Observer bias

Ethnographers in general, and observation-based researchers in particular, are frequently criticized for the subjectivity that informs their work. Even the most apparently unobtrusive observation can have unintended ‘observer effects’ – the tendency of people to change their behavior because they know they are being observed. Most contemporary researchers would agree that it is inadvis­able to seek to avoid all remnants of observer effects, since the only way to do so would be to return to the covert tactics of the ‘complete observer’ role, which has been widely criticized as potentially unethical. Nevertheless, there are some ways to minimize the bias that almost always enters into observational research:

• It might be said that the very naturalness of observation provides some inoc­ulation against bias, since the observer (unlike the interviewer, for example) is usually not demanding that people do anything out of the ordinary. It is hoped that in time his or her presence will no longer even be a matter of note and that people will simply go about their business.
• Observational research is emergent, which in this context means that it has great potential for creativity. Observational researchers can, if they so choose, eschew predetermined categories; at any point in the process out­lined above, the researcher can shift the question(s) he or she is pursuing. Observation has the potential to yield new insights as ‘reality’ comes into clearer focus as the result of experience in the field setting.
• Observational research combines well with other techniques for the collec­tion of information. Laboratory or clinical experiments, for example, lack the natural setting and context of occurrence; they generate ‘data’ that are self- contained and from which all ‘extraneous’ variables have been rigorously excluded. But field-based ethnography is rarely constructed around a self- contained observational ‘experiment’. Rather, observations are made of life as it is lived in the natural setting, and observational findings are constantly being cross-checked with information coming from interviews, archival searches, and so forth. This process of triangulation, which as we have seen is intrinsic to ethnography in general, is a good hedge against the biases that may result from ‘pure’ observation (see also Flick, 2007b).

Source: Angrosino Michael (2008), Doing Ethnographic and Observational Research, SAGE Publications Ltd; 1st edition.