Observation and self-report: similarities and Differences

Both self-report (as in questionnaires and interviews) and observation techniques can differ in their degree of struc­ture. The first of these may involve short, prescriptive question-and-answer formats that are easily analysed in quantitative terms. Using structured observation schedules which involve the recording and counting of pre-determined categories of behaviour is an almost exact analogy.

On the other hand, interviewing can be largely unstructured, apart from identifying the broad area of interest, following directions determined mainly by those being interviewed. Observation can be similarly open-ended, recording what turns up at a level of fine detail, with analysis deferred (as well as final purposes, research questions, theoretical interpretation etc.).

Whether surveys or various methods of observation are employed, the differing levels of structure lead to different kinds of data being collected. For the researcher the ques­tion becomes: what kind of data do I need? Table 1.1 pro­vides a simple contrast.

This polarized distinction is false in two respects:

• the extremes can be used in combination to complement each other
• the extremes represent points on a graded scale because there are techniques with varying degrees of structure (as is the case with interviews).

A structured observation schedule has to be preceded by an unstructured phase – otherwise how would the researcher know what to focus on? However, this kind of unstructured observation is more likely to be of the detached, non­participant variety (see page 23 for more on this ‘fly-on-the- wall’ technique).

The need for a combination of structured and unstruc­tured methods commonly arises when the researcher derives an unsatisfactorily impressionistic judgement from partici­pant observation alone. Conversely, it may be that struc­tured, detached data raise more questions than they answer (as in surveys based on questionnaires) so that supplemen­tary participant observation is required to get at what the data actually mean (the qualitative dimension).

What do structured and unstructured methods have in common ?

The broad common characteristic is that both are systematic, because in taking this approach we come to see things, events and the connections between them that would either be overlooked or not recognized as significant. Detailed recording, in words or images, brings into focus what is there to be seen or understood.

It is a fallacy to suppose that because something (an object, human behaviour) is within our field of view that we ‘see it’. We have all had the experience of ‘not seeing’ something we were looking for even though it was there to be seen all the time (I commonly ‘lose’ something on my desk). But there is another and more profound point: we tend not to see things that we do not know about or understand, that we cannot relate to our existing knowl­edge. Systematic observation: checking, specifying and seeking explanation forces us out of our preconceptions, leading to a shift in understanding.

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