Tutoring medical students for their physics exam

[magnetic flux]

I am a pure physicist (M.Sc.) and on the side I do tutoring. Right now I have medicine students who have to take a physics course as part of their curriculum. Their course consists of one lecture (2 h/week) and a lab course with 10 experiments. The exam in the end are 20 questions where only the answer counts, it's hit or miss. The derivation of the result does not matter at all, there is no partial credit. Also no use of a book of formulas or a calculator.

My feeling is that this form of the exam was chosen after a couple lawsuits. Students used to bring in a hand-written cheat sheet, but that got phased out. Not giving partial credit means that there is no leeway which could be used in a lawsuit against an “unfair” grading. So they are super strict now, at the expense of students who don't sue the university after a failed exam.

There are lots of questions from old exams from the web, students taking the exam have written them down from memory. We have access to these lists and as I myself have had good experience with studying by solving problems, I do the same while tutoring. Letting them solve the problems and talk about the stuff that they had difficulties with, sometimes doing a bit of background.

Some sample questions are:

• Give the unit of viscosity in base SI units.
  Answer: $\mathrm{kg \, m^{-1} \, s^{-1}}$
• A bat picks up a moth at a 45° angle on the same height, its ears are 1.41 cm apart. What is the path difference between the sound waves to its ears?
  Answer: $1\,\mathrm{cm}$
• What is the smallest object size that bats can resolve at a speed of sound of
• $c = 300 \, \mathrm{m\,s^{-1}}$? Hint: What is the highest pitch that one can resolve?
  Answer: 15 mm
• Transition from glass to air, make a sketch of the light ray.
• Determine the voltage in an X-ray tube when the limiting wavelength is 15 pm.

Except for the second one where it took me a while to reverse-engineer the question to figure it out, these are all things that I can easily do in my head or with little envelope calculations. But that is likely because I have a solid foundation in mechanics, electromagnetism and geometrical optics from three terms with lectures (2 + 2 h/week) and tutorials (3 h/week) in the bachelor course. Of course there was also theoretical physics and much more, I'm not counting that.

The medicine students go through these topics in a much more superficial fashion, just from lecture time it is a factor 6. Of course they do not need to go as deep, but I feel hard pressed to study physics superficially and still retain some knowledge. I need to connect it with other things, weave a dense mesh of understanding; I cannot do random facts. But they only have one more week before the exam, it is just a side lecture that they will not *strictly* need any more after it. All they want to do is pass it.

When I tutor I find them struggling with these type of questions as they cannot recall the whole setup of an X-ray tube and the associated equations with that and therefore feel lost with the fifth question. I then go and draw an X-ray tube, go through the steps and make sure that they understand each of them. After that they seem to understand it. But in the next tutoring session virtually everything is lost again because it was not build on a solid foundation. A widespread strategy around medicine students therefore is to learn stuff by heart like “use equation $U = h c / e \lambda$
for the question with the maximum wavelength on the X-ray tube”. It feels wrong to do physics this way, but they are not studying physics.

Given the time constraints and their motivation of just wanting to get over with physics altogether, how can I best teach them such that they get most at least half of the questions right?

 

[Mister T]

I have taught a MCAT preparation class a few times and had similar experiences. You can't teach students who don't want to learn. Keep in mind that the most lucid explanations delivered in the most skillful manner will have little impact on all but a few of the students' understanding. They must be given the opportunity to explain things themselves with the guidance of an instructor. Do a search for Peer Instruction for one example of this.

I understand and appreciate the difficulty of the task. Many if not most students are seeking answer-making strategies while the instructor is emphasizing sense-making strategies. Explain to your students that the test questions are designed to see if students understand and that therefore their attempts at answer-making strategies will not lead to success.

 

[magnetic flux]

Letting students explain it themselves is really powerful. I have learned a lot while teaching, so this certainly is true.

The strange thing is that other medicine students that have passed this course said that just learning all the problems by heart has actually allowed them to pass. And they say that they are more compatible with that approach, so I really think that going for understanding is not strictly needed to pass the course.

 

[Mister T]

The strange thing is that other medicine students that have passed this course said that just learning all the problems by heart has actually allowed them to pass.

I've heard the same anecdote. The evidence indicates that that strategy doesn't work for the vast majority of students.

To really succeed in their careers they need to understand the physics.

 

[kuruman]

Do you write the exam questions or are they handed to you? If handed to you, do you also get a set of grading criteria? I suspect that your University may not entirely avoid lawsuits with this new style of exam, especially with questions that are ambiguous or poorly written and do not have a single answer.

Example 1

What is the smallest object size that bats can resolve at a speed of sound of
Answer: 15 mm

Here the student is expected to know the frequency of h sound emitted by the bat (20 kHz) and then divide that into the given speed of sound of 300 m/s to get the answer of 15 mm. However, the web informs me that the frequency that bats use is in the range 20-200 kHz. Therefore, the smallest object size that bats can resolve is, arguably, 1.5 mm. Would you grade 1.5 mm as incorrect? Had you been able to see the student's work, any answer between 15 mm and 1.5 mm should be correct. Conversely, the student who puts down 1.5 mm can successfully argue in a court of law (as I did above) that the answer ws wrongfully marked incorrect.

Example 2

Transition from glass to air, make a sketch of the light ray.

All three sketches shown are correct. However, as a grader of this I would have a hard time deciding whether a student who sketched C is completely clueless about refraction or understands it so well that he/she drew the case of internal reflection. The ambiguity rests with the phrasing of the question: the word "transition" implies that the ray is moving (transiting) towards the interface from the glass side but is not strong enough to denote that the ray actually exits the glass.

Mind you, I am not attacking your exam. I am trying to point out that when constructing exams "where only the answer counts" one must be extra super clairvoyant to foresee where and how the questions may be misleading and misinterpreted or else, oops here comes another lawsuit.

I taught quite a number of medical students before I retired and I found them to be very competitive. A method that worked very well for me was "The booklet". It was a collection of almost 200 questions (of the show your work and reasoning kind) covering the entire course and posted on the web. Numerical answers to the problems were provided so that students who worked on them would know if they did them right. They were told that a certain number of questions on the current material will be picked at random, there will be no numbers, just symbols, and put on the test. I gave four hourly tests which meant they had to study and learn how to do about 50 problems. My message was simple, "I am not out to get anyone, there will be no surprises and if you know how to do all the problems, you have nothing to fear." It worked well and I could see who was ambitious and who was not. There were always those who didn't care, but those who did care, did well. Without a doubt, anyone who could do these 200 problems achieved the goals of the course.

 

[Michael Price]

Do you write the exam questions or are they handed to you? If handed to you, do you also get a set of grading criteria? I suspect that your University may not entirely avoid lawsuits with this new style of exam, especially with questions that are ambiguous or poorly written and do not have a single answer.

Example 1
Here the student is expected to know the frequency of h sound emitted by the bat (20 kHz) and then divide that into the given speed of sound of 300 m/s to get the answer of 15 mm. However, the web informs me that the frequency that bats use is in the range 20-200 kHz. Therefore, the smallest object size that bats can resolve is, arguably, 1.5 mm. Would you grade 1.5 mm as incorrect? Had you been able to see the student's work, any answer between 15 mm and 1.5 mm should be correct. Conversely, the student who puts down 1.5 mm can successfully argue in a court of law (as I did above) that the answer ws wrongfully marked incorrect.

Example 2

View attachment 247233
All three sketches shown are correct. However, as a grader of this I would have a hard time deciding whether a student who sketched C is completely clueless about refraction or understands it so well that he/she drew the case of internal reflection. The ambiguity rests with the phrasing of the question: the word "transition" implies that the ray is moving (transiting) towards the interface from the glass side but is not strong enough to denote that the ray actually exits the glass.

Mind you, I am not attacking your exam. I am trying to point out that when constructing exams "where only the answer counts" one must be extra super clairvoyant to foresee where and how the questions may be misleading and misinterpreted or else, oops here comes another lawsuit.

I taught quite a number of medical students before I retired and I found them to be very competitive. A method that worked very well for me was "The booklet". It was a collection of almost 200 questions (of the show your work and reasoning kind) covering the entire course and posted on the web. Numerical answers to the problems were provided so that students who worked on them would know if they did them right. They were told that a certain number of questions on the current material will be picked at random, there will be no numbers, just symbols, and put on the test. I gave four hourly tests which meant they had to study and learn how to do about 50 problems. My message was simple, "I am not out to get anyone, there will be no surprises and if you know how to do all the problems, you have nothing to fear." It worked well and I could see who was ambitious and who was not. There were always those who didn't care, but those who did care, did well. Without a doubt, anyone who could do these 200 problems achieved the goals of the course.

Actually B is not possible, despite it often being an exam answer.
Hint: it is not reversible!

 

[magnetic flux]

Do you write the exam questions or are they handed to you? If handed to you, do you also get a set of grading criteria?

I am not involved with that lecture in any way, otherwise I would not feel comfortable tutoring it on the side. So these exam questions are handed to me by the students I tutor, and they get them from the memory of other students. After exams they post all they can remember to a popular website. Often questions are incomplete or contradicting. I often need to peek at the answer to figure out the question.

As far as I know from my ex girlfriend officially tutoring that lecture, the grading criteria is just whether the right final result is in the box. If you know the answer by heart and put it in you get the point. If you do an almost perfect derivation but screw up a constant of nature, you don't get the point.

Given the strange nature of the questions and that they are non-obvious to me as a physicist, it feels that these medicine students will have a hard time. But perhaps they are worded a bit clearer than the account from memory of non-physics students makes them appear.

Example 1
Here the student is expected to know the frequency of h sound emitted by the bat (20 kHz) and then divide that into the given speed of sound of 300 m/s to get the answer of 15 mm. However, the web informs me that the frequency that bats use is in the range 20-200 kHz. Therefore, the smallest object size that bats can resolve is, arguably, 1.5 mm. Would you grade 1.5 mm as incorrect? Had you been able to see the student's work, any answer between 15 mm and 1.5 mm should be correct. Conversely, the student who puts down 1.5 mm can successfully argue in a court of law (as I did above) that the answer ws wrongfully marked incorrect.

I also found this very peculiar, but apparently there was no frequency given and just this hint of the highest pitch that one can hear.

Overall I get the feeling that they want to shield themselves from debates which non-physicists can pull off, like formal stuff. Maybe I should accompany them after the exam to take a look at it as some sort of expert witness.

Example 2
All three sketches shown are correct. However, as a grader of this I would have a hard time deciding whether a student who sketched C is completely clueless about refraction or understands it so well that he/she drew the case of internal reflection. The ambiguity rests with the phrasing of the question: the word "transition" implies that the ray is moving (transiting) towards the interface from the glass side but is not strong enough to denote that the ray actually exits the glass.

I think that everyone just draws A and that this works out somewhat fine. But yes, we know that more cases exist and that also makes the answer ambiguous.

Mind you, I am not attacking your exam. I am trying to point out that when constructing exams "where only the answer counts" one must be extra super clairvoyant to foresee where and how the questions may be misleading and misinterpreted or else, oops here comes another lawsuit.

Feel free to be honest about the exam, even if it was mine. Unfortunately I have no influence over the exam and I struggle to see how to prepare for this. I mean I would easily pass it because I know a lot more than they are trying to ask and I have enough practice with algebra that I don't do silly mistakes on that level. But for somebody at that level one wrong cancellation in the derivation means zero points and I find that utterly harsh.

Apparently they did not get burned by ambiguous questions to a point where they would avoid them. Perhaps the questions are clear and all the students just remember them incorrectly. There is too much hearsay to really criticize the exam from my point of view, but I am not thrilled by it.

I taught quite a number of medical students before I retired and I found them to be very competitive. A method that worked very well for me was "The booklet". […]

This pretty much happens with that website of remembered questions and answers. Still it seems that the exam is so answer-focused and not understanding-focused. Therefore I feel I have a hard time getting buy-in for actual understanding that is sufficiently deep.

So as an outsider I cannot change the system and the only thing that I can apparently try is to give them opportunity to exercise such that they can solve as many problems as possible without wasting time trying to explain the concepts in detail myself.

 

[gleem]

You need to remember that med students are up to their necks in information overload which they are trained to be able to recall quickly. They are trained to work under pressure and under conditions not conducive to much contemplation. They need to know that which will take them to the next step as quickly as possible. I have taught radiation oncology residents physics for their certification exams and have worked with radiation oncologists for years. They only learn enough physics for that exam. The last thing I think a normal practicing radiation oncologist remembers is Bragg-Gray cavity theory with is the basis for measuring radiation dose and which they were taught. It isn't their responsibility to know that. It is their responsibility to know how many Gray to prescribe for a given stage tumor and the tolerance doses for any tissue that is concomitantly irradiated.

Sometimes you do get a student that is actually interested in physics. The down side is when they get into practice they can be a pain in the neck for a medical physicist; you know the saying about a little knowledge.

 

[Mister T]

Given the strange nature of the questions and that they are non-obvious to me as a physicist, it feels that these medicine students will have a hard time. But perhaps they are worded a bit clearer than the account from memory of non-physics students makes them appear.

In the USA the exam pre-med students must take is the MCAT. The test questions are often terrible. Do your best to explain sense-making strategies rather than answer-making strategies. That's my advice. Otherwise you are doing nothing to advance the careers of these students. They have to understand basic physics to be able to grasp the concepts they'll encounter in med school. I hope I never have a doctor treat me who "studied" physics the way these students are trying to do it.

 

[vela]

Given the time constraints and their motivation of just wanting to get over with physics altogether, how can I best teach them such that they get most at least half of the questions right?

With just one week left, I don't think there's much you can do.

One common problem I saw with many pre-med students was memorizing formulas with little or no appreciation that they actually mean something. When presented with a problem, instead of identifying the concept involved, selecting the appropriate formula, and plugging in the right numbers, a student might instead think, "Energy is mentioned. Find all the formulas with an E in it," eventually settling on a formula that has some place to plug in all the given information.

You might try is having students reverse-engineer the formulas or explain where they come from. Or have them make a concept map, where they make diagrams identifying the key concepts (as they see them) and showing the relationships between them. Have them show where the various formulas fit on their diagrams.

 

[Auto-Didact]

This is part of a recurring difficult issue which I used to frequently run into while I was in the early stages of medical training, and today it is a problem I still sometimes run into when training junior physicians to think clinically in the way that I envision is ideal, which is an idiosyncratic of expert medical reasoning but then heavily biased of course due to my training in physics and mathematics.

Because the goal of medical education ultimately is to train effective clinicians, I think the correct answer is to allow them their memorization, because 1) it is what practically all physicians do as well, and 2) in the students' later career a (simplified) physical understanding will more likely only serve as an idealized impediment which cannot be generalized to meet the requirements of the rest of their thinking.

In practice, we already know the outcome of this educational experiment because it has been carried out for decades in many countries: a weak mathematical basis is basically irrelevant to becoming a clinical physician and any applied statistics deficiencies are easily fixed in specialized courses for applied statistics and/or by using computerized statistical software packages.

 

[atyy]

Given the time constraints and their motivation of just wanting to get over with physics altogether, how can I best teach them such that they get most at least half of the questions right?

If one used dimensional analysis, would be one get at least half the questions right?

Then one could teach dimensional analysis, which is physics.

And one could also teach when dimensional analysis fails, which can even include an introduction to the renormalization group, which is also physics.

 

[Auto-Didact]

If one used dimensional analysis, would be one get at least half the questions right?

Then one could teach dimensional analysis, which is physics.

And one could also teach when dimensional analysis fails, which can even include an introduction to the renormalization group, which is also physics.

I've tried teaching colleagues and some students dimensional analysis: some find it a nice but abstract technique, while most typically report that it is too foreign from the way they usually think, as well as remark that they have forgotten the prerequisite knowledge to use it effectively, i.e. knowledge of dimensions and respective units.

After learning dimensional analysis and brushing up on prerequisites, most admit they will not explicitly use it in practice, because - apart from making quick guesstimations - the actual clinical use cases which are proposed are too specific and too sparse to be useful generally; on the other hand, the general scientific use cases is an entirely different question altogether.

Also the guesstimations that are to be made clinically tend to be very limited in terms of different possibilities, so most routine outcomes automatically get memorized, and there actually already exist validated clinical tools which are able to achieve similar results with less thought.