Sign in

← All revision topics

English variety

A complete statistical investigation: plan, collect, analyse, draw conclusions and report

Machine translation

Around age 11 · Free

Learn

  1. A statistical investigation answers a question using data. It has six stages, one after another.

    1. Ask the question
    2. Plan the investigation
    3. Collect the data
    4. Organise and represent
    5. Analyse and draw conclusions
    6. Report
    The six stages of an investigation
  2. Asking the question means choosing a question that the data can answer. It is a good idea to write a hypothesis too: what you think will happen. At the end you check whether you were right.

    Planning means deciding before you start which variable you measure, who you ask and how you record the answers. A good question has only a few possible answers and does not push people towards one of them.

    What we decideIn our investigation
    QuestionWhich do you prefer to play?
    AnswersFootball, skipping, cards, chat
    Who to askThe 20 children in the class
    How to record itTally with tally marks
    HypothesisFootball will get the most votes
    The plan for an investigation about playtime
  3. Collecting means carrying out the survey, asking each person only once, and writing down each answer. Organising means putting the tally into a frequency table. The absolute frequencies add up to the total; the relative frequencies add up to 100%.

    GameAbsolute%
    Football840%
    Skipping525%
    Cards420%
    Chatting315%
    Total20100%
    Answers from the 20 children
  4. Representing means drawing the data. A bar chart compares the categories well. Analysing means reading the chart to answer the question: which got the most votes? How many more than the next one? What part of the total is it? Watch out: the most popular is not always the majority. The majority is more than half the total.

    02468FootballSkippingCardsChattingChildren
    Football is the most popular, but 8 is not half of 20
  5. Drawing a conclusion means answering the question with the data and saying whether the hypothesis holds. You do not say more than the data say: the conclusion applies to the people who answered. Reporting means telling it clearly: title, question, how it was collected, table, chart and conclusion.

Worked examples

Example 1

Tony thinks that more than half of the 24 children in his class have a pet. Plan his investigation.

What we decideIn our investigation
QuestionDo you have a pet?
AnswersYes or no
Who to askThe 24 children in the class
How to record itTally with tally marks
HypothesisMore than 12 will say yes
  1. Ask a question that is answered with only a few answers: “Do you have a pet?”. The answers are yes or no.
  2. Decide who you ask: the 24 children in the class, each one only once.
  3. Decide how you record: a tally with two rows, yes and no.
  4. Write the hypothesis with a number: “More than half” means more than 12 children.
  5. Once the plan is written, you can start collecting the data.

The plan is the table: question, answers, who to ask, how to record it and the hypothesis.

Watch out! Here many people start asking questions without writing the plan, and later they do not know how many data they expected.

Common mistake

Example 2

Ellie asks 25 customers which bread they prefer. She thinks that more than half choose the rod. This is her frequency table. Does her hypothesis hold? What can she conclude?

BreadAbsolute%
Rod936%
Ring832%
Wholemeal520%
Plait312%
Total25100%
  1. Check the table: 9 + 8 + 5 + 3 = 25 customers, the same as the data.
  2. Find the part for the rod: 9 ÷ 25 = 0.36, that is, 36%.
  3. Half is 50%. 36% is less than half, so the hypothesis does not hold.
  4. Compare with the next one: the rod beats the ring by only 9 − 8 = 1 customer.
  5. Write the conclusion with the data.

It does not hold. The rod is the most popular bread, with 36%, but it does not reach half and almost ties with the ring.

Watch out! Here many say it does hold, because the rod is the most popular, even though it does not reach half.

Common mistake

Example 3

Gus writes in his report: “All the children in the school like football best.” He only asked the 20 children in his class. In the pictogram, each picture is worth 2 children. What needs correcting?

FootballSkippingCardsChatting= 2
  1. Look at who was asked: the 20 children in one class, not the whole school.
  2. Look at the data: football has 4 pictures of 2 children, that is, 8 children out of 20. That is 40%, not “all”.
  3. Write only what the data say: what happens in the class and how many children.
  4. Check the new sentence with the data: 8 ÷ 20 = 0.4 = 40%.
  5. Also put in the report who you asked, so that anyone reading it knows who it applies to.

“Of the 20 children in the class, 8 (40%) prefer football. It is the most popular game.”

Watch out! Here many believe that what happens in one class happens the same in the whole school.

Common mistake

A challenge to think about

Carry out a complete investigation with your class or your family. Choose a question with only a few possible answers and write your hypothesis. Plan who you ask and how you record it. Collect the data with a tally, make the frequency table and a bar chart. Write the conclusion, say whether your hypothesis holds and present your report on a sheet of paper.

This topic has no questions yet.