Research Methods and Data Collection
The Range of Research Methods
Psychologists choose a research method based on the question they want to answer. Some questions ask whether one thing causes another — these need an experiment, where the researcher deliberately changes an independent variable (IV) and measures its effect on a dependent variable (DV). Other questions ask about people’s experiences, unusual cases, or the relationship between two variables that cannot be manipulated — these suit interviews, questionnaires, case studies, observations or correlations instead.
Research methods fall into two broad groups: experimental methods (laboratory, field and natural experiments), which can establish cause and effect, and non-experimental methods (interviews, questionnaires, case studies, observations and correlations), which cannot establish causation but suit questions experiments cannot answer.
Laboratory, Field and Natural Experiments
A laboratory experiment takes place in a controlled, artificial setting such as a research lab. The researcher manipulates the independent variable and can control extraneous variables tightly, making cause and effect easier to establish. A field experiment also involves the researcher manipulating the IV, but it takes place in a real-world setting such as a school or workplace rather than a lab.
Laboratory experiments have high control, which increases internal validity — researchers can be more confident that the IV, not another variable, caused any change in the DV. However, the artificial setting can produce demand characteristics, where participants guess the aim and change their behaviour, and lower ecological validity, meaning results may not reflect real-world behaviour.
Field experiments have higher ecological validity because behaviour is observed in a natural setting, and participants are often unaware they are being studied, reducing demand characteristics. The trade-off is less control over extraneous variables, making it harder to be certain the IV caused any change in the DV, and ethical concerns can arise if participants have not given informed consent.
A natural experiment is used when the IV changes naturally rather than being manipulated by the researcher — for example, studying the psychological effects of a natural disaster or a change in government policy. The researcher cannot control who experiences the change in IV, so participants cannot be randomly allocated to conditions.
Strength: natural experiments allow research into situations that would be unethical or impossible to manipulate deliberately, such as extreme deprivation. Weakness: because the researcher has no control over the IV or allocation to conditions, it is harder to establish cause and effect, and the situation may be rare, making replication difficult.
Interviews
An interview involves a researcher asking questions and recording the participant’s spoken responses. A structured interview uses the same predetermined questions asked in the same order for every participant, producing data that is easy to compare. An unstructured interview has no fixed questions and develops as an open conversation, producing detailed qualitative data. A semi-structured interview sits between the two, using some fixed questions alongside the flexibility to follow up interesting responses.
Strength: interviews can explore complex experiences in depth, and unstructured or semi-structured formats allow unexpected themes to emerge. Weakness: interviewer bias — the researcher’s tone, appearance or phrasing can influence how a participant answers — and unstructured interviews are time-consuming to analyse consistently.
Questionnaires
A questionnaire is a set of written questions completed by participants, usually without the researcher present. Closed-ended questions restrict participants to a fixed set of responses, such as ticking a box or rating a scale, and produce quantitative data that is easy to analyse. Open-ended questions allow participants to answer in their own words, producing richer qualitative data.
Strength: questionnaires can reach large numbers of participants quickly and cheaply, and there is no interviewer present to bias responses. Weakness: participants may not answer honestly, particularly if a question feels sensitive, and closed questions can force a response that does not truly reflect the participant’s view.
Correlation
A correlation measures the association between two variables without either being manipulated. It tells us whether, as one variable increases, the other tends to increase (a positive correlation), decrease (a negative correlation), or shows no consistent pattern. Correlations are displayed using scatter diagrams, where each participant contributes one dot based on their two scores.
A correlation shows association, not causation. Even a strong correlation does not prove that one variable causes the other — a third variable problem may exist, where an unmeasured factor causes both. For example, ice cream sales and drowning rates rise together in summer, but ice cream does not cause drowning — hot weather drives both.
Case Studies
A case study is an in-depth investigation of a single individual, group or event, often combining several methods such as interviews, observation and testing to build a detailed picture. Case studies are particularly valuable when the case is rare or extreme, since it may reveal something that could not be studied any other way.
Strength: case studies provide rich, detailed data and can generate hypotheses for further research; they allow study of situations that could never be created or manipulated experimentally. Weakness: findings from a single, possibly unique, case cannot be generalised to other people, and the researcher may become subjectively close to the participant over a long study.
Observation Studies
An observation study involves the researcher watching and recording behaviour without manipulating any variable. To keep recording objective, researchers use categories of behaviour — a predetermined list of specific, clearly defined actions to record, rather than vague labels. Consistency between observers is checked using interobserver reliability, where two or more observers record the same event independently and their records are compared.
Strength: observations, especially in a natural setting, have high ecological validity because real behaviour is recorded. Weakness: observer bias means what one researcher notices may differ from another, and if participants know they are being watched, they may change their behaviour.
Across all these methods, the choice depends on the question being asked. Experiments are the only method that can establish cause and effect, because only the IV is manipulated. Non-experimental methods — interviews, questionnaires, case studies, observations and correlations — cannot prove causation but let psychologists study questions that experiments cannot ethically or practically answer.