Primary and Secondary Methods in Rural Fieldwork

To conduct an Environmental Quality Survey (EQS) step-by-step:

• First, select your survey locations and identify specific rural indicators to assess, such as the condition of buildings, litter levels, or natural versus traffic noise.
• Next, design a bipolar scale to rate these indicators from a negative to a positive value (e.g., -2 to +2), where 0 represents a neutral midpoint.
• Then, survey residents or observe the site, recording the scores for each indicator to generate a total environmental quality score.

Calculate the Mean EQ Score for litter at a specific site:

$$\text{Mean EQ Score} = \frac{\text{Total sum of scores}}{\text{Number of respondents}}$$

Five residents provide scores of 3, 4, 2, 4, and 3 for litter levels.

• $\text{Mean} = 16 \div 5 = 3.2$

Measuring Flows of People

Counting the number of people walking down a street might sound simple, but doing it accurately requires a strict scientific method. A pedestrian count is a primary method used to collect quantitative data about the pedestrian flow in a settlement. To observe spatial changes, sites are typically chosen along a transect moving from the expected high density of the village core to the lower density of the outskirts.

To perform a pedestrian count step-by-step:

• First, establish a fixed time interval at every site (e.g., exactly 5 minutes) to ensure your data is comparable across different locations.
• Next, use a tally chart (the "five-bar gate" system) or a hand-held clicker to record pedestrians passing your observation point.
• Finally, to improve accuracy, use two people to count simultaneous flows in opposite directions. Be sure to manage rural risks by staying on pavements to avoid farm traffic.

Calculate the Mean Pedestrian Flow across multiple sites:

$$\text{Mean} = \frac{\text{Total number of pedestrians}}{\text{Number of sites}}$$

At Site A (Village Square), 42 pedestrians were counted. At Site B (Edge), 4 pedestrians were counted.

• $\text{Mean} = (42 + 4) \div 2 = 23 \text{ pedestrians per site}$

Using Secondary Data and GIS

Before stepping foot in a village, you can uncover its hidden characteristics using data collected by others. You must use at least two sources of secondary data, and one of these must be from the ONS Census. Census data provides essential demographic context at the LSOA level, showing age structures, employment rates, and housing types.

Secondary data provides the background needed to explain anomalies in your primary findings. For example, an unusually high pedestrian count might be explained by local news reports of a farmers' market. You can also use Geographic Information Systems (GIS) to analyze your data spatially.

By taking georeferenced primary data (like EQ scores) and layering it over secondary maps (such as the Index of Multiple Deprivation), you can spot geographical correlations. You can then visit the site to check if the secondary maps match the physical reality, a verification process known as ground-truthing.

Calculate the Ground-Truthing Accuracy of an old map:

$$\text{Accuracy \%} = \left( \frac{\text{Number of correct map features}}{\text{Total features checked}} \right) \times 100$$

You check 20 farm buildings shown on a map, but find only 16 are still used for farming.

• $\text{Accuracy \%} = (16 \div 20) \times 100 = 80\%$

Sampling Strategies for Rural Investigations

You cannot possibly survey every single person or street in a village, so how do you decide who or where to test? You must select locations and people using a clear sampling frame to ensure your data is valid.

There are four main sampling strategies to choose from:

• Systematic sampling: Data is collected at regular intervals, such as taking an EQS every 200m along a transect. This ensures even spatial coverage.
• Random sampling: Sites or people are selected using a random number generator. Every member has an equal chance of selection, which removes human bias, though it can cause accidental clustering of data.
• Stratified sampling: The study area is divided into sub-sets based on secondary Census data. This is often the most representative method because it proportionally matches the known population structure (e.g., age groups).
• Opportunistic sampling: Choosing the nearest accessible site when a pre-selected site is unreachable, such as when a path is blocked by livestock or private land.

Calculate a Systematic Sampling Interval:

$$\text{Interval} = \frac{\text{Total Length of Transect}}{\text{Number of Samples Required}}$$

You need 20 samples along a 1000m transect.

• $\text{Interval} = 1000 \div 20 = 50\text{m interval}$