Fieldwork supports geographical understanding — Fieldwork connects theory with direct observation, measurement and situated experience outside the classroom. It can investigate physical or human processes using quantitative and qualitative approaches appropriate to the question. A visit alone does not establish the meaning of a pattern: preparation, systematic evidence, analysis and reflection turn experience into a geographical enquiry.
Four days span physical and human processes — The A-level course requires four days of fieldwork across physical and human geography, which can be arranged through full or part days under the centre's provision. This course requirement is distinct from the individual investigation's focus. Do not infer that all four days must be spent collecting the same investigation's data or that one online exercise certifies attendance.
The investigation needs a significant field element — The independent investigation may address human geography, physical geography or a combination and must include a significant element of fieldwork. Relevant secondary sources can support the study but do not remove that requirement. The centre confirms actual fieldwork arrangements; this revision resource provides general preparation rather than evidence that a learner has completed them.
The student defines the research question — An individual question or issue should connect with specification content, a theoretical or comparative context and a manageable location. The learner develops it rather than choosing a ready-made title supplied here. Broad themes can support exploration, while the final scope, aims and justified approach remain the student's work. This chapter supplies no candidate-specific title or model report.
Independence applies to more than final wording — Students independently develop the question, source relevant research, contextualise and analyse findings, present evidence and draw conclusions under the assessment arrangements. Copying a peer's structure and merely changing place names does not establish independent reasoning. Keep an honest record of how decisions and interpretations developed.
Collaboration has permitted boundaries — Exploring possible broad focuses and collecting field evidence can be collaborative under AQA's arrangements. Each learner must still independently contextualise, analyse and report their work, with an individual title. Shared collection does not authorise jointly written analysis or exchanged independently sourced materials. Follow the centre's actual instructions and record the origin of collaborative observations.
General guidance differs from supplied assessed work — Teachers can discuss broad feasibility, methods, equipment, safety and ethics and review proposals within the rules. Additional assistance has recording and assessment consequences. General teaching of a technique is different from supplying a specific student's title, dataset, paragraph or improvements. Ask the centre how to record assistance and use approved current guidance; this resource does not certify an assessed draft.
Keep practice and authentic evidence separate — Invented numerical examples in these notes develop methods and must remain labelled practice. They cannot be relabelled as a real local measurement or submitted as personal field observations. Document actual sources, collection roles, dates and decisions for genuine work. If planned collection fails, report the gap and discuss feasible next steps rather than fabricating a complete-looking table.
Framing, theory and feasible design
A focused enquiry connects concept and evidence — A manageable enquiry defines what is being investigated, where, at what scale and through which observable or interpretable evidence. A broad topic can contain several processes that cannot all be tested in the available time. Refine the conceptual relationship and evidence needs without assuming a particular expected answer. A question should allow a reasoned conclusion even if the initial expectation is not supported.
Theory gives a reason to investigate — Relevant theory, concepts or comparative research help explain why a relationship or meaning might be expected and what alternatives matter. Use literature critically rather than adding names without applying ideas. A local study can assess how far a general account fits its conditions, while recognising that a model simplifies reality and may not predict every observation.
A hypothesis is not always the right format — A testable prediction can suit measured relationships, but exploratory or interpretive questions about meaning may need open aims rather than a forced numerical hypothesis. If using a hypothesis, define variables and conditions and distinguish it from the statistical null used later. Choose an approach that fits the enquiry rather than treating all geographical knowledge as one correlation exercise.
Literature research needs provenance and selection — Record full source details, relevant claims, methods, date, location and limitations. Compare more than one credible perspective and explain relevance to the chosen context. Quoting a general process does not establish its local operation; a promotional source may reveal an actor's view without proving the claimed outcome. Avoid treating search order as a ranking of evidence quality.
Location needs a geographical justification — Explain how the actual setting relates to the process, contrast, access and scale of the question. Map the study area with suitable context and boundaries, while protecting sensitive sites or participants. A location chosen only because it is convenient may still be usable if the resulting limitations are acknowledged, but convenience alone does not show that it tests the intended relationship well.
Operationalise concepts without losing meaning — Define what each measured variable, score or interview theme represents and why the method addresses it. A surface infiltration measurement, for example, is not automatically the total flood risk of a whole basin. A place image score is not every aspect of lived well-being. A proxy is an indirect indicator used to stand for something harder to measure. Keep the limits between proxy and concept explicit and use complementary evidence where justified.
Feasibility includes access timing and resources — Assess permissions, transport, equipment, available time, season, likely conditions, participant access and analysis demands. A technically impressive design may be unsuitable if it cannot be implemented safely or produce interpretable evidence. Plan reasonable contingencies without changing the question after seeing results merely to guarantee a preferred finding.
Pilot findings should inform justified revisions — A small trial can reveal inaccessible sites, ambiguous questions, instrument problems or a sample interval mismatched to the phenomenon. Record what changed and why before full collection where possible. A pilot is not a promise that the final design is perfect. Distinguish genuine methodological learning from deleting an inconvenient result after the fact.
Collection, sampling and practical field quality
Match collection to the actual process — Select observations and methods that can address the question at its spatial and temporal scale. Physical processes may need repeated instrument readings and environmental context; human meanings may need interviews, observation and representations. Use an appropriate combination rather than assuming more instruments or more questionnaires automatically mean stronger evidence.
Justify the sample population and allocation — Define the population of sites, times, people or events and explain the frame, selection method and sample size. Random, systematic, stratified and purposeful approaches have different uses and limits. Equal numbers in two groups can aid comparison without representing their true population shares. Explain what the sample can support and whose experience or conditions it omits.This schematic shows an available frame containing only part of a target population, with a sample selected inside that frame. Units outside the frame cannot be selected by repeating selection from the same list. Box sizes do not represent measured proportions; real frames can also contain ineligible units that require separate checks. It is general practice teaching, not a candidate-specific sampling plan.
Frequency and timing need a mechanism — Choose times and intervals that capture relevant variation, such as a changing flow, seasonal conditions, commuter activity or different user groups. A brief observation at one hour may be appropriate for a narrow question but not an all-day or all-season claim. Record exceptional conditions and avoid treating a convenient time as automatically typical.
Repeated readings need consistent procedures — Use clear instructions, comparable equipment positions and standardised recording where consistency matters. Repeats can characterise variation and reduce some random uncertainty; calibration and observer agreement address different issues. A repeat at the same site is not a new independent location. Preserve units, instrument resolution and any method change in the record.
Physical measurements require contextual limitations — A channel cross-section and velocity estimate can inform discharge only under its method's assumptions; a float speed may differ from depth-mean velocity. A soil test samples particular depth, area and conditions rather than the whole catchment. Explain the technique actually used, its calibration and environmental limits instead of applying a generic correction without justification.
Human evidence requires a clear relationship with participants — Explain the purpose, voluntary nature, relevant information use and arrangements for privacy. Use accessible, non-leading questions and allow non-response without pressure. Recruitment through one place or time can exclude groups with different routines. An interview's depth and a survey's coverage are different strengths; neither can be assumed from sample count alone.
Observation and recording should preserve context — Record location, date, time, conditions, method, instrument, observer and relevant notes alongside the value or account. Retain unusual or missing results with explanations where available. Photographs and sketches need viewpoint and annotation. A clean spreadsheet made later should be traceable to authentic field records rather than replace their provenance.
Group collection needs transparent roles — Record who gathered which observations and how consistency was maintained across observers or equipment. Check shared definitions before merging files and preserve individual provenance. Collaborative collection does not erase uncertainty or make every member's later analysis identical. The learner's independent interpretation must remain distinguishable from shared raw recording.
Safety and permissions shape the design — Use the centre's actual risk assessment, supervision and access arrangements, considering weather, water, traffic, terrain, lone work and participant situations relevant to the activity. Do not enter unsafe or unauthorised settings to preserve a planned sample. A cancelled or restricted observation is a methodological limitation to report and manage, not a reason to invent data or bypass precautions.
Analysis, interpretation and geographical explanation
Organise and check without concealing evidence — Preserve raw files, document cleaning, reconcile units and category definitions and check missing-value codes before calculation. Keep paired measurements aligned when sorting. Investigate implausible values against field notes; an extreme observation can be real and should not be removed solely because it weakens the anticipated relationship.
Choose presentation to support a specific claim — Use maps for spatial relationships, graphs for appropriate numerical patterns and contextualised quotations or representations for meanings. Include titles, scales, units, dates, sample bases and source details. A collection of attractive displays without explanation is not analysis. Refer to each display for the particular evidence it provides and the uncertainty it retains.
Analyse patterns before asserting causes — Describe the location, direction, magnitude, variability, timing and unusual cases supported by the data. Then connect these with relevant processes, literature and alternative explanations. Coincident change may have a common driver or result from sampling differences. A participant's causal account is one perspective to assess, not automatic proof of a whole-area mechanism.
Statistical tests must fit the design — Use methods appropriate to the variables, sample and independence assumptions, including the correct handling of ranks, categories and expected frequencies. Define the null, alternative, threshold and test variant. A significant association does not prove causation, and failing to reject does not prove no geographical relationship. Report magnitude and practical meaning as well as the decision.
Qualitative analysis needs transparent interpretation — Explain recruitment, transcription, coding and researcher position, preserve contradictory accounts and use extracts in context. Interpret how meanings relate to identity, history, access and representation rather than merely counting positive and negative words. A frequency within the sample is not a population percentage without a justified design.
Triangulation can reveal disagreement — Different methods may corroborate a pattern, illuminate different dimensions or conflict. Examine whether they measure the same concept, area and time before calling one wrong. A promotional image and a commuter interview can legitimately describe different experiences. Use disagreement to refine the explanation and its limits rather than forcing a single uniform narrative.
Apply theory critically to the actual context — Explain which observations support, qualify or challenge the theoretical expectation and why scale, history, management or environment might matter. A theory's name is not itself analysis, and one local exception does not necessarily refute every version of a broad model. Connect the actual evidence with its relevant assumptions and wider geographical significance.
Conclusions, evaluation, presentation and authenticity
Answer the actual question with weighted evidence — A conclusion should address the enquiry's scope and use the strongest relevant evidence rather than simply repeat every result. State whether support is mixed, limited or conditional and distinguish an unsupported expectation from a failed investigation. Evidence against an initial idea can produce a strong geographical conclusion if interpreted honestly.
Separate finding, interpretation and wider inference — A measured pattern describes the sampled context; an explanation links it with a process; a broader inference extends beyond the observations. Each requires additional justification. State the places, people, times and conditions covered, and avoid presenting a brief local sample as a universal law or a set of interviews as every resident's view.A finding describes the sampled evidence, a process explanation requires an examined mechanism and alternatives, and a wider inference needs justified extension beyond the observations. The arrows identify additional reasoning demands rather than automatic proof. No actual learner finding, report structure or assessed conclusion is supplied.
Evaluation identifies consequential limitations — Explain how a particular calibration, sampling, timing, access or interpretive issue could change a conclusion's direction, magnitude or scope. “Human error” or “more data needed” is too vague without a mechanism. Distinguish uncertainty that threatens the central claim from a minor imperfection that does not materially affect it.
Improvements must target the identified problem — Propose a feasible change and explain what uncertainty it would address, how and at what cost. Extra repeats do not cure a systematically excluded population; a larger sample at the same unrepresentative time can repeat the same bias. Compare practical alternatives and recognise limits that cannot be removed within the available enquiry.
Reflect on ethical and socio-political consequences — Evaluate not only consent and safe practice but also who is represented, whose categories are used and how conclusions or maps might affect communities. Avoid presenting disadvantage as a defining trait of every resident or revealing sensitive sites through combined sources. Reflect on the investigator's influence and how a different perspective might alter the account.
Write a coherent report rather than an evidence dump — Organise a logical argument from context and methods through analysis to conclusions and evaluation. Integrate relevant maps, tables, graphs and extracts with readable captions and references. The advisory report length is 3,000–4,000 words under the specification's counting guidance; it is not a licence to pad or omit essential reasoning. Check current centre instructions for presentation and submission.
Reference sources and assistance transparently — Distinguish the learner's observations and analysis from quoted, paraphrased, adapted or shared material, with traceable references and a bibliography. A downloadable resource does not become the learner's original work when reworded. Keep collection roles, software transformations and assistance records honest, following the centre's authentication procedures and current guidance.
Completion needs authentic independent work — The candidate and teacher use the required record and declarations to confirm the work and conditions. A revision quiz, checklist tick or polished practice export does not certify that fieldwork, independence or assessment criteria have been met. This resource teaches general enquiry knowledge; the student's real question, evidence, analysis, judgement and centre assessment remain their own.
Test yourself
50 questions · Random sets of 10. These quick checks support revision; practise longer explanations and justified judgements too.
Mr Kidman's Classroom
