still 3/4 (or any amount decided by the administration) of a persons grade up to the profs interpretation of other work. Under this system a student could flunk the tests and still get a C in the class.

If you want to be tough, especially when one is talking about a place with as much talent as Williams… then you have to be fair. The only way to be fair is to test students against each other. If profs are giving all A’s and Bs, then there is no one who is average.

Let me ask you all this… are you looking for grades that evaluate students against their peers? If you want grades that evaluate them against every average college student in the nation, then the answer is that every Williams student gets an A. If you want grades that evaluate students against their peers at Williams, then the average grade should be average… in the B- to C range.

A person could get 90% right on a multiple choice bell curve test, and still flunk it. That is the beauty of it. No matter how hard everyone tries, no matter how bright the students are, some of them will fail. Failure is a part of life.

I have visions in my mind of “brains” positioning themselves in classes that are known to be frequented by jocks (tips). Screwing up the curve and getting the easy A.

lol. Life is great.

PTC,

You are still assuming that multiple choice tests, even if only 1/8 of the final grade, are somehow a more fair and accurate assessment of how a given student is doing in a class than other methods. That simply isn’t so for at least most humanities classes at a place like Williams. The point of most classes at Williams is not simply learning facts, but learning how facts fit into particular contexts and play a part in particular arguments. The ability to reason and express one’s chain of reasoning clearly is far more important and simply can’t be tested by multiple choice. If you wanted to see who was a better basketball player, would you use their free throw percentage as the only true and “fair” measure of their ability? Of course not.

Some good points.

“If you wanted to see who was a better basketball player, would you use their free throw percentage as the only true and “fair” measure of their ability? Of course not.”

Nor is that what I am suggesting. Free throw % would factor in somewhere, however. It is quite possible, using the method that I have suggested, that the best “test takers” have lower grades than others in the class.

“That (learning facts) simply isn’t so for at least most humanities classes at a place like Williams. The point of most classes at Williams is not simply learning facts, but learning how facts fit into particular contexts and play a part in particular arguments.”

I disagree. I believe that multiple choice tests can also measure (to an extent) “learning how facts fit into particular contexts and play a part in particular arguments.” Additionally, facts still are a part of learning. I am having a hard time thinking of any liberal arts academic subject that does not use facts as a part of the course? Even the humanities have facts that can be tested and probed using a multiple choice test. History, math, science? Remember, I am suggesting that only a small % of the final grade is calculated using a students ability (and/or hard work in studying) to take a bell curve test against his peers.

“The idea that students should “study harder and learn more data” is exactly what’s wrong with standardized testing / No Child Left Behind…”

The “boogie man” no child left behind is so irrelevant that I am having a hard time commenting without being sarcastic. How about the SAT then, is that also “unfair and irrelevant”?

NCLB is a massive gov program that applies universally to the entire country. NCLB involves a massive national bureaucracy that is not nimble enough to make changes. NCLB also tests kids who are in varying stages of development- universally across all demographics….etc, etc, etc. I find it hard to imagine that Profs at Williams would have a hard time putting together multiple choice tests that accurately test the depth and scope of knowledge in the subjects they are teaching their students?

I am starting to think that I am arguing against the “easy A” here, rather than for a method that could fairly change the system that makes everyone an “all star”. If you want to make it fair, test students and make them compete. There is still plenty of room for other types of analysis (grading) in order to formulate a final grade.

“Do the grades on a Williams’s transcript mean anything today?”

I believe David’s point is that when everyone gets an “A”, then the “A” means nothing. The firms that are looking at graduates records in order to hire have nothing to go on…. graduate schools looking at the transcripts see nothing but great grades… which sets into motion an odd paradox. “Grade inflation” forces post grad programs to rely more heavily on testing to “break students out”- because everyone is getting great grades. Why not do some competitive testing at Williams, and make a “B” at Williams worth an “A” at Harvard?

I suspect that a system like I am suggesting, would force Profs to give a more fair analysis of students abilities and work in every venue of grading.

Another option is to make a mandatory range for averages. Make it so Profs have to have an average grade range not higher than 3.3… or something to that effect.

One way to combat grade inflation would be to do what the British used to do (and still do to some extent), which is to have the person setting and marking the exams or papers be different from the person teaching the course. Then the teacher for the course no longer can try to improve teaching evaluations by making the course easier.

As I understand it, they also used to have a system of external review committees to equalize grades between different universities.

On a different note - employers simply do not have enough data (or, equivalently, are too cheap to generate enough data) to evaluate potential employees without input from their professors in college. If they cannot get this input in the form of grades, they will get it in the form of recommendations.

A system where employers hire based on who gets an A or a B or a C is far more fair and transparent than a system where employers hire based on confidential phone conversations (of which no records are kept to avoid potential lawsuits) with professors about the student’s performance.

Some points:

First, everyone at Williams is not getting As. In fact, the vast majority of students are not getting As or A-s. See above in this post for numbers.

Second, objective measures of performance are great. I’m a teacher, however, and I feel strongly that multiple choice tests (especially in the humanities) are poor indicators of knowledge. Even in courses where crucial knowledge may be answered with a single line, a multiple choice test will not necessarily demonstrate important mastery. Many students (and I was one of these) become very good at figuring out which is the correct answer from a variety of choices. However, a much more important skill is recalling information, rather than simply being able to identify a correct answer from a list. I would discourage just about any teacher in any situation from using multiple choice tests as an assessment. There are other, better, ways to “standardize” grades across courses (notably, rubric-based assessments). However, I think that the very notion that grades should be curved is faulty. Grades should measure performance, not relative standing. Williams is a small school with small classes. Individual class dynamic and quality can, and I am sure does vary from year to year. Forcing teachers to curve even a portion of their grade would hurt students who are misfortunate to end up in a stronger class, and would act as a disincentive for cooperative learning (which we should encourage as much as possible).

Third of all, I think we’re getting carried away with the idea that there isn’t enough distinction between students’ grades at Williams. Williams doesn’t grade on an A-B-C-D-F scale, they grade on an A+-A-A–B+…etc scale. In other words, professors have something like 12 different grades to choose between when assigning students a grade. Even assuming that most profs at Williams grade just about all of their students on the C–>A range and give out no A+s, there are still 8 different options for profs. As someone who grades students now, I can confidently say that this is more than enough; if anything, there is a fairly high degree of arbitrariness between 1/3 grades (an A- and B+, for example). You seem to believe that grades are all “crammed” into the high range, and that profs are “forced” to give two students with different performance the same grade. I assure you that grading is nowhere close to exact enough of a science, even in the best of circumstances, to create that problem. The difference between an A- and a B+, rather than being larger than it is (as you imply by calling for increasingly spread out grades), is borderline arbitrary, performance-wise. To increase the appearance of that slight difference in performance by creating grade separation using curves or enforced grade deflation would only serve to magnify fairly arbitrary performance differences.

What we have at Williams is a system where students are largely (but not exclusively) assessed by their performance as it stands alone. The consequence of this system is that most students typically receive grades that range from B-s to As (given that their performance reflects these sorts of grades). If the difference between a student with a 3.4 gpa and a 3.5 gpa seems arbitrary, well, it is. We shouldn’t be working to exaggerate this difference! Rather, the system we currently have encourages employers to rely a bit more on recommendations (which capture the important sides of a student’s performance) and a bit less on grades (which do not).

Finally, I think we’re still ignoring the main issue at stake in all of this. What are GPAs primarily used for? The answer is NOT for comparing Ephs to Ephs, but rather for comparing Ephs to Jeffs, Terrapins, Buckeyes, Yalies, etc, etc. Given this, we need to ask ourselves, how accurately do Eph grades reflect Eph performance on a national scale? There is no question in my mind that in this light, Eph grades are fairly severally deflated.

*severely, not severally. There are probably other typos as well–sorry, this was typed quickly.

For consideration, I’ll forward the argument (perhaps Toomajian’s “equity concern,” certainly influenced by CT) that when I was assigning a Berkeley student a B- in “Reason and Argumentation”– following a department-imposed strict curve that balanced grades “on a stairstep” with other sections of the course– the same student would have received a B+ or A- or A-bar if at Williams.

N.B. I generally graded those papers in Williamstown and shared them with Williams profs who had similar concerns.

To c.e.:

Thanks so much for your thoughtful recent comments running through several threads. I might otherwise, believe that my frosh was on ‘Academic Easy Street” at Williams and that the grades purportedly worked hard for, mean very little.

Do I remember correctly that you have actually finished at Williams? Are you in grad school or teaching full time? Also, what are you teaching?

I’m teaching High School English through Teach For America right now. And yea, I’m sorry, I do realize that my name is somewhat misleading…but given that I’m not using my real name I opted for consistency over accuracy in my pseudonym.

kthomas–that’s much more of a concern for Berkeley than for Williams given that both schools, when compared to the national pool of universities and colleges, will likely exhibit grade deflation.

current eph:

“High School English through Teach For America”! That must be challenging…and rewarding. Would you consider posting about it here? I, for one, would love to hear more.

Again, your thoughtful comments show a level head and a generosity of spirit…and add a lot to this site. Very much appreciated.