For a study, we sent out a questionnaire to members of the general public. The questionnaire was containing items related to well-being and happiness, among other items that are completely related to my field (computer science).

One respondent wrote about having suicidal thoughts because of personal issues. No signs of seeking treatment in the reply.

The participant left their e-mail address in the field "contact me back for the results and/or further clarifications".

What is my role here as a researcher? My first thought was to contact the participant and pointing them to seeking help (maybe also providing links to their national suicide prevention programs).

  • 5
    Given that the participant specifically provided an email address I would assume that mailing back to him should not be a big ethical quarrel. Of course this does not mean that (s)he will necessarily want to hear what you are saying or listen to it. There is also the question whether you would have expected "signs of seeking treatment" in your survey - this was a computer science survey after all, I see no reason to put this info into the survey even if the participant would be undergoing treatment since months. – xLeitix Aug 2 '16 at 9:58
  • 3
    Also, it seems to me that there are psychologists that have to deal with this situation regularly. What about trying to contact those? – xLeitix Aug 2 '16 at 10:00
  • 5
    @101010111100 "Let's not get involved with this person that seems to be suffering - maybe she has a valid reason for wanting to die." I hope you are trolling. – xLeitix Aug 2 '16 at 10:30
  • 7
    @xLeitix Not really, no. It's about your perspective and worldview. There is a myriad of reasons why a person may decide they don't want to live anymore, and many of those reasons cannot be (practically) remedied. Selfishly forcing people to live is truly disgusting. – 101010111100 Aug 2 '16 at 10:38
  • 5
    @101010111100 What do you believe the ratio is between those who are suicidal and it would be selfish to help them versus those who are suicidal and it would be selfish not to help them? Most people don't want to die when they're functioning properly. – Jessica B Aug 2 '16 at 14:27

I know I've read something in a research textbook about unexpected ethics problems, but I don't remember where or what it said.

I think you need to contact your ethics committee (or equivalent), preferably quickly, for guidance. If there were enough questions about welfare on the questionnaire to be likely to bring up some sort of similar response, you should probably think about pre-empting it next time and automatically including links to suitable resources. It may be that now you can send out details to all participants (those who left email addresses, or perhaps the original distribution list, depending on how participants were recruited), as a way of offering help without singling them out. If the wording of the email address field requires it (I can't quite make sense of it), you could perhaps do so together with the results of the research, or a preliminary summary of the results...

| improve this answer | |

Your work should have been approved (including being classified as exempt) by an institutional review board (IRB). You say

no IRB/ERB at the institution I worked for when the study was executed. It is actually quite rare to find one in small and/or technical universities.

This is actually not a valid excuse. You should have gone to an outside IRB if your university does not have one.

I would classify this a response indicating suicidal tendencies as an adverse event. Adverse events include both "events" like a subject dyeing and abnormal findings on tests (e.g., a blood test or an eye test). Further, since it is life threatening, it is a serious adverse event. Finally, since you did not predict this outcome, it would be classified as an unexpected adverse event. The safest route in the face of an unexpected serious adverse event is to immediately halt the study and contact the IRB.

You are also required to provide subjects who experience an adverse event (both expected or unexpected) with appropriate care (i.e., if a subject has a heart attack, you need to call an ambulance). In this case, this means contacting someone in mental health.

| improve this answer | |
  • 1
    Thank you for the extensive reply. Is there such a thing as an external ERB? It is really not common in my field to have an internal one or to "use" one at all. At least in several universities in Europe (again, specifically to my field). – user7112 Aug 4 '16 at 20:34
  • 1
    @dgraziotin in the US IRB approval is mandated for almost all research involving human subjects that is funded by the government. Many universities and hospitals outsource the IRB process to for-profit IRB companies. – StrongBad Aug 4 '16 at 20:45
  • You mean dying, presumably. I don't think this would be covered by the duty of care, because the problem was not a result of the study, it was just noticed by it. This person would have known of their feelings without the study. To say harm was caused by the study, you'd need to say that answering the questions made things worse, but believe the current consensus is that that is not the case for suicidal thoughts. – Jessica B Aug 6 '16 at 6:26
  • 3
    @StrongBad: "Your work should have been approved", "in the US IRB approval is mandated for almost all research involving human subjects" - that may well be the case in the U.S., but regulations differ wildly around the world, the expectation of approval by any IRB-like entity may either not exist, not considered warranting any ecpenses ("for-profit IRB companies), or be deemed not applicable outside of medical or other ... – O. R. Mapper Aug 10 '16 at 7:38
  • 1
    ... specific branches of research, the protection granted by U.S. IRB oversight might overlap with protections granted by other laws in other jurisdictions, and as a result, IRB-like entities that CS researchers could direct individual inquiries to may just not be a thing in many places. – O. R. Mapper Aug 10 '16 at 7:38

There is clearly a sign that this person is deliberately reaching out to you... Why else would they say what they did and include an email? I agree in getting advice from your ethics people, but don't waste time... However, keep in mind you are committing yourself once you contact this individual so be prepared to offer resources for them, such as a way to contact a free mental health clinic in their area....

| improve this answer | |
  • 3
    “Why else would they say what they did and include an email?” — because there was a survey question asking about their happiness and well-being, and question at the end of the survey asking for (optionally) an email address? I’m not saying the person wasn’t deliberately reaching out — but I don’t think it’s obvious from the description whether they were. – PLL Aug 4 '16 at 17:19
  • 1
    @PLL Exactly! I don't think we can infer intention here, at least not without knowing how the question was worded. Many people occasionally have suicidal thoughts, but have no suicidal intentions. That is, they might wish they weren't alive for a few hours on a bad day, but never go beyond that to even think about making a plan. People like that may benefit from counseling, but an overly dramatic reaction could make things worse by making them feel labeled as "crazy". So I do think it best to ask the ethics people how to proceed. – mhwombat Aug 10 '16 at 13:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy