Both Leon Festinger et al (1956) studying a religious sect that had predicted the imminent end of the world, and Jason Ditton (1977) studying theft among bread deliverymen, had to use toilets as a place for recording their observations in secret
Covert participant observation raises serious ethical (moral) issues. These often conflict with the practical advantage it brings of observing natural behaviour
Researchers should obtain the informed consent of the subjects, and reveal the purpose of the study and the uses to which its findings will be put. With covert observation, this cannot normally be done, at least until the end of the research
Covert observers may have to participate in immoral or illegal activities as part of their 'cover' role, and may also be witnesses to such activities, for which they may have a moral or legal duty to intervene or to report them to the police
What people say they do when asked in a questionnaire and what they actually do in real life, are not always the same thing. Participant observation can obtain rich qualitative data that provides a picture of how people really live
Participant observation allows the researcher to gain empathy through personal experience, and to gain insight into the group's way of life, their meanings and viewpoints, their values and problems
Participant observation gives the researcher a more open-ended approach because it shows us what people do rather than simply what they say they do. It allows the researcher to discover new explanations and change direction to follow them up
Participant observation enables the sociologist to build a rapport with the group and gain its trust, which has proved a successful method of studying delinquent gangs, football hooligans, thieves, drug users, religious sects and other 'outsider groups'
Participant observation can also be used in situations where questioning would be ineffective, such as studying how police and probation officers categorise juveniles by making unconscious assumptions
Participant observation is time-consuming, requires the researcher to be trained to observe sociologically significant aspects, and can be personally stressful and demanding
Covert participant observation in particular raises serious ethical difficulties, including deceiving people in order to obtain information about them and participating in illegal or immoral activities
In participant observation studies, the group studied is usually very small and the sample is often selected haphazardly, which does not provide a sound basis for making generalisations
Participant observation usually produces qualitative data, which can make comparisons with other studies difficult and is unlikely to produce reliable data