What’s in a Good Question?


As a science student, I’ve taken many tests over the years. The staple of a science class is the tests that are spread out over a semester, so they are to be expected. Consequently, I’ve answered many questions on tests, and so I have a fair idea about which questions are actually good questions to ask on tests, and which ones seem like there is no point.

Why have tests?

First of all, I should establish why I think we even have tests. I’m pretty sure it has to do with the fact that most science classes are lecture based, meaning there is a lot of information coming in to the student. To be sure that this information is retained by the student, tests are given (and are the easiest way to do so).

I have no problem with the concept of taking tests. However, I do take issue with the way some of the tests are structured. Certain types of questions fit better in certain situations than other types.

Multiple choice

The almost ever-present feature of any test is the multiple choice section. Many questions are asked, and the only thing that is graded are the answers.

First of all, it seems to me that this is an incredibly dumb way to administer tests in subjects like mathematics or physics (at least, most of the time). The reason is that teachers from very early on tell students to show their work. This is always supposed to be heart of a solution, the process of how one got to an answer. Then, teachers turn around and give their students multiple choice questions where they cannot show their work as they’ve been told numerous times to do. It’s a bit of a confusing situation. Therefore, the multiple choice format does not really fit the objective of the test. When a multiple choice question is simply a short answer question with four options for answers listed, the question fails in my mind. It’s leaving out some students who may do everything right, yet make a small calculator error. And if we still want to take away points for that, then we have to consider the way we correct all the other questions too.

On the flip side, there are instances where multiple choice questions work a bit better. The example I can think of is my biology class, where the theory wasn’t usually mathematical, but more about reasoning with information that is given. In this case, I found the multiple choice format worked well, since the question doesn’t have much “work” that can be shown. Sure, you could argue that the logical reasoning could be shown, but I think that is stretching the idea of showing one’s work a bit much. For these types of questions, it’s okay to use multiple choice.

Here’s a quick example of such a question:

Given a situation X, what is the likely type of microevolution that occurred with this species?

In the above example, the question doesn’t really require work. It simply asks one to recognize the type of microevolution taking place based on the given situation.

Therefore, the multiple choice format can work, but it should only be for certain questions. They don’t serve any purpose (except for allowing the teacher to correct less and give more questions) when the format is used for a question involving some sort of calculation. Additionally, there’s also the possibility that someone who uses the correct reasoning but makes a small mistake gets the incorrect answer, while the one who doesn’t have a clue how to solve the question guesses an answer and gets it right. I’m not sure how frequently guessing works, but I assume it should work a quarter of the time (if there are four possible answers). This is obviously something we do not want, so this is another reason to use multiple choice as little as possible.

Can you actually solve the problem at hand?

Another issue I have with questions is the fact that there are times when questions require one to have knowledge that doesn’t really relate to the specific problem at hand, but nonetheless needs to be known if the problem is going to be answered. I’ve written about this before (link), but this usually comes up in the form of formulas.

The prime example is in my mathematics classes, where formulas were never (or rarely) given. We needed to memorize everything, or else there was no hope for us on the test. This was obviously frustrating, because it created situations in which one would know what needs to be done, yet they couldn’t actually do the problem because they did not precisely remember the formula.

I remember this happening to me while I was taking a linear algebra class. The question was about the Cauchy-Schwarz inequality, and I just couldn’t remember if the two terms multiplied together or if they were summed. I knew exactly how to answer the question if I did know the formula, but since I didn’t all I could do was go with the one I thought was right (I was wrong).

In this situation, I would have never gotten into this situation if the formulas were provided to me. And why shouldn’t they be? The important part of the question was not to test if I knew what the inequality was. It was to see if I could use it, a subtle difference. Therefore, giving me the formula wouldn’t have solved the problem for me, and it would have allowed me to actually solve it correctly (like I knew how to do).

This is the problem with many tests I’ve taken. What is being tested isn’t the only thing required on the test, even though it should be. I just gave an example of not remembering a formula that I believe should have been provided, but there are other examples as well.

Here is another situation. In my physics class, I took a test in which there was a lot of work to be done. The class only lasts fifty-ish minutes, and everyone was working right up into the time limit. Basically, no one was really finished, and so there was frustration all around. (Since my physics teacher was very nice and accommodating, she adjusted the grades of all the students, giving us a bit of a boost since the test was so long.)

What wasn’t fun in this situation was that the questions weren’t all that difficult to solve. Instead, it was the sheer amount of work required combined with the time constraints, that made the test difficult. This is why the test was barely finished by anyone. We simply did not have the time.

In this situation, we couldn’t answer the questions to the best of our abilities because of the time limit. I’m not saying that every student should have an unlimited time to ponder over the questions, but I definitely think that students should have an amount of time that allows one to think through each question.

Put differently: a test does not have to “fill up” the fifty minutes of class with questions.

This was brilliantly done by my astrophysics teacher during our final exam. We had three hours to complete the exam, but in reality it only took myself about an hour and a half. Assuming I went a little bit quicker than the rest of my classmates, it could have taken them two hours to finish. Despite the longer time, they would still have an extra hour to go back over the exam.

In my mind, this is a fantastic idea. Just because there’s a lot of time in some exams, there doesn’t need to be question that fill that time gap. Instead, teachers need to think about what they want out of their students. Do they want well written solutions? Or do they want hastily scrawled answers because of the time constraints?

It’s a no-brainer, really, yet I still don’t see this happening in many classrooms. Instead, I see tests being jammed to the brim with questions, as if testing students with more content is going to be useful.

The pointless question

Finally, there is a category of question that I just despise: the definition or concept question. Basically, this sort of question entails one to respond by giving a definition or some other fact that could be easily looked up in a textbook.

For example, asking students to memorize the first twenty elements of the periodic table is just a waste of time. Sure, it’s important. But it’s also something that anyone working in a scientific field can look up at any moment in time. Furthermore, one will naturally become familiar with the periodic table as one works on questions concerning it, meaning there is no use to having students memorize the table. If they work with it a lot, they will remember it out of convenience. If they don’t remember, they can simply look it up.

An even better example is in one of my experimental physics quizzes. The question I had was: what does the term laser stand for? This was such a useless question. Who cares if it is light amplified by stimulated emission radiation? Knowing the name certainly wasn’t useful to anything I did concerning lasers (which has holography). The question was just a waste of time.

These kinds of questions aren’t always present on tests, but I’ve seen them enough that they deserve a mention. A question about a meaningless fact won’t give us much knowledge about science.


We need to be honest with ourselves about the kinds of questions we want to ask students. The truth is that many of the questions used today just don’t have a purpose for the student. The multiple choice format is a prime example of this, being used in a manner which is often not helpful to the student.

Additionally, exams tend to be constructed with the idea that one must fill the time up with questions, or else the test is too easy. However, if the goal of the test is to see if a student can clearly write a solution to a question, imposing a strong time limit is counterproductive to this goal. By all means, introduce a time limit. Just don’t make the stress of the exam be the time allotted to write the exam.

Finally, questions posed just to make students memorize useless facts aren’t helpful either. They just award grades for items that aren’t important, and therefore shouldn’t be kept.

By keeping these principles in mind, I think it’s fair to say that we can improve the questions science students have to answer. It might be tempting to include a lot of questions on a test, or to throw in questions about abbreviations and such, but they don’t actually help the student learn something about science. They are just little tidbits of information that must be recalled until the test, before they are forgotten.

Instead, let’s focus on giving students questions that challenge their problem solving skills, and make them think about what they are doing.

That would be a lot more helpful than a bunch of multiple choice questions about definitions.

Related Posts

The Grit to Push Through

Behind the Equations

Quantities in Context

Black Boxes

The Priority of Education

A Splash of Colour

Outside the Curriculum

Through the Minefield

Visuals in Mathematics

The Necessary Details