<p>For those of you who are interested, you can download the podcast of NPR's Science Friday with Smoot as a guest. He actually says some pretty interesting stuff and one really funny thing.</p>
<p>It's worth a listen.</p>
<p>
[quote]
I have to disagree. You are basing this because you think Princeton students will be hurt by the grade deflation? Well Princeton students who receive 3.5s and 3.6s nevertheless get into professional schools, added to the fact that the number of As given out only dropped by 5.1% after the policy, leds me to think that this won't hurt Princeton students very much.
[/quote]
</p>
<p>The problem is that Princeton compares itself to schools that have not, and probably will not, enact grade deflation. So, compared to those schools, Princeton has now (sadly) put itself at a disadvantage. </p>
<p>Look, let's be honest. Princeton is in a dogfight for yield vs. schools like Harvard, Yale, and Stanford. And, frankly, it's usually a losing fight. According to the Hoxby revealed preferences ranking, Princeton is less preferred to the other 3, and obviously is less preferred to Harvard by a wide margin (as everybody loses to Harvard by a wide margin). Knowing that Princeton has cut back on grade inflation when the other schools haven't just gives people another reason to prefer HYS over Princeton. </p>
<p>Let me put it to you this way. Let's say you got into Harvard and Princeton. You (like most people) would probably prefer Harvard just for the prestige of the brand name. But now you hear that Harvard also grades easier than Princeton. Well, now you REALLY will probably prefer Harvard. So Princeton will be able to win even fewer students away from Harvard than it could before. </p>
<p>The point is, it's all relative. Sure, you might say that Princeton won't be hurt very much by the change in grading, and that's true. After all, Princeton grading will still be more inflated than most other schools out there, including Berkeley. But that's not the point. The point is that, compared to its competing schools (HYS), Princeton will lose out on a relative basis. </p>
<p>
[quote]
On a similar note, what are your thoughts on Harvard's ending early action? Princeton followed, and later UVa. I think Cornell's graduate department is also trying to pressure Cornell into dropping its ED program. Do you think others will follow (so far it seems like few are following) or do you think this is the same as grade deflation?
[/quote]
</p>
<p>Ah, but you see, this is different? Why? Simple - Harvard is Harvard. Harvard has the clout to make a change and force others to follow. For example, if Harvard were to institute grade deflation, it would lose some students, but it would still most likely have the highest revealed preference ranking in the country. After all, Harvard is Harvard. Whether we like it or not, Harvard is the biggest brand name in the world when it comes to education. </p>
<p>The truth of the matter is, whether we like it or not, Harvard sets the stage for American education in a way that Princeton does not, because Harvard is in the leadership position. Princeton is not. And, frankly, Berkeley isn't either. </p>
<p>
[quote]
So, why can't we just try to make the different classes grade more similarly in difficulty? Unless I'm missing something, you haven't pointed out what's wrong with this idea. For example, we can do what Princeton did and limit the number of As, and also Ds and Fs, and have that be a campus-wide policy.
[/quote]
</p>
<p>You can try all you want, but I doubt it's going to really happen anytime soon, because like I said, there would be too many moving parts. The truth is, profs at any school, including Berkeley, jealously guard their prerogatives and want to run their courses the way that they want to run them, and rarely cotton to anybody telling them how they should teach or (especially) how they should grade. This is especially true of tenured profs who are unfireable and know that they can run their classes however way they want because they are protected by tenure. </p>
<p>Let me give you an example. You mentioned that recently a prof in Math 1B landed on probation because he failed 40% of the class. Let me guess. The prof was Hung-Hsi Wu, wasn't it? If it was (and I strongly suspect it was), then let me tell you that he landed on probation about 15 years ago for the EXACT SAME THING - failing too many people in Math 1B. Heck, one of my friends, who is a pretty smart guy who eventually got his PhD, got a D from Math 1B with Wu. In fact, I think Wu lands on 'probation' every few years for failing too many of his students. </p>
<p>But don't you see the problem with this? What's the point of even putting a prof on probation if he is just going to go back and do the exact same thing when he teaches the class again? This is just a stupid merry-go-round. He teaches the class. He fails too many students. So he lands on probation. He gets off probation. He teaches the class again. He fails too many students again He gets put on probation again. Etc. etc. etc. Honestly, why even have the concept of 'probation' if it doesn't change anybody's behavior? If the guy is just going to keep doing the same thing over and over again anyway, then why even bother to put the guy on probation? </p>
<p>And even in the unlikely event that it isn't Wu that you're talking about, but rather some other prof, that still portrays a weakness in the system (albeit a lesser one). In this case, you would have a prof (either Wu or somebody else) still failing too many students. But at the end of the day, the results are still the same. </p>
<p>Hence, either Wu (or the other prof) is just completely uncontrollable, or the math department doesn't want to exert its control over its profs. But either way, the end result is the same - that individual profs have the power to basically do whatever they want to do, and if they insist on using a grading scheme that is beyond the pale, very little can be done about it. That's the power of being a tenured prof - you can basically run your classes however way you want, and you know that you can't be fired. Nor is this specific to the math department. There are profs in other departments that have used grading schemes that differ greatly from the norm (either they grade too easily or too harshly). </p>
<p>Hence, this is a major reason why I don't think you can just simply standardize grading. It would be nice to do, but I doubt it will happen. You can make all of the standardization rules that you want. But what is a tenured prof doesn't want to follow it? What are you doing to do about it? Answer - nothing. He's tenured, so he's untouchable. Tenured profs have the power to grade their classes however they want.</p>
<p>Now, I know what you are going to say. You are going to ask "Well, what about Princeton?". First off, there are Princeton profs who have stated that they will continue to grade however way they want. In fact, the latest reports indicate that Princeton profs still have not met the guidelines set forth under the grade deflation policiees (although there has been changes). But more importantly, the stakes are obviously a lot lower at Princeton. After all, Princeton doesn't have any impacted majors. Princeton also doesn't have a culture of weeding - basically, everybody at Princeton is going to graduate, whereas not everybody at Berkeley is going to graduate. Again, I think this has to do with the fact that Princeton admits a very high caliber of student such that they don't really need to do weeding or impaction. If Berkeley had a student body quality that was as high as Princeton's, then, frankly, Berkeley probably wouldn't have to weed or institute impaction either. </p>
<p>The point is, at Berkeley, the stakes are high. Weeding and impaction decides who will be able to get into a particular major, and in certain cases, who will be able to graduate at all. Hence, if the stakes are this high, then we need to institute a fair mechanism for deciding how - a mechanism that has as few moving parts as possible. Trying to standardize the grading policies of a bunch of profs is hard, particularly when you have some recalcitrant profs (like Wu) who are just going to grade however way they want regardless of what you tell them. Hence, since changing grading policies is hard (as it basically involves changing the culture of the professorial ranks), then what is easier to do is to use a standardized test. </p>
<p>Profs can still grade any way that they want, but the effect of that grading is reduced. Put another way, what Wu has done is completely screw over a disproportionate percentage of his students from entering impacted majors. By using a standardized test, you have taken that power away from Wu. He can still grade however he wants, but he no longer will have the power to damage people's chances of entering impacted majors. You can get a terrible grade from Wu but as long as you can pass the entrance exam, you can get into the major you want.</p>
<p>
[quote]
Me neither, but hey, getting rid of affimative action has elicited firestorms of protest before...so I'm not saying it's hopeless.
[/quote]
</p>
<p>It elicited firestorms of protest, but there was also clearly a rather 'silent majority' who benefitted from the elimination of affirmative action and hence campaigned for its removal. Namely, whites and (especially at Berkeley) Asians. I remember back when 209 was being discussed, many URM's were obviously outraged. But many Asians, who are also minorities (but are not UNDERREPRESENTED minorities) were quietly pushing for its passage. Not all Asians, obviously, but many (probably most) Asians at Berkeley. Simply put, I think most Asians at Berkeley realized that they were getting screwed by AA. Come on, seriously, when the official UC policy was to automatically subtract points from the academic index or the applications of all Asians just because UC has "too many" Asians, that was obviously a practice that is extremely discriminatory towards Asians. </p>
<p>Hence, Prop 209 as it related to Berkeley was, if nothing else, at least it was 'democratic', because the beneficiaries were the majority. After all, the majority of students at Berkeley are white or Asian. I don't use the word 'democratic' in a normative sense, because just because something is democratic does not always make it right. But the point is, at least it had political support from a lot of people, and that gave the policy a base of political legitimacy.</p>
<p>In the case of raising admissions standards, this probably won't be the case. After all, the benefits are highly diffuse. If you raise admissions standards, then you can reduce weeding and impaction. But the fact is, many students take easy majors that don't have weeding and impaction anyway. So they get no benefit. </p>
<p>The point is, a political base existed for Prop 209 that doesn't exist for raising admissions standards. That is why 209 can pass, but raising admissions standards is unlikely. It's sad, because frankly, I think that having high admissions standards is MORE important than bickering over AA. But that's the way that politics works. Raising admissions standards has no natural political constituency to strongly support it. I wish there was, but there isn't.</p>
<p>Well, I'd still like to think that professors like Wu are in the minority and that if some system is instilled it would help balance the grading out a lot better than it is now, even if it's not perfect.</p>
<p>And what I think will happen is Princeton will keep its system while other schools may follow (probably unlikely as no school has so far) or may not. But we will see.</p>
<p>Thanks for your explanations. What a screwed up system. I suppose in lieu of increasing selectivity we can aim for increasing the quality of the student body by attracting better students, and increasing graduation rate (not sure how to do that with weeders still around). Actually I think the 6-year graduation rate has been steadily climbing over the past several years so that's a good sign.</p>
<p>Sakky, instead of having one huge, terrifying exam; wouldn't be easier to normalize each class exam? That is, have the department make up all the exams for ALL the sections of a particular class. That would resolve many of issues you explained before. </p>
<p>Also, you can't guarantee that the professors taught all the material for the exam. I think a "qualifying" exam would be counter productive. Instead of students learning the material in anticipation for future courses, they will just learn the material for "the test." This really destroys what college is about. The days of over-stressing for the SAT's are over, the focus should now be how you learn, not how well you do on a exam. </p>
<p>I also can't agree with your comparisn with the PHd qualifying exam. This is a very special exam made up by your committee and is different from candidate to candidate. The test you are proposing will be one massive exam given to everyone and will not show the particular strengths of each student.</p>
<p>
[quote]
I also can't agree with your comparisn with the PHd qualifying exam. This is a very special exam made up by your committee and is different from candidate to candidate. The test you are proposing will be one massive exam given to everyone and will not show the particular strengths of each student.
[/quote]
</p>
<p>If you want to get into the nitty-gritty details of how the PhD exams are run, then let's do so.</p>
<p>It is true that every doctoral program at every school has a different method of running its quals. But let's talk about one of the more common formats that is used by many departments in many schools. In this format, the qual exams consist of 2 parts. There is the more general written exam, which basically covers a wide range of topics. This exam exists to ensure that you are reasonable knowledgeable about most of the general events in your discipline. And then there is a second qual exam, which may be written or oral, that is specific to your particular field. Basically, the qual exam is a 2-tier format, and you are only referring to the 2nd specialized exam of the 2 tiers.</p>
<p>I am going to use MIT as an example, mostly because I not only know where to find the data, but the data is also available to the public (i.e. I can't really use Harvard as an example because much of Harvard's data regarding quals is restricted to the Harvard community). This is how MIT runs its PhD quals in many disciplines. For example, this is how things are run in chemistry, EECS, physics, mechanical engineering, Engineering Systems, political science, and Economics (the ones I checked). In fact, I would strongly suspect that this 2-tier system is representative of the vast majority of the PhD programs at MIT that have a decent number of students (some MIT departments have only a handful of PhD students, and so they can give each of them a customized exam). </p>
<p><a href="http://web.mit.edu/chemistry/www/academic/cumulative.html%5B/url%5D">http://web.mit.edu/chemistry/www/academic/cumulative.html</a>
<a href="http://www.eecs.mit.edu/grad/3800.pdf%5B/url%5D">http://www.eecs.mit.edu/grad/3800.pdf</a>
<a href="http://web.mit.edu/physics/graduate/current/generalexams.html%5B/url%5D">http://web.mit.edu/physics/graduate/current/generalexams.html</a>
<a href="http://www.me.mit.edu/GradProgram/GradGuide.htm#Section10%5B/url%5D">http://www.me.mit.edu/GradProgram/GradGuide.htm#Section10</a>
<a href="http://esd.mit.edu/academic/exam_faqs.html%5B/url%5D">http://esd.mit.edu/academic/exam_faqs.html</a>
<a href="http://econ-www.mit.edu/graduate/special.htm%5B/url%5D">http://econ-www.mit.edu/graduate/special.htm</a>
<a href="http://web.mit.edu/polisci/grad/doctoralprogram.html%5B/url%5D">http://web.mit.edu/polisci/grad/doctoralprogram.html</a></p>
<p>While I don't want to check around all of Berkeley, I know that the Berkeley math PhD program uses a similar procedure of a 2-tier exam (a "preliminary" exam and then a "qualifying exam"). I'm sure that other departments at Berkeley follow a similar procedure.</p>
<p>The point is that using a standardized test to determine whether you can advance is a mainstream technique that is used by many of the best PhD programs in the world. I am not proposing something that is way off-base here. This is a well understood and widely utilized method.</p>
<p>
[quote]
Sakky, instead of having one huge, terrifying exam; wouldn't be easier to normalize each class exam? That is, have the department make up all the exams for ALL the sections of a particular class. That would resolve many of issues you explained before.
[/quote]
</p>
<p>No, that actually doesn't resolve anything at all. The issue is not so much about what the CONTENTS of the exam are. The issue is the GRADING of the exams. Like I said, you can be one of the top students Wu's Math 1B class, and still get a terrible grade. That's what happened to my friend. He was good enough to later get a PhD in engineering, so he clearly wasn't dumb. But he got a D from Wu in Math 1B, because LOTS of people got bad grades from Wu. Conversely, you can be a mediocre student and still get a great grade. </p>
<p>The main problem is that some profs are very easy graders and others are very harsh graders. THAT'S the problem. And profs generally consider it their prerogative to grade their classes however way they want. If a prof wants to give out all A's to all of his students, including the terrible ones who did no work, nobody can really stop that from happening. Conversely, if a guy (like Wu) wants to flunk 40% of his class, nobody can really stop that either. Yeah, you can put him on probation, but like I said, that apparently doesn't accomplish much, because once he gets off probation, he's going to continue to do the same exact thing anyway. </p>
<p>So the real issue is what vicissitudes and I have pointed out - that the weeding is unfair and uneven. You can study extremely hard, know the material extremely well, and get quite good scores on your exams and homeworks, and STILL end up with a worse letter grade than somebody who was mediocre, just because you happened to have a tough-grading prof and the other guy happened to have an easy-grading prof. </p>
<p>
[quote]
I think a "qualifying" exam would be counter productive. Instead of students learning the material in anticipation for future courses, they will just learn the material for "the test." This really destroys what college is about. The days of over-stressing for the SAT's are over, the focus should now be how you learn, not how well you do on a exam.
[/quote]
</p>
<p>Uh, I think you have a highly romanticized view of the way that things work now. Let's face it. RIGHT NOW, you have people just learning the material for 'the test'. It just happens that 'the test' in question is basically the midterm exams and final exams of the classes they are taking. Let's be perfectly honest here. Right now, there are a lot of undergrads at Berkeley and every other school who, frankly aren't REALLY that interested in learning. They're just interested in getting good grades by getting a high score on the test. That is happening RIGHT NOW. </p>
<p>Hence, I don't see how my proposal makes things any worse than how things are right now. </p>
<p>
[quote]
Also, you can't guarantee that the professors taught all the material for the exam.
[/quote]
</p>
<p>By the same token, MIT PhD students can't be guaranteed that their profs will have taught them all they need to know to pass their quals. In fact, some of the MIT quals specifically say that they are not going to follow the specific formats of any class and that you would be expected to pick up some material via self-study. </p>
<p>But so what? What's so bad about self-study? I don't know that you should necessarily have the expectation that you are going to be able to learn absolutely everything that you need to know within a class. Self-study should be a technique that every decent student should be expected to know how to use. Seriously, if you are completely unable to learn anything by self-study, you got bigger problems to worry about. Whatever it is that you choose to do in life, you're not going to have somebody teaching how to do everything that you need to know. In fact, you will only rarely have such a teacher. Most things in life, you are going to have to teach yourself via self-study. </p>
<p>So if the prof didn't teach you about topic X, then you should go to the library and read about it yourself. Is that really such an outrageous proposal? Are you really saying that you can't get the book and read it yourself, and that you always have to have somebody telling what to do?</p>