Researching Families
Why Is Researching Families Challenging?
Families are one of the most private social institutions. Unlike schools, workplaces or public spaces, families are largely closed to outside observation. This creates significant practical issues for sociologists who wish to study family life: gaining access, maintaining trust and capturing honest behaviour. At the same time, families involve deeply personal relationships — raising sensitive ethical issues around consent, confidentiality and the potential for harm. AQA requires students to evaluate research methods applied specifically to families, not just in the abstract.
Surveys and Questionnaires: Studying the Domestic Division of Labour
Surveys using questionnaires have been widely used to study the domestic division of labour and conjugal roles. Willmott and Young’s research for The Symmetrical Family (1973) used a large-scale survey to gather data about how domestic tasks were divided. Nearly 2,000 people in Greater London and the surrounding area were interviewed. This quantitative approach allowed them to identify broad patterns and trends across social classes.
However, surveys have important limitations when used to study family life. Questionnaires produce data that is easily comparable and statistically analysable but may lack depth. Respondents may give socially desirable answers — describing a more equal division of tasks than actually exists — rather than recording their genuine behaviour. Ann Oakley criticised this limitation directly in her critique of Willmott and Young: their survey data on male domestic participation was based on a low threshold, and respondents may have overstated male involvement. Qualitative methods may better capture the complexity of domestic life than tick-box questionnaires.
Interviews: Gaining Depth on Family Experience
Oakley’s own research on housework used in-depth interviews with women about their experience of domestic labour. She used unstructured interviews, which allowed participants to speak freely in their own words about their experiences and feelings. This generated rich qualitative data about women’s attitudes to housework, their sense of identity and the emotional demands of domestic labour — data that a survey could not have captured.
Interviews have significant advantages for researching sensitive aspects of family life, such as experiences of domestic violence, parenting practices or the experience of divorce. They allow the researcher to probe for detail, follow up unexpected responses and build rapport with participants. However, they are also time-consuming and expensive, produce data that is difficult to generalise from a small sample, and rely on participants reporting their experiences honestly — which may be particularly problematic for sensitive topics within families.
Willmott and Young used large-scale surveys (quantitative) to study the domestic division of labour — allowing broad patterns to be identified but risking socially desirable responses. Oakley used in-depth unstructured interviews (qualitative) with women about housework — generating rich, detailed data but with a smaller, less generalisable sample.
Observation of Family Life
Observation of family life raises acute practical and ethical challenges. Observing family behaviour in its natural setting — the home — could provide highly valid data about how families actually behave (rather than what they say they do). However, gaining access to private family homes for observation is extremely difficult. Even when access is granted, families are likely to change their behaviour when they know they are being observed, creating a Hawthorne effect.
Participant observation — where the researcher lives with or becomes part of the family for a period — could generate highly naturalistic data, but raises serious ethical issues about intrusion into private life, the boundaries of the researcher’s role, and the potential for family members to feel pressured to participate. The Rapoports’ research on family diversity involved extensive contact with families over time, using a mixed methods approach combining interviews with observational elements.
Official Statistics on Families
Official statistics are a major source of data for studying family patterns. The Office for National Statistics (ONS) publishes annual data on marriage rates, divorce rates, cohabitation, household composition, lone parent families and birth rates. The census collects detailed data on household composition every ten years.
Official statistics on divorce, marriage and family forms are quantitative, nationally representative and freely available, making them highly convenient for sociologists. However, they have significant limitations. Official statistics on families count legal arrangements (married couples, registered civil partnerships, legal divorces) but miss cohabiting couples who separate without legal proceedings and families that do not conform to standard administrative categories. Marriage and divorce statistics also change with legal reform (as the dramatic effect of the Divorce Reform Act 1969 demonstrates) — meaning trends in statistics may partly reflect changes in law rather than changes in family behaviour.
Practical issues include: access — families are private institutions and researchers cannot simply enter without invitation. Recruiting family research participants often requires working through schools, community centres, online networks or snowball sampling. Time and cost — longitudinal studies of families (tracking the same households over years or decades) are expensive and resource-intensive, though they yield valuable data on how families change. Representativeness — families that agree to participate in research may not be typical of all families (those experiencing serious domestic problems, for example, are less likely to volunteer for a study of “family life”). Changing composition — families change over time through births, deaths, divorce, remarriage and children leaving home, making longitudinal family research methodologically complex.
The AQA specification requires students to know three ethical issues in research: consent, confidentiality and harm. In family research, these apply with particular force: family members must all give informed consent; data about family relationships must be kept confidential; and researchers must consider the potential for their presence or questions to cause distress, particularly when studying sensitive topics such as domestic abuse, divorce or child neglect.
Informed consent is particularly complex when researching families with children. Children below the age of consent cannot legally give consent themselves, so parents or guardians must consent on their behalf. However, sociologists debate whether parental consent alone is sufficient — should children also give their assent? Children may feel unable to refuse to participate if their parents have agreed. Confidentiality is also challenging: if a researcher discovers evidence of child abuse or neglect during family research, they face a conflict between their duty of confidentiality to participants and their legal and moral obligation to protect the child from harm. AQA requires students to think about how ethical issues “can be addressed” — not just identified. Possible approaches include: briefing participants clearly about the limits of confidentiality before the study begins; using independent child welfare advisers; ensuring children can speak to researchers separately from parents.