Geographical Investigations » Collecting Data for an Investigation

Designing a data collection sheet

Paper 4 asks how you would collect data to test a hypothesis. A good plan starts with a data collection sheet: a table you fill in out in the field. The techniques and equipment are in Fieldwork Equipment and Physical Techniques and Human Fieldwork Techniques. This lesson is about planning the collection so the data can be trusted.

Key terms:

• Data collection sheet: a prepared table or form for recording results in the field.
• Accuracy: how close a measurement is to the true value.
• Reliability: whether repeating the collection would give similar results.
• Sample size: how many measurements, people or sites you collect data from.
• Control: a site or group kept the same, or without the factor being tested, so you can see the effect of that factor.

Control groups help you test a cause. Suppose you test whether a pedestrian zone increases shoppers. You would count in that street and also in a similar street with no pedestrian zone, as the control. Reliability improves when you repeat measurements, use a good sample size and keep conditions the same.

Questionnaires

A questionnaire can be oral or written. Use one for topics such as spheres of influence, use of services, a farm or factory study, leisure, tourism or people's attitudes to a development.

Designing it well:

• Layout: a title, a short introduction, a logical order, space to answer and tick boxes where possible.
• Format of questions: closed questions have fixed answers to tick, such as age groups, and are quick to count. Open questions let people answer in their own words and give richer detail, but are harder to analyse.
• Wording: short, simple and neutral. "Do you agree that the noisy traffic ruins this street?" pushes people to say yes. "How would you rate the noise level here?" is fairer.
• Number of questions: few enough that people finish, and every one linked to the hypothesis.

Practical considerations: you cannot ask everyone, so you ask a sample. There are four sampling methods.

Sampling method	How it works	Example
Random	Everyone has an equal chance, chosen by chance such as random numbers	Using random numbers to pick which house numbers to visit
Systematic	Taking every nth person or point at regular intervals	Asking every 10th person who passes, or sampling every 50 m along a path
Stratified	Sampling each group in proportion to its size	If 20% of the town is aged over 65, then 20% of your sample is aged over 65
Opportunity	Asking whoever is available and willing	Asking the first people you meet outside the shop

• Pilot survey: a small trial run, for example with five people, to find confusing questions before the full survey.
• Location and timing: choose a place where the right people are, such as the town centre for shoppers. Ask at different times of day, because different people are there.

Opportunity sampling is quick but can be biased. Stratified sampling is fairer but needs more planning.

Observations and counts

Observations record what you see, for example land use in a town or features of a river or coast. You can use maps, recording sheets, field sketches and annotated photographs. A recording sheet lets you tick the land use of each building, so the results can be compared and added up. Label and annotate sketches and photographs so they show the feature you are studying.

Counts, such as pedestrian and traffic counts, need a tally recording sheet with:

• a clear layout, with columns for each time period, direction or vehicle type;
• simple instructions, such as "count for 5 minutes at the start of each hour";
• the details to identify the sheet afterwards: time, date, location and name of recorder.

Without that information, a sheet found later cannot be used.

Measurements

When recording measurements, plan three things: the layout of the recording sheet, where you site the instruments and the sampling methods. This is what makes the data reliable.

Human survey strategies include pedestrian and traffic counts and questionnaires. Use the same time and day at every site, and a sample size that you can complete.

Exam-style question

A group tested the hypothesis "People travel further to the large supermarket than to the corner shop." They used a questionnaire, asking every 10th person leaving each shop. They also took depth readings across a river channel on another day. Fig. 1.1 shows made-up depth readings, in cm.

Fig. 1.1	Reading 1	Reading 2	Reading 3	Reading 4	Reading 5
Depth (cm)	12	18	25	21	14

(a) Identify the sampling method used for the questionnaire. [1 mark]

(b) Give one reason for carrying out a pilot survey. [1 mark]

(c) Suggest why the question "Don't you agree that this busy shop is much too crowded?" is badly worded. [2 marks]

(d) Calculate the mean depth shown in Fig. 1.1. [1 mark]

(e) Plan how the group could collect reliable data to compare a river channel at two sites. [6 marks]