Sampling and Bias in Samples
Judge whether a sample can speak for a population, and why bigger is not always better
A taste of a lesson
Our app has 40,000 reviews averaging 4.6 stars. Isn't that a reliable measure of user satisfaction?
It is a big sample, but size does not fix who chose to write a review. People who review are often very happy or very annoyed, and many apps prompt reviews right after a good moment, which tilts ratings up. Users who quietly stopped using the app never appear at all. So 4.6 tells you about reviewers, not all users. A small random survey of active and lapsed users could be more representative. Which group do you think is missing most from those 40,000?
Written by the teacher as an example. In your lesson the tutor answers your own questions, and like any AI it can be wrong.
What you will be able to do
- Distinguish random sampling error from bias
- Explain why larger samples reduce error but not bias
- Identify selection, non response and survivorship bias
- Describe simple, stratified and cluster sampling designs
- Write an honest note on a sample's limitations
Lesson plan
- 1 Population, frame and sample Identify the population, the sampling frame and the actual sample. Start
- 2 Random error versus bias Understand which errors shrink with more data and which do not. Start
- 3 Who chose to answer? Spot voluntary response and non response bias. Start
- 4 Survivors and convenience Recognise survivorship and convenience sampling. Start
- 5 Better sampling designs Compare simple random, stratified and cluster sampling. Start
- 6 Describing a sample honestly Write a limitations note that tells readers what the sample can support. Start
Try asking
About this tutor
A beginner tutor on the question behind every survey, poll and dataset: who is in the sample, and who is missing? You will learn the difference between random error and bias, why sample size reduces one but not the other, and how selection, non response, survivorship and convenience sampling distort results. Lessons introduce simple, stratified and cluster sampling and show how to describe a sample's limits honestly. Examples include customer feedback forms, app store reviews, online polls and datasets scraped from the web for AI training.
Reviews
4.5
2 ratingsSample
- Ama O.Sample
The guest list framing is so simple and so useful. We changed how we sample our customer survey after this.
- Henrik S.Sample
Good explanation of why our huge feedback dataset was biased. The AI training data example was a nice touch.
About the teacher
Statistics in plain language, from averages to Bayesian reasoning
9 tutors 350 lessons taught Sample
I teach statistics to people who were put off by it the first time. My approach is to start from a question someone actually has, simulate or count our way to an answer, and only then name the formula. I have worked as an analyst on survey and health research projects, so I have a soft spot for messy samples,...
See Lina's profile and tutorsMore like this
Other tutors on the same or nearby topics.