Researching Education

Why Does Research Method Matter in Education?

All the sociological knowledge discussed in this unit — from Bowles and Gintis on the correspondence principle to Willis on counter-school cultures, from Hargreaves on deviant labelling to Ball on teacher expectations — rests on research conducted in and about schools. The choice of research method shapes what a sociologist can find out. Different methods provide different types of data, carry different practical difficulties, and raise different ethical issues — particularly when the research subjects are children. Research methods permeate the whole of GCSE Sociology: every topic must be studied not just as content, but as something that has been researched in particular ways.

Key Fact

Sociologists researching education use: observation (participant or non-participant), interviews (structured or unstructured), questionnaires and secondary data such as official statistics, league tables and Ofsted reports.

Observation in Schools

Participant observation involves the researcher joining the group being studied — sitting in classrooms, spending time in corridors and building relationships with pupils and teachers over an extended period. Willis’s Learning to Labour (1977) is the most celebrated example: Willis spent months observing twelve working-class boys at a West Midlands comprehensive, attending lessons and informal gatherings. The strength of this method is validity: by immersing himself in the lads’ world, Willis gained rich, in-depth insight into their values and daily lives that a questionnaire could not have produced.

Non-participant observation involves watching what happens in classrooms without joining the group. This is less likely to change participants’ behaviour, but generates less detailed insight into meanings and motivations. Ball’s Beachside Comprehensive research used observation alongside interviews to investigate how banding shaped teacher expectations and pupil trajectories. A key weakness of any observation in schools is the risk of researcher influence: the presence of an observer changes the dynamic between teachers and pupils, potentially undermining validity.

Interviews and Questionnaires in Education Research

Unstructured interviews — open-ended conversations guided by broad topics rather than fixed questions — are particularly effective for exploring sensitive subjects such as the experience of being labelled by a teacher, or attitudes to school among disengaged pupils. They produce rich qualitative data. Ball’s Beachside Comprehensive research combined classroom observation with interviews, using triangulation to increase confidence in the conclusions.

Questionnaires allow large-scale data collection from many respondents quickly and at low cost. They generate quantitative data that can be compared across schools and over time, producing high reliability. However, questionnaires are less effective at exploring why pupils hold particular attitudes or behave in particular ways — the “why” questions that interviews address better.

Official Statistics and League Tables

Sociologists researching educational achievement frequently use official statistics: GCSE and A-level results published by the Department for Education, Ofsted inspection reports, and national league tables. These sources are large-scale, longitudinal and freely available, allowing identification of trends in achievement across social class, gender and ethnicity over time. However, official statistics must be treated with critical caution. They are socially constructed — they reflect decisions about what to measure and how. League tables measure exam results but not the value a school adds to pupils’ progress. A school serving a disadvantaged intake may appear low-achieving by league table standards but be more educationally effective than a school in a wealthy catchment area.

Gaining access to schools is a major practical challenge. Schools are hierarchical institutions with multiple levels of gatekeepers: the headteacher must agree to the research, then individual teachers must open their classrooms, and parents must give consent for their children to participate. Each level of gatekeeping can limit or skew the data. Headteachers may only allow observation of “good” lessons, meaning the data does not represent everyday school life.

Researcher influence is a specific issue in schools: the presence of an unknown adult in the classroom changes pupil and teacher behaviour. Teachers may perform more carefully; pupils may behave differently in front of an observer. This reduces validity. Time and cost are further practical constraints: qualitative studies such as Willis’s required months of fieldwork, making them expensive and difficult to replicate on a larger scale.

Informed consent is particularly complex when research subjects are children. Sociologists must obtain consent from both the child and their parent or guardian. This requirement can introduce class bias: requiring parental written consent may reduce the sample to pupils whose parents are engaged and able to respond. Children must also be told clearly what the research involves and that they can withdraw at any time without penalty.

Confidentiality means protecting participants’ identities. Willis called his school “Hammertown”; Ball used the pseudonym “Beachside”. Pupils who share sensitive information — about bullying, poverty, family difficulties or experiences of labelling — must be assured that their identities are protected. However, confidentiality must be balanced against the duty to report serious safeguarding concerns. A researcher who discovers evidence of abuse cannot remain silent — the ethical obligation to protect the child overrides confidentiality.

Harm to participants includes psychological distress: asking pupils to describe experiences of labelling, exclusion or racism may cause upset. Researchers should use sensitive interview techniques, provide debriefing and ensure pupils know they can stop at any point. These strategies for addressing ethical issues are not a checklist to recite — they must be reasoned about in the specific research context.