AQA GCSE Sociology (8192): Sociological research methods – Study Guide

This guide covers section 3.7, Sociological research methods, of the AQA GCSE Sociology (8192) specification, Version 1.2 (June 2026), for GCSE exams 2019 onwards. Section 3.7 has seven content rows: research design, qualitative and quantitative methods, different types of data, primary and secondary sources, interpretation of data, practical issues and ethical issues. It is not a separate paper. Paper 1 (The sociology of families and education) and Paper 2 (The sociology of crime and deviance and social stratification) both assess “relevant areas of social theory and methodology”; each is a 1 hour 45 minute written exam worth 100 marks and 50% of the GCSE.

Use this guide with the research methods revision notes and the research methods practice questions. Our shorter research methods overview gives a quick summary, and the exam preparation guide covers technique. Track the course on the AQA GCSE Sociology hub and the 8192 checklist.

Coverage table

Syllabus reference	What you must know or be able to do (from the specification)	Where assessed
3.7 Research design	Describe and explain the processes involved in research design: appropriate aims and relevant hypotheses, pilot studies, appropriate sampling methods, analysis of data	Papers 1 and 2
3.7 Qualitative and quantitative methods	Describe and explain questionnaires, interviews and observations; assess their value, application, strengths and weaknesses; assess the usefulness of the mixed methods approach	Papers 1 and 2
3.7 Different types of data	Assess the usefulness of qualitative and quantitative data, and official and non-official statistics	Papers 1 and 2
3.7 Primary and secondary sources	Describe and explain primary and secondary sources of data	Papers 1 and 2
3.7 Interpretation of data	Interpret graphs, diagrams, charts and tables to discern patterns and trends in statistical data	Papers 1 and 2
3.7 Practical issues	Time, cost and access	Papers 1 and 2
3.7 Ethical issues	Consent, confidentiality and harm to participants, and how the issues can be addressed	Papers 1 and 2
3.7 skills	Do all of the above “in the context of the various social structures, social processes and social issues detailed in the specification” (families, education, crime and deviance, social stratification), for “a specific area of research”	Papers 1 and 2

Key terms

Pilot study comes from the research design content row; the other terms appear in the specification’s key terms appendix (Appendix A).

Term	Meaning
Hypothesis	A testable statement that predicts a relationship, e.g. “boys in lower sets are more likely to be excluded”
Population and sample	The whole group to be studied, and the smaller group actually studied
Sampling frame	A list of everyone in the population, from which a sample is drawn
Representative sample	A sample with the same characteristics as the population, so findings can be generalised
Pilot study	A small trial run of the research before the main study
Respondent	A person who answers a questionnaire or interview
Reliability	Whether another researcher using the same method would get the same results
Validity	Whether the data give a true picture of what is being studied
Bias	Anything that distorts results, such as leading questions or an unrepresentative sample
Triangulation	Using more than one method or source to check that findings agree
Selective use of data	Choosing only the evidence that supports an argument

Research design

Research design is the plan for a study. The specification names four stages.

1. Aims and hypotheses. An aim is a general statement of what you want to find out: “to investigate how couples divide housework”. A hypothesis is narrower and testable: “dual-career couples share housework more equally than couples with one earner”. Quantitative research usually tests a hypothesis. Qualitative research often starts with an open aim, because the researcher does not want to decide the answer in advance.

2. Pilot study. A trial on a few people shows whether questions are confusing, whether the method takes too long and whether people will take part.

3. Sampling. A sampling frame is needed for most methods. School rolls, the electoral register and GP lists are examples, though access to some lists is restricted. The key terms appendix lists four sampling methods:

• Random sample: every member of the population has an equal chance of selection, e.g. names drawn by computer. Fair, but a small random sample can still be unrepresentative by chance.
• Systematic sample: every nth name on the list. To pick 100 pupils from a roll of 1,200, take every 12th name. Quick, but a list ordered in a pattern can create bias.
• Quota sample: the researcher fills set numbers for each group to match the population. If a town is 52% female, a sample of 150 would include 78 women and 72 men. No sampling frame is needed, but the researcher chooses who fills each quota, which can bring in bias.
• Snowball sample: one participant introduces the next. This reaches hidden groups, such as people in an illegal activity, but contacts tend to be similar, so the sample is rarely representative.

4. Analysis of data. Quantitative data are counted and shown in tables and charts. Qualitative data are sorted into themes, with quotations.

Evidence from the specification’s named studies

The specification’s text summaries show these choices in real research:

• Halsey, Heath and Ridge (Origins and Destinations, 1980) used a sample of over 8,000 males born between 1913 and 1952. The size gives confidence in the class patterns. The specification notes the research excluded females, which “might have made a significant difference to the findings”. Large is not the same as representative.
• Carlen (Women, Crime and Poverty, 1988) carried out in-depth, unstructured interviews with 39 women aged 15 to 46 who had been convicted of crimes. The specification calls this “a relatively small sample”. The depth is high, but generalising is risky.

Worked example paragraph

How might a sociologist design a study of whether school setting affects pupils’ self-image? The researcher could set the hypothesis that pupils in bottom sets have lower self-esteem than pupils in top sets (knowledge). In a school with 1,200 pupils, a systematic sample of every 12th name on the roll would give 100 pupils from all sets (application). A pilot with a few pupils would show whether questions about self-esteem are understood and not upsetting, which matters with young respondents (analysis). However, one school cannot represent all schools, and pupils may give answers they think teachers want, which lowers validity (evaluation).

Qualitative and quantitative methods

Quantitative methods produce numbers. Qualitative methods produce words, meanings and descriptions. The specification names three main methods.

Questionnaires

A list of written questions. Closed questions give fixed answers (yes/no, tick boxes, rating scales). Open questions let the respondent write freely. An attitude survey measures opinions, often on an agree/disagree scale. A survey is any study that gathers the same information from a large sample.

• Strengths: cheap and fast for large samples; easy to compare; high reliability; anonymous, so people may be more honest about private matters.
• Weaknesses: low response rates for postal or online forms; closed questions force answers into the researcher’s categories; no chance to explain a question; people may give socially acceptable answers.
• Example: the specification’s summary of Townsend (Poverty in the United Kingdom, 1979) says his research was based on questionnaires issued to over 2,000 households and more than 6,000 individuals. He built a deprivation index from items such as diet, fuel, clothing and housing. Critics argued some items might reflect choice rather than poverty, a validity problem.

Interviews

A conversation in which the researcher asks questions. A structured interview is a questionnaire read aloud: standard questions, in order. An unstructured interview is more like a guided conversation; the interviewer follows up interesting answers. A focus group is a group interview where participants discuss a topic together.

• Strengths: structured interviews have good response rates and reliability; unstructured interviews give depth and validity and build rapport, useful for sensitive topics such as Carlen’s.
• Weaknesses: time-consuming and costly; interviewer bias (the interviewer’s manner, age or gender may shape answers); unstructured interviews are hard to compare or repeat; focus groups can be dominated by one voice.
• Example: the specification’s summary of Young and Willmott (The Symmetrical Family, 1973) describes a large-scale social survey in which nearly 2,000 people were interviewed in Greater London and surrounding areas. Devine (Affluent Workers Revisited, 1992) interviewed male manual workers at the Vauxhall car plant in Luton and their wives in the late 1980s.

Observations

The researcher watches behaviour. In participant observation the researcher joins the group. In non-participant observation they watch without joining. Covert observation is secret; in overt observation the group knows. An ethnography is a detailed study of a group’s way of life, usually by participant observation over a long period.

• Strengths: sees what people do, not what they say; high validity; reaches groups that would not fill in forms.
• Weaknesses: the observer may change the group’s behaviour if overt; covert work raises consent problems; the researcher may lose objectivity; one small group may not be typical; hard to repeat, so low reliability.
• Example: Ball (Beachside Comprehensive, 1981) carried out three years of fieldwork as a participant observer in a south coast comprehensive school. Willis (Learning to Labour, 1977) used observation, participant observation, recorded group discussions, informal interviews and diaries to understand schooling from the students’ point of view.

Other designs in the key terms

A case study looks in depth at one setting, such as Ball’s school. A longitudinal study follows the same group over time. Content analysis examines documents or media content.

Mixed methods

Mixed methods research combines quantitative and qualitative methods in one study, for example a questionnaire to find how common something is, then interviews to find out why.

• Usefulness: combines reliability and validity; numbers show the pattern, words explain it; allows triangulation.
• Limitations: costs more time and money; results from different methods may clash and be hard to combine; the researcher needs skills in both.

Different types of data

Quantitative data are numerical: they show patterns and trends but can hide meaning (a divorce rate tells you how many, not why). Qualitative data are descriptive: they give depth and the participant’s view but are harder to compare.

Official statistics are numerical data produced by government and its agencies, such as the census or figures on exam results, exclusions and recorded crime. Non-official statistics are produced by others, such as charities, universities, pressure groups or businesses.

	Strengths	Limitations
Official statistics	Large, often national coverage; free to use; collected regularly, so trends show; the census covers the whole population	Definitions are set by government and can change; some events go unrecorded; interactionists see them as social constructions, the product of decisions by officials
Non-official statistics	Can cover topics the state ignores; may use different definitions	May be smaller or less representative; the producer may have an interest, so check for bias and selective use of data

Townsend’s work shows how definitions change the numbers. He believed more than 22% of the population was in poverty in 1968-69 on his relative deprivation measure, compared with just over 6% on the state standard and a little over 9% on relative income. One society, three figures.

Primary and secondary sources

Primary data are collected first-hand by the researcher for their own study: their questionnaires, interviews and observations. Secondary data already exist, produced by someone else: official statistics, earlier studies, diaries, letters, school records, websites and media content.

• Primary data fit the aim exactly, but cost time and money.
• Secondary data are quick and cheap and may cover the past or a whole country, but were made for another purpose.

Devine is a useful example of combining the two: by returning to Luton, she compared her new primary data with Goldthorpe and Lockwood’s 1960s research, which was by then secondary data.

Interpretation of data

The specification expects you to read graphs, diagrams, charts and tables to “discern patterns and trends”. A trend is a general direction over time.

For any data item: read the title, units, dates and source; describe the overall pattern; quote figures with units; note exceptions; ask what the data cannot tell you.

Worked example (fictional data). A survey in the fictional town of Dunmarsh found that 28% of couples said they shared housework equally in 2005, and 42% in 2025.

• The rise is 42 - 28 = 14 percentage points.
• As a proportion of the starting figure, 14 / 28 x 100 = a 50% increase.

“It rose by 14%” would be wrong. Then evaluate: “said they shared” is self-reported, and two years cannot show what happened in between.

Practical issues

The specification names time, cost and access.

• Time: participant observation can take years (Ball’s fieldwork lasted three); a questionnaire can be done in weeks.
• Cost: large samples, travel, interviewer wages and transcription all cost money. Secondary data are usually cheap.
• Access: some groups are hard to reach. Research in a school usually needs the head teacher’s agreement, and parental consent is normally sought for pupils; prisons and courts are tightly controlled; private family life happens behind closed doors. Some of Carlen’s interviewees were in prison or youth custody at the time.

Ethical issues

The specification names three issues and asks how each “can be addressed”. UK sociologists also follow the British Sociological Association’s Statement of Ethical Practice.

Issue	What it means	How to address it
Consent	Participants should agree to take part, knowing what the research involves (informed consent)	Explain aims, give information sheets, get written consent; parental consent for children; tell people they can withdraw
Confidentiality	Participants’ identities and information are kept private	Anonymise names and places; store data securely in line with data protection law; report findings so individuals cannot be identified
Harm to participants	Research must not damage people physically, emotionally or socially	Avoid distressing questions; allow people to stop; offer support contacts; debrief after the study

Covert observation is the classic dilemma. It may give the most valid picture of a deviant group, but it breaks informed consent. Some sociologists argue it is justified when no other method could reach the group and identities are protected; others say deception is never acceptable. Present both views.

Common errors

• Writing “a random sample is representative”. It gives everyone an equal chance; it does not guarantee representativeness.
• Confusing reliability (can it be repeated with the same results?) with validity (is it true to life?).
• Calling all official statistics “secondary data and therefore useless”. Usefulness depends on the topic and definitions.
• Treating “percentage” and “percentage points” as the same.
• Naming a method without saying how it fits the topic in the question: a questionnaire on domestic violence raises different issues from one on homework.
• Describing a named study’s findings when the question wants its method.

Where to go next

Test yourself with the revision notes, then the practice questions. For methods in topic context, see Families and Education and Crime and Stratification.

Official syllabus

AQA GCSE Sociology (8192) specification, Version 1.2 (June 2026), for GCSE exams 2019 onwards: section 3.7 Sociological research methods, with the research methods terms in Appendix A and the study summaries in Appendix B.