Academia Stack Exchange is a question and answer site for academics and those enrolled in higher education. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm a repeat Teaching Assistant in a fairly large (100-200 students/semester) core, required CS course at a large, private US research university. I'm very interested in researching what factors influence success in the course, possibly with an eye toward SIGCSE.

I'm concerned about how to ethically conduct research on students.

Yes, I know a little about IRB, and I assume at some point I'd need to get IRB approval. However I'm so unsure about this, I'm not even sure what to ask the IRB yet. I'm hoping y'all can provide me some guidance and advice as I begin to plan for next year.

I assume I have to ask for explicit, opt-in consent? Obviously, participation or non-participation couldn't affect their grade. However, if only a very biased subset bother to opt in, the research might not be worth conducting. Can I compel students to choose to opt in, or opt out?

Additionally, the course is very similar year to year, so the extent that it would be very useful to be able to include previous semesters. I imagine this is even trickier.

I have the full encouragement and cooperation of the faculty member, but working with human subjects, much less students, if totally outside their expertise.

This is similar to this question, but slightly different and, I think, more difficult.

share|improve this question
up vote 21 down vote accepted

I conduct biology education research at my large, state research university, so I have experience with this process.

An IRB is necessary if you plan to make the results public. Educational research is considered "exempt" research under IRB regulations, but this really just means you have fewer flaming hoops, not NO flaming hoops. You will likely be able to convince your IRB that students do not need to sign consent forms, since you will post a study information sheet and they have the opportunity to opt out to someone else without it affecting their grade. You may also be able to get student data from previous years with a retroactive IRB.

An additional hurdle is that you will probably need the approval of your FERPA specialist in addition to IRB approval, as FERPA is very concerned with the sharing of private student information. Our FERPA guy is in our registrar's office.

As you can see, I'm being a bit vague. Every institution interprets the human subjects regulations a bit differently, and sometimes schools are relaxed... and sometimes they are not.

So, it's doable, but not at all trivial. If you still want to move ahead, I recommend you call your institution's IRB office and say you are considering education research and would like to meet with an analyst. And the first question to ask the analyst is, "who else on this campus has ever done education research?" and get the real skinny from that research group.

Then, the basic steps will be:

  1. Design the experiment, do a lit review, write up the IRB proposal and study information sheet
  2. Send the proposal and SIS to the IRB analyst, and they will help you make changes
  3. Send the proposal and SIS to the FERPA specialist and they will send an approval letter to IRB (in a month?)
  4. Wait for the IRB subcommittee to approve the proposal (another month?)
  5. Post the study information sheet on the class website, and get another grad student or postdoc to be the "safe" researcher that students can contact to opt out. This researcher will make a short announcement on the first day of class.
  6. Teach using your fabulous new technique
  7. Ask the registrar for grades for previous years (if you've worked this out)
  8. Do analysis, write it up, present it. Fair warning - most of the time we find that the first try is the beta test, and the SECOND try generates better results.
share|improve this answer
This is amazingly helpful, thank you so much. I totally understand, re, beta testing. I've been doing a bit of informal poking around at the data on past semesters. Hopefully that means I only need 1 beta test ;) – chmullig May 10 '14 at 23:10
You're very welcome, and I hope your results are significant AND have awesome effect sizes. – Adrienne May 11 '14 at 4:07

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.