Psychology Research Methods

Psychologists use a variety of research methods. Each method has its own strengths, weaknesses and ethical issues that are considered when deciding which method to use.


Experiments fall into three main categories:

  1. Laboratory experiments
  • They are conducted in a laboratory.
  • The independent variable is manipulated by researchers.


  • There is tight control over variables.
  • Cause and effect (causal relationship) is easily identifiable.
  • They are usually replicable.
  • They are usually less time-consuming and less costly.
  • Scientific equipment is used.


  • The artificial environment means there is low realism.
  • Demand characteristics and experimenter bias may occur.
  • There is usually low ecological validity as it is an unrealistic setting.
  • A laboratory means limited space; you can’t fit a big sample into a laboratory at once.

2. Field experiments

  • They are conducted in a natural environment.
  • The independent variable is manipulated by researchers.


  • There is higher realism as the environment is natural.
  • Behaviour is more likely to be natural, therefore the ecological validity is also high.


  • There is less control over extraneous variables (e.g. weather conditions).
  • It is difficult to replicate them completely.
  • They may be time-consuming and costly.

3. Quasi experiments

  • They are conducted in a natural environment.
  • The independent variable is not manipulated by researchers; it occurs naturally.


  • There is a very low chance of demand characteristics and experimenter bias occurring.
  • The natural environment and lack of manipulation means that everything is natural and realistic, therefore there is high ecological validity.
  • They are useful in situations where it would be unethical for the experiment to manipulate a variable.


  • The experiment has no control over any variables.
  • It is very difficult to replicate so reliability is difficult to check.
  • They can be very time-consuming.

Case Studies

These are investigations about a single factor, for example: a person, an illness, etc.


  • They are very detailed and produce a lot of information.
  • They apply a range of methods, such as: interviews, observations and questionnaires.
  • The researcher may become close to a person who is at the centre of the case study, thus allowing them access to more privileged and confidential information.


  • They are very time-consuming and may be costly.
  • There may be a lot of subjectivity, especially if the data is mostly qualitative.
  • Results are rarely generalisable because they focus on one person or thing.
  • The method is extremely difficult to replicate.
  • While closeness between the researcher and sample has its own advantages, it may lead to bias.


Observation is when the sample is observed. Generally, observations alone are done when natural behaviour is being examined; this is called “naturalistic observation”.

“Controlled observations” are done under controlled conditions, usually in a laboratory.

Observations may be covert or overt. Covert observations are done secretly, such as behind a two-way mirror without the sample knowing. Overt observation is done out in the open.

“Participant observation” is when the observer is part of a group that is being studied. Non-participant observation is the opposite; to observe without being part of the sample under observation.


  • People are more likely to display natural, realistic behaviour when they are unaware they are under observation.
  • They allow us to observe animals that cannot be kept captive.
  • They allow us to observe natural situations without an artificial setting.


  • The observer, if detected, may distort the observations by influencing the sample.
  • Observation without consent may raise ethical issues.
  • A single observer is not reliable enough; more observers would be needed.
  • This can get time-consuming.
  • It is very difficult to replicate.


There are different self-reporting research methods: surveys, questionnaires and interviews.


  • These are lists of written questions that are completed by participants.
  • It may have open-ended questions (elaboration is allowed) or close-ended questions (fixed choice).


  • It allows a large sample to be tested at the same time so it saves time.
  • It is not as costly as other methods.
  • It can generate qualitative and quantitative data.
  • They are very easy to replicate and convenient to use.
  • Subjectivity can be avoided if close-ended questions are used.
  • Participants can take part anonymously and be ensured confidentiality.


  • The response rate of postal questionnaires is generally quite low.
  • There might be misunderstandings of participants when they read the questions, especially with unfamiliar terminology.
  • With postal questionnaires, there is no guarantee that right person has filled it in.
  • If the experimenter is present or if there is no anonymity, demand characteristics may occur.


  • Interviews are usually face-to-face (or on the phone) conversations.
  • They may be ‘structured’ with pre-determined questions, ‘unstructured’ with no pre-determined questions, or ‘semi-structured’ with some pre-determined questions.


  • Face-to-face interviews allow the experimenter to also observe body language.
  • They can generate quantitative or qualitative data.
  • Semi-structured and unstructured interviews allow the interviewee to give extra, more-detailed information.
  • Structured interviews are easily replicable.
  • Misunderstandings can be cleared out on-the-spot.
  • The intended interviewee will definitely be the one answering questions.


  • Demand characteristics are very likely to occur, especially during sensitive questions.
  • They are very time-consuming and can be costly.
  • Interviewer effects may take place (attractive interviewers, awkward interviewers, impolite interviewers; they all have their own effects on the interviewee).
  • Qualitative data from unstructured interviews may be difficult and time-consuming to analyse.