We're currently redeveloping our website and you may notice some inconsistencies with our designs.

Structured observation and data

main content

Descriptive, inferential and evaluative observational techniques

One of the first considerations when using observations is the issue of how the observations will be interpreted. Have a look at this webpage, and especially the paragraph about observational variables (descriptive, inferential, and evaluative).

Task 1

Read the article by Lynn Ogden (2000). Discuss the nature and purpose of observation as a method of data collection.

Structured Observation

Structured observations are focused, look selectively at the social phenomena, and can be used to test hypothesis.

When conducting a structured observation, the focus of the observations has been determined beforehand. This type of observation follows the principles and assumptions of quantitative research: the focus of the observation is fragmented into predetermined, smaller, more manageable pieces of information (behaviours, events etc.) that can be aggregated into variables.

Of course, there are different levels of “structure” that structured observations take. For example, in a highly structured observation the researcher has decided in a rather precise and mutually exclusive way the observation categories in advance. In a semi- structured observation, the researcher starts with an agenda of what will be observed and how, but collecting the data with observations is done in a less systematic or predetermined way.

Collection of data by observations can be conducted on facts (e.g., the number of students in a classroom), events (e.g., the amount of collaborative work taking place between students in the classroom) or behaviours (e.g., the number of incidents of antisocial behaviour in a classroom). Observations can look at verbal or non-verbal behaviour.

Thinking about these categories, we will discuss some of the decisions the researcher needs to take while designing an observational study, and highlight some points for consideration.

Natural versus artificial settings

A distinction can be drawn between observations conducted in natural versus artificial settings. Much observation is conducted in real world contexts or, in other words, natural settings. However, observation is a legitimate method of data collection within the experimental research tradition. In experimental research, the relevant conditions (independent variables) are manipulated or contrived in systematic ways and the effect of these conditions on specified behaviours (dependent variables) is measured. In some sense at least, since a change is being introduced into the contextual conditions, the setting for this kind of intervention research can be described as artificial. The extent to which artificiality is introduced into the setting represents a challenge to the authenticity or validity of the research. Observation like all research methods can be seen as always involving some kind of balance or compromise between the interests of validity, reliability and feasibility.