Conditional Probability — SL 4.4 Conditional probability and independence - IB Mathematics Analysis and Approaches

SECTION 1

Historical Context & Motivation

Probability has been studied for centuries, but the idea of conditional probability — how the likelihood of an event changes when you already know something else has happened — transformed it from a gambler's tool into a rigorous branch of mathematics. Early thinkers realized that context matters: drawing a second card from a deck after seeing the first card is fundamentally different from drawing with no prior knowledge. This insight drives medical testing, weather forecasting, criminal investigations, and machine learning today.

1654 • Pascal & Fermat Correspondence

Blaise Pascal and Pierre de Fermat exchanged letters about gambling problems, laying the groundwork for probability theory and the idea that prior outcomes influence future chances.

1763 • Bayes' Theorem Published

Thomas Bayes' posthumous essay introduced the first formal framework for updating probabilities based on new evidence — the mathematical heart of conditional probability.

1812 • Laplace's Théorie Analytique

Pierre-Simon Laplace extended and systematized Bayes' ideas, defining conditional probability with the formula still used today and applying it to astronomy and demographics.

1933 • Kolmogorov's Axioms

Andrey Kolmogorov published his axiomatic foundation for probability, formally defining conditional probability and independence as cornerstones of modern statistics.

The central question this topic addresses is deceptively simple: if you already know that one event has occurred, how does that change the probability of another event? Understanding this question, and knowing when the answer is "it doesn't change it at all" (independence), is one of the most powerful tools in the IB Mathematics course.

SECTION 2

Core Principles & Definitions

Before diving into formulas, it helps to build strong intuition for the key ideas. Conditional probability is about restricting the sample space. When you learn that a certain event has already happened, you are effectively shrinking the universe of possible outcomes to only those that are consistent with that information. The probability of another event is then recalculated within this smaller universe.

Conditional Probability P(A | B)

The probability that event A occurs given that event B has already occurred. Read as "the probability of A given B."

Joint Probability P(A ∩ B)

The probability that both events A and B occur together. This is the overlap region in a Venn diagram and is essential for computing conditional probabilities.

Independent Events

Two events are independent if the occurrence of one does not affect the probability of the other. Mathematically, P(A | B) = P(A), which leads to P(A ∩ B) = P(A) × P(B).

Dependent Events

Two events are dependent if knowing one has occurred changes the probability of the other. Drawing cards without replacement is a classic example of dependence.

Mutually Exclusive Events

Events that cannot happen at the same time — P(A ∩ B) = 0. Mutually exclusive events are always dependent (unless one has probability zero), because if A happened, B definitely did not.

✦

KEY TAKEAWAY

Think of conditional probability like a filter on a search engine. If you search "shoes" you get millions of results (the full sample space). But if you filter by "red" first, you've restricted the results to only red items — and now the proportion that are shoes is different. P(A | B) is the proportion of event B's outcomes that also belong to event A.

SECTION 3

Visual Explanation — Venn Diagram

A Venn diagram is one of the clearest ways to see how conditional probability works. The entire rectangle represents the sample space S. Two overlapping circles represent events A and B. The overlap is the intersection A ∩ B. When we compute P(A | B), we zoom in on circle B and ask: what fraction of B is also in A?

The pink overlap region represents A ∩ B. To find P(A | B), we divide the probability of the overlap (0.10) by the total probability of B (0.40), yielding 0.25. Visually, we are asking: "what fraction of circle B is pink?"

Notice how the conditional probability P(A | B) = 0.25 is different from the unconditional probability P(A) = 0.35. This tells us that A and B are dependent events — knowing that B occurred actually lowers the chance of A in this example. If A and B were independent, we would have P(A | B) = P(A) = 0.35, and the overlap would equal P(A) × P(B) = 0.35 × 0.40 = 0.14 instead of 0.10.

SECTION 4

Mathematical Framework

The formulas for conditional probability and independence are concise but powerful. Make sure you understand what each variable represents before memorizing the equations.

CONDITIONAL PROBABILITY FORMULA

P(A | B) = P(A ∩ B) / P(B), where P(B) > 0

P(A | B) = probability of A given B has occurred; P(A ∩ B) = probability of both A and B occurring; P(B) = probability of event B.

MULTIPLICATION RULE (REARRANGED)

P(A ∩ B) = P(A | B) × P(B)

This rearrangement is extremely useful. It lets you find the joint probability when you know the conditional probability. It also works as P(A ∩ B) = P(B | A) × P(A).

TEST FOR INDEPENDENCE

A and B are independent ⟺ P(A ∩ B) = P(A) × P(B)

Equivalently, A and B are independent if and only if P(A | B) = P(A). If this equality holds, knowing B gives you no additional information about A.

COMPLEMENT RULE FOR CONDITIONAL PROBABILITY

P(A′ | B) = 1 − P(A | B)

Within the restricted sample space of B, the probabilities of A and not-A still sum to 1. This is handy when it's easier to calculate the complement.

⚠

Common Mistake

Students often confuse P(A | B) with P(B | A). These are generally not equal. "The probability it rains given that there are clouds" is very different from "the probability there are clouds given that it rains." Always identify which event is the 'given' (the condition) and which is the event whose probability you seek.

SECTION 5

Tree Diagrams & Two-Way Tables

Two of the most useful tools for organizing conditional probability problems are tree diagrams and two-way tables. A tree diagram shows sequential events as branches, where the probability on each branch is a conditional probability given the path taken so far. You multiply along branches to get joint probabilities and add across final outcomes to get total probabilities.

A tree diagram for a medical test. D = has disease (2% prevalence), T⁺ = tests positive. The test has 95% sensitivity (correctly detects disease) and 3% false-positive rate. Multiply along each branch to get joint probabilities on the right.

The tree diagram above illustrates a classic scenario that reveals a surprising result. Even though the test is 95% accurate at detecting the disease, a person who tests positive actually has a probability of only P(D | T⁺) = 0.019 / (0.019 + 0.0294) ≈ 0.392 of truly having the disease. This is because the disease is rare (2% prevalence), so the false positives from the large healthy population outnumber the true positives. This counterintuitive result, sometimes called the base-rate fallacy, shows why understanding conditional probability is so important in real life.

	T⁺ (Positive)	T⁻ (Negative)	Total
D (Disease)	0.019	0.001	0.020
D′ (No Disease)	0.0294	0.9506	0.980
Total	0.0484	0.9516	1.000

Two-way table equivalent of the tree diagram above. Each cell shows a joint probability.

SECTION 6

Worked Example

Let's work through a full problem that combines conditional probability and testing for independence — the type of question that commonly appears on IB exams.

Are Two Events Independent?

1
Step 1 — Read the Problem

In a class of 30 students, 18 study Biology (B), 12 study Chemistry (C), and 6 study both Biology and Chemistry. A student is chosen at random. Find P(B | C) and determine whether B and C are independent events.
2
Step 2 — Identify the Given Values

P(B) = 18/30 = 3/5, P(C) = 12/30 = 2/5, P(B ∩ C) = 6/30 = 1/5.

P(B) = 0.6, P(C) = 0.4, P(B ∩ C) = 0.2
3
Step 3 — Apply the Conditional Probability Formula

P(B | C) = P(B ∩ C) / P(C) = (1/5) / (2/5) = (1/5) × (5/2) = 1/2.

P(B | C) = 0.5
4
Step 4 — Test for Independence

For independence we need P(B | C) = P(B). We found P(B | C) = 0.5 but P(B) = 0.6. Since 0.5 ≠ 0.6, the events are not independent. We can verify with the product rule: P(B) × P(C) = 0.6 × 0.4 = 0.24 ≠ 0.2 = P(B ∩ C).

B and C are dependent events.
5
Step 5 — Interpret the Result

Knowing a student studies Chemistry slightly reduces the probability that they also study Biology (from 60% to 50%). This makes intuitive sense: students who chose Chemistry may be less likely to also take Biology due to timetable clashes or subject preferences.

SECTION 7

Common Pitfalls & Comparisons

Many mistakes in conditional probability come from confusing similar-sounding concepts. The table below highlights the most frequent errors and how to avoid them.

Pitfall	What Students Do	What They Should Do
Swapping conditions	Treat P(A \| B) as the same as P(B \| A).	Carefully identify which event is 'given.' The denominator in the formula is always P(given event).
Confusing independent and mutually exclusive	Assume that events that can't both happen are independent.	Mutually exclusive events have P(A ∩ B) = 0, so P(A \| B) = 0 ≠ P(A) — they are dependent (unless one has probability 0).
Using P(A) × P(B) without checking	Multiply probabilities as if events are always independent.	Only use P(A ∩ B) = P(A) × P(B) after verifying independence. Otherwise use the multiplication rule P(A ∩ B) = P(A \| B) × P(B).
Forgetting the restricted sample space	Compute P(A \| B) using the full sample space instead of restricting to B.	When B is given, only outcomes inside B matter. The denominator must be P(B), not 1.

✦

KEY TAKEAWAY

Independence and mutual exclusivity are opposites in spirit. Independent events don't affect each other at all — knowing one tells you nothing about the other. Mutually exclusive events affect each other maximally — knowing one happened guarantees the other didn't. Think of it like two friends at a party: independent means they came separately and don't interact, while mutually exclusive means only one of them was invited.

SECTION 8

Connection to Advanced Theory

The conditional probability formula you've learned in SL 4.4 is the foundation for several more advanced topics. As you continue in mathematics and statistics, these ideas become increasingly powerful.

SL 4.4 Concept	Advanced Extension	Where It Leads
P(A \| B) = P(A ∩ B) / P(B)	Bayes' Theorem: P(A \| B) = P(B \| A) × P(A) / P(B)	HL Mathematics, medical diagnostics, spam filters, AI and machine learning
Independence: P(A ∩ B) = P(A) × P(B)	χ² test for independence	Testing whether two categorical variables are associated using real data (IB HL Topic 4)
Tree diagrams with conditional branches	Markov chains	Modeling systems where the next state depends only on the current state — used in Google PageRank, genetics, economics
Multiplication rule for dependent events	Conditional distributions	University-level probability: defining entire probability distributions conditional on other random variables

The most immediate extension you might encounter is Bayes' Theorem, which lets you "reverse" a conditional probability. In the medical test tree diagram from Section 5, we used Bayes' Theorem (without naming it) when we calculated P(D | T⁺) from P(T⁺ | D). If you study HL Mathematics or take statistics at university, Bayes' Theorem will become one of your most-used tools.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A fair coin is tossed twice. Event A is "the first toss is heads" and event B is "both tosses give the same result." Are A and B independent? Explain your reasoning without performing any calculations.

Show answer

Yes, A and B are independent. The sample space is {HH, HT, TH, TT}. P(A) = 2/4 = 1/2 and P(B) = 2/4 = 1/2 (the matching outcomes are HH and TT). P(A ∩ B) = P(HH) = 1/4. Since P(A) × P(B) = 1/2 × 1/2 = 1/4 = P(A ∩ B), the events are independent. Intuitively, knowing the first toss is heads doesn't change the chance of matching — it's still 50% because the second toss is equally likely to be H or T.

Question 2: BASIC CALCULATION

In a group of 50 students, 30 play football (F) and 20 play basketball (B). 10 students play both sports. Find P(F | B).

Show answer

Using the conditional probability formula: P(F | B) = P(F ∩ B) / P(B). We have P(F ∩ B) = 10/50 = 0.2 and P(B) = 20/50 = 0.4. Therefore P(F | B) = 0.2 / 0.4 = 0.5. This means that among the students who play basketball, exactly half of them also play football. Note that this is different from P(F) = 30/50 = 0.6, confirming that F and B are dependent events.

Question 3: INTERMEDIATE

A bag contains 5 red and 3 blue marbles. Two marbles are drawn without replacement. Find the probability that the second marble is red given that the first marble was blue.

Show answer

Let R₂ = "second marble is red" and B₁ = "first marble is blue." After drawing a blue marble first, the bag contains 5 red and 2 blue marbles (7 total). So P(R₂ | B₁) = 5/7 ≈ 0.714. This is a straightforward application of conditional probability where the sample space shrinks after each draw. Compare this to the unconditional probability of drawing red from the original bag: 5/8 = 0.625. The probability increased because removing a blue marble shifts the composition toward red.

Question 4: APPLIED

A factory has two machines. Machine A produces 60% of all items and Machine B produces 40%. The defect rate for Machine A is 3% and for Machine B is 5%. An item is selected at random and found to be defective. What is the probability it was produced by Machine A?

Show answer

Let A = "produced by Machine A", D = "defective." We need P(A | D). First, find P(D) using the law of total probability: P(D) = P(D | A) × P(A) + P(D | B) × P(B) = 0.03 × 0.60 + 0.05 × 0.40 = 0.018 + 0.020 = 0.038. Then apply Bayes' Theorem (which is the conditional probability formula rearranged): P(A | D) = P(D | A) × P(A) / P(D) = 0.018 / 0.038 = 9/19 ≈ 0.474. Despite Machine A producing more items overall, it is actually less likely that a defective item came from Machine A because Machine A has a lower defect rate. This problem connects directly to the tree diagram method covered in Section 5.

Question 5: CRITICAL THINKING

Events A and B satisfy P(A) = 0.4, P(B) = 0.5, and P(A ∪ B) = 0.7. (a) Find P(A ∩ B). (b) Determine whether A and B are independent. (c) Find P(A′ | B′) — the probability that A does not occur given that B does not occur.

Show answer

(a) Using the addition rule: P(A ∪ B) = P(A) + P(B) − P(A ∩ B), so 0.7 = 0.4 + 0.5 − P(A ∩ B), giving P(A ∩ B) = 0.2. (b) Check: P(A) × P(B) = 0.4 × 0.5 = 0.2 = P(A ∩ B). Since the product equals the joint probability, A and B are independent. (c) P(B′) = 1 − 0.5 = 0.5. P(A′ ∩ B′) = P((A ∪ B)′) = 1 − P(A ∪ B) = 1 − 0.7 = 0.3. Therefore P(A′ | B′) = 0.3 / 0.5 = 0.6 = P(A′). Notice that since A and B are independent, A′ and B′ are also independent, which is confirmed by P(A′ | B′) = P(A′) = 0.6. This is an important property: independence of complements follows from independence of the original events.

SUMMARY

Lesson Summary

Summary

Conditional probability measures how the probability of an event changes when you know another event has occurred. The formula P(A | B) = P(A ∩ B) / P(B) captures this by restricting the sample space to only the outcomes where B is true. The multiplication rule rearranges this to P(A ∩ B) = P(A | B) × P(B), which is essential for tree diagram calculations. Two events are independent if and only if P(A | B) = P(A), equivalently P(A ∩ B) = P(A) × P(B). Independence means one event gives no information about the other.

Key tools for solving problems include Venn diagrams (for visualizing overlaps), tree diagrams (for sequential events with branches showing conditional probabilities), and two-way tables (for organizing joint and marginal probabilities). Always be careful not to confuse P(A | B) with P(B | A), and remember that mutually exclusive events are dependent, not independent. These concepts form the basis for Bayes' Theorem, the chi-squared test for independence, and much of university-level statistics.