Data Types, Planning and Primary/Secondary Sources

Start with good data

Before you can draw a chart or work out an average, you must collect data well. Choosing the right type, source and plan is what makes a statistical enquiry trustworthy. This module covers the types of data, where it comes from, and how to plan an investigation - ending with you writing a plan of your own.

The two families of data

Every variable you collect is either words or numbers, and each family splits again. The split you choose decides which charts and averages you are allowed to use later:

Match each variable to its data type

Recorded to the nearest second

A cafe times how long each customer waits to be served, and writes each one down to the nearest second. What type of data is the waiting time?

• Continuous
• Discrete, because the recorded values are whole seconds
• Ordinal
• Qualitative

Primary and secondary sources

Data also differs by who collected it, and what for: - Primary data is collected by you, for your purpose (a survey or experiment you run). - Secondary data was collected by someone else, or for a different purpose. It must be acknowledged, and you should check how reliable and accurate it is. The Office for National Statistics census is secondary data for you. So are last year's figures from your own school office: they are second-hand to your enquiry even though they came from inside your building. The test is not where the data lives, but who gathered it and why.

Whose data is it?

For your enquiry you email a local sports centre, and they send you their own attendance figures from last year. Is that primary or secondary data?

• Secondary, because the centre collected it, not you
• Primary, because you requested it yourself for your project
• Primary, because it came directly from the source rather than a website
• Neither, because attendance figures are not statistical data

Reliability and validity

Two words the examiners define precisely, and the pair most often confused: - Reliability is the extent to which repeated measurements yield similar results. - Validity is the extent to which a test measures what was intended. A bathroom scale that always reads 2 kg heavy is reliable but not valid: perfectly consistent, and consistently wrong. Reliability is about agreement between repeats; validity is about whether you measured the right thing at all. A measurement can have either without the other.

Reliable or valid?

Each line describes something going wrong in a study. Tap the TWO that are problems with VALIDITY rather than reliability.

• The same stopwatch gives a slightly different reading every time the same race is timed.
• A reading test is printed so small that it partly measures how good the pupils' eyesight is.
• Two people counting the traffic on the same road at the same time get very different totals.
• A survey meant to find out what all pupils think of school meals is given only to pupils queuing for school meals.
• A thermometer has to be left for a different length of time at each reading.

Planning an enquiry

A statistical enquiry is planned before any data is collected. Four words do the work:

Plan an enquiry

Priya thinks Year 11 pupils sleep less than Year 7 pupils. That testable statement is her _____. She will ask pupils herself, so her data will be _____. First she tries her questionnaire on ten pupils to find out which questions confuse them, which is a _____ study. She is careful not to ask "Do you agree that Year 11 are exhausted?", because that wording would introduce _____.

Which one can you test?

Which of these would work as a hypothesis for a statistical enquiry?

• Year 11 pupils spend more hours on homework each week than Year 7 pupils
• Older pupils are better at school
• How much homework do Year 11 pupils do?
• Schools should set less homework

Plan an investigation

You want to investigate whether taller students have longer arm spans. Work through the plan.

• The statement "taller students have longer arm spans" is your...
• You measure the students yourself. This data is...
• You measure ten students twice and get noticeably different arm spans the second time. What is the problem?

Write your plan

Your school wants to know whether pupils who walk to school arrive more awake than pupils who come by car or bus. Write the plan for the enquiry.

• State a hypothesis: a testable statement, not a question or an opinion
• Say what you will collect and what data type each variable is
• Say whether your data is primary or secondary, and why
• Name one constraint you must plan around, and one thing you would do to avoid bias

Data collection summary

Numerical data is _____, while descriptive data is qualitative. Data you collect yourself is primary, while data collected by somebody else is _____ and must be acknowledged. _____ is the extent to which repeated measurements agree, while _____ is the extent to which a test measures what was intended.