Presenting and Analysing Your Data

📊 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.