You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
The analysis section is the heart of your language investigation. It is where you apply linguistic methods to your own data, and it is where the bulk of the AO1, AO2 and AO3 marks are earned. The defining feature of strong analysis is that it goes beyond describing what you found — it interprets the data, links it cautiously to relevant concepts and prior research, and evaluates what the findings do and do not show. This lesson covers how to structure the report within ~2,000 words, how to analyse data systematically at multiple language levels, how to integrate theory without "bolting it on," how to present data and reference sources, and how to draw the cautious, evidence-based conclusions that distinguish top-band work.
The AQA specification does not impose a rigid template, but the following structure mirrors academic linguistic writing and works well within the ~2,000-word allowance (your data sit in appendices and do not count toward it).
| Section | Indicative words | Purpose |
|---|---|---|
| Introduction | 200–300 | Introduce the topic, state the research question or hypothesis, give the linguistic context, explain why it matters |
| Methodology | 300–400 | How you collected the data, why that method, and the ethics |
| Analysis | 900–1,100 | Present and analyse the data in depth, across language levels, linked to theory |
| Conclusion | 200–300 | Summarise findings, evaluate the investigation, suggest further research |
Key Definition: Language levels — the layers at which language can be analysed: lexis and semantics (vocabulary and meaning), grammar (morphology and syntax), phonology (sounds, including prosody), pragmatics (meaning in context), discourse (the structure and organisation of whole texts or interactions), and graphology (visual presentation). A top-band investigation works at several of these levels, not one.
A good introduction does four things, efficiently:
This investigation examines the politeness strategies used by male and female participants in a small mixed-gender study group. Drawing on Brown and Levinson's (1987) politeness theory, it tests the hypothesis that female participants will mitigate disagreement with more positive-politeness and hedging strategies, while male participants will more often produce bald, on-record face-threatening acts. It builds on Holmes's (1995) argument that hedging frequently serves a relational rather than a deficit function, and remains alert to Cameron's (2007) caution that gendered patterns are performed in context rather than fixed traits.
This states the topic, the hypothesis and the theoretical frame in a few sentences, attributes each idea accurately, and signals from the outset the cautious, theory-aware stance the analysis will sustain.
The commonest weakness in student work is feature-spotting — noticing things in the data but not analysing them. The remedy is to work systematically across several language levels, always asking "so what does this do?"
Top-band investigations interpret data through relevant frameworks; they do not park theory in a separate section or list theorists for show. The strongest writing weaves a concept into the very sentence that analyses a feature — and then evaluates the fit.
Weak (feature-spot plus name-drop):
Speaker A uses "perhaps" three times. Lakoff said women use more hedging.
Strong (feature → function → theory → evaluation):
Speaker A's repeated epistemic hedge "perhaps" might at first appear to confirm Lakoff's (1975) characterisation of women's language as tentative. Closer reading of the surrounding turns, however, shows "perhaps" prefacing disagreements with the tutor — softening a face-threatening act rather than signalling genuine uncertainty. This supports Holmes's (1995) relational account of hedging over a straightforward deficit reading, and is consistent with Cameron's (2007) view that such features are strategic performances rather than fixed markers of gender.
The strong version identifies the feature, explains its function in context (AO3), connects it to accurately attributed theory (AO2), and evaluates the connection rather than asserting it. That move — from "what" to "so what," tested against theory — is the single clearest marker of a high-band investigation.
Important — fabrication guard: attribute every concept to the right scholar and only to claims they actually made. If you are unsure who originated an idea, attribute it cautiously ("often associated with…") or omit the name and analyse the feature on its own terms. A confident but wrong attribution is worse than none.
The phrase "be more analytical" is easy to give and hard to act on, so it helps to see analysis as a ladder with distinct rungs. The same observation can be expressed at four ascending levels, and you can train yourself to keep climbing until you reach the top two — the only rungs that earn high-band credit.
| Rung | What it does | Example on the same feature |
|---|---|---|
| 1. Identify | Names a feature (feature-spotting) | "Speaker B uses the imperative 'sit down'." |
| 2. Describe | Adds detail about the feature | "Speaker B uses several bald imperatives such as 'sit down' and 'open your books'." |
| 3. Analyse | Explains the feature's function in context | "These bald imperatives assert the teacher's authority directly, leaving the pupil no face-saving room to refuse and so foregrounding the asymmetry of the classroom." |
| 4. Evaluate | Tests the reading against theory and weighs alternatives | "This direct, on-record strategy fits Brown and Levinson's account of an unmitigated face-threatening act, though its acceptability here suggests that institutional role licenses directness that would be impolite between equals — a reminder that politeness is context-relative, not absolute." |
Most students stop at rung 2, occasionally reaching rung 3 for a sentence or two. The discriminator that separates a top-band investigation is habitually reaching rungs 3 and 4 — turning every salient feature into an account of what it does and how that reading stands up to scrutiny. A useful self-check during drafting is to read each analytical sentence and ask, "which rung is this?" If the answer is 1 or 2, push it upward by adding "which functions to…" (rung 3) and then "though this might also be read as…" or "consistent with X's claim that…" (rung 4).
Coursework Tip: Climbing the ladder does not mean writing about more features; it means writing more deeply about fewer. It is far better to take three or four well-chosen features all the way to rung 4 than to list a dozen features that never rise above rung 2. Depth, not coverage, is the currency of the high bands.
A recurring question in the analysis section is how to combine the numbers you have counted with the interpretation you want to offer, because handled clumsily the two pull apart: the report becomes a frequency table followed by some unrelated commentary, or a stream of interpretation with no evidential backbone. The strongest analyses integrate the two so that each supports the other — the quantitative locating a pattern, the qualitative explaining what it means.
The basic move is a three-step rhythm repeated through the analysis. First, establish the pattern quantitatively: "across the two extracts, the tutor produced 24 directives, of which 18 were mitigated as interrogatives." Second, illustrate it with a specific, quoted example: "for instance, rather than 'open your books' the tutor asks 'shall we open our books?'" Third, interpret the example's function and test it against theory: "this preference for the interrogative form preserves the pupil's negative face (Brown and Levinson, 1987) while still securing compliance — authority exercised through mitigation rather than command." Numbers, evidence and interpretation arrive as a single integrated unit, and the reader can see both that the pattern holds and why it matters.
This rhythm also protects you from the two opposite failures. A report that is all numbers — table after table with thin commentary — satisfies the descriptive part of AO1 but never reaches the analytical heart of AO3, because it never asks what the counts mean. A report that is all interpretation — confident claims about face-work and power with no counts or quoted data beneath them — reads as assertion rather than analysis, because the reader cannot see the evidence the claims rest on. Interleaving the two, claim by claim, is what makes an analysis feel both grounded and insightful.
A further refinement is to let the qualitative evidence complicate the quantitative where the data warrant it, because real data rarely fall into tidy categories. Suppose your counts show the tutor's directives are "mostly mitigated," but close reading reveals that the unmitigated ones cluster at moments of misbehaviour. The number alone would flatten that; the qualitative reading restores it, yielding a richer claim — "directives are mitigated by default but become bald precisely when behaviour, rather than learning, is at stake, suggesting the teacher reserves overt face-threat for discipline." That interplay, where the qualitative reading nuances rather than merely echoes the quantitative one, is exactly the kind of analytical sophistication that lifts a piece into the top band.
Coursework Tip: As you draft each analytical paragraph, check that it contains all three elements — a pattern, a quoted example, and an interpretation — and that the numbers and the prose are talking to each other rather than sitting in separate sections. If a paragraph has a count but no example, or an interpretation but no evidence, it is not yet doing the integrated work the band descriptors reward.
To see multi-level analysis in action, consider a short, invented transcript extract from a study of classroom power. (Your own data would, of course, be real and ethically collected; this is a teaching illustration.)
T: right (.) so who can tell me what a SUBORDINATE clause is (1.0) anyone (.) P3? P3: is it um (.) like a bit you can take out= T: =good (.) it's a START but I want the proper TERM
A single observation can now be analysed across several levels at once, which is precisely what lifts work above lexis-only treatments:
A strong write-up would then integrate and evaluate: the convergence of turn-control, directive grammar and evaluative feedback all points to authority exercised through routine discourse structure rather than overt command — a textbook instance of Fairclough's "power in discourse" — while the latching ("=") and the rapid "good… but" keep the exchange brisk and teacher-driven. It would close with a note of caution: a three-line extract can illustrate a mechanism but cannot establish how typical it is, even of this one teacher.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.