Human Geography Fieldwork

🔍 The Underpinning Theory: Burgess's Concentric Zone Model

Before collecting any data, good fieldwork starts with a theory — a prediction about what you expect to find and why. For human fieldwork in urban areas, that theory is almost always the Concentric Zone Model, developed by Ernest Burgess in 1924 from his study of Chicago. Even though the model is a century old, it still provides the framework for most urban fieldwork hypotheses at GCSE.

Burgess argued that cities grow outward from the centre in a series of concentric rings, each dominated by a different land use and population group. Understanding the model tells you what to expect at each point along your transect — and allows you to explain your results rather than just describe them.

The model makes a clear prediction: as you move outward from the CBD, environmental quality should generally increase — a pattern geographers call a negative correlation between distance from the centre and deprivation. Your fieldwork tests whether this prediction holds in reality, and you are expected to explain any anomalies (places where the pattern breaks down) using your knowledge of local context.

Limitations of the model to know for evaluation: Burgess developed it from 1920s Chicago, a rapidly growing North American city. UK cities — with their medieval cores, post-war council estates, and recent gentrification — do not always follow the same pattern. The model is a simplified abstraction, not a literal description of any real city.

📋 Method in Detail: The Environmental Quality Survey (EQS)

The Environmental Quality Survey is the most common data collection method in human fieldwork — and the one most likely to appear in exam questions. You must be able to describe it precisely, justify why it was used, explain its limitations, and suggest realistic improvements. Vague answers — "I used an EQS to score the environment" — score Level 1. Precise answers score Level 3.

What it is

An EQS is a systematic method for measuring the perceived quality of an environment at multiple sites. You create a scoring sheet listing 8–12 environmental criteria. At each site, every criterion is rated on a numerical scale. The scores are totalled to give an overall Environmental Quality Index (EQI) for each site. Sites can then be compared or plotted against distance from the city centre to test the hypothesis.

Designing the scoring sheet

Choose criteria that are relevant to your hypothesis and observable from the street. A well-designed EQS for an urban transect would include:

Total possible range: 10–50. Higher score = better perceived environmental quality.

Setting up the transect

Choose a main road that runs from the CBD outward through the inner city, inner suburbs, and outer suburbs. Identify 6–10 survey sites at regular intervals — for example, every 200 metres along the transect from the city centre outward. Regular spacing is called systematic sampling, and it ensures you capture data from all urban zones, not just the most accessible or convenient points.

Record the distance of each site from the city centre (using a measuring wheel or OS map). This distance becomes the independent variable on your scatter graph.

Reducing subjectivity

The biggest weakness of the EQS is that it produces qualitative judgements converted into numbers — the scores are only as reliable as the observer's perception. Two people rating "building condition" on the same street may give different scores. This means the data looks quantitative but contains subjective variation. To reduce this problem:

• Photographic benchmarks: Before the survey, prepare photographs showing what each score (1, 2, 3, 4, 5) looks like for each criterion. All surveyors refer to the same photographs. This anchors scores to observable evidence rather than individual judgement.
• Multiple observers: Have 2–3 surveyors independently score each site. Calculate the mean. Where scores differ by more than 1 point, discuss and re-evaluate before recording.
• Training: Practise the survey at a trial site before fieldwork day, so all surveyors calibrate their judgements against each other.

Example results (hypothetical)

Notice that Site 3 scores slightly higher than Site 2 — this is an anomaly. In the exam, you would be expected to explain it: perhaps Site 3 lies near a recently regenerated area, or includes a park, or was surveyed on a quieter side street. Identifying and explaining anomalies is a Level 3 skill.

🚶 Other Data Collection Methods: Pedestrian Counts, Land Use and Questionnaires

Pedestrian Count

A pedestrian count records how many people walk past a fixed point within a set time period — usually 5 minutes. It is a simple, reliable, and genuinely quantitative method (unlike the EQS, there is no judgement involved — you either count a person or you do not). Pedestrian counts test hypotheses about footfall, sphere of influence, and the relative busyness of different zones.

How to conduct it properly:

• Choose a fixed counting position at each site (e.g., stand at the kerb edge, facing the road).
• Count every person who passes in both directions within 5 minutes.
• Repeat 3 times at each site, then calculate the mean of the three counts.
• Conduct all sites at the same time of day — pedestrian flow varies enormously between morning rush hour, lunchtime, and mid-afternoon. If you count Site 1 at 9am and Site 6 at 12pm, the difference in counts may reflect time of day rather than location.

Limitations:

• Time-of-day variation: Even if all sites are surveyed at the same time, that time of day may not represent typical patterns. A 10am count on a Tuesday misses commuters and weekend shoppers.
• Weather effects: Rain significantly reduces pedestrian counts in all zones. If weather changed between surveying different sites, results are not directly comparable.
• Special events: A market, festival, or roadworks near one site will distort the count for that site.
• Repeat surveys: Counting on a single day gives a snapshot, not an average. Ideally, repeat the survey on different days of the week and average the results.

Land Use Survey

A land use survey records the use of every building along the transect — or in a sample area around each site. Buildings are categorised as: retail, office/commercial, residential, industrial/warehouse, vacant/derelict, leisure/hospitality, or other. The results are recorded on a base map and can be presented as a proportional bar chart for each zone, showing how the mix of land uses changes with distance from the CBD.

How to conduct it:

• Use a pre-printed categorisation sheet with a colour code for each land use type.
• Record the ground-floor use of every building in a 50m stretch around each site.
• Note whether ground floor and upper floors differ (e.g., a shop on the ground floor with flats above — this is mixed use, common in the CBD).

Key prediction from Burgess's model: Retail and commercial land use should be highest in the CBD (Zone 1) and decrease with distance. Residential land use should be lowest in the CBD and increase toward the outer suburbs.

Limitations:

• The ground floor does not always represent the building's primary use (e.g., a bank occupying the ground floor of a 12-storey office block).
• Categorisation can be ambiguous — is a nail salon retail or personal service?
• The survey captures a snapshot; land uses change over time (vacant units may have been recently occupied).

Questionnaire / Interview

A questionnaire gathers data about people's perceptions, attitudes, and experiences of an area. This produces qualitative data (open questions: "What do you like most about this area?") and/or quantitative data (closed questions with Likert scales: "Rate this area for cleanliness, where 1 = very dirty and 5 = very clean").

How to conduct it:

• Prepare your questionnaire in advance — do not ask questions on the spot.
• Use a mix of closed and open questions (typically 6–10 questions in total).
• Use random sampling: approach every 5th person who passes, regardless of who they are. This avoids cherry-picking approachable-looking people, which would bias your sample toward certain demographics.
• Aim for a minimum of 20 responses per site. 50+ is better for statistical significance.
• Note the time of day and day of week on every questionnaire — this helps you identify temporal bias when analysing results.

Ethical considerations (essential knowledge):

• Always introduce yourself and explain the purpose of the research before asking questions.
• Participants have the right to refuse — never pressure anyone to respond.
• Do not record names. Keep responses anonymous.
• Do not photograph respondents without explicit consent.
• Store data securely; use it only for the stated purpose.

Limitations:

• Social desirability bias: People may give the answer they think the researcher wants, or the answer that makes them look good, rather than their true view.
• Temporal bias: A weekday morning survey over-represents retired people and carers; a Saturday survey over-represents shoppers. Neither group is representative of all users of the space.
• Response rate: Many people refuse to stop. The sample of people willing to be questioned may not represent the wider population.
• Question wording: Leading questions ("Don't you think this area is unsafe?") influence responses. Questions must be neutral and unambiguous.

📊 Presenting and Analysing Your Data

Choosing the right presentation method for each type of data is itself an exam skill. The rule is: the presentation method should match what the data is testing and make the pattern clearly visible.

EQS scores against distance: scatter graph

Plot EQS total score on the y-axis and distance from CBD on the x-axis. Each survey site becomes one point on the graph. Add a best-fit line. If the line slopes upward (positive gradient), it supports the hypothesis that EQS increases with distance. Calculate Spearman's rank correlation coefficient to quantify the strength of the relationship:

• rs close to +1: strong positive correlation (EQS increases with distance — supports Burgess)
• rs close to 0: no correlation
• rs close to −1: strong negative correlation (EQS decreases with distance — contradicts Burgess)

Compare your rs value against the critical value table at your chosen significance level (usually 95%, with n = number of sites). If rs exceeds the critical value, the correlation is statistically significant — i.e., unlikely to have occurred by chance.

EQS by criterion: bipolar (radar/spider) chart

Plot a separate radar chart for each survey site. Each axis represents one criterion (litter, noise, building condition, etc.), scaled 1–5. Connect the scores with a line. The larger the shaded area, the better the overall environmental quality. Overlay the radar charts for two contrasting sites (e.g., CBD vs outer suburb) on the same diagram to make the comparison visual and immediate. This is a powerful way to show which criteria account for the quality difference, not just the total score.

Pedestrian counts: bar chart or line graph

Bar chart: one bar per site, showing mean pedestrian count. Makes comparisons between sites clear. Line graph: useful if sites are ordered by distance, since it shows whether the decline is gradual or stepwise. Both approaches work — justify your choice in the exam.

Land use: proportional divided bar charts

For each site along the transect, draw a bar whose total height is 100%. Divide each bar into coloured segments proportional to the percentage of buildings in each land use category. Place the bars in order along the transect from CBD to suburbs. The visual transition — from mostly commercial (CBD) to mostly residential (suburbs) — directly illustrates the Burgess model.

Questionnaire results: pie charts and comparative bar charts

For closed questions, draw a pie chart per site per question showing the distribution of responses. For direct comparison between two sites (e.g., CBD vs outer suburb), use a grouped bar chart showing the percentage of respondents giving each score at each site side by side.

Identifying anomalies

An anomaly is a data point that does not fit the expected pattern. On your scatter graph, it appears as a point far from the best-fit line. You are expected to identify it and explain it using local knowledge. Common explanations:

• A site near a recently regenerated area (e.g., new housing development) scores unexpectedly high despite being in the inner city — gentrification effect.
• A site in the outer suburbs near a major road scores unexpectedly low for noise and traffic — proximity to a bypass or arterial route overrides the general pattern.
• A site scored differently by different members of the group — surveyor error, which is a methodological weakness rather than a true geographic anomaly.