The credibility of any sociological enquiry rests on how carefully its evidence is gathered. Techniques of data collection are the practical instruments through which abstract concepts are translated into observable facts, and the choice among them is never merely technical — it reflects deeper commitments about whether social reality is best captured through measurement or through meaning. This note surveys the main techniques an examiner expects you to command, pairing each with a candid appraisal of what it can and cannot deliver.

Primary and secondary data

The first distinction is between primary data, collected first-hand by the researcher for the specific enquiry, and secondary data, already in existence and gathered by others — official statistics, historical records, diaries, media output or earlier studies. Primary data is tailored to the research question and current, but it is costly in time and money. Secondary data is economical and often historical in reach, yet it was produced for other purposes and may be incomplete, biased or non-comparable. Durkheim's Suicide remains the classic demonstration of rigorous secondary analysis, while also warning us of the ecological fallacy — inferring individual behaviour from aggregate figures.

Survey and questionnaire

The social survey administers a standardised questionnaire to a sample in order to generalise to a wider population. Questions may be closed (fixed responses, easily quantified) or open (free responses, richer but harder to code). Its strengths are breadth, comparability, reliability and — through sampling — representativeness, which is why it is the workhorse of positivist and policy research. Its limitations are equally clear: a low response rate threatens validity, respondents may give socially desirable answers, and a pre-set schedule imposes the researcher's categories on respondents, missing the meanings they themselves attach to their conduct.

The interview

Interviews range along a continuum of structure.

  • Structured interviews follow a fixed schedule read verbatim; they behave like a spoken questionnaire — reliable, replicable, quick to analyse, but rigid and shallow.
  • Unstructured interviews resemble a guided conversation; they yield deep, valid, contextual insight favoured by interactionists and feminists such as Ann Oakley, but are time-consuming, hard to compare and vulnerable to interviewer bias.
  • Semi-structured interviews blend a common set of themes with freedom to probe.
  • The focus group interviews several participants together, allowing the researcher to observe how meanings are negotiated collectively — though dominant voices and group conformity can distort responses.

A recurring difficulty across all types is the interview effect — the interviewer's presence, status or manner shaping what is said.

Observation

Participant observation immerses the researcher in the group being studied, either overtly or covertly. Its unrivalled strength is verstehen — a first-hand, empathetic grasp of social life as members experience it, high in validity and capable of discovering the unexpected. Its costs are low reliability, poor representativeness, the ethical hazards of covert work and the risk of going native. Non-participant observation keeps the researcher detached, often using a structured schedule to count behaviours; it is more systematic and reliable but sacrifices depth and may provoke the Hawthorne effect, where people alter conduct because they know they are watched. For the wider debate on positivism and interpretivism that frames these choices, see research methods in sociology and the entry on verstehen.

Case study and content analysis

The case study examines a single unit — a person, group, event or institution — in exhaustive depth. It generates rich hypotheses and holistic understanding and is invaluable where the phenomenon is rare, but findings cannot readily be generalised. Content analysis systematically examines communication — newspapers, advertisements, textbooks, social media. In its quantitative form it counts the frequency of categories, achieving high reliability; in its qualitative form it interprets latent meaning, achieving greater validity. It is unobtrusive and works well on secondary material, but coding categories inevitably reflect the researcher's own assumptions.

Triangulation

Because every technique trades off breadth against depth, and reliability against validity, sociologists increasingly combine methods — triangulation — using one to check and enrich another. A survey may map a pattern that participant observation then explains, marrying the positivist demand for generalisation with the interpretivist demand for meaning.

How to use this in the exam

Examiners reward candidates who can match technique to purpose rather than list definitions. Structure an answer around the twin axes of reliability and validity, and illustrate each technique with a named study — Durkheim, Oakley, Whyte, Goffman. When a question asks you to evaluate, always close with the trade-off and the case for triangulation, showing that methodological choice follows from theoretical stance. A crisp two-line strength-and-limitation for each technique, deployed as evidence, will lift a competent answer into the top band.