There are two different sources researchers get their information from. One is from experiments, where variables are added and taken away as the researchers wish, and the other is from case studies, which is the analysis of an instance that arose by itself.
There are as with everything pros and cons to both these sources.
Take for example Stanford’s prison experiment. Twenty four men were randomly selected, and then randomly split into the two groups of guards and prisoners. The experiment was abandoned 6 days in as the ‘officers’ began to abuse their authority and the participants started to drop out. This was an experiment. Therefore variables were controlled, participants were randomly selected and conditions were monitored. However even though this small group of the population acted this way, does it mean all the population would? Can it be generalised and it be said that all of us, if put in a position of authority would abuse that role? What about the real prison guards. In a real life situation authoritative roles have responsibility and apart from the odd few that authority is not abused. Maybe there were other variables that affected how those men behaved. Maybe one had been abused in early life and took the experiment as a way to release anger by inflicting abuse on someone else. Maybe eleven of the guards had personalities which made them easily influenced and the twelfth guard had an aggressive streak. Maybe if the men were left in those roles for longer they would have maintained a level of decency however the experiment had to be terminated due to ethical reasons.
On the other hand, take the case of Genie. Genie was a girl who spent the first thirteen years of her life locked in a bedroom and strapped to a potty chair where her only her father was allowed to talk to communicate with her and he would only growl and bark at her like a dog to keep her quiet. When genie was discovered she was mute had developed a characteristic “bunny walk” in which she held her hands up in front, like claws and constantly sniffed, spat, and clawed. She provided an excellent case for the study of linguistics and psychologists interested in child isolation and how we learn language at at later age. However Genie was just one child who was put in this situation. The variables were not controlled and Genie herself was not randomly selected. Maybe she was born with a mental disorder anyway that would have made her behave in some of the ways that was out down to be a cause of her isolation. Maybe if another child were put in such horrific circumstances they would have reacted very differently. Of course unless another case study such as Genie presents itself these questions will never be answered. And even if such a case arose, it would be very unlikely that the child’s treatment, environment or even gender would be the same. Obviously such a circumstance can not be replicated in an experiment as that would be unethical.
So basically, experiments give us more control over variables and a bigger subject size, but its limited by ethics and the inability to control all variables for a long period of time. Case studies are valuable as they are genuine and not limited by ethics but are rare to come across and hard to compare to other such cases as they are all very different.
I think its experiments that wins this one. No one (in theory) is harmed by its occurrence and variables can be better controlled.
libbyayres 
I would argue that both methods have definite benefits to the field of psychology. Whilst, as you said, lab experiments etc do provide a great deal of control to allow a researcher to distinguish between variables, without case studies we would not be able to understand certain things about human behaviour because they are unethical. Also if a case study brings to light certain new information, this can potentially lead to the development of new areas experimental research connected to the case study.
In my opinion, experiments provide more useful insight than case studies. As you’ve said, variables can be controlled and manipulated to provide and analyse results, and can go on to assist in the exploration of future hypotheses and experiments. With the Zimbardo experiment, it certainly gave us insight into the theory of deindividuation, where an individual takes on the personality or adapts their behaviour and decisions/opinions to that of a group of people i.e. they disregard their own beliefs in favour of the behaviour or opinions shown by the entire group. Deindividuation has been shown to have a significant on human behaviour, even to the extent shown by the prison guards in the Zimbardo experiment.
The best example of case studies I can think of are those used by Sigmund Freud during the era of the psychodynamic approach. These are mostly disregarded today due to the lack of scientific evidence to support his theories. Whilst this might not be the case with case studies, they may provide some useful insight to individual behaviour. But it is certainly the use of experiments with an adequately sized sample of participants with results that can be generalised to the population that really help in the progression of psychological research.
Both are very important for psychology and provide a different look, at either a hypothesis or a certain situation. However experiments provide a more natural setting even in a laboratory environment and allow the subjects to be viewed by the experimenter. This in comparison with the case studies which is an occurrence and the psychologist dealing with the situation after the incident. Ethics don’t play much of a part during the incident because the psychologist is unable to effect the outcome thus showing the positive side of case studies. Freud favoured these case studies despite his poor choice of candidates.
In my opinion both methods have great importance to the field of psychology. Experiments provide a great deal of control (Lab studies) and allow the research to be conducted in a much more natural environment (Field studies). Whereas case studies provide us with much more detailed and rich research about individuals, towards human behaviour. They allow us to focus on a particular area and study it in detail. Problems with this approach are that ethics don’t particularly play a part as with most case studies psychologists are ealing with the outcome of a situation after an incident has occurred. We are unable to effect the outcome thus highlighting a advantage to case studies. They are much more natural and give us a greater idea about the research. Of course what tends to happen, an example being Freud, is that the results of said research are generalised to the rest of the population which simply cannot be done.