Solving Physics Problems: Strategies & Practice

[americanforest]

How do you guys go about solving a Physics problem. Is there a specific method which most of you follow? I think this is a part of Physics that isn't really dealt with at universities and schools but is very important. When you have a multistep problem, in which a path to the solution is not immediately clear, you can't just dive in. What to you all to do solve such problems?

Also, do you all think that practicing problems is the best way to improve this skill, or is it just a matter of thinking harder?

 

[ranger]

The way I solve physics problems is by writing down all the known parameters and the ones that need to be found are designated with a question mark. If a diagram is needed to clarify the situation, such as problems with kinematics, I'll make a quick sketch. Then all the mathematical relations between the knows and unknowns are written down. Working through the mathematical relations one step at a time and its pretty much straight forward after that.

As with mathematics, the more practice problems you do, the better you will become. Its not so much about thinking "harder", IMO if you really have to think that hard, it just shows that you are unprepared.

 

[cristo]

I'm not sure I can add much to that- but here's my attempt to do so! Like ranger said, the key to attempting problems is to first write down the variables you know. Then write down the ones you are asked to find out. If, as you say, there is no clear route to the problem, then work out what you can. Remember key relations and equations, but at the same time, knowing the concepts behind the equations will help you enormously--if you know the concepts, then you are more likely to pick the correct path to the solution the first time.

Also, do you all think that practicing problems is the best way to improve this skill,

Definitely. Practicing problems will help you identify, in future problems, the correct path to the problem quicker.

Its not so much about thinking "harder", IMO if you really have to think that hard, it just shows that you are unprepared.

Not necessarily. Some questions will make you think harder, and if we are considering an exam type question, then the examiner will throw in some questions that are different enough to all other questions you may have seen, that you have to think about them. This is not a bad thing, however, since it will enable you to show your knowledge.

So, to summarize I would say:
1. Make sure you know equations and the concepts behind these equations.
2. Plan your work well--work in stages, starting with the aim of the question and known variables, and building up to the solution of the problem.
3. Don't panic: if you know 1. and follow 2. then you will be able to solve the problem.

 

[americanforest]

So you suggest writing down all equations that can be deduced from the data before beginning instead of thinking of them as you go along through the problem?

 

[cesiumfrog]

Is there a specific method which most of you follow? I think this is a part of Physics that isn't really dealt with at universities and schools but is very important. When you have a multistep problem, [..] a path to the solution is not immediately clear

That's why you're given multistep problems.

There is a significant and known danger in science teaching, that students will learn the method of solving typical physics problems without understanding the physics concepts. It's a current topic of physics education research.

Classic example: student turns over exam paper, sees circuit diagram, and concludes (before even reading the problem) "must write down kirchhoffs laws". The test problem has nearly zero remaining physics content, it's now just a question of whether the student can step through the "solving simultaneous equations" algorithm (which they memorised when studying math). However, students who get full marks on this type of problem will frequently fail a much simpler problem (but with equivalent physics) if it is worded qualitatively (to employ the physics behind kirchhoffs laws, rather than the quantitative equations that usually follow).

So no, I advise studying the physical concepts rather than recipes for solving textbook problems.

 

[Gilligan]

Well, I just look at the problems. Then I decide what known equations will make this problem easiest to solve with the given values and depending on what type of problem it is. Drawing a picture is often useful.
Physics isn't history or english class. You have to actually think.
I can't really say much more than that without inserting a bunch of verbiage.

 

[cristo]

So you suggest writing down all equations that can be deduced from the data before beginning instead of thinking of them as you go along through the problem?

I never said write down all the equations you know- that would take ages! No, I said make sure you know the relevant equations, then you'll be able to pick the correct one.

 

[Moridin]

I usually do it in steps as well, although I'm not really consciously thinking of them. I guess I do it is something like this:

Understand the Problem - If one does not understand the problem, it is likely that one will not understand the solution either. What is the question? What are they looking for?

Draw a figure - It doesn't matter if it appears to be an easy task, an image will always help, especially when dealing with vectors and directions. It will also give a better general view of the problem and stupid mistakes can be avoided easier.

Think of and state laws, relationships and formulas - Which ones can be used in this situation? Which variables are missing and how can they be incorporated?

Do the necessary calculations - Which variables is needed to get the answer? Follow the direct link between relationships and laws that was given in the previous point. SI-units should always be used when calculating the final answer.

Check the answer and use the right unit - Repeat the logical steps as well as the calculations to see if the answers match. Is the answer in the requested unit?

Is the answer reasonable? - The distance from the Earth to the sun isn't 10 meters. Does the answer fit with reality?

These points also help in writing an adequate and clear solution.

Of course, the easier the problem gets, the more steps are ignored. For instance, if it is something as easy as s = vt, I barely do number 4 on the list.