I could use some advice from more experienced instructors.

My social science course culminates in a term paper based on independent research.

However, in some cases students end up changing the topic of their term paper because they hit a bottleneck.

In the past, I encouraged them to make the most of it, even report null results. But now I wonder if I can be more inclusive if I allow them to abandon a topic, but report on why and then elaborate how they will course-correct and even begin that new line of research, albeit in a preliminary way (because of time constraints).

What do you think of this approach? It is essentially allowing them to combine two papers into one. Anyone tried something similar? Other ideas?

  • Are they changing topics because they're only finding null results, or some other reason, like lack of any research at all? Commented Feb 17, 2022 at 18:19
  • Do they choose their topic originally or you? Do they change with permission or on their own?
    – Buffy
    Commented Feb 17, 2022 at 18:23
  • And, are you the designated instructor of the course or a TA?
    – Buffy
    Commented Feb 17, 2022 at 18:41
  • They sometimes change topics because they lose interest — this term paper may converge with their thesis if they plan wisely
    – Mij
    Commented Feb 17, 2022 at 18:46
  • 1
    When you say "more inclusive", what do you mean?
    – Karl
    Commented Feb 18, 2022 at 11:37

3 Answers 3


You're starting at the wrong end if you ask "Should X be allowed?". Rather, start with "What do I want me students to (i) learn and (ii) demonstrate?" For example, I assume that you want your students to learn research design, the evaluation of data, drawing conclusions from the data, and then putting these conclusions into the context of the literature. For all of this, it is not necessary to have positive results: One can write a cogent term paper about a research design and the evaluation of data so obtained even if the data only supports the null hypothesis.

In other words, if a student can demonstrate what they learned, then the actual form of their paper should not actually matter all that much.

To sum this up: Focus on the "what" questions, rather than the "how" questions. That's because if you know the "what", you often know which "hows" are useful and which are not.

  • 1
    Quite right, but writing a paper that says "we don't have enough data to support any of the alternate hypotheses" is not really possible, because you cannot do the intended evaluation. Can still be a good exercise in error estimation, or in designing a study that would give sufficient data, but thats again a change of topic.
    – Karl
    Commented Feb 18, 2022 at 11:34
  • If you don't have enough data, you get that the likely range of your parameter is 3+/-12. From a learning perspective, being able to derive this is just as useful as someone coming up with 4.22+/-0.03. It also teaches that whether a hypothesis is supported or not depends on your error bars and, consequently, the amount of data. That's why testing of COVID vaccines could be done so quickly: Because it was easy to get huge numbers of cases together in such a short time during a pandemic, as opposed to testing vaccines for rare diseases. Commented Feb 18, 2022 at 19:46

even report null results

Null results are incredibly important in science. There may be institutional and individual biases against null results, but students should not be encouraged to change course merely because their results are "uninteresting". To the contrary, they should be encouraged to understand the importance of null results and learn how to properly interpret these results, which is often much more challenging than interpreting results that clearly favor a particular hypothesis.

changing the topic of their term paper because they hit a bottleneck

Bottlenecks are the hard part of research; if the term paper is meant to give students exposure to what it means to do research, then they should learn to confront and work with these bottlenecks. Hitting a bottleneck should not, in itself, be sufficient reason to change topic. Bottlenecks are distinct from other obstruction in that there is in fact a path through, even if it is narrow or uncomfortable.

allow them to abandon a topic, but report on why and then elaborate how they will course-correct and even begin that new line of research, albeit in a preliminary way

This has been the standard approach in biology courses I've been involved with that have a research or pseudo-research component: when things go wrong, students explain how they've gone wrong, how the things that went wrong impacted their results, what they've learned, and, since usually there is limited time to take the next steps of redoing things incorporating what they've learned, at least outline a plan for what they would do next if they were able to spend more time.

You should choose the approach that best fits the learning goals of your course, however. There may be certain barriers that prevent a student from achieving those goals, and that may require a change in topic, hopefully as early as possible (and you should try to identify and prevent these issues if at all possible at the initial topic selection phase). For example, if your course has a quantitative component, obstacles to collecting any data would preclude a student from doing the quantitative component, so you may consider alternatives such as simulated data. It is certainly true that these types of projects are often very difficult for students, though, and they likely need a lot of guidance in navigating through their projects. Some balance should be reached between letting them flounder a bit and giving recommendations. Definitely avoid letting their evaluation be based on things they could not foresee.


This is from experience with a doctoral dissertation in mathematics, so it may not apply exactly, but sometimes it is necessary to change topics. But it should be done with permission so that you get a chance to advise them.

In my doctoral studies I worked on three problems. The first turned out to be too easy and I was able to develop one or more theorems with proofs every day. It was a bit fun, but also trivial. Hence boring.

The second problem was like a perfect diamond and there were no tools that I could find to "crack" it. I worked a few weeks totally unsuccessfully. Nada. Nothing but frustration.

The third problem (are you getting "Three Bears" here?) was just right. It was significant but approachable. The thesis turned out to be a complete theory, not just some minor results. Some of the parts were easy, some hard, but it was possible to attack both with existing and with new tools.

My advisor was involved through all of this, agreeing when it was time to move on and also making suggestions.

I realize that a term paper is much less consequential, but, perhaps the same constraints should apply. Seek permission to change focus. An give advice when permission is sought.

An orthogonal idea is that, depending on the field, a single university term/semester may not be sufficient to tackle some topics and so you may just be forced to make some accommodation in fairness. It is impossible to schedule results when true research in many fields is undertaken, since it is a look into the unknown. In such cases, an explanation of what was done and how and why it may have failed can be as valuable for learning as something more "tidy".

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .