Planning an Enquiry (Edexcel GCSE Statistics)

Statistical Enquiry Cycle

What is the statistical enquiry cycle?

• The process for statistical investigations in the real world is a cycle, known as the statistical enquiry cycle

  • Being a cycle means there is no simple ‘beginning’ and ‘end’

  • Instead the steps are repeated with improvements made each time

    • An ‘improve-repeat’ process like this is known as an iterative process

• The statistical enquiry cycle is divided into five stages:

  • Hypothesis and Planning

    • Specify a hypothesis to be investigated

    • Plan what data to collect (and how it will be recorded)

    • Plan how you will process the data

    • Plan how the data will be represented (graphs, tables, diagrams, etc.)

    • Reasons should be given for each choice made in a plan

  • Collecting Data

    • Design data collection to minimise bias

    • Be aware of possible issues of sensitivity

    • Collect primary data using an appropriate method

    • Consider using secondary data (but only if it is reliable)

  • Processing and Representing Data

    • Organise the data and process it according to the plan

    • Clean the data if necessary

    • Create diagrams, etc., to represent the data

    • Calculate summary statistics to allow the data to be compared

    • Consider your target audience when presenting the data

    • Acknowledge any sources used (e.g. sources of  secondary data)

    • Use technology where appropriate to save time and avoid errors

  • Interpreting Results

    • Interpret your summary statistics, and your diagrams, etc., in the context of the investigation

    • Draw conclusions that are related to the hypothesis

    • Make any appropriate inferences and predictions

    • Be sure to comment on the reliability of the results 

  • Evaluating

    • Identify any possible issues with how the data was collected, processed and represented

    • Suggest improvements to deal with those issues

    • Reflect on how appropriate the data representation(s) were for the target audience

    • Repeat the process with improvements to investigate the hypothesis further

• An exam question may directly mention one or more stages of the statistical enquiry cycle

  • But you should keep the statistical enquiry cycle process in mind when answering any exam question

Hypotheses & Constraints

What is a hypothesis and how can it be tested?

• A hypothesis is a statement that you would like to test using statistics

  • For example, ‘As cars get older their annual maintenance cost is likely to go up’

• A hypothesis should always be stated at the start of a statistical enquiry

  • Before any data is collected

• Testing a hypothesis requires

  • Collecting valid and relevant data

  • Appropriate analysis of the data collected

What sort of constraints might affect a statistical investigation?

• Constraints are practical limits that affect how an investigation may be conducted

  • You should try to anticipate these at the start of an investigation

  • And include them in your planning

    • ‘Anticipate’ means try to think of what they might be ahead of time

• Time

  • You may only have a limited amount of time for conducting the investigation

  • So you must plan an investigation that can be completed with the time available

• Cost

  • The ‘best’ investigation might cost more than is available

  • You need to plan for the ‘best investigation you can afford’

• Ethical Issues

  • You must always look out for the well-being of any participants in the investigation

• Confidentiality

  • People may not be willing to reveal confidential (‘secret’) information

  • Any confidential information collected must be kept confidential

• Sensitivity

  • People may be uncomfortable discussing sensitive topics

  • Who the data collector is, or how the data is collected, may affect this

• Convenience

  • Some pieces or types of data might be hard to find or collect

What other issues might affect an investigation?

• You should also try to anticipate issues that might arise during the statistical enquiry process

  • And think of proactive ways to deal with these

    • Being proactive means acting ahead of time, instead of only reacting once a problem has appeared

• Some examples might be:

  • Difficulties identifying the population you want to study

  • People may not answer some or all of the questions asked

  • Some responses or outcomes of an experiment may be unexpected