Observation Method in Psychology

When psychologists observe, they watch peoples' behaviour and measure particular aspects in a way which is as precise as possible.

It is usual to have more than one observer because behaviour is complex and the observer may be biased. If the behaviour is videoed, the observers will analyse the behaviour from the video. They need to be trained in how to analyse and measure the behaviour being studied so that they all interpret it in the same way This is called inter-observer reliability.

Inter-observer reliability means that the measurement of behaviour is consistent between observers. To establish inter-observer reliability, both partners should observe the same behaviour independently and compare results afterwards. If the results are very similar, then inter-observer reliability has been established.

Behaviour is noted on an observation schedule (see Fig. 1). The researchers must decide what behaviours are to be noted. This is quite complex, for instance, measures of aggression might involve pushing, kicking, various facial expressions and shouting. Decisions must be made about how the participants will be observed and over what time period. Observers may watch a child for a full playgroup session or they may observe a child for five minutes every 15 minutes. This is called time sampling. If they are interested in behaviour in a particular setting, they will observe children in that setting, such as any child who comes to play in the sand box.

The behaviour may be analysed by being broken down into units, for example the behaviours shown in Figure 1. Here the observer would note each time a child performed one of these behaviours. The observer may have categories of behaviour on the schedule, such as playing alone, playing with one other, playing in a group, inter- acting with an adult. The task here would be to categorise the behaviour and note how much time is spent performing that particular behaviour.

Researchers may run a pilot study because they can then watch the kind of behaviour they will be analysing, and thus devise the most useful measures. Figure 1 shows a simple schedule for observation of-co-operation of children’s play. A number identifies each participant in order to preserve confidentiality.

  Fig. 1 Example of event sampling of co-operation in children’s play.

The observational method can be used for a variety of purposes, as described below

Naturalistic Observation

Here spontaneous behaviour is recorded in a natural setting. For example, Festinger ‘s observation of a cult who believed the earth would be destroyed by a flood (re: cognitive dissonance), and here the observer does not interfere with or manipulate the behaviour being observed. Before starting the study, the observers try to become familiar to those they are observing in order to minimise the effect that their presence may have on them. The observations would generally be non-participant.

  ·  Pro: It is valuable to see how people behave in a natural setting and an observation provides very detailed qualitative information which can be used as a starting point for further, more controlled research.

  ·  Pro: It can be used when other methods might be unethical.

  o Con: The presence of observers could influence the behaviour of those being observed, and it is difficult to control extraneous variables.

  o Con:  It is difficult for another psychologist to replicate the observation.

o Con: Because there are so many variables which could affect behaviour it is not possible to draw conclusions about determine cause and effect (i.e. there is no manipulation of the IV).

Controlled / Experimental Observation

This type of observation is frequently used with children or in social psychology research, and may be one of several methods used in the same investigation. In order that control can be exerted and to make behaviour easier to observe (such as through two-way mirrors), the observations may take place in an artificial setting such as a research laboratory.

For example, Bandura’s study of observational learning and aggression. Therefore, spontaneous behaviour is still observed, but in a controlled situation. Data obtained is quantitative.

·        Pro: By controlling some variables, it is possible to draw conclusions (extraneous variables can also be controlled).

  o       Con: Low ecological validity: The artificial (and unfamiliar) setting may affect participants’ behaviour and make it less natural.

  o       Con: There are also ethical issues relating to informed consent, deception and privacy.