Probability · Lesson 8 of 8
Chapter Summary and Practice
“Connect event sets, probability models, counting, overlap, and complements in complete problem-solving arguments.”
• Connect all major chapter concepts through a single probability-model framework. • Choose between outcome-weight sums, direct enumeration, and combinatorial counting. • Translate verbal conditions accurately using event algebra. • Combine addition and complement rules in mixed applications. • Audit proposed solutions for equal-likelihood assumptions, overlap errors, and inconsistent weights.
One framework connects the chapter
Probability begins with an uncertain experiment, but solving a problem requires a precise description. First decide which results are recorded and place them in the sample space. Then turn the question into an event. Finally, attach the event’s probability by using a valid model of its elementary outcomes.
The chapter’s later methods do not replace these first steps. Counting is an efficient way to sum equal weights. The addition rule keeps shared outcomes from being counted twice. Complement reasoning finds a region by removing its opposite from the total. Each method is grounded in the same ideas of membership, disjoint regions, and total probability one.
| Chapter idea | Meaning or relationship | What to check when using it |
|---|---|---|
| Sample space and occurrence | S lists possible results; E occurs when ω∈E | The outcome records the correct order, objects, or positions |
| Event types | ∅ is impossible; S is sure; simple means one point, compound more than one | Classify by the event set rather than by expectations |
| Event algebra | Not: A′; or: A∪B; and: A∩B; but not: A∩B′ | Inclusive or permits both; the complement is relative to S |
| Event families | Exclusive means empty intersections; exhaustive means union S | Overlap and coverage are separate tests |
| Probability axioms | Nonnegative weight; P(S)=1; disjoint weights add | All outcome weights must sum to one |
| General event probability | P(E) adds the weights of points in E | Do not assume weights are equal |
| Equally likely counting | P(E)=n(E)/n(S) | The counted elementary outcomes must have equal chances |
| Arrangements and selections | P(n,r) records order; C(n,r) ignores order | Match the numerator and denominator conventions |
| Two-event addition | Add both probabilities, subtract the overlap once | A∩B must be known or correctly derived |
| Three-event addition | Singles minus pairs plus the triple | Pair intersections already include triple members |
| Complements and mixed events | P(A′)=1−P(A); neither, not both, only one use different regions | Translate the sentence before doing arithmetic |
A partition provides a particularly useful connection. If a family of events is pairwise disjoint and exhaustive, exactly one category occurs, and their probabilities add to one. The three-toss head-count categories and the four regions of two events are examples. A and A′ are the simplest two-part partition.
The formulas as consequences of the model
Rather than memorising isolated formulas, remember what each one counts. For any finite model, an event’s weight comes from its member outcomes. Equal weights then give a counting fraction. Grouping those outcomes into disjoint regions yields the union and complement rules.
Choosing a method before calculating
Begin by asking whether probabilities are supplied as weights or whether the outcomes are specified as equally likely. If weights are supplied, sum the weights in the required event. If the outcomes are equally likely, count them by listing, a grid, permutations, or combinations. For an at-least-one event, compare that direct count with the simpler no-success complement.
Next, check how events overlap. A union of disjoint cases needs no overlap correction. An overlapping union needs the intersection subtracted once. An exactly-one condition needs the shared region excluded entirely. A final answer should be accompanied by the set or counting reason that makes the arithmetic valid.
Problem
S={1,2,3,4,5,6} has weights 0.05,0.10,0.15,0.20,0.20,0.30. Let A={1,3,5} and B={3,4,5}. Find the union, neither, and exactly-one probabilities.
- 1.The weights are nonnegative and total 1, so the model is valid.
- 2.P(A)=0.05+0.15+0.20=0.40; P(B)=0.15+0.20+0.20=0.55.
- 3.A∩B={3,5}, with probability 0.15+0.20=0.35.
- 4.The union has weight 0.40+0.55−0.35=0.60. Its members {1,3,4,5} directly give the same sum.
- 5.Neither has weight 1−0.60=0.40, belonging to {2,6}. Exactly one has weight 0.40+0.55−2×0.35=0.25, belonging to {1,4}.
- 6.A count such as 4/6 for the union would be invalid because the six elementary weights differ.
Problem
For two fair labelled dice, let A mean the first die is even and B mean the total is 7. Find the chances of at least one condition, exactly one condition, and neither.
- 1.Use 36 equally likely ordered pairs. A has 3×6=18 points. B has six points, one for each first score.
- 2.Their intersection is {(2,5),(4,3),(6,1)}, with three points.
- 3.The union count is 18+6−3=21, so its probability is 21/36=7/12.
- 4.Exactly one has count 18+6−2×3=18, giving 1/2.
- 5.Neither has count 36−21=15, giving 5/12. Its complement check is 7/12+5/12=1.
Problem
Choose three people uniformly from five men and four women. Find the chance of exactly two women and of at least one woman.
- 1.Groups are unordered; there are C(9,3)=84 possible groups.
- 2.Exactly two women requires C(4,2)C(5,1)=6×5=30 groups. Its probability is 30/84=5/14.
- 3.The opposite of at least one woman is no woman. All three must then be chosen from the five men, in C(5,3)=10 ways.
- 4.The at-least-one probability is 1−10/84=74/84=37/42.
- 5.Exactly two and at least one are different events. The second includes one, two, and three women.
Problem
Four cities A,B,C,D are visited in a uniform random order. Find the chance that A is before B but not immediately before B.
- 1.Half of the 24 orders have A before B, so there are 12 such orders.
- 2.Treating AB as a fixed block gives 3!=6 orders with A immediately before B.
- 3.Every immediately-before order is contained in the before event. Removing these six from the twelve leaves six orders with at least one city between A and B.
- 4.The probability is 6/24=1/4. Subtracting event weights is valid here because the smaller event is genuinely a subset of the larger.
Problem
Let three events have probabilities 0.50,0.40,0.30, pairwise intersections 0.20,0.15,0.10, and triple intersection 0.05. Find the chances of at least one and of none.
- 1.The union is 0.50+0.40+0.30−0.20−0.15−0.10+0.05=0.80.
- 2.None is the complement of that union, so its probability is 1−0.80=0.20.
- 3.The triple term must be added back because its points were included in all three pairwise subtraction terms.
- 4.The data are consistent: the three single-only regions are 0.20,0.15,0.10; the three pair-only regions are 0.15,0.10,0.05; and the triple region is 0.05. These sum to 0.80, with outside weight 0.20.
Auditing a probability argument
A numerical answer alone can hide an incorrect sample space. Ask whether the elementary results are genuinely equally likely, whether repeated labels represent different objects, and whether order matters. Then inspect the requested wording: exactly, at least, at most, neither, and not both can describe very different sets.
Look for impossible region weights. An intersection cannot exceed either event probability, a union cannot be smaller than either event, and none of the disjoint pieces may have negative weight. For numerical bounds, check strict inequalities explicitly. For objects selected without replacement, the available counts change after a choice; the same object cannot silently reappear.
A displayed area in a set diagram does not automatically encode probability. Unless an area scale is specified, use it to understand membership and use the assigned outcome weights to calculate chance.
To test an addition or complement answer, reconstruct disjoint cases whose probabilities should sum to one. To test a counting answer, use an alternative but consistent outcome description when it is manageable.
From games of chance to an axiomatic theory
Questions about games of chance helped motivate early probability mathematics. Gerolamo Cardano, also known as Jerome Cardan, wrote Book on Games of Chance; his work was published after his death in 1663. In 1654, a dice question associated with the Chevalier de Méré led to correspondence between Blaise Pascal and Pierre de Fermat. Their work showed how careful counting and reasoning could replace unreliable intuition about chance.
Later developments involved Christiaan Huygens, Jacob Bernoulli, Abraham de Moivre, Pierre-Simon Laplace, Pafnuty Chebyshev, and Andrey Markov. Andrey Kolmogorov’s 1933 Foundations of Probability gave probability its modern axiomatic foundation. The chapter’s set language and rules reflect this development: a probability function must describe events consistently, whether they arise from cards, manufactured objects, selections, or other uncertain situations.
Quiz
Which procedure is valid for an event in a finite unequal-weight model?
What two properties make a family a partition of S?
For three fair tosses, why are head counts 0,1,2,3 not equally likely?
A two-event model has P(A)=0.4, P(B)=0.5, and overlap 0.2. What is the probability of exactly one?
Which selection uses a combination count as its elementary outcome description?
If P(A∪B)=0.72, what is the probability that neither occurs?
Why is the triple intersection added in three-event inclusion–exclusion?
An event has four of six outcomes. What is needed before concluding its probability is 2/3?
Practice Problems
- For three ordered tosses, create one simple event, one compound event, an impossible event, and a sure event. Describe the occurrence test for each.
- For a die, A={1,2,3,4} and B={3,4,5}. Write the union, intersection, A−B, B−A, both complements, and the exactly-one event.
- Construct three events on a die that are exhaustive but not mutually exclusive. Then modify them to form a partition and explain both tests.
- For S={a,b,c,d,e}, propose valid unequal weights. Find an event whose counting fraction differs from its actual probability.
- For the weights 0.2,0.3,0.1,x, find x and the probability of the event containing the first and fourth outcomes. Explain why the model is valid.
- A fair coin is tossed four times. Find the chances of exactly one head, at least one head, and at most one head. Show the favourable sequence counts.
- On two fair labelled dice, let A be total at least 10 and B a 6 on the first die. Find P(A),P(B),P(A∩B),P(A∪B), and the probability of neither.
- Choose three people uniformly from four women and five men. Find the chances of no women, exactly two women, and at least one woman. Check your use of order.
- For a random order of four distinct objects A,B,C,D, find the probabilities that A is first, A is before B, and A is immediately before B. Explain why the last two differ.
- Draw five cards uniformly without replacement from a standard deck. Find the chances of exactly two aces and at least one ace, stating all combination counts.
- Given P(A)=0.45,P(B)=0.55,P(A∩B)=0.20, calculate all four disjoint region weights, the union, neither, and exactly one.
- Check the consistency of P(A)=0.7,P(B)=0.6,P(A∩B)=0.1. Use the addition rule and identify which region would become impossible.
- Derive the three-event addition rule, then find the union for single probabilities 0.5,0.4,0.3, pairwise probabilities 0.2,0.15,0.1, and triple probability 0.05.
- Three objects are placed uniformly into three matching labelled positions. Find the chance that none is correctly placed and the chance that at least one is correctly placed.
- Explain the error in each claim: three die-number labels must have probability 1/3 each; a union always adds probabilities directly; neither equals not both; all four-digit codes have a nonzero first digit.
Key Takeaways
• Define the sample space and event before calculating probability. • The axioms connect outcome weights, disjoint sums, and total weight one. • Equal-likelihood counting is useful only for equally weighted elementary outcomes. • Sets distinguish overlap, exclusivity, exhaustiveness, and complements. • Permutations and combinations must match whether outcomes record order. • Union, complement, and inclusion–exclusion rules combine the same region weights in different ways.
Previous · Lesson 7
Complements and Multi-Step Probability
Next
End of chapter