Examine the arguments for and against the use of experiments in sociological research
Examine the arguments for and against the use of experiments in sociological research.
An experiment is an example of a research method used to collect primary data. Experiments are very rarely used in sociological research, and are far less common than questionnaires, interviews and participant observation.
Laboratory experiments are designed to test an hypothesis, with a view to accepting, modifying or rejecting it. They are far easier to conduct in the natural sciences than in the social sciences. This is because it is necessary to hold all other variables constant while focusing the experiment on those variables that have been put forward as significant in the hypothesis. In normal circumstances this is practically impossible to achieve with human beings in any sociological experiment. Moreover, people tend to act in terms of their definitions of situations - they are likely to define laboratories as artificial situations and act accordingly. As a result, their actions may be very different from their normal behaviour in the real world.
Sociologists sometimes try to get round this problem by conducting 'field experiments'. These are conducted in normal social situations such as the classroom, the factory, the pub or the street corner. For example, in order to test the effect of social class on interaction between strangers, an actor stood outside Paddington Station and asked people for directions. The request never altered, but the actor's dress varied from that of a businessman to that of a labourer. The experiment indicated that people were more helpful to the 'businessman'. So, arguably, they were responding to their perception of social class. But such experiments are bound to be inexact; it is not possible to identify and control all the variables that might affect the results. In the above example, people might have been responding not to social class but rather to the level of confidence presented by the actor, if as is likely he behaved more confidently when acting as the businessman. There is also the problem of unintended experimenter bias here. People act in terms of how they perceive others - age, gender, ethnicity, etc. So their responses will tend to differ accordingly - and they also tend to act in terms of how they think others (in this case, the actor) expect them to act.
People are also often aware that they are the subject of an experiment, which itself is likely to affect their behaviour. This is known as the 'Hawthorne effect', first observed in the 1920s during a study at the Hawthorne works in Chicago. The experiment was designed to discover whether there was a relationship between productivity and variables such as levels of lighting and heating. The results made little sense - for example, productivity seemed to increase whether the workplace temperature was turned up or down. So the only factor that seemed to explain the increase in productivity was the workers' awareness that they were part of an experiment.
Finally, experiments on human beings raise ethical questions. Clearly no sociologist would condone the medical experiments performed on inmates against their will in Nazi concentration camps. Yet few would object to the Paddington Station experiment. The British Psychological Society says that people should be told they are the subject of an experiment - either in advance, or immediately afterwards.
I hope this is helpful.