Collecting and Organising Data
Section: Statistics | Syllabus: Cambridge Primary Mathematics (0845)
Planning a Statistical Investigation
A statistical investigation starts with a question you want to answer using data, followed by a prediction of what you expect to find. Before collecting any data, it helps to know what type of data you are looking for.
- A statistical investigation begins with a question, followed by a prediction of what the data might show
- Data can be grouped into 3 types:
- Categorical - sorted into named categories with no numerical order, e.g. favourite colour, type of pet
- Discrete - counted in whole, separate steps, e.g. number of siblings, number of goals scored
- Continuous - measured, and can take any value within a range, e.g. height, time, temperature
- A set of related statistical questions can build a fuller picture than a single question alone, e.g. "What is the most common pet?" alongside "How many pets do most people have?"
Knowing the type of data helps decide how to collect and represent it
Worked Example: Identifying the Type of Data
- Question: A survey asks people their favourite sport, how many times a week they exercise, and their exact running time in minutes. What type of data is each of these?
- Step 1: Favourite sport - sorted into named categories, no numerical order: categorical
- Step 2: Number of times per week - counted in whole, separate steps: discrete
- Step 3: Exact running time - measured, can be any value: continuous
- Answer: categorical, discrete, continuous
Common Mistakes
MistakeThinking discrete and continuous data are the same because both involve numbers
Fixdiscrete data can only take certain separate values (like whole numbers); continuous data can take any value, including fractions and decimals in between
Interpreting Data and Drawing Conclusions
Once data has been collected and represented, the next step is to look for patterns, draw conclusions, and check whether the original prediction was correct.
- Look for patterns within a single data set (e.g. is one category much bigger than the rest?) and between different data sets (e.g. does one group's data differ from another's?)
- A conclusion should be supported by the data, and should say clearly whether it matches the original prediction
- Variation means the natural differences that occur in data - not every result matches the prediction exactly, and considering why helps explain the data properly
Worked Example: Interpreting Data and Checking a Prediction
- Question: Before a survey, a class predicted that most people's favourite subject would be P.E. The results showed: Maths 6, Science 4, P.E. 12, Art 3. Does the data support the prediction?
- Step 1: Compare the results - P.E. has 12 votes, more than any other subject
- Step 2: Check this against the prediction that P.E. would be the most popular
- Answer: Yes, the data supports the prediction - P.E. was the most common answer
Common Mistakes
MistakeIgnoring the original prediction when writing a conclusion
Fixa good conclusion always compares the data back to the prediction, saying clearly whether it was correct, partly correct, or incorrect
Interactive revision notes, videos and practice questions load below.