A-Level English Language NEA: planning and writing the language investigation

What the investigation is marked on (AQA 7702)

From the AQA 7701/7702 specification. The other 50 NEA marks are for original writing and its commentary. OCR H470 marks its report and poster against its own criteria.

What a good investigation looks like

AQA suggests four kinds of question: genre-based (what are the distinctive features of this type of language?), function-based (what is the language used to do?), attitudes-based (how do people feel about this language?) and user-based (who uses this type of language?). Its example areas include children's language, gendered talk, dialect, the language of new technologies, and code switching between English and other languages, which suits many students in Gulf schools.

The best investigations are narrow. 'Language and gender' is a topic; 'How do boys and girls in one Year 12 group use hedges and tag questions in a group discussion?' is an investigation. Narrow questions give manageable data, clear methods and conclusions you can defend.

AQA sets out the sections: introduction (why this focus, and aims), methodology (an evaluative account of how data was collected and organised, and approaches to analysis), analysis (findings, concepts and issues, contextual influences), conclusion, references, and appendices with clean copies of the data. Students should collect their own data.

A route from idea to final draft

1. Choose an area you can access data for: family conversations, a class group chat (with permission), school newsletters, adverts, children's writing.
2. Write a research question with one variable and one population. Add two or three sub-questions.
3. Read two or three relevant studies or theories so your analysis has concepts to test.
4. Collect data ethically: informed consent, anonymised names, and enough data for patterns but not so much you cannot analyse it.
5. Choose frameworks before analysing (for example lexis, grammar, discourse, pragmatics) and count where counting helps.
6. Draft the analysis by findings, not by text, with quantitative and qualitative evidence.
7. Conclude cautiously: what your data shows, its limits, and what a further study would test.

Worked example: aim and an analysis paragraph

Aim (invented): 'This investigation examines when and why three bilingual Arabic-English speakers aged 16 switch into Arabic during informal conversation about school, testing whether switches cluster around emotional or culturally specific content.'

Analysis paragraph (on invented data): 'Of the 42 switches identified in 25 minutes of transcribed talk, 19 (45%) occurred in evaluative comments, often single words or set phrases expressing approval, surprise or reassurance. This pattern supports the view that bilingual speakers use the second language for affective meaning that feels less natural in the first. However, 11 switches (26%) involved nouns for which English has an equivalent, such as references to food and family members, which suggests a second function: marking shared cultural identity within the group. The context is significant here, since all three speakers attend an English-medium school, and the switches may signal in-group membership in a setting where Arabic is otherwise absent.'

Why it works: specific counts (AO1), a concept tested against data (AO2), contextual explanation (AO3), and cautious wording ('suggests', 'may signal') appropriate to a small sample.

Common mistakes

• A topic that is too broad to analyse in 2,000 words.
• Too much data, or data the student did not collect.
• Methodology that describes but does not evaluate (what went wrong, what was controlled).
• Analysis organised text by text instead of by finding.
• Theory summarised in a block instead of being tested against the data.
• Over-claiming: treating three speakers as proof of how everyone talks.
• Ignoring ethics: no consent, or real names in transcripts.

How one-to-one lessons help with the NEA

The NEA must be the student's own work, and teachers must be able to authenticate it, so a tutor's role is to teach the method, not to write or edit the investigation. Within those rules, lessons are useful at three points. Early on, the tutor helps the student narrow a topic into a workable question and choose analytical frameworks. During analysis, the tutor teaches the terminology and concepts the student needs, using practice data rather than the student's own.

Later, the tutor runs sessions on academic writing: how to evaluate a method, how to hedge conclusions, how to integrate a theory. The student applies those skills to their own draft, and the school's rules on feedback decide what the teacher may comment on.

Self-check

• Can I state my research question in one sentence with a clear variable and population?
• Did I collect the data myself, with consent and anonymity?
• Does my methodology evaluate as well as describe?
• Is my analysis organised by findings, with numbers where useful?
• Do I test at least one concept or study against my data?
• Are my conclusions cautious and linked back to the aim?
• Am I within the word count (AQA 2,000 excluding data)?