Sampling and Ethical Guidelines
Choosing Who to Study: Populations and Samples
Researchers can rarely study everyone they are interested in. Instead, they define a target population — the whole group the study is about — and then select a smaller sample to actually take part. The sample stands in for the target population, so how it is chosen matters enormously for the whole study.
A sample must be representative of its target population for findings to be generalised. A biased sample threatens the validity of any conclusions drawn from a study.
Sample Size, Representativeness and Generalisability
Representativeness depends partly on sample size — a larger sample is usually more likely to cover the range of people in the target population, though a large sample can still be unrepresentative if it is drawn badly. If a sample is representative, the researcher can achieve generalisability — confidence that findings from the sample apply to the whole target population. A small or biased sample limits generalisability, whatever its size.
Four Sampling Methods on the OCR Course
OCR requires you to know four ways of selecting a sample from a target population: random sampling, opportunity sampling, self-selected (volunteer) sampling and snowball sampling. Each has a different balance of practicality and representativeness, and exam questions often ask you to name the method, describe it briefly, then give a strength and a weakness.
Random sampling gives every member of the target population an equal chance of selection, using a lottery or random number generator applied to a full list of the population. Strength: unbiased and representative. Weakness: a complete list of the population is often hard to obtain, and people who decline to take part can still leave the final sample biased.
Opportunity sampling selects whoever happens to be available and willing to take part, such as a researcher approaching students outside the college canteen. Strength: quick, easy and inexpensive. Weakness: biased towards people who are conveniently available at that particular time and place, so the sample may not represent the wider target population.
Self-selected sampling relies on participants volunteering, often by replying to an advertisement or a request for participants. Strength: convenient, and participants are willing, which can improve engagement. Weakness: volunteers often share unusual characteristics — they may be more curious, confident or motivated than the average person — making the sample unrepresentative.
Four sampling methods: random (equal chance, unbiased but needs a full population list), opportunity (quick and easy but biased towards who is available), self-selected (willing volunteers but prone to volunteer bias), snowball (existing participants recruit others; useful for hard-to-reach groups but often unrepresentative).
Ethical Issues in Psychological Research
Psychological research involves real people, so researchers follow ethical guidelines set out by the British Psychological Society. Three key ethical issues named on the OCR specification are lack of informed consent, protection of participants, and deception.
Lack of informed consent occurs when participants are not told enough about a study beforehand to make a free, informed decision about taking part. This is a problem because participants cannot properly agree to something they do not understand.
Protection of participants means participants should not suffer physical or psychological harm as a result of taking part — for example, being made to feel embarrassed, distressed or unusually anxious.
Deception is when a researcher withholds information about the true purpose of a study, or actively misleads participants. Some deception is sometimes unavoidable if revealing the true aim would change how participants behave, but it must be justified and dealt with afterwards.
Debriefing deals with a lack of informed consent and deception. After the study, the researcher explains its true purpose, corrects any deception and answers participants' questions, so they leave fully informed.
Right to withdraw deals with protection of participants. Participants must be told before and during the study that they can stop taking part at any time, without giving a reason, protecting them from ongoing harm.
Confidentiality protects participants' privacy. Researchers avoid naming participants in reports, often using codes such as Participant A, and store personal data securely.
Every ethical issue needs a matching solution: lack of informed consent and deception are addressed through debriefing; protection of participants is addressed through the right to withdraw; risks to privacy are addressed through confidentiality. Naming an issue alone is not enough — you must pair it with how it was resolved.
The BPS Code of Ethics and Conduct
All of these guidelines sit within the BPS Code of Ethics and Conduct, published by the British Psychological Society. It sets the professional standard that any psychologist carrying out research in the UK is expected to follow, covering consent, protection from harm, deception, the right to withdraw and confidentiality together as one connected set of responsibilities to participants.