Advice for Test Taking Struggles at a Difficult University

[Ethan Klein]

I am currently taking upper division physics, math, and materials science courses at a difficult US university. I never feel lost during lecture, read and comprehend the text, go above and beyond on problem sets, and can nearly do any problem in the book. Yet, when it comes time for the exam, I either go to slow, miss connections between or the meaning of the variables, or make careless errors that I rarely make when working through book problems. In a nutshell, I choke.

While I fully comprehend all my exam problems (have not been stumped, yet) my grade just simply does not reflect my level of understanding. I can walk out of an exam (this happened yesterday), get on a chalkboard, and solve the problems in an instant.

I plan to go talk to some advisors and faculty to get their take on the matter, but in the meantime, I'm curious to know if anyone on the forum has had these issues and found a way to work through them. I do plan to apply to graduate school, and absolutely want to accurately display by level of understanding in my application.

 

[fresh_42]

It is usually a matter of practice and / or nervousness. Maybe these tips can help:
https://www.physicsforums.com/insights/10-math-tips-save-time-avoid-mistakes/

 

[Ethan Klein]

It is usually a matter of practice and / or nervousness. Maybe these tips can help:
https://www.physicsforums.com/insights/10-math-tips-save-time-avoid-mistakes/

Those techniques are awesome. I actually use a good number of them (especially writing out which functions are dependent on which variables and using grouping notions like brackets, parenthesis, etc) and will continue to do so.

Ironically, on the midterm I just took, when I went to integrate a gradient (to find potential from force), I did write out all the dependencies and still made a very odd mistake. Like you said though, it's over and I can't change it, and most importantly, I learned from it.

 

[WWGD]

Do you first read each question, get an idea of what you need to do and then start solving the ones that seem easier or do you jump in and try to solve everything right away?

 

[Ethan Klein]

Do you first read each question, get an idea of what you need to do and then start solving the ones that seem easier or do you jump in and try to solve everything right away?

Yes, that's exactly what I do. After reading all the questions, whichever one "speaks to me" first, I go for. If I get stuck halfway through, I either go onto other questions or try to use the latter parts of the problem/given constants as hints.

Sometimes, I will admit, it takes me too long to pull the trigger and move on. And when I do move on, I have recently been slow to start problems. To me, it's not that I don't know how to solve the problem, it's that I want to think slowly and methodically so that I start problems correctly, and don't waste a bunch of time doing math for an incorrect setup.

 

[WWGD]

Yes, that's exactly what I do. After reading all the questions, whichever one "speaks to me" first, I go for. If I get stuck halfway through, I either go onto other questions or try to use the latter parts of the problem/given constants as hints.

Sometimes, I will admit, it takes me too long to pull the trigger and move on. And when I do move on, I have recently been slow to start problems. To me, it's not that I don't know how to solve the problem, it's that I want to think slowly and methodically so that I start problems correctly, and don't waste a bunch of time doing m6ath for an incorrect setup.

I would suggest you consider that the thinking , methodical evaluation are to be done _ before the test_ ( for tests I would consider reasonable) and the test is about responding, answering questions. I think I remember research to the effect that test scores are inversely proportional to brain activity . At some point you need to stop thinking and act, answer questions. Of course, over the longer run you may revisit and improve, deepen your knowledge , but the time of the actual exam is hammer-time, time to execute the material you have already learnt, not to do yet more thinking . Of course, I am assuming these are standard timed exams; it would be different for other types, e.g., take-home exams. Imagine you train yourself to , e.g., fight fires. When the real thing happens your best , maybe even necessary , course of action is to react and put out the fire, not to make a careful evaluation. Thinking everything at every moment will bog you down, maybe paralizing you: paralysis by analysis. Yes, you need to think beforehand, but not all the time. I used to be like you in this regard, but now I see it as a flawed, ineffectual way of doing things. Think, act, take data and tweak, rethink. Of course, there may be extreme cases in which the seriousness of a situation may require constant thinking. Save for these, constant thinking, like most strategies/courses of action, has a point of diminishing returns.
My $.02.

 

[Dr. Courtney]

You need to train how you plan to fight.

Most students I've seen in your shoes benefit from practicing an intentional problem solving approach on every homework problem. Something like IDEA

Interpret:
Understand what the problem is really asking. Identify the important general principles of physics needed to solve it.

Develop:
Draw a picture and articulate a specific plan to solving the problem. No equations yet. But often students need to solve for intermediate variables not asked for explicitly in route to a solution.

Evaluate:
This is where you write down and solve the equations. Delay plugging in numbers until the last step. Include units and do the math on the units also.

Assess:
Does the quantity, sign, and units make sense?

 

[WWGD]

My bottom line view is that tests, time-based tests, measure how well you've learned the material up to a given point, so you need to show what you know, not think it through endlessly. You practice, practice, but at some point, you stop practicing and perform.

 

[Ethan Klein]

I would suggest you consider that the thinking , methodical evaluation are to be done _ before the test_ ( for tests I would consider reasonable) and the test is about responding, answering questions. I think I remember research to the effect that test scores are inversely proportional to brain activity . At some point you need to stop thinking and act, answer questions. Of course, over the longer run you may revisit and improve, deepen your knowledge , but the time of the actual exam is hammer-time, time to execute the material you have already learnt, not to do yet more thinking . Of course, I am assuming these are standard timed exams; it would be different for other types, e.g., take-home exams. Imagine you train yourself to , e.g., fight fires. When the real thing happens your best , maybe even necessary , course of action is to react and put out the fire, not to make a careful evaluation. Thinking everything at every moment will bog you down, maybe paralizing you: paralysis by analysis. Yes, you need to think beforehand, but not all the time. I used to be like you in this regard, but now I see it as a flawed, ineffectual way of doing things. Think, act, take data and tweak, rethink. Of course, there may be extreme cases in which the seriousness of a situation may require constant thinking. Save for these, constant thinking, like most strategies/courses of action, has a point of diminishing returns.
My $.02.

See, while this has my usual approach in years past, I feel that I have gotten away from this mindset. This was refreshing to hear, so thank you.

I will say though, the University that I am at is synonymous with very difficult exams. Luckily, I have only encountered a few tricky questions and nothing too unfair. The midterm I just had was very fair, but I would think too much after I had begun a problem, rather than trusting myself and working from pure ability.

 

[Ethan Klein]

You need to train how you plan to fight.

Most students I've seen in your shoes benefit from practicing an intentional problem solving approach on every homework problem. Something like IDEA

Interpret:
Understand what the problem is really asking. Identify the important general principles of physics needed to solve it.

Develop:
Draw a picture and articulate a specific plan to solving the problem. No equations yet. But often students need to solve for intermediate variables not asked for explicitly in route to a solution.

Evaluate:
This is where you write down and solve the equations. Delay plugging in numbers until the last step. Include units and do the math on the units also.

Assess:
Does the quantity, sign, and units make sense?

Yes this is very good advice, and is the system I use and have used since high school. Unfortunately I've been getting a bit lazy with the last step, which would've saved me from some of my careless errors on my midterm. All I can do now is, live and learn from that mistake and make sure I check units in the future.

 

[Joshy]

I'm not a good test taker. It's probably not great advice, but I hope even a little bit that felt helpful for me might help you along with the other suggestions.

Something I felt like helped me in exams is to write down everything I could during the first 2-5 minutes of the exam (could be a few formulas, the symbols and units of things, a graph or two)- like making a cheat sheet at the top of your exam, but you wrote it down during the exam from your own memory. I noticed that I would get through 1-2 problems full of confidence before stumbling just a tiny bit and totally choke for the remainder of the exam. Have the little cheat sheet like material at the top, beginning or back of my exam would sometimes save me; squeeze out a few more points or have something to work with to make it through.

I scarcely had time to finish the exam, but if I was choking or stuck on something... or lucky enough to finish early, then I would run through sanity checks on my problems. The easy ones for me would be to make values of things go to zero or infinity and see if the result would make sense. I did get burned with that trick once because the zero didn't account for a sign error-- something to be aware of or maybe try deliberately a value like -1 to whichever parameter you're testing.

I didn't just study the materials, but I would look for any shortcuts I could get. For example: That H. Cover method saved my behind on a bunch of exams. I studied electrical engineering and so there was something called Middlebrook extra element theorem professors loved taking well-known circuit problems and breaking them a bit; that formula showed you how to take the known/easy formula and multiple it with a correction factor for the extra elements added to the circuit.

 

[Ethan Klein]

I'm not a good test taker. It's probably not great advice, but I hope even a little bit that felt helpful for me might help you along with the other suggestions.

Something I felt like helped me in exams is to write down everything I could during the first 2-5 minutes of the exam (could be a few formulas, the symbols and units of things, a graph or two)- like making a cheat sheet at the top of your exam, but you wrote it down during the exam from your own memory. I noticed that I would get through 1-2 problems full of confidence before stumbling just a tiny bit and totally choke for the remainder of the exam. Have the little cheat sheet like material at the top, beginning or back of my exam would sometimes save me; squeeze out a few more points or have something to work with to make it through.

I scarcely had time to finish the exam, but if I was choking or stuck on something... or lucky enough to finish early, then I would run through sanity checks on my problems. The easy ones for me would be to make values of things go to zero or infinity and see if the result would make sense. I did get burned with that trick once because the zero didn't account for a sign error-- something to be aware of or maybe try deliberately a value like -1 to whichever parameter you're testing.

I didn't just study the materials, but I would look for any shortcuts I could get. For example: That H. Cover method saved my behind on a bunch of exams. I studied electrical engineering and so there was something called Middlebrook extra element theorem professors loved taking well-known circuit problems and breaking them a bit; that formula showed you how to take the known/easy formula and multiple it with a correction factor for the extra elements added to the circuit.

I love the Heaviside Cover-up method! The math class I'm taking right now is all about applied math and using shortcuts to do problems quickly. While I do not claim to be good at applying any of the techniques I've learned in that class, it absolutely has made computation easier and I will continue to try applying those skills.

Good luck on your exams too, Joshy, sounds like we're in the same boat.