Secondary evidence, source criticism and mixed methods

By UK GCSE Sociology · independent study · myAudience

Original OpenAI imagegen conceptual teaching illustration, generated 4 October 2026; not observational evidence. · AI-generated original asset for this independent guide; no third-party artwork intentionally reproduced.

Secondary evidence, source criticism and mixed methods

Primary data are generated for the present investigation, such as a researcher's interviews. Secondary data already exist and are reused for a new question. The same material can be primary in the original project and secondary in a later one. Official statistics are produced by public bodies; non-official statistics can be produced by charities, companies or researchers. Those labels describe provenance, not automatic truth or falsehood. Evaluate definitions, coverage, collection, missing cases, purpose, date and comparability. A source suited to administering a service may be less suited to explaining experiences of people who never reach it.

The Office for National Statistics' current families and households methods guide is dated 17 April 2026 and describes estimates based on the Labour Force Survey household dataset. Its definitions and coverage matter: private-household estimates do not simply describe every possible living arrangement. The guide explains that communal establishments are outside the survey's coverage and discusses how students in halls are treated. These methodological facts illustrate source criticism; this lesson reproduces no national numerical estimates. An official family definition is useful for a consistent statistical purpose, but it is not a judgment about which relationships are genuine or valuable.

Administrative records can provide continuity and large volumes but reflect institutional recording rules. A school register records attendance categories; it may omit reasons or change practice over time. A charity's user survey may reveal experiences of those it serves but cannot automatically represent people outside the service. Documents, diaries, media and photographs can reveal meanings and representations, yet authors select what to record and audiences shape presentation. An image is not a transparent window into someone's beliefs. Distinguish what a source directly records, what its author interprets and what a later researcher infers.

Classic sociological arguments also need source criticism. Appendix B is an official guide to selected arguments, not a substitute for the entire original text or access to raw study data. When this batch explains Parsons or Ball, it uses those prescribed summaries and labels that evidence basis. It does not invent interview counts, quotations or additional findings. A study from one historical setting can generate a useful concept while having limited scope for claims about current institutions. Critique the sampling and social context where documented; when the method is not available, state that limitation rather than fabricate a flaw.

Mixed methods combine different kinds of evidence in a design. A survey can estimate defined responses within a sample; interviews can explore how respondents interpret the categories; observation can examine routines rather than recalled conduct alone. Triangulation compares evidence about a phenomenon from more than one source or approach. Agreement can strengthen an account if sources have partly independent limitations. Agreement from three reports copying the same record is not three independent confirmations. Disagreement may expose different definitions, periods or populations, or a genuinely different experience. It should prompt analysis rather than deletion of the inconvenient source.

Plan integration explicitly. For a fictional library-hours investigation, begin with a defined user population, examine administrative opening records, use a carefully sampled questionnaire about access, then select contrasting accounts for interviews. Explain why each stage answers a different aspect of the aim. An interview sample chosen for contrasting experiences is not a prevalence estimate. Publish bounded conclusions alongside the sources' limits. Methods become stronger when the design specifies how evidence will support or challenge a mechanism; merely adding another method can increase costs without answering the question. No private records or real participants are used in this lesson.

Worked example: Original fictional source bundle: an annual report says registered library visits increased; a charity's survey of its own clients reports difficulty travelling; three newspaper stories repeat the annual report. A council concludes that access improved for every resident. Critically synthesise the evidence and propose one useful additional source.

Separate visits from unique visitors and from all residents. Increased recorded visits could reflect repeated use by a subset or changed recording.

Assess the charity sample's scope. It may illuminate relevant barriers among clients without estimating their prevalence in the whole town.

Identify dependence: the newspaper stories repeat the same evidence and do not provide independent corroboration.

Propose a population-relevant source, such as a defensibly sampled access survey with non-user options, and explain how its findings would interact with the existing sources.

Answer: The annual report supports a claim about recorded visits if definitions are comparable, but it cannot show access improved for every resident. The charity survey offers evidence about barriers within its client group and could explain why some residents remain absent. Newspaper repetition adds visibility rather than independent evidence. A broader survey including non-users could examine whether opening times, travel or other conditions restrict access. A combined report should retain both increased use and remaining barriers if supported, rather than force the sources into a single universal claim.

Check your understanding: Extended transfer: a fictional researcher compares two school attendance reports from different years and interviews five learners selected for contrasting experiences. Write a source-criticism plan identifying three comparability checks, the proper role of the interviews and how a contradiction would be investigated. Indicative rubric: 12 points: checks 6, interview scope 3, contradiction analysis 3.

Answer: Check definitions of absence, denominators and populations, plus whether recording rules or periods changed. Use interviews to explore meanings and mechanisms, not to estimate a school-wide percentage. If an interview describes barriers while the report shows higher attendance, investigate whether these concern different groups or periods, and whether attendance conceals difficult journeys. Both can be true. Revise the conclusion to match the comparable evidence.

Why: Each check needs a reason: altered definitions can create artificial trends, altered populations change composition and altered periods affect comparison. A high-quality plan distinguishes evidence integration from majority voting among sources. An unexplained contradiction should remain unresolved in the report. The ONS example is a verified methodological illustration; every school and library bundle here is invented.