Assess the strengths and weaknesses of using experiments in investigating power and authority in prisons (15 marks)

There are two types of experiments which are used by sociologists to study various causes and effects of variables within settings and situations, these include laboratory experiments and field experiments. As favoured by positivists, the laboratory experiments are artificial environments where the researcher controls variables to discover their effect, with the aim to discover a causal law. However, sociologists sometimes use field experiments to overcome the lack of validity of laboratory experiments. Field experiments differ from lab experiments as they take place in the participants' natural surroundings, and the participants do not know they are in an experiment.

Researchers could use laboratory experiments to create a prison situation amongst participants and observe their behaviour in terms of power and authority. However, if prisons itself were to be studied, many practical issues may arise. For example, prisons are closed organisations which mean that access to them may not be gained, and even so, prison guards and other authority figures could be protective over their 'canteen' like culture and thus may deny requests for access. Despite this, previous research has shown that researchers have easier access to organisations such as prisons if they are encouraged by the government. For example, researchers of the 'short, sharp shock' approach gained access by easier means than they would have had without having the support from greater authorities.

In an exception, if access is permitted and a laboratory study is carried out within a prison, guards would be aware of the research taking place thus they may produce behaviour which is not natural or usual to them in order to protect their professionalism. The occurrence of Hawthorne effect would undermine the findings as they would not be valid due to the awareness of researcher presence. However, it can be argued that such studies would be valid because Milgram was able to create an artificial setting and saw significant fulfilment of roles amongst guards and prisoners even though the participants were aware they were being observed, which suggests that the behaviour of subjects may not be affected. Even if field experiment was used to disregard any influence of the Hawthorne effect, any change to the prisoners' routine would be noticed quickly thus they may be likely to change their behaviour if a variable is manipulated.

Furthermore, laboratory studies tend to consist of samples which are small because otherwise they would be time consuming and difficult in comparing experimental group with the control group. Also, individuals are unique and different, therefore it would be difficult to 'match' members of the control and experimental groups exactly. In fact, prisons are single sex thus any research carried out in one prison would not represent the findings of another prison of the opposite gender. This means that the findings may not be representative to the wider populations which suggests that the study cannot be generalised. However, it could be suggested that prisoners lack activities thus they would be willing to participate in any experiment just to entertain themselves, although the sample size may be improved, this could affect the findings as they would produce behaviour which they think is useful for the researcher's hypothesis.

Despite this, laboratory experiments consist of extensive controls of variables and the settings altogether, meaning that the researcher is able to manipulate all aspects of the study. Therefore suggesting that the procedure is standardised allowing the research to be replicated and tested for reliability. For example, Zimbardo's study procedures were clear and thus the research could be replicated, although the extent to how ethical it would be to do so can be argued. However, in comparison to field experiments, laboratory experiments are much more reliable because they can be easily replicated due to high controls, this suggests that laboratory experiments have advantages over other research methods.

Considering ethical issues, even if the experiment is carried out in a prison, it is essential to gain informed consent and have the participants to agree to take part, having first explained nature, purpose and risks of the experiment. Additionally, the research should not harm the participants without a compelling justification, and any avoidance of psychological or physical harm should be made. As shown by Zimbardo, those who are not in the possession of power and authority can experience harm, as guards had become extremely violent and abusive towards the prisoners.

Alternatively to experiments, researchers could use observations which are favoured by interpretivists as they produce qualitative data. Although similar problems may arise, such as difficulties in gaining access, danger and lack of safety for the researcher. Also, if the observation is overt, the participants may produced desirable behaviour thus evidently suggesting that the findings will be influenced by Hawthorne effect. However, as suggested by interpretivists observations have high validity because they observe behaviour which is naturally occurring and is not affected by any variable manipulation. In conclusion, although laboratory experiments may have weaknesses in terms of practical concerns, they are arguably they most appropriate for closes systems such as prisons, because in an open system where countless factors are at work, making it impossible for the researcher to identify, let alone control, all the relevant variables.