What this lesson teaches
I can plan a fieldwork investigation with a clear question that can be tested in the time and with the tools available.
Syllabus K229, Topic 1.3, Key Question 1. How to design fieldwork?
I can organise fieldwork data and analyse its patterns with counts, averages and comparisons.
Syllabus K229, Topic 1.3, Key Question 3. How to process and analyse data?
I can present fieldwork findings with clear maps, graphs, sketches and text.
Syllabus K229, Topic 1.3, Key Question 4. How to present findings?
I can judge the reliability and validity of an investigation and say how to improve it.
Syllabus K229, Topic 1.3, Key Question 3 (part). How to process and analyse data?
Make a guess
A survey asks a leading question the same way to every person. Is the data reliable, valid, or both?
- Neither, because survey answers change from person to person.
- Reliable but not valid.
- Both, because the method was the same each time.
- Valid but not reliable.
Show the answer
Reliable but not valid.
Asking the same way every time gives consistent results, so it is reliable. But the leading wording stops it measuring real views, so it is not valid.
A good conclusion answers the question with evidence, admits its limits and suggests a fix for a real weakness.
A conclusion answers your original question with evidence. Say whether the results support the hypothesis, give relevant figures or quotes, and explain exceptions. "Our six practice cases generally show fewer visits at greater distances" fits the scatter graph. "Distance is the only reason people visit parks" goes beyond the evidence.
Reliability means consistency. Repeating observations in similar conditions checks whether measurements and results are stable. Validity means the method measures what the question asks and supports the conclusion. A leading question can be asked the same way every time, so it is reliable, yet still be invalid for finding people's real views.
For each weakness, explain its effect and suggest a matching fix. Missing non-users needs a sample that includes them. Unclear answer options need rewording and a pilot. A bigger sample does not fix every problem.
Full Geography includes a longer fieldwork project over 10 weeks. One possible plan: weeks 1-2 for the question, background reading, ethics and risk planning; weeks 3-4 for sampling and a pilot; weeks 5-6 for approved data collection; weeks 7-8 for processing and analysis; and weeks 9-10 for conclusions, presentation and reflection. Adapt this with your teacher.
Put these in order
Put the stages of a fieldwork inquiry in order.
- Run a pilot to catch unclear questions and practical problems.
- Collect the data at approved sites.
- Write a conclusion, then evaluate its limits and suggest fixes.
- Ask a specific question and write a hypothesis you can test.
- Plan the sample, the methods, ethics and risks.
- Process and analyse the data with graphs and statistics.
Show the answer
- Ask a specific question and write a hypothesis you can test. Once you know what to measure, you can plan whom to sample and how to stay safe and ethical.
- Plan the sample, the methods, ethics and risks. A plan can still have unclear questions or practical problems, so test it on a small scale first.
- Run a pilot to catch unclear questions and practical problems. Fixing problems in the pilot means the real data are worth collecting, so collection can start.
- Collect the data at approved sites. Raw counts and answers cannot answer the question yet, so they must be processed and analysed.
- Process and analyse the data with graphs and statistics. Patterns in the data can now be compared with the hypothesis, so you can write a conclusion.
- Write a conclusion, then evaluate its limits and suggest fixes.
Elective Geography uses shorter investigations within topics, such as designing a questionnaire or mapping a local area. This lesson explains the fieldwork process, but your school's actual investigation still needs its own plan and teacher guidance.
Sort these
Students surveyed 12 residents at the park entrance one weekday morning. Tag each part of their method.
Groups: Sampling flaw, Timing flaw, Strength.
- The survey took place on one weekday morning.
- Everyone was asked at the park entrance.
- Each resident was asked the same questions in the same order.
- Only 12 residents were asked.
- Students recorded how far each resident lived from the park.
Show the answer
Sampling flaw
- Everyone was asked at the park entrance. Residents who never visit the park were missed, so the study cannot fairly test how distance affects visits.
- Only 12 residents were asked. With so few people, one or two unusual answers can change the pattern, so the results are less reliable.
Timing flaw
- The survey took place on one weekday morning. Weekend, evening and after-school visitors were missed, so the sample may not be typical of park users.
Strength
- Each resident was asked the same questions in the same order. Asking everyone in the same way makes the answers consistent and comparable, which improves reliability.
- Students recorded how far each resident lived from the park. The hypothesis is about distance, so measuring it for every person makes the method valid for the question.
- Reliability
- Consistent results when repeated.
- Validity
- Measures what the question asks.
- Full vs Elective
- 10-week project vs short tasks within topics.
Step by step
Answer with evidence
Use figures and exceptions.
Check the conclusion
Reliability, validity and missing groups.
Fix a weakness
Match the fix to the problem.
Worked example: A complete limitation
"We only surveyed one weekday lunchtime, so evening users may be missing. We should repeat the survey on some evenings and weekends." This names the weakness, its effect and a fix. "We needed more time" does not.
Watch out for this
A large dataset makes any conclusion valid.
Size cannot fix a wrong measure, a missing group or a claim of cause without evidence.
Check your understanding
Which improvement fixes a leading questionnaire item?
- Rewrite it neutrally, split combined ideas, pilot it.
- Ask the same item to twice as many people, so the sample is larger.
- Add a 'don't know' option after the same leading question.
Show the answer
Rewrite it neutrally, split combined ideas, pilot it.
Yes. This fixes the bias itself.