Psychology uses scientific methods to support hypothesis, using these methods correctly ensures that the results are reliable, valid and objective. If a hypothesis is tested, and the result is supported, then it’s a valid theory as opposed to an opinion.
An experiment is when an independent variable is manipulated to measure the effects of the dependent variable. In an experiment the situation is controlled, and there are as little confounding variables as possible. This way we can observe the cause and effect relationship of both variables. If the experiment is repeated enough the result is considered reliable, as long as the results are similar. Without methods of testing theories then there is no way of knowing what is a tested theory, and what is opinion.
The main advantage of experiments is that it’s very likely that it’s the independent variable that is manipulating the dependent variable, not a confounding variable. Anything that manipulates the dependent variable, that isn’t the independent variable is considered a confounding variable. In an experiment, the cause and effect relationship is clear as there are as little variables manipulating the dependent variable as possible.
Some may say experiments are too controlled. The results from that experiment may only apply to that one situation; therefore it would be a mistake to apply the results to all situations. The way one group of people in a controlled environment behave is different to the way another group behave in another environment. The participant in an experiment may alter their behaviour, they might change their behaviour or answers to the ones the researcher might find preferable just because they’re in an experiment. This problem could be resolved if the person doing the experiment doesn’t let on which answer is preferable. A double blind study would also be a way around this problem, this would mean that the researcher would have no contact with the participants, and would have someone else who has no idea of the expected result of the experiment conduct it, so the data collected would not be biased.
There are also field experiments, this would be the ideal solution if being in an experiment modifies behaviour, or to have the results reflect everyday behaviour. A field experiment would be conducted in a real world surrounding, so the participants would feel more at ease and their behaviour in real life situation would be shown. The disadvantage to field experiments is that there are more confounding variables and less control of the situation, so we might not be sure if it’s really the independent variable effecting the dependent variable, or if there’s a confounding variable.
It could be argued that simplifying behaviour into two variables might be over simplifying a complex thing, and the results might not be valid. For example if GCSE results were used as a measure of intelligence this would be over simplifying, as effort in school and home life might have an effect on the grade, not just intelligence. If this isn’t a valid measure, then it would effect the validity of the result. However we can overcome this by cross referencing the GCSE results to another measure of intelligence, such as an IQ test.
Experiments are replicable. If the researcher shows the methods clearly when publishing the research, other psychologists are able to do the experiment for themselves. This is an advantage as it’s a way of knowing if the results gained were because of chance. If other researcher get the same results then they are reliable, and scientific data is gained to support the theory.