Kablow!

Sampling and populations

Apprends en jouant

Réponds à ces questions pour gagner de l'énergie, puis pêche et explore. Sans compte.

Pour les enseignants : diapos de leçon, notes de révision prêts à l'emploi pour Sampling and populations (Mathematics, Statistics) — utilise-les en cours, ou lance le thème en activité de classe interactive que tes apprenants jouent en direct.

Notes de leçon

Population vs Sample

  • A population is the entire group you want to draw conclusions about (e.g. all students in a school).
  • A sample is a subset of the population that is actually surveyed or measured.
  • Sampling is used because testing the whole population is often too expensive, time-consuming, or impractical (e.g. testing every light bulb would destroy them).
  • The sample should be representative of the population so that conclusions drawn from it can be generalised back to the whole population.

Why Sampling is Used

  • It is cheaper and faster than a census (a survey of the whole population).
  • It may be impossible to reach every member of the population (e.g. all fish in the ocean).
  • Testing can be destructive (e.g. checking matches or fireworks).
  • A well-chosen sample can give accurate enough results without the cost of a full census.

Representative Samples

  • A representative sample reflects the key characteristics of the population (e.g. same proportion of boys/girls, ages, or locations).
  • A simple random sample gives each member an equal selection chance. It helps avoid deliberate selection bias but does not guarantee a perfectly representative realised sample.
  • Use a complete sampling frame and account for nonresponse. Equal selection chances alone cannot remove every source of survey bias.
  • A larger sample is generally more reliable than a smaller one, but only if it is still representative.

Bias in Sampling

  • Bias occurs when the sample is not representative, leading to conclusions that are not true for the whole population.
  • Bias can arise from the sampling method (e.g. only asking people in a shopping centre, which excludes those who don't go there).
  • Bias can arise from the location of the survey (e.g. asking people outside a gym about exercise habits).
  • Bias can arise from the time of the survey (e.g. asking commuters at 8am misses night-shift workers).
  • A convenience sample (e.g. asking friends or people nearby) is often biased because it does not represent the whole population.
  • A larger sample does NOT remove bias – if the sampling method is biased, increasing the sample size still gives biased results.

Sample Size and Reliability

  • A larger sample tends to give more reliable estimates because it reduces the effect of random variation.
  • A small sample may not reflect the population well, so conclusions drawn from it are less reliable.
  • However, a large sample does not guarantee accuracy if the sample is biased (e.g. only asking one age group).
  • When commenting on reliability, mention both the sample size and the sampling method.

Estimating Population Totals

  • To estimate a population total from a sample, scale up the sample result proportionally.
  • For example, if 30 out of 200 sampled students own a bike, and the school has 1000 students, estimate: \frac{30}{200} \times 1000 = 150 bike owners.
  • This scaling assumes the sample is representative of the whole population.
  • Always state the assumption: the sample is representative and unbiased.

Common Misconceptions

  • Misconception: A larger sample always removes bias. Correction: A biased method stays biased regardless of sample size.
  • Misconception: A convenience sample is random. Correction: Convenience samples are not random and are often biased.
  • Misconception: You can generalise from a sample taken from one narrow group. Correction: The sample must represent the whole population.
  • Misconception: Confusing the sample with the population in calculations. Correction: Always know which group your data comes from.
  • Misconception: Criticising a survey without explaining how it introduces bias. Correction: Always link the criticism to a specific source of bias.

Typical Exam Questions

  • Explain why a sample may be biased – identify the source (method, location, time) and suggest an improvement (e.g. use a random sample from the whole population).
  • Use sample data to estimate a population total – scale up proportionally and state the assumption.
  • Comment on reliability – discuss sample size and representativeness.
  • For a simple random sample, assign unique IDs and generate the required number of distinct random IDs. For systematic sampling, use a random starting position then a fixed interval, checking for periodic patterns.
  • Identify a leading question – a question that suggests a particular answer (e.g. "Don't you agree that school lunches are terrible?") is biased.

Proportional stratified allocation

  • For a population of 500 with year-group sizes 200, 150 and 150, a sample of 20 allocates 8, 6 and 6.
  • Select the allocated number at random within each year group. Round coherently when needed so the final sample size is correct.

A proportional stratified sample

A proportional stratified sampleGroupPopulationSample of 20Year 72008Year 81506Year 91506Total50020

Diapos

Sign up free to view the lesson slides

Step through every slide for this topic — plus flashcards and revision notes — with a free account.

Questions d'entraînement

Aperçu gratuit — 8 sur 63 questions. Inscris-toi pour toutes les voir.
  1. 1.What is the difference between a population and a sample?

    Easy
    • AA population is the entire set of items of interest, and a sample is a subset of that population.
    • BA population is a subset of a sample, and a sample is the entire set of items of interest.
    • CA population is a larger sample, and a sample is a smaller population.
    • DThere is no difference; they are the same thing.
  2. 2.Why do we use samples instead of the whole population?

    Easy
    • ATo save time and money while still getting information about the population.
    • BTo make results more accurate than using the whole population.
    • CBecause samples always give the same results as the population.
    • DTo avoid having to define the population.
  3. 3.A school has 1200 students. A sample of 60 students is taken to estimate the mean number of hours spent on homework. What is the population?

    Easy
    • AThe 60 students sampled
    • BAll 1200 students
    • CThe mean number of hours
    • DHomework hours
  4. 4.A larger sample always removes bias from a survey.

    Easy

    True or false?

  5. 5.A survey about school meals is conducted by asking students in the canteen at lunchtime. Which of the following is the most likely source of bias?

    Medium
    • AThe sample is only taken at lunchtime, so it may miss students who do not eat school meals.
    • BThe sample is too large.
    • CThe sample is too small.
    • DThere is no bias in this method.
  6. 6.Which of the following are ways to reduce bias in a survey? (select all that apply)

    Medium
    • AUse a random sampling method
    • BEnsure the sample is representative of the population
    • CUse a larger sample, even if the method is biased
    • DAvoid leading questions
    • EOnly survey people who are easy to reach
  7. 7.Match each term with its definition.

    Medium
    • Population
    • Sample
    • Bias
    • A part of the population selected for study
    • The entire group of interest
    • A systematic error that makes results unrepresentative
  8. 8.Arrange the following steps for selecting a random sample of 20 students from a school of 500 in the correct order.

    Medium
    • Assign each student a number from 1 to 500
    • Use a random number generator to pick 20 distinct numbers
    • Select the students with those numbers

Unlock all 63 questions & more

Crée un compte gratuit pour voir toutes les questions, les diapos, les cartes mémo et les notes de révision de ce thème.

Annales

Les annales d'entraînement pour ce thème arrivent bientôt.
Bientôt disponible