The language investigation: framing a focused research question, collecting and handling data ethically, applying analytical methods, and writing up findings.

What this dot point is asking

Edexcel wants you to plan and write the coursework language investigation: a focused, independent study of real language data in which you frame a narrow research question, gather data ethically, apply analytical frameworks rigorously, and write up findings that reach an evidence-led conclusion. It is the research half of the non-exam assessment, and it is the place where the analytical toolkit from the rest of the course is put to independent work. The question being tested is not "do you know about language and gender" but "can you turn a curiosity into a rigorous, answerable investigation".

The answer

A successful investigation is built as a small piece of empirical research. You choose a narrow research question in an area of language study (variation, change, gender and power, occupation, child language, or the language of a genre), collect authentic data ethically, and apply the language levels plus relevant theory as the analytical frameworks. The write-up follows a clear research structure (introduction and aim, methodology, analysis, conclusion), and the marks reward a tight focus, a systematic method, and analysis that reaches conclusions rather than narrating the data.

Framing the question

The most common cause of a weak investigation is a question that is too broad. "Language and power" or "how children learn to talk" are fields, not questions; they cannot be answered with a small data set. The fix is to narrow on three axes at once: a single feature (tag questions, interruptions, lexical density, modal verbs), a defined context (one radio interview, one set of estate-agent listings, one parent-child play session), and a clear comparison or pattern to test (does the feature differ between speakers, registers, or time periods).

A well-formed question is a hypothesis in disguise. "How do tag questions function in this radio interview?" can be sharpened to "Do the host's tag questions serve facilitative or epistemic functions, and does this differ from the guest's?" Now the data has something to confirm or disconfirm.

Collecting and handling data ethically

Data must be authentic (real language, not invented) and gathered ethically. The non-negotiables are informed consent (participants know they are being recorded and how the data will be used), anonymisation (no names or identifying details in the transcript), and special care with vulnerable participants, above all children, where consent comes from a parent or guardian.

The analytical pipeline: from data to conclusion

A rigorous investigation moves through a fixed pipeline. First, quantify: count the target feature and tabulate it, giving an objective evidence base (for example, hedges per 100 words by speaker). Second, analyse qualitatively: use the language levels to explain the pattern, quoting representative examples and naming the features precisely. Third, interpret through theory: ask whether the data supports, complicates or contradicts the theory you set out to test. Fourth, conclude and evaluate: answer the question directly, and state honestly what the data cannot show.

Writing it up

The write-up mirrors a research report: an introduction stating the aim and question and reviewing the relevant theory; a methodology justifying the data and method and addressing ethics; an analysis working systematically through the evidence with quantitative support and language-level commentary; and a conclusion that answers the question and evaluates limitations. Throughout, the prose should analyse and argue, not describe.

Examples in context

An occupational-register investigation. A student investigates "how does estate-agent property-listing language construct desirability?" using twenty online listings as data. The quantitative stage counts evaluative pre-modifiers ("stunning", "deceptively spacious", "characterful") and euphemisms ("compact", "in need of modernisation"); the qualitative stage analyses these at the lexis and pragmatics levels, showing how euphemism manages negative information through implicature. The investigation interprets the pattern through Fairclough's concept of synthetic personalisation (the "your dream home" address) and concludes that the register persuades by combining hyperbole with deniable euphemism, while noting the limit that twenty listings from one website cannot represent the whole genre.

A child-language investigation. A student records a three-year-old over two play sessions to investigate overregularisation. The data shows forms like "goed" and "foots". The analysis applies the grammar and morphology level and interprets the errors through Chomsky's notion of an innate rule-forming capacity and the U-shaped learning curve, arguing the errors are evidence of rule application, not failure. The conclusion answers the question (the child is overgeneralising the regular past-tense and plural rules) and evaluates the limits honestly: one child, two sessions, and a setting that may not be fully naturalistic.

Try this

Q1. Why must a language-investigation research question be narrow? [2 marks]

• What the marker wants. That a narrow, single-feature, single-context focus allows depth and a genuinely answerable hypothesis within the word limit, where a broad field cannot be analysed with a small data set.

Q2. State two ethical requirements and one methodological limitation typical of an investigation. [3 marks]

• What the marker wants. Ethics: informed consent and anonymisation (especially with children); limitation: small sample size, single context, or the observer's paradox.

Q3. Explain how an investigation should move from data to an evaluated conclusion. [16 marks]

• What the marker wants. The pipeline: quantify and tabulate, analyse qualitatively with the language levels, interpret through named theory as hypothesis-testing, then conclude by answering the question and evaluating limitations honestly.

A note on sources

This guide is AI-written and not individually human-reviewed. It reflects the Pearson Edexcel A-Level English Language (9EN0) non-exam assessment requirements and published examiner guidance. Verify current coursework rules, word limits and ethical requirements against the official Pearson specification before relying on it.