Define the place, people, variables and time before choosing a method.
"Is our city liveable?" is too broad for a short investigation. "Which barriers do residents aged 65 and over report on walking routes to Clinic A?" identifies a group, place and observable issue. Choose a boundary you can cover with the time, people and equipment available.
A geographical question asks what, how or why. A hypothesis predicts a relationship you can test, such as "Reported walking difficulty increases with route distance to the clinic." State how you will measure distance and difficulty. A vague term such as accessibility needs indicators: route length, steps, crossing delays or residents' ratings.
Plan what evidence could challenge the prediction as well as support it. Distance may matter less than shade or mobility. A measurable association does not by itself show that distance causes difficulty. Keep the original question visible when collecting and analysing data.
Step by step
Set the study area
Choose a defined neighbourhood, population and period.
Make the variables observable
Record route distance in metres and use the same difficulty scale, with each response label explained.
Allow disagreement
Include routes and residents likely to reveal exceptions, not only convenient supporting cases.
Worked example: Question and hypothesis
Question: How does route distance relate to reported walking difficulty? Hypothesis: Longer routes have higher difficulty ratings. Record distance and rating for the same journey; do not pair one resident's rating with a different route.
Watch out for this
A good investigation proves the hypothesis.
Test the hypothesis fairly. Evidence that challenges it can still be a useful finding.
Check your understanding
Which change makes "Do residents respond to climate change?" researchable?
- Use more complicated language.
- Define the neighbourhood, response and time period.
- Ask only people attending a climate event.