I'm not really sure about your question, but if you are just talking to people in the street, you can simply explain what you are doing and ask them if they are willing to answer a few questions. Also, much of this depends on the context. I do research in Japan where written informed consent often is not possible. Older people, in particular, have been the targets of scams that get them to sign things they shouldn't and as a result if I ask people to sign something, I've pretty much ended any chance of collecting data. Instead, I basically read to them my consent statement while recording and ask them if they are ok with that. Another issue is the content of the research you are doing. Is it exempt from review? Many qualitative studies are. This is actually a fairly difficult question to answer.

I agree with John that it really depends on your specific project, and its local and national context. I have some experience with struggling to get ethical approval for an ethnographic fieldwork study in a hospital in Norway. I ended up having to abandon the project because the medical ethics committee who were the delegated authority defined "informed consent" as "written informed consent of all/any patients I would come across in the field", meaning I would have become a very conspicuous and pretty annoying presence on any hospital ward. So, an important question, Sunil, but more detail is needed in order to give good advice.

I agree that the question is not very clearly stated. If you are stopping people and interviewing them in the street or public places, formal informed consent would certainly be required in our setting. Such consent dcould be written or an REC might approve witnessed verbal consent in special circumstances. However, if you are just observing people's behaviour in a setting where they have no reasonable expectation of privacy (e.g. pedestrian behaviour, car park behaviour/public transport passenger behaviour) informed consent would probably be waived by the REC/IRB. If people have a reasonable expectation of privacy (e.g. in a hotel room, public restroom etc), informed consent would be required.

Thank you for both of you for your feedback. If I further clarify my question, as an example, If I want to observe natural incidents such as how to teach, nursing practice or how people behave in different events in the society. In some cases we can get the informed consent in advance and then develop rapport and gradually can start the observation process. But some incidence happen in the society (How people behave in road traffic accident) can not get the consent in advance. Then as you mention ethical committees will question and reluctant to give the approval in such cases. Even researcher act as a 'participant observer' can the researcher reveal he/she is observing others? Then it will not be a natural setting. In that case what will be the ethical concerns?

As mentioned above, most RECs/IRBs (depending on applicable local ethics guidance), will decide depending on whether people in the setting in question have a reasonable expectation of privacy. There are also other creative and ethical strategies to get consent but not affect the behaviour being observed - e.g. by getting consent for the observation including a statement that the observation could occur at any unspecified time in a specified period. (E.g. oberving health care workers' handwashing behaviours in a clinical setting requires consent but the observation could occur at a non-specified time point in a 6-12 month period - such consent is unlikely to affect the observed behaviour). Permissions from the gatekeepers from any non-public settings is also usually required by the RECs

Brigitta raises a problem indirectly that I think is a widespread issue with IRBs in the US, at least: the process is driven by medical research practices and medical ethics. Unfortunately, those practices are quite different from research approaches in other fields. Ethnography is a very good example of this, as it involves a very different approach to gathering data from most medical research (there are medical anthropologists who do ethnography, of course) and, thus, requires a different approach to getting informed consent. As an ethnographer who works in Japan, if I ask everyone I talk to for written informed consent, then I'd never actually do any research. It would be a show-stopper right from the start. Ethnographers often make use of casual conversation as a data source, and it simply is not possible to ask for consent every time you say hello and chat about the weather or the latest baseball game.

That said, it is still quite important that people who interact with me recognize that I'm doing research and that I am always on the clock. This gives them the opportunity to avoid talking to me if they so desire. When I did long-term ethnographic research in the 1990s, I actually had the local village newsletter publish a short article explaining what I was doing and letting residents know that I might ask them questions. The newsletter went to every household, so it was unlikely that someone with whom I would talk would not know about what I was doing. This provided them with the opportunity to avoid talking to me--and some residents did refuse to talk. I've spent a lot of time thinking about these things, because I was at one time Chair of an IRB and I teach research methods and ethics is an important component of what I teach. I think it is very unfortunate when a project cannot be completed because an IRB is unwilling or unable to negotiate issues like how to obtain informed consent--or what informed consent means in a particular context.

I forgot to respond to the issue of observing what happens in something like a traffic accident. Is your work strictly observational? If you are not going to talk to anyone and you have no identifying information about the people you are observing, then that has an influence on how this would work out. You might want to take a look at this article: Luke Gelinas, Alan Wertheimer, and Franklin G. Miller, “When and Why Is Research without Consent Permissible?” Hastings Center Report 46 (2016): 1‐9.

In the abstract, the author states that the argument of the article is that, "if it stands to infringe no right of the participants and obtaining consent is impracticable, or if the gravity of the rights infringement is minor and outweighed by the expected social value of the research and obtaining consent is impracticable."

I think this is a very good way to think about this issue.

Thank you all for very valuable resources and clarifications. As a developing country with medical dominance, the most of the IRB committee members try to see these proposal in quantitative angle. Hence they always less flexible and do not want to see the problem qualitatively.

If I get your poser correctly, qualitative research into the review of patients' health records may be an example?

In this case, ethical requirements have it that IRBs (ethics committee) can stand in for the patients and grant approval when obtaining individual patient's consent is not practicable. E.g a retrospective review of patients' health records for data quality.