MATH19661 Chap.11 Probability and Conditional Probability
Probability and Conditional Probability
Probability and Conditional Probability covers the Week 11 lecture notes and the final question of each January paper. It begins with the language of experiments: a sample space listing every outcome, events as subsets of it, and the set operations union, intersection and complement read as OR, AND and NOT.
Kolmogorov's three axioms follow, together with their consequences, including the complement rule and the union rule that subtracts the overlap once. When outcomes are equally likely, probability reduces to counting, and the chapter shows how to count ordered pairs from two dice without gaps or repeats.
Outcomes that are not equally likely, such as a biased coin, are handled by multiplying the probabilities of independent stages. Conditional probability then restricts attention to the part of the sample space where a known event has happened, which leads to a precise test for independence.
The chapter ends with total probability and Bayes' theorem, using a tree to reverse a conditional in a defect-testing setting, and it highlights the reasons examiners give for lost marks on these parts, from an independence verdict with no reason to an answer that never writes a valid conditional-probability expression.
What this chapter covers
- 01
Sample spaces, outcomes and events
- 02
Union, intersection, complement and disjoint events
- 03
Kolmogorov's axioms and the union rule
- 04
Counting with equally likely outcomes
- 05
Conditional probability
- 06
Testing independence with a reason
- 07
Total probability and Bayes' theorem on a tree
Worked example · free
Use a probability tree to reverse a conditional after a positive test
- 1Define events: L = the valve leaks, T = the test is positive.
- 1Translate the data: P(L) = 0.10, P(T | L) = 0.85, P(T | L′) = 0.15.
- 1Branch products: P(T ∩ L) = 0.10 × 0.85 = 0.085 and P(T ∩ L′) = 0.90 × 0.15 = 0.135.
- 1Total probability: P(T) = 0.085 + 0.135 = 0.220.
- 1Bayes' theorem: P(L | T) = 0.085 / 0.220 ≈ 0.386.
Key terms
- Sample space
- The set of every possible outcome of an experiment, from which all events are formed as subsets.
- Conditional probability
- The probability of one event once another is known to have occurred, found by dividing the probability of both by the probability of the known event.
- Independent events
- Two events for which the probability of both equals the product of their separate probabilities.
- Bayes' theorem
- A formula that reverses a conditional probability, expressing P(D | X) through P(X | D), P(D) and P(X).
- Disjoint events
- Events with no outcome in common, so their intersection is the empty set.
Probability and Conditional Probability FAQ
How do I show that two events are independent?
Compute the product of the two separate probabilities and compare it with the given or calculated probability of both events. Write both numbers and state whether they match. The official 2025 solutions gave no credit to a plain yes, so the comparison itself is the answer the mark rewards.
Are disjoint events the same as independent events?
No. Events that are disjoint cannot happen together, so the probability of both is zero. If each has a positive probability, the product is positive, which means they are not independent: learning that one occurred tells you the other did not. Keep the two ideas separate.
When can I just count outcomes to get a probability?
Only when every outcome in the sample space is equally likely, as with fair coins, fair dice or a well-shuffled pack. With a biased coin or weighted die, find the probability of each outcome first: for a weighted die from weights that sum to 1, and for repeated trials by multiplying independent stages, and then add the outcomes that make up the event.
What is total probability and why does Bayes need it?
Total probability splits an event across cases that cover everything, such as faulty and sound items, and adds the branch products. Bayes' theorem divides one branch product by that total, so without the total you cannot reverse the conditional. A tree diagram keeps both steps visible.
Why does a positive test often give a small probability of a fault?
When faults are rare, the false positives from the large group of sound items can outnumber the true positives from the small faulty group. In the chapter example only about 37% of positives are faulty, even though the test catches 90% of faults, because sound items dominate.
Exam move
Start each session by writing out a sample space by hand, whether three coin tosses or two dice, and counting events from it, since the feedback links correct answers to an itemised sample space. Next, drill the three-number set: given P(A), P(B) and one of the intersection or union, find the other, the conditional and an independence verdict with the comparison written out.
Then practise the tree routine on one percentage problem per day, labelling each given figure as a conditional before drawing. Check every tree by confirming that the tip probabilities total 1. Finish with the probability questions from the 2024 and 2025 papers, comparing your layout with the official solutions so each step that earns a mark is visible.
Working through Probability and Conditional Probability in MATH19661? Sia is AskSia’s AI Mathematics tutor — ask any MATH19661 Probability and Conditional Probability question and get a clear, step-by-step explanation grounded in how MATH19661 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.