The main aim of research is to look at a certain phenomena; describe, evaluate and thus explain it. When research is conducted within psychology the researchers try to choose a method to suit their need, with each giving a different type of data.

A good knowledge of research methodology and the reasons behind it is needed so that any student or researcher may employ them appropriately within a study. A knowledge of how to correctly modify several of the methods so that they are applicable to a specific area is essential as well as being able to identify possible variables.


The survey is one of the most widely used of the psychological research methods. It involves questioning a large section of the population about whatever the research is about and can be done via phone, post or in person.

A survey should have standardised questions and could include both open ( where the participant is free in how they respond to the question, the outcome of which is qualitative data ) and closed ( where there is a set of answers which a participant must choose from, giving quantitative data ) questions. The questions should be simple or worded carefully so as not to cause confusion.

The Likert scale is used for many closed question surveys. This scale gives the participant the allowance to give their strength of feeling on a subject on a scale of 1-5 or 1-7 etc. with a neutral option always present. Another must for a survey is a representative sample. If a study was being carried out on how many of the population fear dogs, doing a survey at a dog show would not give a representative result.

Surveys are cheap and can provide a basis on which researchers may study further. They also provide a large amount of information with a relatively small amount of work and most of the information gained (with closed question surveys) is easily turned into statistics.

However, surveys are limited in the volume of qualitative data which they can produce and are slightly restrictive. There can also be problems with people trying to "please" a researcher i.e. giving them the answer they think the researcher wants.


Interviews are face to face questioning. They, like a survey, come in two forms. Structured interviews are pre-planned and consist of a set group of questions which the participant may respond to. Unstructured interviews are unplanned (or mostly so) and involve allowing the interview to go wherever the answers may lead. This method is thus very popular with clinical psychologists as it allows them to gain a large volume of qualitative data, and when standardised, allows diagnosis of many conditions.

Interviews almost always have open questions, but due to this fact the time involved is much greater than the time taken to do a survey. Participants in interviews are also subject to the same demand characteristics of those completing surveys - they wish to "please" the researcher by giving them the answers they want.


Experiments are based on manipulating an independent variable whilst recording the effect of that on a dependant variable. This method is scientific and is used to test a specific hypothesis. An experiment must take place in an area which can be controlled by the researchers as much as possible so that confounding variables are eliminated.

All experiments will have what is known as a "control" group, which is usually similar to the group which is being tested. Both groups are told that they are participating in the experiment, though, in many cases so that the placeboeffect can be shown not to be a variable.

Experiments are use to establish a cause and effect relationship, and are designed to be easily repeated from documentation. Experiments give quantitative data, but due to the laboratory setting lose a little validity ( as it is an artificial environment ). Experiments in psychology can also involve a little deception, as participants cannot always be told the purpose of the experiment if this will bias the result. This then brings in ethical issues which must be resolved before an experiment is used.

There are also, though, field and natural/quasi experiments, in which the independent variable can be manipulated without the knowledge of those who are "participating", or is already naturally available. Much of the time people are not aware that they are being involved in such experiments, and although confounding variables might be higher, the results show a much more naturalistic representation. There is no bias in these due to demand characteristics but they are also hard to apply across the population.

Naturalistic Observation

Observation allows for people to be observed without knowing and without manipulation by a researcher. Observation can also be divided into two sections.

Participant observation

This occurs when a researcher joins a group and observes from within.

Non-participant observation

This is when the researcher does not take part in the situation, instead remaining detached. The participants may or may not be informed.

One of the greatest pluses of this method is that it allows naturalistic behavior to be looked into when observation is covert. It also allows researchers to look into things which would not be ethically possible to manipulate within an experiment.

Observation, however, offers very little control for a researcher as variables are increased. If the participants are aware of the researcher they may also modify their behaviour so that the results of such observation would be invalid.

Case Studies

These are the in depth study of a particular individual. A researcher using this method will look into the person’s past, all their records and makes use of frequent interviews to establish a full picture of the person and their behaviour (it can be used also on small groups). Case studies gain highly qualitative data and give far more detail about a single person (or group) than any of the other methods.

Case studies take a far longer time to conduct and bias might become involved if the researcher, due to the vast amount of time spent with and studying them, becomes close to the subject. There is also the problem that the data may only be applied to that individual and is unreplicable.


Correlation looks at the relationship between variables. As it does not establish cause and effect the variables are known as "co-variables" The result of a correlation can be none, positive or negative. Positive correlations occur when, as one variable goes up, the other one goes up. Negative correlation, on the other hand, occurs when one variable goes up and the other goes down. This method does not show a "cause" for something, it merely shows two variables and how they appear to relate to each other. Just because there is a positive or negative correlation does not mean the variables are directly linked.

Whichever method is used, the purpose of the research as well as the advantages and disadvantages of the research methods must be taken into consideration so as to provide a valid set of results.

Log in or register to write something here or to contact authors.