Loading page…
Loading page…
A short H2 Maths revision video on H2 Maths 6.1 - Conditional Probability via Tree Diagram, built for quick recap before tutorial practice or exam revision.
Read through the explanation after watching, or jump straight to the step you want to replay.
Step 1 - Set up the problem
A disease affects two percent of the population.
Step 1 - Set up the problem
A screening test correctly identifies ninety percent of infected people.
Step 1 - Set up the problem
But the test also returns a false positive five percent of the time for healthy people.
Step 1 - Set up the problem
Given a positive result, find the probability that the person actually has the disease.
Step 2 - Label the tree branches
Draw two first-stage branches: disease, with probability 0.02, and no disease, with probability 0.98.
Step 2 - Label the tree branches
From each node, draw two second-stage branches: positive or negative test result.
Step 2 - Label the tree branches
Write the conditional probabilities beside each second-stage branch.
Step 3 - Multiply along branches
Multiply the probabilities along each path to get the joint probability at each leaf.
Step 3 - Multiply along branches
Disease and positive: 0.02 times 0.90 equals 0.018.
Step 3 - Multiply along branches
No disease and positive: 0.98 times 0.05 equals 0.049.
Step 3 - Multiply along branches
The other two leaves cover the negative results.
Step 4 - Find P(+) using the total probability rule
A positive result can come from two paths: disease and positive, or no disease and positive.
Step 4 - Find P(+) using the total probability rule
Add those two joint probabilities to get the total probability of a positive test.
Step 4 - Find P(+) using the total probability rule
0.018 plus 0.049 gives 0.067.
Step 5 - Apply conditional probability and read the result
Now divide by .
Step 5 - Apply conditional probability and read the result
0.018 divided by 0.067 gives approximately 0.269.
Step 5 - Apply conditional probability and read the result
So despite a positive result, there is only a 27 percent chance the person has the disease.
Step 5 - Apply conditional probability and read the result
Common pitfall: do not use P of D alone - always condition on the positive test by dividing by P of plus.