The Institutional Review Board doesn't seem to be much of an "ethical board" given they don't consider the stress this causes the human side of the website operators. How did they conclude this doesn't count as "human subjects research" given the expectation is to measure a human response to an email (framed as from a person)?
1. does not disclose that this is a research survey;
2. does not inform the recipient that they can decline to participate without any adverse consequence (as they state in their study FAQ), instead doing entirely the opposite - ending the mail with an assertion (untrue for most recipients) that they have a legal duty to respond;
3. lies about the origin of the email, inventing a fake persona;
4. lies about what the responses will be used for.
You are the first person in this thread to concisely state the major ethical failings. These 4 things are universally required in human subjects research (or IRB-exempt HSR) and it's black-and-white unethical to omit them.
Yeah this is the big one for me. Coming from the context of medical research it is shocking to me how many commenters think this is okay. If this happened to be a disguised medical research survey the ethics board would never have let this past. And if they did it would bring down some serious bureaucracy on their heads.
I guess software and law just think they can test whatever they want and get away with it. Who cares if a few people end up emotionally scarred?
5. which caused people harm, in the form of time, stress, and in apparently a fair number of cases, money spent consulting a lawyer in the belief that they were about to be sued.
> The Institutional Review Board doesn't seem to be much of an "ethical board"
YES! I know you intended this as an insult, but it is actually one of the most insightful comments in the thread. An IRB is not supposed to be an ethical board. They have a much more limited scope of overseeing human subject research.
> How did they conclude this doesn't count as "human subjects research" given the expectation is to measure a human response to an email (framed as from a person)?
Because there's a precise definition of "human subject research" which is actually a lot more limited than many people think.
Then it sounds like the blame is solely on the researchers for thinking IRB approval == greenlight to do whatever the hell we want. Or we have a mismatch in the social contract of research.
> Because there's a precise definition of "human subject research" which is actually a lot more limited than many people think.
We've had two different cases now where that disconnect has led to studies which cause stress and other harm to human subjects, because the study was classified as "not human subject research" despite directly involving humans at every step of the way. Perhaps that definition needs to be rewritten.
