Presenting evidence in cartographic, diagrammatic and graphical form

Overview

Choosing a technique to fit the data, worked through pedestrian counts, a land use transect, channel velocity, shopper origins and soil composition, and what ICT and GIS add.

What you need to know

Answer each one before you reveal it.

Questions from this subtopic

Each one with the model answer underneath.

1. 1

   How do you decide which presentation technique suits a set of fieldwork data?

   Model answer

   Start from the type of data and the question. If location matters, the technique should be a map; if the interest is a relationship between two variables, it is a scatter graph; if it is change over a continuous scale such as time or distance, a line graph; if it is comparison between discrete categories, a bar chart; if it is the composition of a whole, a pie chart, compound bar or triangular graph; if it is movement between places, flow lines. Then check that the data are the right kind for the technique - rates rather than totals for a choropleth, three components summing to a hundred for a triangular graph - and that the result will be readable.

2. 2

   How would you present pedestrian counts recorded at ten sites across a town centre, and why that way?

   Model answer

   As proportional symbols, most simply circles, drawn on a base map of the town at the position of each counting site, with the area of each circle scaled to the count and the key stating the values. This is the right choice because both the size of the flow and its location matter, and a map lets the pattern be related to the streets, the shops and the transport nodes. A bar chart would compare the sites but throw away the geography; a choropleth would be wrong because the data are counts at points, not rates for areas.

3. 3

   How would you present land use recorded along a transect from a city centre?

   Model answer

   As a transect diagram or a strip map: a long narrow strip drawn to scale along the route, divided by distance from the centre and shaded by land use category with a key, so the sequence of zones is visible in order. A compound bar chart divided by distance band works as an alternative, showing the changing mix of uses with distance. Either way the distance scale must be shown, since the whole point is how the mix changes as distance increases, and a north arrow orients the transect.

4. 4

   How would you present velocity measurements taken at points across a river channel?

   Model answer

   As an annotated cross-section: plot channel width along the horizontal axis and depth below the water surface on the vertical axis to show the shape of the channel, then add the velocity recorded at each measuring point. Where enough readings have been taken at different depths and positions, lines joining points of equal velocity can be drawn across the section, which is an isoline treatment and shows immediately where the fastest flow lies and how friction slows the water near the bed and banks. The section needs a title, both axes labelled with units, and any vertical exaggeration stated.

5. 5

   How would you present the home locations of shoppers surveyed at a town centre?

   Model answer

   As desire lines or flow lines drawn on a base map from each origin to the centre, with line width proportional to the number of shoppers from that place, which shows both where the trade comes from and how much comes from each direction. A dot map of respondents' home addresses is an alternative that shows the density of the catchment, and drawing a boundary enclosing most of the origins delimits the sphere of influence. The key must state what a given line width represents, and the map needs a scale and north arrow so distances can be judged.

6. 6

   How would you present the sand, silt and clay proportions of soil samples from twenty sites?

   Model answer

   On a single triangular graph, with each site plotted as one point from its three percentages. That puts all twenty samples on one diagram, so groupings of similar soils and any outlying sample are immediately visible, and comparison between sites needs no cross-referencing between charts. It only works because the three components are percentages of one whole and sum to a hundred; twenty separate pie charts would carry the same information but make comparison across sites almost impossible.

Other ways to work this

The same content, four other ways.

What is inside Levo

Six ways to work the same material.