Statistical Enquiry Cycle: B Data collection — Edexcel A-Level Statistics
Test yourself on Statistical Enquiry Cycle: B Data collection with PEARSON EDEXCEL A-Level practice questions.
7 days Premium · Then free forever · No card, no charge
Statistical Enquiry Cycle: B Data collection explained
In the data collection stage of the statistical enquiry cycle, primary sample data are collected first-hand for the current enquiry.
Read the full explanation
Designing unbiased collection methods means choosing procedures that give every intended unit a fair chance of inclusion and that record responses accurately. Key decisions include the sampling method (simple random, systematic, stratified or cluster), the sampling frame, the data collection instrument (questionnaire, interview, sensor, observation) and the administration protocol. To avoid selection bias, use a probabilistic method with a complete frame. To avoid non-response bias, plan follow-ups and compare respondents with non-respondents. To avoid response and measurement bias, pilot the instrument, use neutral wording, standardise conditions and calibrate equipment. Document the protocol so the method can be repeated and audited.
Your focus
- Select and justify a probabilistic sampling method that gives every intended unit a known chance of selection.
- Design a data collection instrument and protocol that minimise response, non-response and measurement bias.
- Evaluate a collection method for sources of bias and propose specific improvements.
Statistical Enquiry Cycle: B Data collection exam tips
Marking Points
- Selects a probabilistic sampling method appropriate to the population and explains why it gives each unit a known, non-zero chance of selection.
- Uses a sampling frame that matches the target population and identifies any coverage gaps that could introduce bias.
- Designs the data collection instrument to avoid leading, ambiguous or sensitive wording that could cause response bias.
- Plans procedures to reduce non-response, such as reminders or call-backs, and to monitor differences between respondents and non-respondents.
- Standardises measurement conditions and calibrates instruments to reduce measurement bias, and records the protocol for reproducibility.
Examiner Tips
- 💡Match the sampling method to the structure of the population, for example stratified sampling when identifiable subgroups exist and differ.
- 💡State the practical steps that reduce non-response, such as reminders, incentives or call-backs, rather than merely naming the problem.
- 💡Describe how the instrument was piloted and how wording was checked for neutrality, since this is direct evidence of bias control.
Common Mistakes
- Assuming any random-looking selection is unbiased: convenience sampling such as choosing the first fifty people available is not probabilistic and introduces selection bias.
- Ignoring non-response: a low response rate can bias results if non-respondents differ systematically from respondents, so follow-up procedures must be planned.
- Using leading or double-barrelled questions: these push respondents towards particular answers and create response bias, so questions must be neutral and single-issue.