What you'll learn

• How sociologists classify sources of data: primary, secondary, quantitative and qualitative.
• The main features of questionnaires, interviews, observation, experiments, documents and official statistics.
• How to evaluate each source using practical, ethical and theoretical issues.
• How to apply methods to UK examples involving identity, culture, power and inequality.

First: what counts as data?

In Sociology, data means the evidence researchers use to describe, explain or understand social life. A source of data is where that evidence comes from. A research method is the technique used to collect or analyse it.

For example, if you study class differences in school achievement, your data might come from pupil interviews, classroom observation, DfE statistics, school behaviour records or newspaper reports about education policy.

A helpful way to map sources of data is to place them on two axes: primary versus secondary, and quantitative versus qualitative.

The evaluation toolkit

When you evaluate any source of data, use three big categories.

Practical issues

Practical issues are about whether the research is feasible: time, cost, access, researcher skill, sample size and response rate.

Ethical issues

Ethical issues are about protecting people’s rights and wellbeing. Key ideas include informed consent: participants should understand the research before agreeing; confidentiality: personal information should not be revealed; and avoiding harm.

Theoretical issues

Theoretical issues are about the kind of knowledge the method produces.

Questionnaires

A questionnaire is a list of written questions completed by respondents. A respondent is someone who answers research questions.

Questionnaires often use closed questions, where people choose from fixed answers, such as “strongly agree” to “strongly disagree”. These produce quantitative data. They may also include open questions, where respondents write their own answers, producing limited qualitative data.

Questionnaires are often favoured by positivists: sociologists who believe society can be studied scientifically by collecting objective, measurable data. Because the same questions are asked in the same way, questionnaires can be reliable and useful for comparing large groups.

They are practical too. Online questionnaires can be cheap, quick and distributed widely. They can also feel anonymous, which may help when researching sensitive topics such as experiences of poverty, racism or family conflict.

However, questionnaires can lack validity. Respondents may misunderstand questions, rush answers, or give socially desirable responses. Low response rates can also damage representativeness, especially if marginalised groups are less likely to reply.

Interviews

An interview is a research conversation where an interviewer asks questions and records answers.

There are three main types. A structured interview uses fixed questions, like a spoken questionnaire. A semi-structured interview has prepared questions but allows follow-up discussion. An unstructured interview is more flexible and participant-led.

Structured interviews tend to be more reliable because each participant is asked the same questions. Unstructured interviews often have higher validity because they allow people to explain meanings in their own words.

Interpretivists often prefer unstructured interviews. Interpretivism is the view that sociology should understand social action from the point of view of the people involved. Weber called this verstehen, meaning empathetic understanding.

Interviews can be especially useful for studying socialisation, culture and identity: for example, how young people describe masculinity, religious identity or experiences of racism. They can also reveal power inequalities, such as how benefit claimants experience stigma.

But interviews have limits. The interviewer effect means the interviewer’s age, gender, ethnicity, accent or behaviour may influence responses. Participants may also exaggerate, hide information or tell the interviewer what they think is acceptable.

Observation

Observation means watching social behaviour directly. It is useful when researchers want to see what people do, not just what they say they do.

In participant observation, the researcher joins in with the group being studied. In non-participant observation, the researcher watches without taking part. Observation may be overt, where people know they are being studied, or covert, where they do not.

Participant observation can produce rich qualitative data and high validity. Willis’s study of working-class “lads” in Learning to Labour (1977) used ethnographic methods to show how anti-school culture shaped boys’ identities and future class positions.

However, observation is often time-consuming and difficult to access. Covert observation raises serious consent issues. Overt observation may create the Hawthorne effect, where people change their behaviour because they know they are being watched.

Experiments

An experiment investigates cause and effect by changing one factor and observing the result. The factor changed is the independent variable; the outcome measured is the dependent variable. A control group is a comparison group that does not receive the experimental change.

Laboratory experiments are rare in Sociology because social life is hard to control and artificial settings may reduce validity. Field experiments happen in real-world settings. Natural experiments compare groups affected by real social changes, such as policy reforms.

Experiments appeal to positivists because they can be reliable and can test causal relationships. Rosenthal and Jacobson’s school study (1968), often linked to teacher expectations, is useful for thinking about how labels may affect pupil performance.

Field experiments are also used to study discrimination. For example, researchers may send identical job applications with different names to test whether employers treat ethnic groups differently.

Documents

Documents are existing written, visual or digital materials. They can include diaries, letters, school records, government reports, newspapers, films, websites and social media posts.

Documents may be analysed quantitatively through content analysis, where researchers count categories or themes. They may also be analysed qualitatively to interpret meanings, representations and discourse.

Scott (1990) suggested four tests for documents: authenticity: is it genuine? Credibility: is it accurate? Representativeness: is it typical? Meaning: what does it mean in context?

Documents are useful because they are often cheap and non-reactive: people do not change their behaviour for the researcher because the document already exists. They can also show how culture and identity are represented, such as media portrayals of migrants, benefits claimants or young people.

But documents are not neutral. They are produced by people and institutions with particular interests. A school behaviour record, for example, may reflect official power as much as pupil behaviour.

Official statistics

Official statistics are numerical data produced by government departments or public bodies. Examples include the Census, ONS data, police recorded crime, the Crime Survey for England and Wales, DfE education statistics, NHS data and unemployment figures.

Official statistics are often large-scale, cheap to access and useful for identifying trends. The Census, for example, helps sociologists analyse changes in family structure, religion, ethnicity, housing and migration in the UK.

Durkheim used official suicide statistics to argue that suicide rates were shaped by social integration and regulation. However, Atkinson criticised this approach, arguing that suicide statistics are socially constructed through decisions by coroners, doctors, families and officials.

Some statistics are “harder” than others. Birth and death statistics are usually seen as relatively reliable. Crime, unemployment and suicide statistics are “softer” because they depend more heavily on reporting, definitions and recording rules.

Linking methods to theory

Different sociological perspectives tend to prefer different sources of data.

Positivists usually favour questionnaires, structured interviews, experiments and official statistics because these can produce reliable, comparable, quantitative data.

Interpretivists usually favour unstructured interviews, participant observation and personal documents because these can produce valid qualitative data about meanings and identity.

Feminist researchers often focus on power in the research relationship. Oakley criticised hierarchical interviewing and argued for more equal, supportive interviews. However, feminists may also use official statistics to expose gender inequality, such as pay gaps, domestic labour or violence against women.