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 final stage — preparing your "Language in Action" folder for submission — matters as much as the research, writing and analysis that came before it. A polished submission signals the professionalism, attention to detail and self-discipline that the level descriptors reward. Keep the frame in view as you finish: the NEA is worth 100 marks and 20% of the A-level, totals 3,500 words, and is assessed by your teachers and moderated by AQA. It comprises two parts — a language investigation of ~2,000 words excluding data, and original writing with a reflective commentary totalling 1,500 words combined (the commentary is part of that 1,500, not a separate 750-word piece). This lesson is a thorough guide to editing, proofreading, formatting, referencing and submitting the folder, with a final-error checklist — all anchored to the verified word allowances so you do not finish strong on the wrong target.
These are different jobs, done in order. Editing makes substantive changes to content, structure and expression. Proofreading is the final sweep for surface error. Edit first; proofread last.
| Question | What to look for |
|---|---|
| Is the research question clear in the introduction? | The reader should know exactly what you investigate by the end of the first paragraph |
| Does the methodology explain and justify the approach? | Not just what you did but why |
| Does the analysis cover multiple language levels? | Lexis, grammar, pragmatics, discourse — at minimum |
| Is theory integrated throughout? | Used as you analyse, not bolted on at the end |
| Are data examples well chosen and clearly presented? | Each followed by analysis, not left to speak for itself |
| Does the conclusion answer the research question? | It summarises findings; it does not introduce new analysis |
| Have you evaluated honestly? | Limitations and alternative readings acknowledged |
| Is every linguistic term accurate? | Double-check any you are unsure of — a wrong term costs AO1 |
| Question | What to look for |
|---|---|
| Does the piece demonstrate genre conventions? | Would a reader recognise the genre unprompted? |
| Is the voice consistent? | Any jarring shifts of register or tone? |
| Is sentence structure varied? | Monotony — long or short — is a common weakness |
| Does every paragraph earn its place? | Cut anything not contributing to the whole |
| Is the opening engaging and the ending resonant? | Hook at the start; a lasting impression at the close |
| Are word choices precise? | Could any word be replaced with a sharper one? |
| Question | What to look for |
|---|---|
| Accurate metalanguage throughout (AO1)? | Specific labels, not "I used descriptive language" |
| A genuine, specific style-model link? | Named techniques, not "I was inspired by..." |
| Audience and purpose shown, not just stated (AO3)? | Explain how they shaped particular choices |
| Multi-level analysis? | Lexis, grammar, discourse and, if relevant, phonology |
| Analysis over recap? | Focus on HOW and WHY, never WHAT |
After editing, proofread carefully. Errors in spelling, punctuation and grammar weaken your written expression — which the AO1 descriptors reward in both the investigation and the commentary — and dent the whole folder's quality.
| Common error | Example | Correction |
|---|---|---|
| its / it's | "The language lost it's formality" | "...lost its formality" (possessive takes no apostrophe) |
| their / there / they're | "Their are many examples" | "There are many examples" |
| affect / effect | "This had a big affect" | "...a big effect" (usually: effect = noun, affect = verb) |
| comma splice | "The data shows a pattern, this fits the theory" | Use a semicolon, full stop or connective |
| apostrophe slips | "The speaker's use hedging" | "The speakers use hedging" or "the speaker's use of hedging" |
| subject–verb agreement | "the data shows" vs "the data show" | Both are defensible; be consistent |
| dangling modifier | "Having analysed the data, the results were interesting" | "Having analysed the data, I found the results interesting" |
Coursework Tip: Ask a fresh reader — parent, sibling, friend — to proofread. They will catch errors you have gone blind to. But they must only flag errors, never rewrite: the folder must be your own work, and your teacher has to be able to authenticate it as such.
Key Definition: Authentication — the requirement that your centre confirms the NEA is your own independent work. Wide reading and accurate citation support it; unattributed borrowing or having others redraft passages for you undermine it and can jeopardise the qualification.
Clear presentation aids readability and creates a positive impression for the moderator, who must navigate your folder without ever having met you. There is no single mandated format, but these conventions help.
Your investigation must include a bibliography of every source you reference, and it does not count toward your word allowance. A consistent author–date system such as Harvard works well.
| Source type | Format |
|---|---|
| Book | Cameron, D. (2007) The Myth of Mars and Venus. Oxford: Oxford University Press. |
| Journal article | Holmes, J. (1995) 'Women, men and politeness strategies', Journal of Pragmatics, 23(4), pp. 461–483. |
| Website | Crystal, D. (2011) 'Internet linguistics'. Available at: (URL) (Accessed: 15 January 2026). |
| Chapter in an edited book | Lakoff, R. (1975) 'Language and Woman's Place', in Thorne, B. and Henley, N. (eds.) Language and Sex. Rowley, MA: Newbury House, pp. 45–79. |
Key Definition: Harvard referencing — an author–date system in which sources are cited in the text by surname and year (e.g. Cameron, 2007) and listed in full in a bibliography. Accurate referencing both supports the conceptual engagement the AOs reward and evidences that the folder is authentically yours.
Important — verify before you cite. Reference only sources you have genuinely consulted, and attribute concepts to the right scholar. Do not pad a bibliography with works you have not read or invent plausible-looking entries: a fabricated or mis-attributed citation is an academic-integrity risk and is easily spotted. If you are unsure who originated an idea, hedge ("an idea often associated with...") rather than guess confidently.
Observing the allowances is itself a sign of the control the descriptors reward. The verified figures are:
| Part | Word allowance | What counts | What is excluded |
|---|---|---|---|
| Language investigation | ~2,000 words | Introduction, methodology, analysis, conclusion | Data, transcripts, appendices, consent forms, bibliography |
| Original writing + commentary | 1,500 words combined | The writing and the commentary together (a typical split is ~1,000 + ~500) | — |
State your word count at the end of each part. A small margin of flexibility is normal, but big deviations signal weak control: a folder well over length usually pads description or fails to push raw data into appendices; one well under length usually signals thin analysis or development. Note especially that the original writing and commentary share a single 1,500-word envelope — there is no separate 750-word commentary allowance.
Coursework Tip: Decide early what is "report" and what is "data," because mis-sorting them is the commonest word-count error. Quote only the short extracts you actually analyse in the investigation body; send full transcripts to an appendix and refer to them by line or item number. That keeps your ~2,000 words for analysis, where the marks are.
Confirm your centre's exact requirements, but a complete folder typically contains:
Most centres accept digital submission, but check. Whether digital or paper:
| Error | Consequence | Prevention |
|---|---|---|
| Submitting without proofreading | Surface errors weaken written expression (AO1) | Proofread several times, using different strategies |
| Missing or thin bibliography | Weakens conceptual engagement; raises integrity concerns | Compile references as you write, not at the end |
| One-language-level analysis | Limits depth; weakens AO1 and AO3 | Work systematically through several levels |
| Commentary without metalanguage | Loses the AO1 dimension the commentary draws on | Ensure each paragraph uses at least one precise term |
| Treating the commentary as a separate 750-word task | Mis-targets the allowance; risks overrunning | Plan the writing + commentary to 1,500 words combined |
| Inconsistent formatting | Poor impression; harder for the moderator | Use a template; check before submitting |
| No stated word count | Looks unprofessional | Add counts to the end of each part |
| Plagiarism or undue help | Threatens authentication and the qualification | Keep the work your own; reference every source |
It helps to see the final review in action rather than as an abstract list. Imagine a candidate, two weeks before their centre's deadline, opening a near-complete folder.
They begin with the word allowances, because mis-targeting these is the commonest avoidable error. The investigation runs to 2,180 words — comfortably near ~2,000 once they confirm that the two long transcript stretches still sitting in the analysis section are moved to an appendix, where they belong and do not count. The original writing and commentary together total 1,610 words; a quick trim of an overwritten paragraph in the story brings the combined figure back toward 1,500. Crucially, the candidate does not try to hit a separate "750-word commentary" target — there is no such target; the writing and commentary share one envelope.
Next they read the level descriptors with their own work open alongside, asking of each section not "did I work hard?" but "can a stranger see on the page that this meets the band I want?" They notice the conclusion currently only summarises findings; they push two sentences up the ladder to evaluate, acknowledging a limitation and an alternative reading. They check the commentary completes the feature–effect–justification move rather than merely naming techniques, and tighten one paragraph that had slipped into plot recap.
They then verify the bibliography — every in-text citation appears in the list, the formatting is consistent, and, importantly, no entry has been padded with a work they did not actually read. They confirm the appendices are labelled and referred to in the main text, that candidate and centre numbers sit on every page, and that the style model is included or clearly referenced. Finally they proofread aloud, on paper, catching two comma splices and a missing word that the screen had hidden.
Coursework Tip: Do this walkthrough at least a week before the deadline, not the night before. Every problem it surfaces — a bloated word count, a descriptive conclusion, an unlabelled appendix — takes time to fix properly. The candidates who score highest are almost never the most gifted; they are the ones who left themselves room to act on their own final review.
When a draft runs over its allowance — the ~2,000-word investigation or the 1,500-word writing-plus-commentary — the instinct is often to cut the wrong things: to trim the analysis or the model-answer depth that actually earns the marks, while leaving the padding that caused the overrun. Disciplined cutting targets the padding first.
In the investigation, the most reliable savings come from moving raw data into appendices (it should never have been in the body), deleting summary that merely restates rather than analyses, and removing an over-long introduction or methodology that has crowded the analysis. What you should not sacrifice is the depth of the analysis itself — the multi-level engagement, the integrated theory, the evaluation — because that is precisely where the marks live. If the analysis is strong and the piece is still over length, look again at the description-to-analysis ratio: descriptive sentences that name without analysing are both weak and expendable.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.