Human Subjects Research

What Does it Mean for my Government Thesis Project?

What is the IRB?  What is CUHS?  Are they the same thing? (Yes.)

The IRB stands for “Institutional Review Board,” a generic term for the board that reviews human subjects research.  At Harvard, our IRB is called the Committee on the Use of Human Subjects in Research, or CUHS. See its website.

All research that constitutes “human subjects research” must have advance (as in, prior to conducting research!) approval by CUHS.  This means that you have to submit a proposal to CUHS before you interview or survey anyone about their experience(s) or opinion(s) and before you even recruit anyone to interview or survey. If you are planning to do a survey or interviews for your thesis over the summer, you may very well need to get approval from CUHS first.

What is human subjects research exactly?

So, you have a research project, and you’re wondering if it might be “human subjects” research.  If it is, you need to get CUHS approval before you do the “human subjects” portion of the data collection. But not all interviews or other research involving human beings are actually “human subjects” research. 

The idea of human subjects research is that it is research about or on a human subject.  If the purpose of the interview (or survey, focus group, ethnography, etc.) is to study or understand what some person thinks, feels, or has experienced, or how he or she responds to a certain stimulus, then your human being is a human subject.  Such people are the subject of your research, in other words. 

It’s entirely possible, however, that you will be surveying or interviewing human beings about something other than themselves.  You could, for example, want to know about a certain government policy and how it works – so you could interview bureaucrats about the policy, and the people you talk to would not be considered “human subjects.”  Or you might want to measure levels of government provision of services, so you want to do a survey of people to ask if their neighborhood has streetlights and public parks.  If the information you are acquiring is strictly factual, and the objective of your research is not to study the people who are giving you data, then it is not human subjects research.  IF, however, you want to know how those people feel about having or not having streetlights in their neighborhood, or if they would want more public parks, then the humans become subjects of your research.  Does that make sense?

Why do I have to get approval to do “human subjects” research?

Imagine that someone in your House who you think is a fellow student secretly records your conversations with other students, eavesdrops on small talk at dinner, and coaxes data out of you and your friends while pretending to be just another house resident. It turns out that she is an anthropology graduate student from the University of Chicago writing a PhD dissertation on the evolving mores of elite college students. She argues that she could not have done the research without deception.  That may be true, but deception means preventing people from consenting to being studied; even if sometimes justified, deception requires particular ethical justification. Most undergrads do not deceive anyone by doing thesis research, but this example may get you to think about why oversight of human subjects research might be a good thing.

Indeed, human beings have done some terrible things to each other in the name of “research” in the past, even with good intentions, and we all share the responsibility to make sure things like that never happen again. 

You need to be very sensitive to the fact that your interest (getting material for your thesis) is different from that of your subjects. Even if you are sympathetic to the population you are studying, and they have good will towards you, they don’t share this interest in data collection. And if you don’t take precautions, they may be hurt by allowing you to interview them or record their preferences.

The goal of the human subjects process is to make sure that you protect your subjects by minimizing any risk to them (embarrassment, discomfort, danger from exposure of their identities) and that you obtain their informed consent.

Harvard is required by law to comply with human subjects regulations; we could lose our Government funding if we don’t do so.

Note that the standards for approval of research on human subjects are becoming tighter, including at Harvard. There is a new administration in the Committee on the Use of Human Subjects which is paying more careful attention to applications and procedures. The Department is working with a number of other DUSs and the office of Undergraduate Education to get clarification on exactly what undergraduates can expect.

What if I’m not sure if my project needs approval?

Happens all the time.  Some projects may not require full Committee review, and others may be exempt from review altogether. Call or email the CUHS with any questions about the need to receive Committee approval or about the application process in general.   Phone: (617) 496- 2847 (CUHS)

Okay, so you are pretty sure your project constitutes “human subjects” research – what now?

Now you have to submit an application to the Committee on the Use of Human Subjects to get approval for your project or a letter from them certifying that your project, as currently conceived, is exempt because of the kind of people you are studying. Typically projects studying public figures (legislators, for example) are exempt, under the assumption that public figures will be able to assess the risks involved with research and will be able to protect themselves.

In order to get an exemption letter, you still have to submit an application to the CUHS—i.e., tell them about your project—so that a determination can be made about whether or not your research is exempt from further review. You must receive an exemption letter or approval for your research BEFORE you conduct the research. You may not your research before you receive approval, even if you have already applied.

Note that the CUHS staff member who initially reviews your application may come back to you with questions about how you have designed the research, about your interview or survey questions, about your consent procedures, or about how you plan to store your data. The point of such questions is to make sure that your project does not inadvertently cause harm to the interests of your human subjects. You must leave enough time for this process of back and forth.

Getting Certified – CITI training

In order to get your project approved, you need to take an online “ethics training” (series of online modules).  This may sound like a pain, but it’s not that bad, and you learn a lot!  It’s very important to know what kinds of harms you want to avoid causing your subjects.

The online training course is called CITI (Collaborative Institutional Training Initiative): What you do is go to this page, create a username and password, and select Harvard University (Cambridge/Allston campus) as your institution.  Then you follow the directions – there are 11 modules, and each takes 15 or so minutes – you don’t have to do them all at once (in fact, it’s good to do a few at a time, save, and exit, then come back another time, to keep your mind fresh.)

Once you complete the training and pass each of the modules, you will receive a certification of completion that you can submit with your application.

You don’t have to have completed the training to submit an application, but your application won’t be approved until CUHS has a confirmation of your CITI training on file.

Note that your adviser will also need to complete CITI training before your application can be approved.

Submitting Your CUHS Application

Okay!  So you’ve determined that you need official approval or exemption and you’re dutifully working through the online ethics training. You’re ready to write up and submit a proposal!  Go to the CUHS website, in the “Forms” section.

Download the basic form called “Initial Application Form MS Word” – also get the “Quick Tips For Completing Application Form,” while you’re there.

Read through both and follow the directions.  Working with your thesis adviser, fill in the form (including your information and information about your project).  If you are doing an interview or survey, CUHS has thoughtfully provided some sample forms that you can use as consent forms.

  • If you don’t have a thesis adviser yet, contact Cheryl Welch, DUS or George Soroka, ADUS to help you answer all the questions in the form and accurately describe your plans for data collection. 
  • If your adviser is not a voting member of the faculty (i.e., an assistant, associate, or full professor or a senior lecturer at Harvard) then Cheryl Welch will need to be the official faculty sponsor for your application, in addition to the lecturer, instructor, or graduate student who is actually advising your thesis. CW has already completed CITI training, but will need to review your application, as she will be adding her support to it.  She won’t be able to review each application substantively, but will go over it to make sure that you have completed the major parts appropriately. Once she is comfortable, and once she has received an email from your adviser confirming that he or she has reviewed the application and approves, she will send her approval to the CUHS.
  • Be sure to LEAVE PLENTY OF TIME for this process, as you’ll probably need to do a few drafts and also give your adviser time to read your application.

To reiterate the process:

  • Fill out the application and submit it to your adviser. (If your adviser does not have a faculty-level appointment, list Cheryl Welch as the “Faculty Sponsor” and submit the application to your adviser and to CW.) Incorporate any feedback you get from your adviser and/or CW and redraft the application.
  • When the application is ready to be submitted, send it to the CUHS and cc your adviser and CW.  Your adviser can then email his or her approval to CUHS. If your adviser is a lecturer or graduate student, he/she should email the approval to CW, who will email CUHS.
  • The whole process—submission and approvals—should be done electronically.

Once I submit my proposal, then what happens?

For most undergraduate projects, that’s mostly it – if someone from CUHS writes back and tells you to modify your project, you will have to resubmit the application with changes.  You should allow at least four weeks to complete the application process.

If you want to work with vulnerable populations who may not be able to give full consent, then you probably need to allow more time. Examples of vulnerable populations are children, prisoners, people with HIV, people who are doing something illegal, people who are mentally ill, people doing something at particular political risk, etc.  In cases of human subjects such as these, your application will probably need what’s called “full committee review,” where a larger committee of people need to read the application – so that will take longer.  The full Committee does not meet very often, so leave a lot of time for this whole process.  Here is the Committee’s meeting schedule.

What if I change my project after I have received approval?

Once you are in the field, if you make substantial changes to your project (for example, asking a different set of questions, or conducting research with a different population, or changing consent procedures), you need to apply for an amendment to your approved research. This can also be done electronically, but you cannot continue with the research until you have received approval.


CUHS website (with links to forms and peoples’ names)                                    

Useful FAQ page

Helpful guide to the process

Information on process from Harvard Psych Dept

Summary of process from the Harvard Student Handbook