Doing Research: Experiments, Self-Report and Observations
Doing Research: Experiments, Interviews, Questionnaires and Observations
Once a researcher has written a hypothesis and identified their variables, they must decide how to actually collect data. OCR groups this stage as ‘doing research’ and covers three broad approaches: experiments, self-report methods such as interviews and questionnaires, and observations. Each approach has its own strengths and weaknesses, and each suits a different kind of research question.
The three types of experiment are laboratory, field and natural. A laboratory experiment gives the researcher the most control over variables but the least realistic behaviour. A field experiment trades some control for a real-world setting. A natural experiment studies an independent variable that occurs naturally, without any manipulation by the researcher.
Laboratory, Field and Natural Experiments
All three types of experiment involve measuring the effect of an independent variable on a dependent variable. They differ mainly in where they take place and how much control the researcher has over the environment.
A laboratory experiment takes place in a specially controlled setting, such as a university lab. The researcher can control lighting, noise, timing and the exact stimuli each participant sees. Strength: high control makes results easier to replicate and more clearly linked to the IV. Weakness: the artificial setting can reduce how naturally participants behave, so results may not reflect everyday life.
A field experiment takes place in a real-world setting, such as a school or shop, while the researcher still deliberately manipulates the independent variable. Strength: participants behave more naturally because they are in a familiar environment. Weakness: the researcher has less control over other variables that might affect the results.
A natural experiment studies an independent variable that would have changed anyway, such as comparing behaviour before and after a change in law. Strength: it allows psychologists to study variables that would be unethical or impossible to manipulate directly. Weakness: participants cannot be randomly allocated to conditions, so it is harder to be sure the IV caused any difference found.
Interviews
A structured interview asks every participant the same fixed questions, in the same order. Strength: because every participant answers identical questions, the interview is easy to repeat and answers are simple to compare. Weakness: the interviewer cannot follow up on an interesting or unexpected answer, so some detail may be lost.
An unstructured interview has no fixed questions. It works more like a guided conversation, with the interviewer free to ask follow-up questions based on what the participant says. Strength: it can produce rich, detailed qualitative data and can explore topics a structured interview would miss. Weakness: interviews are hard to repeat in exactly the same way, and the interviewer’s own views may unintentionally influence the questions asked.
Questionnaires: Open, Closed and Rating-Scale Questions
Closed questions give the respondent a fixed set of possible answers, such as ‘yes/no’ or a list of options. They produce quantitative data that is quick to score and easy to compare across large numbers of respondents.
Open questions have no fixed answers — the respondent replies in their own words. They produce qualitative data with more depth and detail, but the answers are harder to analyse and compare because every respondent may word their answer differently.
A rating scale asks respondents to indicate the strength of an opinion on a numbered range, for example from 1 (strongly disagree) to 5 (strongly agree). This turns an attitude into a number, so it can be added up and compared statistically, though respondents may avoid the extreme ends of the scale.
Closed questions and rating scales produce quantitative data that is quick to analyse; open questions produce qualitative data with more depth but less ease of comparison. Choosing between them depends on whether the researcher wants breadth (many respondents, simple answers) or depth (fewer respondents, detailed answers).
Observations
Observations involve watching and recording behaviour rather than asking about it. OCR describes observations using three separate pairs of contrasting terms: the setting (naturalistic or controlled), whether participants know they are being watched (overt or covert), and whether the researcher takes part (participant or non-participant). A single study combines one option from each pair.
A naturalistic observation watches behaviour in its normal, everyday setting, with no interference from the researcher. Strength: high ecological validity because behaviour is unaffected by the research process. A controlled observation takes place in a setting where the researcher has altered some conditions, such as observing children through a one-way mirror in a set-up room. Strength: easier to control extraneous variables and to repeat the observation exactly. Weakness: the artificial setting may reduce how naturally participants behave.
In an overt observation, participants know they are being watched, which respects their right to consent but can cause them to change their behaviour because they are aware of being observed. In a covert observation, participants do not know they are being watched, so behaviour is more natural, but this raises ethical concerns about consent and privacy.
In participant observation, the researcher joins in with the activity or group being studied, which can give unique insight into a group’s behaviour from the inside. In non-participant observation, the researcher stays separate and simply watches, which keeps their judgement more objective but may miss detail only visible from within the group.
Observations vary along three independent dimensions — setting (naturalistic or controlled), participant awareness (overt or covert), and researcher involvement (participant or non-participant). A single study can combine any of these, for example a covert, non-participant, naturalistic observation of shoppers in a supermarket.
Choosing the Right Method
The right choice always depends on the research question. A topic needing strict control, such as testing a precise cognitive theory, suits a laboratory experiment. A topic needing natural behaviour, such as helping behaviour in public, suits a field experiment or naturalistic observation. A topic needing personal detail and explanation, such as someone’s experience of a mental health problem, suits an unstructured interview or open questions. In every case, name the specific method precisely and explain why it fits the scenario given, not just describe it in general terms.