We've delved into the examiner reports for Papers 1 and 2 – but what about the NEA?

AQA's 2026 Principal Moderators' Report for the Independent Investigation contains a huge amount of useful feedback for teachers and students. Unlike an examiner report focused on individual questions, this gives us an insight into the whole NEA journey: from choosing a title and developing hypotheses, through methods and data analysis, to evaluation and conclusions.

And there is plenty to be positive about.

AQA reports that the overall standard remains broadly comparable with recent years, with encouraging improvements in the clarity of aims, justification of methods and integration of theoretical concepts into analysis. Several students achieved the full 60 marks, while most produced viable investigations meeting the specification's minimum requirements.

At the top end, the characteristics are clear: strong conceptual understanding, well-justified methods and confident interpretation of data. At the lower end, three issues crop up repeatedly: insufficient data, unfocused aims and descriptive rather than analytical writing.

So, what can we learn from this year's moderation?

Overall summary

What was done well?

AQA highlights three particular strengths across the 2026 cohort:

• investigations were generally clearly structured and professionally presented
• most students could identify methodological limitations and propose improvements
• conclusions were generally strong and evidence-based, with the best synthesising findings effectively across the investigation.

Where did students struggle?

Again, there are some very clear themes:

• making meaningful links to wider geographical context
• demonstrating how their geographical understanding had improved
• understanding and implementing sampling strategies correctly
• collecting enough data to support meaningful analysis
• integrating geographical theory into analysis rather than confining it to the literature review.

Perhaps the biggest overarching message from the report is this - a successful NEA needs to work as one connected geographical investigation.

The title influences the hypotheses → the hypotheses determine the methods → the amount and quality of data affects the analysis →he theoretical framework established near the beginning needs to reappear when interpreting results → and all of that ultimately determines how convincing the conclusion can be.

Weakness early in the process can therefore have consequences much further down the line

Students need to start with a question that is actually investigable!

The good news is that most students produced clearly stated, appropriately focused aims and hypotheses.

AQA notes that the familiar model of one overarching aim supported by three sub-questions or hypotheses continues to work effectively, helping students maintain a clear line of enquiry.

Better titles and aims...

• were focused and manageable
• identified specific locations
• investigated measurable concepts
• were clearly grounded in geographical thinking
• created opportunities for analysis and evaluation.

AQA gives examples including: 'To what extent has gentrification altered the socio-economic profile of Stokes Croft, Bristol?' and investigations into vegetation, soil and biodiversity across a psammosere; environmental quality and lived experience across a rural-urban fringe; perceptions of safety between different groups; and relationships between building height, albedo and microclimate.

Weaker titles such as 'A study of sand dunes' or 'Environmental quality in my local area' identify a topic, but they don't really establish something to investigate.

There's a useful warning here about doing too much as well. One exemplar used five separate research questions. Although all were relevant, AQA suggests this was ambitious for a 3,000-4,000-word investigation and risked losing coherence.

So perhaps the mantra should be - focused enough to investigate deeply, broad enough to generate meaningful geography.

Don't just copy and paste the specification!

Most students linked their investigations to the specification, but AQA identifies an important difference between mentioning the specification and actually demonstrating how an enquiry connects to it. Some weaker responses essentially copied specification statements. Stronger work selected the geographical concepts that genuinely mattered to the investigation and explained the connection.

The psammosere exemplar is revealing here. The student identified relevant-sounding concepts including succession, coastal deposition and carbon cycling, but didn't explain why those ideas were important for understanding the particular biotic and abiotic variations being investigated. The result was descriptive rather than purposeful.

That's a useful distinction to make with students: 'This is on the specification' isn't the same as 'this geographical concept helps me investigate my question'.

The literature review

The literature review needs to do some work - this is one of the strongest messages in the whole report. Most students included a literature review or theoretical framework, but quality varied considerably...

Stronger literature reviews...

• used contemporary academic sources, journal articles and reputable materials
• explained relevant geographical concepts and theories
• connected theory explicitly to aims and hypotheses
• helped provide a rationale for the investigation
• considered relevant comparative context.

The strong exemplar on pages 11-12 is worth looking at. It doesn't simply report what different sources say. The student explains a regeneration framework and then applies it to Longbridge, alongside academic definitions, council documentation and the process of deindustrialisation. AQA describes it as a well-targeted section connecting the literature directly to the enquiry's aims.

Likewise, the exemplar on page 13 uses previous research into social inequality in coastal and inland towns as genuine comparative context, connecting those studies explicitly to the student's hypotheses.

Weaker literature reviews...

• became collections of background information
• relied heavily on textbooks or general descriptions
• summarised sources one after another
• used tables containing little more than 'source/what it says/link to my study'
• failed to establish geographical theory that could subsequently be applied.

The low-Level 2 exemplar on page 14 demonstrates the problem particularly clearly. It contains sources, but there's no synthesis of the ideas and little meaningful explanation of succession, soil development, vegetation zonation or geomorphological change.

And this isn't just an Area 1 problem - AQA explicitly notes that students who don't establish a strong theoretical framework here often subsequently struggle to justify methods in Area 2, interpret patterns in Area 3 and produce evaluative conclusions in Area 4. In other words, the literature review shouldn't be something students write and then forget about - it should give them geographical ideas they can use later.

Methods - could somebody actually replicate what they did?

There is plenty of positive feedback here - most students justified their methods appropriately and step-by-step descriptions were more common this year. The strongest work explained what was collected, where, when and how, providing sufficient detail for the investigation to be replicated. Annotated recording sheets, photographs of collection locations and explanations of measurements could all strengthen this.

But there are two important distinctions...

Firstly, saying 'I used this method because it helps answer Hypothesis 2' isn't necessarily sufficient justification - the strongest students also considered why a method would generate valid, reliable and representative data.

Secondly, methods need to be operational rather than simply named...

• The weaker exemplar on page 17 identifies several techniques but provides too little information on sampling, sample size, timing or frequency for someone to replicate the investigation. Crucially, there is also very little explanation of why each method is appropriate.
• One nice idea appears on pages 18-19: annotated recording sheets/questionnaires. Rather than simply inserting a blank questionnaire, the annotations explain how individual questions relate to hypotheses, socio-economic groups and expected patterns. This demonstrates that the instrument has been designed purposefully rather than assembled generically.

Sampling needs some attention!

This is probably one of the clearest practical teaching points from the report.

Students frequently...

• confused stratified and systematic sampling
• misapplied random sampling
• named a sampling technique without explaining how it was implemented
• failed to justify why that strategy was appropriate.

It's not enough for a methodology to contain the sentence 'I used systematic sampling' - students need to explain what they actually did and why that approach generated an appropriate sample. This is one area where a relatively small amount of explicit teaching before students begin their independent investigations could potentially make a substantial difference.

Data - quality vs quantity

AQA identifies data quantity as a significant issue - More data isn't automatically better, but too little data is definitely a problem. Some students attempted to build investigations around short transects, very small questionnaire samples or minimal sediment samples. The consequences appeared later: limited patterns, difficulty applying statistical tests and weaker conclusions. The interconnected nature of the NEA means that weak data collection puts a ceiling on what students can subsequently do with the data. AQA's advice for future cohorts explicitly recommends longer transects, larger questionnaire samples and more robust physical measurements where appropriate.

Presentation methods should be revealing!

Stronger students selected sophisticated and spatially explicit techniques including:

• proportional circles
• divided bar charts
• choropleth maps
• kite diagrams
• beach profiles
• scatter graphs with lines of best fit.

But the key isn't simply to use the fanciest graph possible - the best presentation methods were selected because they helped reveal patterns, relationships or spatial variation relevant to the investigation. The example on page 23 makes this very tangible - an ArcGIS map combines IMD, housing-quality and house-price data so that contrasts, clusters and an anomaly can be seen spatially. AQA's commentary emphasises that this communicates relationships more effectively than text alone. At the other end, repeated basic bar charts, pie charts and tables sometimes created repetition rather than additional insight.

So the question shouldn't be 'How many different presentation techniques have I used?' It should be 'What does this technique allow me to see?'

Statistics - use them because they're useful, not because it's an NEA!

Students need to be careful here - it is more important that their statistics are useful, not high in quantity because they think that's what an NEA should have. AQA reports effective use of Spearman's rank, Pearson's correlation coefficient, Chi-squared and, less frequently, Mann-Whitney U.

But there are some important caveats. Statistical tests were sometimes attempted with inappropriate sample sizes - particularly Spearman's rank with fewer than the recommended ten data pairs. Some students also omitted full workings or failed to test for significance.

The strong Chi-squared example on page 22 demonstrates what statistical testing should achieve. The student calculates observed and expected values, degrees of freedom and a critical value, before using the result to reach a justified decision. The statistic therefore moves the analysis beyond description.

Again, the message is purpose - don't include a statistical test to tick a box. Use one because it helps answer the geographical question.

Analysis was a strength - but description still isn't enough

AQA describes data interrogation as one of the strongest sections of the NEA.

Most students could describe patterns, identify trends and recognise anomalies. Higher-level responses went further by:

• explaining patterns
• linking findings together
• exploring unexpected results
• considering accuracy, reliability and representativeness
• cross-referencing primary and secondary data.

There's also a useful structural recommendation - some students organised their analysis by method – e.g. 'questionnaire results', followed by 'EQS results', followed by 'pedestrian counts', etc.

AQA suggests that organising it around hypotheses or sub-questions can produce a much more coherent line of enquiry because students can bring different datasets together to answer the same geographical question.

That feels like a potentially significant piece of advice - analyse the question, not the method!

Bring the theory back!

This is particularly important for those students aiming for the top grades

Area 3c (Application of Geographical Theory) remains challenging. Many students introduced geographical theory in Area 1 but then barely mentioned it again. Stronger students did something very different. They revisited the literature, using geographical concepts and theory to interpret the patterns they had actually found and comparing their results with established geographical understanding.

The Level 4 example on page 25 shows this really well. Rather than simply stating differences in EQS and building-decay scores, the student links those spatial patterns to deprivation, housing age and social-housing distribution, while applying ideas around 'Left Behind Places' and IMD.

Theory shouldn't therefore live in an isolated section called Literature Review. A strong NEA might work more like this: Theory → question → method → evidence → analysis → theory revisited → judgement.

'My geographical understanding has improved because...'

One particularly interesting point in the report concerns 3b(ii) - AQA identifies this as one of the least well-addressed strands: students often don't explicitly demonstrate how their geographical understanding has improved through undertaking the investigation. This isn't simply about saying "I learnt how to conduct a questionnaire." The focus is geographical understanding.

What does the evidence now make the student understand differently about regeneration, place identity, coastal processes, succession, deprivation, microclimate – or whatever geographical phenomenon they've investigated? This feels like something that could easily be overlooked unless students are explicitly encouraged to reflect upon it.

Evaluation - go beyond 'we should have collected more data'

Most students could identify methodological limitations and suggest improvements - that's encouraging. But the stronger evaluations went one important stage further: they considered what those limitations actually meant for the findings.

For example: Limitation → consequence for reliability/representativeness → impact on conclusion → realistic improvement. That's considerably more powerful than simply producing a table headed Limitation/ Improvement.

The Level 2 example on pages 28-29 demonstrates the difference. There is personal reflection, but many comments are generic rather than being connected directly to the methods and evidence. AQA says stronger evaluation would consider how reliability, representativeness and the methods chosen affected the strength of the conclusions.

Don't forget ethics

Ethical considerations were generally handled thoughtfully, particularly in human geography investigations. The stronger example on page 30 considers anonymity, voluntary participation, non-intrusive observation, environmental care and responsible use of public space - all in relation to the actual investigation. Ethics shouldn't be a generic paragraph that could be pasted into anybody's NEA. It needs to relate to what the student actually did.

Wider context - a big opportunity for improvement

Alongside theoretical application, this is one of the biggest weaknesses identified in the report. AQA says wider context remains the least well-addressed aspect of Area 4. Students often offered only brief or superficial comments about how their findings related to broader geographical issues or comparable locations.

The stronger partial evaluation on pages 30-31 shows what this might look like. The student connects their findings to broader environmental change and land-use pressures and suggests realistic future data collection that would strengthen reliability.

So students need to move from - 'This is what I found in my study area' towards 'What might my findings tell us about this geographical issue beyond my particular investigation?'

Conclusions should synthesise, not repeat

Conclusions were generally a strength. Higher-level students explicitly answered their research questions, judged the extent to which aims had been met and supported those judgements with evidence.

The Level 4 Bee Network exemplar across pages 32-34 is particularly instructive. It considers each hypothesis using quantitative and qualitative evidence – including questionnaires, observational counts, interviews and secondary sources - before reaching an overarching judgement grounded in the cumulative findings.

Contrast that with the low-Level 2 example. Each hypothesis is briefly addressed, but there's no final synthesis bringing the strands together into an overall judgement.

So the conclusion isn't simply: Hypothesis 1 → Hypothesis 2 → Hypothesis 3 → finished. Students need to return to the overall enquiry and ask what all of their evidence, taken together, allows them to conclude.

Independence really does matter

AQA's report also contains an important message for centres. Most students demonstrated appropriate independence, and moderators particularly praise some large centres for enabling substantial cohorts to produce imaginative, personalised investigations rooted in local issues. In other words, having a large cohort isn't an excuse for identical NEAs.

Concern arose where cohorts had very similar titles, methods or analytical approaches, sometimes differing only by coordinates or minor wording. Shared fieldwork is perfectly acceptable, but students still need to demonstrate their own decision-making. There's a similar warning about centre-provided methodology tables. These can help with structure but may restrict individuality and therefore limit access to the higher levels.

And a word on AI...

Perhaps unsurprisingly, AI appears explicitly in the 2026 report...

Malpractice cases remained low, but moderators did identify occasional concerns around suspected AI-generated text. Warning signs included generic phrasing, lack of specificity to the study area and inconsistencies between sections.

AQA also notes cases where AI appeared to have been used to produce background information or theoretical summaries. These were often overly general, poorly connected to the aims and resembled textbook explanations rather than applied geographical thinking.

AQA's position in the report is clear: AI may support understanding, but must not generate NEA content, and the submitted written report must be the student's own work. Interestingly, there's a wider geographical lesson here too. Generic writing isn't just an authenticity issue - it's often weak NEA writing anyway.

The best work in almost every section of this report is characterised by specificity: this place, this hypothesis, this method, this evidence, this geographical theory.

A few practical reminders for centres

There are some less glamorous (but important) messages from the moderation process too...

• Best practice for submission is one combined PDF per student, scanned in colour, clearly named, with a complete CRF and detailed commentary. Recurring problems included multiple files, black-and-white scans obscuring maps, upside-down or incomplete pages, work attached to the wrong student and missing CRFs.
• AQA also encourages clear annotation and CRF comments that identify where marks have been awarded and explain level decisions. Moderators found that some centres were over-generous at the top end, while others were over-harsh with weaker investigations.
• And one potentially useful time-saving reminder: risk assessments are not required for assessment purposes and shouldn't distract from clear methodological explanation.

So, what should we take into the classroom?

AQA finishes with a useful set of recommendations for future cohorts. Pulling those together with the detailed moderator commentary, these feel like the big messages for 2027:

• Get the title right at the start - a focused, viable, genuinely geographical question makes everything that follows easier.
• Keep it individual - shared fieldwork doesn't mean shared investigations.
• Make the literature review useful - students need theory they can subsequently apply, not pages of background information.
• Teach sampling properly - students need to know what random, systematic and stratified sampling actually mean, and how and why they used them.
• Collect enough data - a sophisticated analytical technique can't compensate for an inadequate dataset.
• Justify methods geographically - think validity, reliability and representativeness, not simply 'this answers my hypothesis'.
• Choose presentation techniques for a reason - maps, graphs and statistics should reveal something useful about the data.
• Analyse around hypotheses rather than methods - bring different sources of evidence together.
• Bring theory back into the analysis - Area 1 shouldn't be the last time geographical concepts appear.
• Explain how geographical understanding has developed - how has this changed over the investigation.
• Evaluate consequences, not just limitations - what does the weakness mean for the reliability of the evidence and confidence in the conclusion?
• Think beyond the study area - wider geographical context remains a major opportunity to access higher levels.
• Synthesise in the conclusion - answer the overall enquiry rather than simply repeating earlier analysis.
• Keep the NEA authentic - the specificity and individual decision-making that demonstrate independence are also characteristics of the strongest geographical investigations.

AQA's final advice places particular emphasis on proposal-stage guidance, methodological rigour, theoretical integration and evaluative depth, and gives us the central message from the 2026 NEA report: The strongest NEAs aren't collections of separate sections - they're coherent geographical arguments!

A good investigation has a thread running all the way through it. The question leads to the theory. The theory informs the methods. The methods generate appropriate evidence. The evidence is interrogated using geographical understanding. And the conclusion answers the question that started the whole process.

If students can maintain that thread - and keep asking 'how does this help me answer my geographical enquiry?' - they're much more likely to produce an NEA that feels purposeful, independent and genuinely geographical.