Collecting data
Conducting an investigation
Before you collect a single number, two questions decide everything: what kind of data do I need, and how will I get it?
Statistical questions
A statistical question is one you answer by collecting DATA. For example: ā¢ How many brothers do the learners in your class have? ā¢ What is the average mass of a baby born in your country? ā¢ What sports do learners in your school like to watch? Each one needs different data, collected in a different way.
Three types of data
DISCRETE data is counted, and the answers are whole numbers. The number of brothers you have is discrete ā you cannot have 2.4 brothers. CONTINUOUS data is measured, and can take any value in between. The mass of a baby is continuous ā it could be 3.2 kg, or 3.24 kg, or anything between. CATEGORICAL data is a name or a category rather than a number. A favourite sport is categorical. The quick test: can you count it (discrete), measure it (continuous), or neither (categorical)?
Choosing how to collect it
The method has to suit the question. ā¢ A company wanting customers' opinions could use an online questionnaire. ā¢ A drug company could interview people using a new drug about their experience. ā¢ Teachers worried about speeding could record the speeds of cars passing the school. Asking people works for opinions; measuring works for quantities. Choosing the wrong method gives you data that cannot answer your question.
Planning
Before you start, decide: ā¢ what sort of data you want to collect ā¢ how you will collect it ā¢ what you predict you will find ā¢ how you will analyse the results A prediction written BEFORE collecting is what makes it a test rather than just a survey.
Worked examples
'How many brothers do learners in your class have?'
- The answers are counted, in whole numbers.
- You cannot have 2.4 brothers.
Answer: Discrete data
'What is the average mass of a newborn baby?'
- Mass is measured, not counted.
- It could be 3.2 kg or 3.24 kg or anything between.
Answer: Continuous data
'What sports do learners like to watch?'
- The answer is a name, not a number.
Answer: Categorical data
Collecting customers' opinions
- Opinions have to be asked for, not measured.
- An online questionnaire reaches many people.
Answer: A questionnaire
Finding out whether cars speed past a school
- Speed is a quantity, so measure it.
- Asking drivers would not be reliable.
Answer: Record the speeds of passing cars
What to decide before collecting anything
- What data you need and how to collect it.
- What you predict.
- How you will analyse it.
Answer: A prediction written first makes it a test, not just a survey
Where you see this in real life
- A shop asking 'how satisfied are you?' collects categorical data; asking 'how much did you spend?' collects continuous data. Same customers, different questions.
- Recording car speeds outside a school gives continuous data ā which is exactly what you need to check against a speed limit.
Key words
- Data
- the information you collect
- Discrete data
- data you count, always whole numbers
- Continuous data
- data you measure, which can take any value
- Categorical data
- data that is a name or category, not a number
- Questionnaire
- a set of questions used to collect data
- Statistical question
- a question you answer by collecting data
- Prediction
- what you expect to find, written before you collect
- Analyse
- work out what the data is telling you