The Limitations of Observation Method

In Chapter 8 we considered the role of self-observation as a research method in its own right, particularly direct access to ‘internal’ events, though these can to some extent be obtained at second-hand, by interview as well as by diary techniques. By such means all aspects of behaviour can be reported; which brings us to the contrary point that large areas of human activity cannot be witnessed. These include the more mundane aspects of our domestic existence (except where there is a ‘live-in’ researcher, as Whyte was early in his research project). But such areas as domestic violence, intimate personal confidences and sexual beha­viour are almost inaccessible except through interview or questionnaire methods. And even here, it is only compara­tively recently that these topics have been systematically researched. The first comprehensive investigation of sexual behaviour in the UK was published as recently as 1994, a project vigorously opposed at the funding stage by the then Prime Minister, Margaret Thatcher, as an intrusion into personal privacy (Johnson et al., 1994).

1. The evolution of behaviour in its cultural context

But perhaps the greatest limitation of observation, which cannot be made up for by other concurrent methods (i.e. at the same point in time), is the evolution or history of the behaviour and context observed. How did it arise?

How we behave is part of our current social system – our relationships with other people, the roles we inhabit, the norms and expectations of the ‘local’ culture. So how we are is as much maintained in the present as caused by events in the past. But our personality, the way in which we char­acteristically behave, has a history within which many of these elements originated. Any complete understanding of the present has to take account of the past; more than that we are part of an evolving system. This dimension may not be apparent from brief observation; indeed, may only emerge over a period of time.

So there is a need to consider the origins of what one observes and the process of change over time, particularly how these changes come about. Even when someone appears to be ‘always the same’ – like Ian the Glasgow street beggar – one still has to ask (for adequate understanding): how did he get to that apparently unchanging state? And what would it take to disrupt it?

2. The need for complementary research methods

Observation cannot tell the whole story; and even when extended over time it can only incorporate a narrow section of the evolution of a group, a culture, or an individual. Of course, what you hear or are told will often have a bearing on, or elucidate, the past. Something like a biographical approach may be required: every culture, every institution, like every individual has a history. What was this community (school, prison, company – or whatever) like twenty, thirty, forty years ago? What is the continuity? What has led to improvement (or decay)? What maintenance factors have been constant?

In the ‘cultures’ of large, established organizations – hospitals, schools, courts of law, social work services, for example – there will be extensive documentation of the formal kind: policy documents, regulations, reports, minutes of meetings and so on, but these are only one perspective on the past. Their ‘official’ character, often a rather self­conscious one, will not display the informal realities which make up the origin of most of what one finds from current observation.

At best such documents are going to require interpreta­tion in the light of the experiences of the doctors and nurses, the teachers, the social workers who have to operate the services represented in this way. Their experiences are likely to be unrecorded in any systematic fashion and it is here that ‘life history’ interviewing is a possible com­plementary method (see Gillham, 2005, Chapter 7). Multi­ple individual interviews are probably the only way to build up the background picture to what is observed in the pre­sent. Formal documentation has, at best, to be interpreted and translated by those to whom it presumably relates; and it may, in fact, have had no active relationship at all. This is the dilemma for social historians – how to construct a picture of the social reality which runs alongside what has been documented.

‘Interviewing’ of one kind or another is often concurrent with observation even if the questions asked are limited and fragmentary. Taylor (1993) made a point of interviewing the women drug users she had been studying at the end of her investigation when she knew what she needed (additionally) to find out; and what would only be uncovered in that more systematic way. And the women involved were only likely to be willing to be interviewed once Taylor had established herself with the group and become trusted over time.

3. The contribution of observation to theory

Ethnographic studies do not usually adopt a theoretical perspective. This is partly because of the strong emphasis on description and the often explicit attempt not to impose an explanation (theory) on what is observed, but to interpret after the event.

This stance is similar to what Glaser and Strauss (1967) describe as grounded theory, where explanation emerges from the evidence uncovered. But here we have a conundrum: what determines the selection of observed elements ? Is there no prior intuitive ‘theory’ there? Scientists traditionally work from a base of existing theory on the principle that how else would one know what to look for?

In real-world social research the answer is that the moti­vation of the researcher has to be in terms of human values:. those aspects of human existence which most people would agree to be of importance. Indeed much of the appeal of social research is precisely because its applications (and content) do reflect human concerns about individuals and social issues.

Our concern to do something about society’s ills – the difficulties of the disabled, of single mothers, of the home­less, or the unemployed, of those who are ‘antisocial’ or illicit drug-dependent – presents us with the challenge of understanding. And this is where appreciative research may have a contribution to make to social action, by developing from that perspective explanations (theories) different from those implicit in the shorthand judgements of those who are more advantaged (but sometimes feel threatened). Taylor provides a good example of this in her study of female drug users: that their ‘criminal’ activities were an inevitable con­sequence of their need to fund their drug habit – and where regulated drug provision would have removed the need for shoplifting and prostitution. This is not a new argument, but her detailed account makes the case much better than any amount of rhetoric. It is a case based on a reasonable interpretation of her evidence.

Glaser and Strauss (1967) make a distinction between substantive theory – explanation of what you have found – and formal theory which has a more extensive and not just academic application. One of the values of observational research, concerned with social action or social attitudes (which are often ill-informed), is that it may make such a contribution beyond the level of its immediate purpose.

Source: Gillham Bill (2008), Observation Techniques: Structured to Unstructured, Continuum; Illustrated edition.