1. "I supervised an MA student a few years ago whose ethics proposal was sent back seven times. By the end they were asking questions like "what if you become really famous from this research and then the police decided to subpeona your records and the confidentiality of your sources was comprised?". Sadly, that's not even a made up question (and this was for an MA in English lit, btw). Obviously, the student was freaked and the process took seven months and almost scuttled her plans. We were drawing up a Plan B for her thesis to become about the review process, since she wouldn't have time to do the actual research. In the end she got what she wanted, did an excellent thesis and kissed academia good-bye rather than pursue a PhD since she thought the whole process of doing research is clearly deranged."
2. "IRBs are better some places than others, and it depends on the discipline. At my current institution, humanities scholars are subject to an IRB that only makes sense for scientists collecting blood and doing life-threatening experiments on small children. Yet, most of the major ethical concerns that are well known and of the greatest importance and concern in my discipline, those aspects of tangible risk resulting from research, are outside the purview of the IRB forms/processes. The policies change every year without notice, meaning it's very hard to teach our grad students how to successfully navigate the IRB process. It's become an arbitrary affair."
3. "ARGH YES. A network ("snowball") sample (so I don't know exactly who or how many yet!), semi-structured exploratory interview protocol that will evolve during the course of data collection--it was just too much for them. In principle, I'm OK with them watching out for ethical stuff, of course, but when they start questioning my research methods, they've gone too far. Back off. You're a chemist. The advice to be as vague as possible--yes. What I really want is exempt status when I'm talking to professionals about their professional roles--I've heard of it at other universities."
4. "My subjects are often powerful people doing bad things. It is quite reasonable to suppose that as a result of my research they could lose their jobs (it hasn't happened so far, but it could). And I wouldn't necessarily think that was a bad thing. From my own perspective, research ethics means making sure interviewees understand the implications of being interviewed, know the potential uses to which the data could be put, and that any deals I make with them about anonymization, or not writing about certain topics, or whatever, are kept. If they don't want to be interviewed, sometimes I use other ways to find out about what they are up to, and write about them anyways. Which often makes them prefer to be interviewed, because at least then they can put their view across.
"It seems to me that in an environment where the standard assumption is that no harm can come to research subjects as a result of the research, this kind of research is right out. Which pretty much means that I can't come back to the USA. Which is ok, since its not like there are any jobs there anyways."
5. "A new pernicious IRB trend is to need to get targeted organizational sign off on research about an organization, even if you interview individuals who work for it in their homes or in a coffee shop off campus. Clearly, this is to protect one's own campus from lawsuits, but I'm not sympathetic. (Key is the more important and latent function of IRBs-- protecting one's organization, not the subjects.)"
The discussion also features some defenses of IRBs, but they are vaguer and less eloquent. In particular, none tells a story of an IRB review that proved necessary.