What you'll learn
Sociological research methods are the techniques sociologists use to collect and analyse evidence about society, such as questionnaires, interviews, observation and statistics. Choosing a method always involves trade-offs between how much data you can gather, how deep it goes, how much it costs and whether it is ethical. In AQA GCSE Sociology (8192), research methods are tested on both papers: each paper (1 hour 45 minutes, 100 marks, 50% of the GCSE) includes methods questions set in the context of its topics (families and education on Paper 1, crime and deviance and social stratification on Paper 2). This guide covers research design, qualitative and quantitative methods, types of data, primary and secondary sources, interpreting data, and the practical and ethical issues the specification lists, with worked examples of the kind of methods questions you will meet.
Key terms and definitions
Aim — A general statement of what a piece of research is trying to find out.
Hypothesis — A testable statement that predicts a relationship between two or more factors, for example "Students in lower sets are more likely to have negative attitudes to school."
Pilot study — A small trial run of the research, used to test questions and procedures before the main study.
Population — The whole group the researcher is interested in.
Sampling frame — A list of all members of the population from which a sample is chosen, such as a school register.
Sample — The smaller group selected to take part in the research.
Representative — A sample is representative when it has the same characteristics, in the same proportions, as the wider population.
Generalise — To apply the findings from a sample to the whole population.
Reliability — Research is reliable if another researcher using the same method would get the same results.
Validity — Research is valid if it gives a true, accurate picture of what it claims to measure.
Quantitative data — Numerical data, such as statistics, percentages and counts.
Qualitative data — Descriptive, non-numerical data, such as quotations, accounts and observations of meaning and feeling.
Primary data — Data collected first-hand by the researcher.
Secondary data — Data that already exists, collected by someone else, such as official statistics, diaries or previous studies.
Mixed methods — Using more than one method in the same study, often combining quantitative and qualitative approaches. Using several methods to check findings is sometimes called triangulation.
Informed consent — Participants agree to take part after being told what the research involves and that they can withdraw.
Confidentiality — Keeping participants' identities and personal information private.
Core concepts
Research design
Research begins with an aim or a hypothesis. A researcher then decides on methods, conducts a pilot study to iron out problems (such as confusing or leading questions), selects a sample, collects data and analyses it to draw conclusions.
The main sampling methods are:
- Random sampling — every member of the sampling frame has an equal chance of selection, for example drawing names at random. It avoids researcher bias but may by chance be unrepresentative.
- Systematic sampling — choosing every nth name from a list, for example every tenth student on the register.
- Stratified sampling — dividing the population into groups (strata) such as gender or age, then sampling randomly from each group in proportion to its size. This gives a more representative sample.
- Quota sampling — the researcher fills set numbers (quotas) of people with certain characteristics, such as 50 men and 50 women, choosing whoever is available. It is quick but not random.
- Snowball sampling — one participant puts the researcher in touch with others. It is useful for hard-to-reach groups, such as gang members or drug users, but is unlikely to be representative.
Quantitative and qualitative methods
Questionnaires are lists of questions given to many people, often by post, email or online. Closed questions have fixed answers that can be counted, producing quantitative data; open questions let respondents answer in their own words. Questionnaires are cheap, quick and reliable, can reach large samples and allow anonymity, but response rates can be low, respondents cannot ask what a question means, and answers may lack depth or honesty.
Interviews involve the researcher asking questions face to face or remotely. Structured interviews use a fixed list of questions in the same order, giving reliable, comparable data. Unstructured interviews are more like a guided conversation, allowing the researcher to explore issues in depth and gain valid, qualitative data, but they are time-consuming, hard to compare and open to interviewer bias, where the researcher's presence or manner affects the answers. Semi-structured interviews combine set questions with freedom to follow up answers. Group interviews gather several people together and can reveal shared views, though some members may dominate.
Observation involves watching behaviour. In participant observation the researcher joins the group; in non-participant observation the researcher watches without taking part. Observation can be overt (the group knows they are being studied) or covert (they do not). Participant observation can give rich, valid insight into how people actually behave, as in Willis's study of "the lads", but it takes a long time, the researcher may "go native" and lose objectivity, findings are hard to repeat, and covert observation raises ethical problems because there is no informed consent. Overt observation risks the Hawthorne effect, where people change their behaviour because they know they are being watched.
Each method has value for particular research questions. Quantitative methods suit measuring patterns and trends across large groups; qualitative methods suit understanding meanings and experiences. A mixed methods approach can combine the strengths of both, for example a questionnaire to identify patterns followed by interviews to explain them.
Different types of data
Quantitative data allows comparison and the identification of trends, but may miss meanings. Qualitative data gives depth and validity but is harder to analyse and generalise.
Official statistics are produced by government agencies, such as crime figures, exam results, marriage and divorce statistics and the census. They cover large populations, are often free and allow trends to be tracked over time, but they may be collected for the government's purposes rather than the sociologist's, and some, such as crime statistics, are shaped by how data is recorded. Non-official statistics are produced by other organisations such as charities, businesses or researchers.
Primary and secondary sources
Primary sources let researchers collect exactly the data they need, but this costs time and money. Secondary sources (official statistics, previous studies, letters, diaries, media content) are cheaper and may cover the past or very large populations, but may not be accurate, complete or relevant to the researcher's question.
Interpretation of data
You need to read graphs, charts, diagrams and tables to identify patterns (differences between groups at one time) and trends (changes over time). Always quote figures accurately, give the units (percentages, thousands), state the time period, and describe the overall direction before mentioning exceptions.
Practical and ethical issues
Practical issues are time, cost and access. Large-scale questionnaires may be cheap per person; long-term participant observation is costly in time; some groups, such as elites or criminals, are hard to access.
Ethical issues include informed consent, confidentiality and avoiding harm to participants, whether physical or psychological. Researchers deal with these by explaining the research and obtaining consent (from parents or guardians for children), anonymising data, allowing participants to withdraw, and following the ethical guidelines of professional bodies. Research with vulnerable groups or on sensitive topics, such as domestic abuse, requires extra care.
Worked examples
Example 1: Choosing a method
Question: A researcher wants to find out whether students in different sets have different attitudes to school. Identify and explain one advantage of using a questionnaire (4 marks). A questionnaire can be given to every student in each set, producing a large sample. Closed questions give quantitative data that can be compared between sets, for example the percentage in each set who agree that school is useful. This makes the findings reliable and easier to generalise.
Example 2: Choosing a sample
To study the views of a school of 1,000 students, of whom 60% are girls and 40% boys, a researcher wants a sample of 100. A stratified sample would randomly choose 60 girls and 40 boys, so the sample matches the population's gender balance and is more representative.
Example 3: Ethical issues
Question: Explain one ethical issue in a study of young offenders (4 marks). Young offenders are a vulnerable group, so there is a risk of harm, for example if discussing their offences causes distress. The researcher should gain informed consent from both the participant and a parent or guardian, keep their identities confidential, and let them stop at any time.
Example 4: Interpreting data
A table shows recorded divorces falling over two decades while the number of marriages also falls. A good interpretation states the trend in divorces, links it to the falling number of marriages, and notes that fewer divorces does not necessarily mean happier marriages, because fewer people are marrying in the first place.
Common mistakes and how to avoid them
Do not confuse reliability (the same results if repeated) with validity (a true picture of reality).
Do not say a large sample is automatically representative. Representativeness depends on how the sample is chosen.
Avoid describing a method without applying it to the research context in the question. "Interviews give in-depth data" earns less than "unstructured interviews would let the researcher explore why girls feel more motivated in lessons".
Remember that covert observation raises ethical problems of consent and deception, while overt observation risks the Hawthorne effect.
Do not call official statistics primary data. They are secondary data for a sociologist who uses them.
Exam technique for "Research methods"
Methods questions are usually set in a context, so always tie your answer to that context: name the group, the topic and why the method suits it. When asked to "evaluate" a method, give at least one strength and one weakness using the specification's criteria: reliability, validity, representativeness, practical issues (time, cost, access) and ethical issues (consent, confidentiality, harm). If you are asked to design research, state the aim or hypothesis, the method, the sampling method and how you would deal with one practical and one ethical issue.
Quick revision summary
- Research design: aim or hypothesis, pilot study, sampling, data collection, analysis.
- Sampling: random, systematic, stratified, quota, snowball.
- Questionnaires: large samples, reliable, quantitative; low depth.
- Interviews: structured (reliable) to unstructured (valid, in depth); interviewer bias.
- Observation: participant or non-participant; overt or covert; Hawthorne effect.
- Quantitative = numbers; qualitative = words and meanings; mixed methods combine both.
- Primary data is collected first-hand; secondary data already exists.
- Practical issues: time, cost, access. Ethical issues: consent, confidentiality, harm.
- Patterns compare groups; trends show change over time.