Questions & Answers: Week 2

Here are some of the many comments which were sent to me after the third exam. Some of them I find insightful. Others I wish to respond to. In any case I hope that you find them interesting. As people continue to send me comments I plan to append them to the end of this document.

I think a major problem with 8.02 is that it assumes a lot of prior knowledge. Important subjects such as capacitance were glossed over quickly, and Exam 3 clearly indicates that many of us don't quite understand the basics. Another thing -- 8.02 also assumes prior knowledge of circuits. My partner and I have absolutely no experience with circuits and how to hook things up. Absolutely no idea. It would be nice if we spent a little more time going over the basics.: This is one of several letters which said that we assume lots of prior knowledge. Unfortunately none of them explain why this seems to be the case (please feel free to let me know if you feel this way). I assure you, I personally assume no prior knowledge of E&M, and only some prior knowledge of things that you really should have learned in 8.01 (so, for example, I don't assume that you remember simple harmonic motion). In particular I assume that very few people have any experience with circuits (this is a relatively new phenomenom though -- in years past most students would have built crystal radios, for example, and played with basic circuit kits).

What I do assume, and maybe this is where the breakdown occurs, is that you read before coming to class. We give you prelabs, both reading and questions on the psets, in order to make your experience in the lab more pleasant and to make it a more useful learning experience. As for spending more time going over the basics, I am of two minds on this, since I'm not exactly sure what you mean by "basics." In one sense, the most basic information we expect you to get through reading and doing mastering physics before class. This should prepare you by helping you know what you find confusing. In this way, you should come to class prepared to ask questions. By the way, this is one of those transferrable skills. I have a feeling many of you didn't use it to full advantage in 8.02, but I encourage you to try it out again in future classes. Read a little ahead. Think about what you will be discussing, and what confuses you. You might be surprised how much it helps.

On the other hand, I don't know that we spend time doing anything but the basics in class. I mostly reserve more difficult problems for the problem sets. In class what I hope you get is a discussion of how to think about the concepts, and practice working through basic problem solving skills.

I don't think the failure to realize the trick of making up the charge/current on problem 4a/b demonstrates significantly less understanding of the course materials. A suggestion for that method could have been included and then a majority of the students would have been able to calculate the relevant quantities. On the other hand, we had the impression from the class that inductors(coils) had inductance and capacitors (parallel conductors) had capacitance. We didn't do any problem that suggested otherwise. So when we got zero value for inductance, it seemed to make sense along that line.: This is another common statement -- that these problems had tricks and hence did not test understanding of
My suggestion, for grading, is to give people their grades as is. In my own opinion, if people have done poorly on this exam, its more a representation of their own effort on personally learning the knowledge than on the teaching of it. As students its our responsibility to learn the material and here at MIT we have plenty of resources available to make sure that this is done. Daniel Guetta not only had a review session but also sent out compiled comprehensive review notes. I hope that does not sound unduly harsh, coming from a student, but I did poorly on the 1st exam myself and have made a point of ensuring that I would do better on the exams that came later.: This was not the only email expressing this opinion, but it probably was in the minority. In some respects I agree. I think that many students at this point in the class were confused about the material, and so the exam, in that sense, was an honest assemesment of knowledge. At the same time, I don't think the first exam can be explained away in this way (hence the +2 pts on the grade break points). More importantly, given that this is the case, I want to take some action that will lead to the most students improving their understanding of the material the most, before the class ends. By allowing the final exam to weigh more heavily I hope that many of you will be encouraged to attend office hours, ask questions, study with your peers, and really learn the material before the final exam.
I just wanted to clarify how the 18.01/18.02 do over exams work, as they inspired a humble thought for the present 8.02 situation. If a student achieves an exam score that is below the minimum passing score (C-), he or she is entitled to take a makeup exam. The highest possible score on the makeup, however, is the minimum passing - even if a student gets a 100 on the makeup test, it is only recorded as whatever that minimum C- grade was on the real test. This both limits the number of test takers (for the sake of the PhD student proctor, who has to grade all of the makeup tests in his/her room) and respects the fact that the student already used up their first chance to achieve a good score. It simply allows a student to demonstrate enough competency on those specific exam topics to prevent a failing score from being recorded. I do understand, of course, for the reasons you listed that such a makeup exam would be impractical in this case. Even if the ability to take a makeup test were limited to only those students who received below a C- (63), the number of eligible students would still be overwhelming, given that the mean was 65. My idea is this: allow the submission of supplemental work to "bridge" some or all of the gap between a student's actual (poor) exam score and some passing threshold that you could set. Perhaps this work could consist of challenge problems from the course notes (which lack solutions), or perhaps some special MP assignments, or perhaps active/proctored participation in the office hours and weekend study sessions for material that focuses on topics covered by the exam. That way, deficient students would have the opportunity to demonstrate competency in the Exam 3 topics, which the teaching staff feels is critical to reaching a passing grade in the course. At the same time, capping the maximum possible grade benefit from this supplemental work would limit the amount of credit earned outside of an exam setting. (Now that I think about it, such a supplemental work scheme could provide a structured way to prepare for the final exam, just as an added benefit.): First of all, thank you for filling me in on the details. It is a little embarassing how little faculty in one department generally know of practices in other departments, so I am very glad that we have students who can help us share good ideas (like the essay questions on each week's mastering physics, which we borrowed from Course 16 Mud cards -- not quite there yet, as the TA who was going to be doing this broke his shoulder and couldn't type, but we are building up). Please don't hesitate to tell faculty of good things that are done in other classes. I can't speak for all faculty of course, but personally I really appreciate learning about what is used. Another example -- students have told me that in some classes homework is not graded but rather used as a study guide and tested. I like this idea, and plan to test it out in 8.02 in the fall. If you have any thoughts of whether this did or didn't work for you if you have been in class like this before, please share them with me.

Secondly, although we ended up not going this way I wanted to say that I really liked the idea, and we strongly considered it (mostly in the context of "bonus" questions on the final which would be applied to midterm grades). In the end I decided that a comprehensive final, allowed to replace the lowest midterm grade, would achieve the same ends (while being more generous, I guess, since I'm not capping it).

It is difficult to do further parts of a multi-part problem if you cannot do the first part for two reasons: a) you usually need your answers to the first part (or they are at least helpful in gleaning insight into later sections of the problem) and b) psychologically, it is extremely difficult to move on and continue a problem you couldn't do part A of. .: This is certainly true. Unfortunately, I'm not sure how to get around this issue if I want to ask more involved questions (which I certainly do). For example, in this problem there is simply no way to ask about the Poynting vector without first asking about E and B fields. I do want to report though that a large number of students managed to get the last part (the meaning of the Poynting vector) correct (to full credit) without getting the E or B fields correct. So I strongly encourage you to push ahead in problems even when you do get tied up on part A. In fact, some times you can work backwards. In this exam, a few students knew they wanted to get that the integral of the Poynting vector was e^2/R (power dissipated by the resistor) so worked backward to calculate the B field. Very clever.
The particularly low score for this exam is probably because that it covers too much materials as we don't know what to study for. I pulled a whole night the day before to study for this exam, just to understand better about the interference and diffraction problems, because although I learned the idea from the lab, I still wasn't able to do the film problems. So I spent a lot of time reading the course notes and researched online to learn about it. I also spent a lot of time learning about the wave function related problems, which turned out to be on the exam. But I did not realize that the capacitor/inductor problem would be a 50% portion of the exam, so I did not realize I should spend most of my time to master these materials.: The third exam was in many senses a mini-final. It was comprehensive and synthetic. So, think of it as a preview of what is to come on the final. My hope was that the distribution of problems on the sample exams I provided would mimic that on the real exam (I'm pretty sure they did). I will do this again in repackaging the old final exams into sample exams for you.
In my opinion, and that of many of my peers in 8.02, the issue with the first and third exam was not their length but familiarity with the material covered. I understand that the purpose of the exam was to test how well we've learned the material and its concepts; however, when the exam contains material that we've never seen before (as in the rolled/folded capacitor of the first exam and calculating the magnetic field, inductance, and poynting vector of a rectangular parallel plate capacitor), the purpose of the exam changes. I think including that final question in the test was somewhat unfair given that the stated purpose of the exams was to provide an opportunity for us "demonstrate mastery of the material," but the exam included material that was not covered. We never dealt with or covered rectangular parallel plate capacitors in this unit of the class but instead limited our examples in problem sets and group problems in class and the questions in previous exams to situations involving circular capacitors or solenoids. I think the second exam was more appropriate because the questions it contained were those similar to class problems and problems on previous exams. Questions like that should instead be on the problem sets, to challenge and push the boundaries of what we've learned. Exams should simply be questions that demonstrate our understanding of what has been covered in class. A valid question, I suppose, would be how you expect us to study? If our problem sets, group problems, and previous exams cannot prepare us adequately for the exam, what is it really testing? I would point to 18.02 with Professor Auroux as a fantastic example of how material should be delineated between problem sets and exams.: A comment with which I must completely disagree. On the specifics, it is rather weak -- calculating the capacitance of a rectangular parallel plate capacitor is no different than that of a circular parallel plate capacitor. I also framed the problem as closely as possible to the coaxial cable problem on the sample exams so as to suggest the methodology for solving it. But that is not really the important point. Much more significantly, I have absolutely no desire to teach you how to calculate the capacitance of a circular parallel plate capacitor. This knowledge is something that I am sure will flee your head shortly after the final exam (hopefully not beforehand). What I want to teach you is an important concept -- how to understand capacitance -- and a much more general skill -- even broader than how to calculate capacitance in general, how to look at a problem and decide that calculating the capacitance is something that you want to do. That is, I want you to learn abstraction -- how to look at a complex problem and boil it down to its essential physical components, then how to think about those parts. This is why my favorite questions in the class are the back of the envelope questions. They are hard, yes, because they make you do something that you aren't challenged nearly often enough to do -- to try to figure out the right questions to ask, the right assumptions to make. At the same time, these are much more the skills that you will need to rely upon in your life than any individual calculation that we make you do in the class.

So, to specifically address your comment and question: exams should NOT simply be questions that demonstrate your understanding of what has been covered in class. They should be questions that allow you demonstrate your understanding of the material which we have asked you to learn (some of which we want you to learn outside of class), including, in particular, the ability to extend techniques to novel situations. How do I expect you to study? By doing problems that get you to do the same. By asking yourself at every juncture, "What if this was different?" By realizing that memorization is not going to get you very far in this class, and that you need to actually understand the concepts and techniques in general, not just for specific situations.
Hello! I know you've heard a lot of feedback about the tests and scores and I've read some responses and your responses provided on the front page. I know you said you preferred non-anonymous responses because you can have a discussion, but this makes me feel safer to fully express myself. I want to be a physics major, and I've come close to failing all of the tests given. I love physics and the material and understand it extremely well - I commonly teach people things, help them with psets, etc. This makes me think there is something wrong with the teaching methods and/or tests. Personally, the TEAL method of teaching isn't teaching or really technologically advanced or whatever the goal was supposed to be. It's professors using a power point and being able to project themselves onto multiple screens. All it really does is allow me to see them better. Explanations are minimal, practice problems are almost nonexistent, and I feel like I don't understand any of the material covered in class. Everything I have learned is from high school AP Physics and my own work. It's like two hours of lecture with no recitation time to ask questions, work through harder problems, and really examine the material in depth. Sunday sessions are where people go to get pset problems worked out for them. The whole general method of the class fails me and many other people. Personally, I think there should be a lecture/recitation offered next year for people who choose it. As for the tests, the second one was a very fair test. There was a comment on the student response page that I agreed with but that I think you unfairly shot down. Tests are supposed to be an opportunity for us to show our mastery of a subject. Problem sets are supposed to take what we learned in class and apply it to generally harder problems that make us think more. Tests are also time constrained while Psets are not. Therefore, I would expect the tests to contain more basic problems as well as less than psets. The second is adheredto, the first is not. I understand that tests can be used to apply concepts to 'novel' problems, but, for example, the first test - the rolled/folded capacitor - requires a LOT of thought and time to figure it out. That problem is definitely more of a problem set problem. There has to be a line between testing us on what we know and pushing out brains to apply it to 'novel' ideas. Another small problem I had with grading on the third test is units. On problem 3 when we had to calculate all those values (wavelength, frequency, period, etc.) - a point was taken off for every value where there were no units. Clearly, forgetting to put units does not merit 10 points off because clearly we understood what we were doing - we just made a mistake to not put units. Yes, units are important, but 10 points? All on one problem? That's a bit ridiculous. Points/problem should probably have a set 1-2 points allocated for units and the rest should be for understanding of the calculations/material. That's almost a whole letter grade for a silly mistake on one problem. Personally, this class has definitely lowered my liking and appreciation of physics, which is a shame because I really adore it. I realize this was a really negative email, but I really want to add that I HIGHLY appreciate the effort you are putting in to listening to the students responses and finding a fair way to make our grades represent our true knowledge. It's definitely something that as a student makes me really trust that you have our best interest in minds. I guess my final say is that I had a few little complaints (units grading, hardness of test vs. psets), and really generally think my lack of understanding/doing well is purely based off of the TEAL teaching method.: Two (or more) comments for the price of one. I'll start by addressing the exam comments. First of all, for what it is worth, in my conversations with other faculty (outside of the course) when I mention the mean on our first exam their response is "perfect." Of course, our grading scheme changes the effects of a more standard (lower) mean, but please do be aware as you go on to other classes that exams are often designed to challenge you, and that most faculty (myself included) view this as a good thing. So, to address your comment that the test should contain more "basic" problems, in a sense that is true (I try to make them mathematically less sophisticated and to require less writing to solve), but I would argue that conceptually this should not be the case. Demonstrating how to apply your knowledge to 'novel' situations is EXACTLY what an exam should be testing, because in physics we don't care about your ability to memorize some set of standard problems but rather about your ability to understand and broadly apply what you have learned. So I, and I think most of the teaching staff, think that conceptually the folded vs rolled capacitor for example was actually a pretty good exam problem. It was a little wordy, and hence daunting. But I absolutely disagree that it was too complex for an exam, and frankly think it is too easy to be a problem set question.

Your next comment regards units. Here I strenuously disagree with you. "Forgetting" to write units demonstrates a huge hole in understanding of an important concept that is one of the most basic in science overall. A number without a unit when describing a physical value (aside from the few which don't take units) is completely meaningless. Otherwise, if I asked any question, such as "what is the radius of an atom," you could just answer "1." The implied units, you might argue, are the radii of atoms. Frankly, I think we were overly generous in our partial credit of problem 3 -- we gave 1 point each for choosing reasonable variable names. So do I think that in that problem that leaving off units is an error worthy of a 6 point deduction? Absolutely -- that you were able six times to write a number for a physical value without once thinking of writing the units is a huge mistake.

Finally, you lay your problems in the class at the foot of the TEAL teaching method. If I were to accept that hypothesis, could I then argue that the lecture-recitation format was responsible for the 15-20% of the students who failed in the former class style? The fact is, TEAL has led to documented (published) learning gains when compared to the lecture/recitation format. These gains are in line with those found at institutes across the country which have moved from passive to active learning methods. You state that "explanations are minimal [and] practice problems are almost nonexistent." But, at the very least, one hour a week (Fridays) is nothing but practice problems. And although I haven't sat in on every class of every section, I would be shocked to learn that they hadn't done any PRS questions or group problems, which are also practice problems. I'll let the rest of my feelings on this matter be summed up by a letter from another student:
I say this as a student who barely passed 8.01 (by barely I mean I got a D and had to take the IAP remedial class in order to change my grade to a C). I learned a lot from this experience and now know that anyone can do well in 8.01/8.02 by actually proactively reading the course notes, taking mastering physics seriously, and spending a decent amount of time on the psets. These may be things that you may think are very obvious to success, but I needed that rude awakening from 8.01 to realize it.: Read before class? Take time to think about the assignments? Great ideas! I want to temper this enthusiasm a little -- I think that 8.02 is a hard class. The material is very non-intuitive, and not something that you have daily experience with so it is often difficult to conceptualize what is going on. There are also strong intimidation factors. For example, we launch into vector calculus in the first week of class, which is, I think, terrifying for many who have not seen it in a math class before. There are also many students who have taken AP physics before, so they seem so far ahead to those who have not taken physics in high school (a sizeable number of students had never taken ANY physics), that it must seem impossible that they could ever "catch up." I call these intimidation factors, because I think in reality it isn't that bad -- though we use the concepts of vector calculus, we fairly rarely demand any real use of it (certainly not beyond a level that can't be picked up in a hour or two of office hours), and the fact that there are students who know way more is irrelevant to your ability to learn (actually it is a good thing as they can help). None-the-less, this is hard, and every semester there are a number of students who really have to struggle to pass, even some of whom are doing the reading and mastering physics. That said, I think the number of students who find themselves in that situation would be significantly less if more students followed your simple advice.

Exam 3 & Course Comments