Conditional Probability Problems & Exercises | Practice

Conditional probability is the likelihood of an event occurring given that another event has already occurred. Formally, for two events \(A\) and \(B\) with \(P(B) \neq 0\), the conditional probability of \(A\) given \(B\) is defined by the formula \( P(A \mid B) = \dfrac{P(A \cap B)}{P(B)} \). Mastering this concept is essential for understanding Bayes’ theorem, the multiplication rule, independent events, tree diagrams, Venn diagrams, and the law of total probability. The problems below are organized by theme and increase progressively in difficulty from easy to hard.

Direct Application of the Conditional Probability Formula

These problems build fluency with the core definition \( P(A \mid B) = P(A \cap B) / P(B) \). You will practice identifying the condition, the event of interest, and the relevant probabilities — the essential first step before tackling more complex scenarios.

Problem 1: Numerical Probability Values

Easy

Let \(A\) and \(B\) be two events defined on the same sample space. You are given that \( P(A \cap B) = 0.12 \) and \( P(B) = 0.40 \).

  1. Calculate \( P(A \mid B) \).
  2. If additionally \( P(A) = 0.30 \), determine whether events \(A\) and \(B\) are independent. Justify your answer.
Hint
Apply the conditional probability formula directly. For independence, recall that \(A\) and \(B\) are independent if and only if \( P(A \mid B) = P(A) \).
View Solution
Solution to question 1:
\[
P(A \mid B) = \frac{P(A \cap B)}{P(B)} = \frac{0.12}{0.40} = 0.30
\]
Solution to question 2:

Since \( P(A \mid B) = 0.30 = P(A) \), the occurrence of \(B\) does not affect the probability of \(A\). Therefore, events \(A\) and \(B\) are independent.

Problem 2: Drawing Cards Without Replacement

Easy

A standard deck of 52 cards is well shuffled. Two cards are drawn one at a time without replacement.

  1. What is the probability that the second card is a king, given that the first card drawn was also a king?
  2. What is the probability that the second card is a heart, given that the first card drawn was the king of hearts?
Hint
After the first card is removed, the deck has only 51 cards. Update both the number of favorable outcomes and the total sample space accordingly.
View Solution
Solution to question 1:

There are 4 kings in the deck. After one king is drawn and not replaced, 3 kings remain in a deck of 51 cards.

\[
P(\text{2nd king} \mid \text{1st king}) = \frac{3}{51} = \frac{1}{17} \approx 0.0588
\]
Solution to question 2:

The king of hearts is a heart. After removing it, 12 hearts remain among 51 cards.

\[
P(\text{2nd heart} \mid \text{1st is king of hearts}) = \frac{12}{51} = \frac{4}{17} \approx 0.2353
\]

Problem 3: Conditional Probability from a Two-Way Table

Medium

A survey of 200 university students recorded their major (Science or Arts) and whether they use the library daily. The results are shown below.

Daily Library UserNot a Daily UserTotal
Science5466120
Arts324880
Total86114200
  1. Find the probability that a randomly selected student is a daily library user given they study Science.
  2. Find the probability that a randomly selected student studies Arts given they are not a daily library user.
  3. Are the events “studies Science” and “daily library user” independent? Justify using probabilities.
Hint
For each conditional probability, identify the row or column that defines the condition, then restrict your count to that subgroup. For independence, compare \( P(\text{Science} \cap \text{Daily}) \) with \( P(\text{Science}) \times P(\text{Daily}) \).
View Solution
Solution to question 1:
\[
P(\text{Daily} \mid \text{Science}) = \frac{54}{120} = \frac{9}{20} = 0.45
\]
Solution to question 2:
\[
P(\text{Arts} \mid \text{Not Daily}) = \frac{48}{114} = \frac{8}{19} \approx 0.421
\]
Solution to question 3:

Check independence: \( P(\text{Science}) = \frac{120}{200} = 0.60 \) and \( P(\text{Daily}) = \frac{86}{200} = 0.43 \).

\[
P(\text{Science}) \times P(\text{Daily}) = 0.60 \times 0.43 = 0.258
\]
\[
P(\text{Science} \cap \text{Daily}) = \frac{54}{200} = 0.27
\]

Since \( 0.27 \neq 0.258 \), the two events are not independent; knowing a student studies Science slightly increases the probability they use the library daily.

Multiplication Rule and Dependent Events

The multiplication rule connects conditional probability to joint probability: \( P(A \cap B) = P(B) \cdot P(A \mid B) \). This section covers sequential experiments, sampling without replacement, and event chains where each outcome changes the conditions for the next.

Problem 4: Colored Balls in a Bag

Easy

A bag contains 5 red balls and 7 blue balls. Two balls are drawn one after the other without replacement.

  1. Find the probability that both balls drawn are red.
  2. Find the probability that the first ball is red and the second ball is blue.
Hint
Use the multiplication rule: \( P(A \cap B) = P(A) \cdot P(B \mid A) \). After drawing the first ball, the total count decreases by 1, and the composition of the bag changes.
View Solution
Solution to question 1:
\[
P(\text{R}_1 \cap \text{R}_2) = P(\text{R}_1) \cdot P(\text{R}_2 \mid \text{R}_1) = \frac{5}{12} \cdot \frac{4}{11} = \frac{20}{132} = \frac{5}{33} \approx 0.152
\]
Solution to question 2:
\[
P(\text{R}_1 \cap \text{B}_2) = P(\text{R}_1) \cdot P(\text{B}_2 \mid \text{R}_1) = \frac{5}{12} \cdot \frac{7}{11} = \frac{35}{132} \approx 0.265
\]

Problem 5: Quality Control on a Production Line

Medium

A factory ships boxes of components. Each box contains 20 components, of which 3 are defective. A quality inspector draws components one at a time without replacement.

  1. Find the probability that the first two components drawn are both defective.
  2. Find the probability that exactly one of the first two components drawn is defective.
  3. Given that at least one of the first two components is defective, find the probability that both are defective.
Hint
For part 2, consider the two disjoint cases: (defective then good) and (good then defective). For part 3, use the definition of conditional probability with the events “both defective” and “at least one defective.”
View Solution
Solution to question 1:
\[
P(D_1 \cap D_2) = \frac{3}{20} \cdot \frac{2}{19} = \frac{6}{380} = \frac{3}{190} \approx 0.0158
\]
Solution to question 2:
\[
P(\text{exactly one defective}) = P(D_1 \cap G_2) + P(G_1 \cap D_2)
\]
\[
= \frac{3}{20} \cdot \frac{17}{19} + \frac{17}{20} \cdot \frac{3}{19} = \frac{51}{380} + \frac{51}{380} = \frac{102}{380} = \frac{51}{190} \approx 0.268
\]
Solution to question 3:

Let \(E\) = “both defective” and \(F\) = “at least one defective”. Note \(E \subseteq F\), so \(P(E \cap F) = P(E)\).

\[
P(F) = P(\text{exactly one}) + P(\text{both}) = \frac{102}{380} + \frac{6}{380} = \frac{108}{380} = \frac{27}{95}
\]
\[
P(E \mid F) = \frac{P(E)}{P(F)} = \frac{6/380}{108/380} = \frac{6}{108} = \frac{1}{18} \approx 0.0556
\]

Tree Diagrams and Sequential Experiments

Tree diagrams are a powerful visual tool for organizing multi-stage probability problems. Each branch represents a conditional probability, and the probability of any complete path is found by multiplying the branch probabilities along that path. These exercises develop the systematic thinking required for complex real-world scenarios.

Problem 6: Weather and Commuting Delay

Easy

On any given weekday, the probability of rain is \(0.30\). If it rains, the probability of a commuting delay is \(0.70\). If it does not rain, the probability of a commuting delay is \(0.15\).

  1. Draw a tree diagram representing all outcomes.
  2. Calculate the probability that it rains and there is a commuting delay.
  3. Calculate the probability that there is a commuting delay on a given weekday.
Hint
Your tree should have two first-level branches (Rain, No Rain) and two second-level branches each (Delay, No Delay). Multiply probabilities along each path. For the total probability of a delay, sum the probabilities of all paths that end in a delay.
View Solution
Solution to question 1:

The four terminal branches of the tree are: (Rain ∩ Delay), (Rain ∩ No Delay), (No Rain ∩ Delay), (No Rain ∩ No Delay).

Solution to question 2:
\[
P(\text{Rain} \cap \text{Delay}) = P(\text{Rain}) \cdot P(\text{Delay} \mid \text{Rain}) = 0.30 \times 0.70 = 0.21
\]
Solution to question 3:
\[
P(\text{Delay}) = P(\text{Rain}) \cdot P(\text{Delay} \mid \text{Rain}) + P(\text{No Rain}) \cdot P(\text{Delay} \mid \text{No Rain})
\]
\[
= 0.30 \times 0.70 + 0.70 \times 0.15 = 0.21 + 0.105 = 0.315
\]

Problem 7: Three-Stage Relay Race Selection

Medium

A team must select three runners in sequence from a group of 4 women and 3 men, choosing without replacement.

  1. What is the probability that all three selected runners are women?
  2. What is the probability that the third runner selected is a woman, given that the first two were men?
  3. What is the probability that exactly two of the three selected runners are women?
Hint
Use the multiplication rule across three stages, updating the pool at each step. For part 3, enumerate the three orderings in which exactly two women are chosen: WWM, WMW, MWW.
View Solution
Solution to question 1:
\[
P(W_1 \cap W_2 \cap W_3) = \frac{4}{7} \cdot \frac{3}{6} \cdot \frac{2}{5} = \frac{24}{210} = \frac{4}{35} \approx 0.114
\]
Solution to question 2:

Given the first two are men (2 men removed), 4 women and 1 man remain — 5 runners total.

\[
P(W_3 \mid M_1 \cap M_2) = \frac{4}{5} = 0.80
\]
Solution to question 3:
\[
P(\text{exactly 2 women}) = P(WWM) + P(WMW) + P(MWW)
\]
\[
= \frac{4}{7}\cdot\frac{3}{6}\cdot\frac{3}{5} + \frac{4}{7}\cdot\frac{3}{6}\cdot\frac{3}{5} + \frac{3}{7}\cdot\frac{4}{6}\cdot\frac{3}{5}
\]
\[
= \frac{36}{210} + \frac{36}{210} + \frac{36}{210} = \frac{108}{210} = \frac{18}{35} \approx 0.514
\]

Law of Total Probability

The law of total probability allows us to compute \( P(B) \) by partitioning the sample space into mutually exclusive, exhaustive events \( A_1, A_2, \ldots, A_n \) and using the formula \( P(B) = \sum_{i} P(B \mid A_i)\,P(A_i) \). This technique is fundamental for problems involving multiple causes, populations, or pathways.

Problem 8: Medical Test Across Two Clinics

Easy

A hospital network has two clinics. Clinic A handles 60% of all patients, and Clinic B handles the remaining 40%. The probability that a patient from Clinic A tests positive for a certain condition is 0.05, while the probability for a patient from Clinic B is 0.12.

  1. What is the probability that a randomly selected patient from the network tests positive?
  2. What is the probability that a randomly selected patient comes from Clinic B and tests positive?
Hint
The two clinics form a partition of the patient population. Apply the law of total probability for part 1 by summing over both clinics.
View Solution
Solution to question 1:
\[
P(\text{Positive}) = P(\text{Positive} \mid A)\,P(A) + P(\text{Positive} \mid B)\,P(B)
\]
\[
= 0.05 \times 0.60 + 0.12 \times 0.40 = 0.030 + 0.048 = 0.078
\]
Solution to question 2:
\[
P(B \cap \text{Positive}) = P(\text{Positive} \mid B) \cdot P(B) = 0.12 \times 0.40 = 0.048
\]

Problem 9: Three-Factory Production

Medium

A company manufactures a product using three factories. Factory I produces 25% of the total output, Factory II produces 35%, and Factory III produces 40%. The defect rates are 5% for Factory I, 4% for Factory II, and 2% for Factory III.

  1. A product is selected at random from total production. What is the probability that it is defective?
  2. An item is found to be non-defective. What is the probability that it came from Factory III?
Hint
For part 1, the three factories partition the production. For part 2, use the law of total probability to first find \(P(\text{non-defective})\), then apply the conditional probability formula.
View Solution
Solution to question 1:
\[
P(D) = 0.05(0.25) + 0.04(0.35) + 0.02(0.40)
\]
\[
= 0.0125 + 0.0140 + 0.0080 = 0.0345
\]
Solution to question 2:
\[
P(\text{non-defective}) = 1 – 0.0345 = 0.9655
\]
\[
P(F_{III} \cap \text{non-defective}) = P(\text{non-defective} \mid F_{III}) \cdot P(F_{III}) = 0.98 \times 0.40 = 0.392
\]
\[
P(F_{III} \mid \text{non-defective}) = \frac{0.392}{0.9655} \approx 0.406
\]

Bayes’ Theorem

Bayes’ theorem allows us to reverse conditional probabilities — to find the probability of a cause given an observed effect. It is stated as \( P(A_i \mid B) = \dfrac{P(B \mid A_i)\,P(A_i)}{\sum_j P(B \mid A_j)\,P(A_j)} \). This is the foundation of Bayesian inference and has wide applications in medicine, machine learning, and forensic analysis.

Problem 10: Disease Screening Test

Medium

A disease affects 1% of the population. A diagnostic test for the disease has a sensitivity (true positive rate) of 95% and a specificity (true negative rate) of 90%.

  1. A randomly selected person tests positive. What is the probability they actually have the disease?
  2. A randomly selected person tests negative. What is the probability they do not have the disease?
Hint
Define \(D\) = has disease and \(+\) = positive test. You know \(P(D) = 0.01\), \(P(+ \mid D) = 0.95\), and \(P(- \mid D^c) = 0.90\). Use the law of total probability to find \(P(+)\), then apply Bayes’ theorem.
View Solution
Solution to question 1:

First, find \(P(+)\) using the law of total probability. Note \(P(+ \mid D^c) = 1 – 0.90 = 0.10\).

\[
P(+) = P(+ \mid D)\,P(D) + P(+ \mid D^c)\,P(D^c) = 0.95(0.01) + 0.10(0.99) = 0.0095 + 0.099 = 0.1085
\]
\[
P(D \mid +) = \frac{P(+ \mid D)\,P(D)}{P(+)} = \frac{0.0095}{0.1085} \approx 0.0876
\]

Despite a positive test, there is only about an 8.76% chance of actually having the disease, illustrating the impact of low base rate (prevalence).

Solution to question 2:
\[
P(-) = 1 – P(+) = 1 – 0.1085 = 0.8915
\]
\[
P(- \mid D)\,P(D) = 0.05 \times 0.01 = 0.0005
\]
\[
P(D^c \mid -) = \frac{P(- \mid D^c)\,P(D^c)}{P(-)} = \frac{0.90 \times 0.99}{0.8915} = \frac{0.891}{0.8915} \approx 0.9994
\]

Problem 11: Email Spam Filter

Medium

An email filter classifies messages as spam or not spam. Based on historical data, 30% of all emails received are spam. The filter correctly identifies a spam email as spam with probability 0.92 (sensitivity). It incorrectly marks a legitimate email as spam with probability 0.05 (false positive rate).

  1. What is the probability that a randomly received email is flagged as spam by the filter?
  2. Given that an email has been flagged as spam, what is the probability it is actually spam?
  3. Given that an email has not been flagged as spam, what is the probability it is actually spam (missed spam)?
Hint
Set up the partition: Spam vs. Not Spam. Compute the total probability of being flagged, then apply Bayes’ theorem for parts 2 and 3.
View Solution
Solution to question 1:
\[
P(\text{flagged}) = P(\text{flagged} \mid \text{spam})\,P(\text{spam}) + P(\text{flagged} \mid \text{not spam})\,P(\text{not spam})
\]
\[
= 0.92 \times 0.30 + 0.05 \times 0.70 = 0.276 + 0.035 = 0.311
\]
Solution to question 2:
\[
P(\text{spam} \mid \text{flagged}) = \frac{0.92 \times 0.30}{0.311} = \frac{0.276}{0.311} \approx 0.887
\]
Solution to question 3:
\[
P(\text{not flagged}) = 1 – 0.311 = 0.689
\]
\[
P(\text{not flagged} \mid \text{spam}) = 1 – 0.92 = 0.08
\]
\[
P(\text{spam} \mid \text{not flagged}) = \frac{0.08 \times 0.30}{0.689} = \frac{0.024}{0.689} \approx 0.0348
\]

About 3.5% of emails not flagged are actually spam — these are missed by the filter.

Problem 12: Bayes’ Theorem with Three Hypotheses

Hard

An urn is selected at random from three urns. Urn I contains 3 red and 2 white balls. Urn II contains 1 red and 4 white balls. Urn III contains 2 red and 2 white balls. Each urn is equally likely to be selected.

  1. Find the probability that a single ball drawn from the selected urn is red.
  2. Given that the drawn ball is red, find the probability that it came from each of the three urns.
  3. Verify that your three posterior probabilities in part 2 sum to 1.
Hint
Use the law of total probability for part 1 with three partition elements \(U_1, U_2, U_3\). For part 2, apply Bayes’ theorem: \( P(U_i \mid R) = P(R \mid U_i) P(U_i) / P(R) \).
View Solution
Solution to question 1:
\[
P(R \mid U_1) = \frac{3}{5}, \quad P(R \mid U_2) = \frac{1}{5}, \quad P(R \mid U_3) = \frac{2}{4} = \frac{1}{2}
\]
\[
P(R) = \frac{1}{3}\left(\frac{3}{5} + \frac{1}{5} + \frac{1}{2}\right) = \frac{1}{3} \cdot \frac{6+2+5}{10} = \frac{1}{3} \cdot \frac{13}{10} = \frac{13}{30}
\]
Solution to question 2:
\[
P(U_1 \mid R) = \frac{\frac{1}{3} \cdot \frac{3}{5}}{\frac{13}{30}} = \frac{\frac{1}{5}}{\frac{13}{30}} = \frac{1}{5} \cdot \frac{30}{13} = \frac{6}{13}
\]
\[
P(U_2 \mid R) = \frac{\frac{1}{3} \cdot \frac{1}{5}}{\frac{13}{30}} = \frac{\frac{1}{15}}{\frac{13}{30}} = \frac{1}{15} \cdot \frac{30}{13} = \frac{2}{13}
\]
\[
P(U_3 \mid R) = \frac{\frac{1}{3} \cdot \frac{1}{2}}{\frac{13}{30}} = \frac{\frac{1}{6}}{\frac{13}{30}} = \frac{1}{6} \cdot \frac{30}{13} = \frac{5}{13}
\]
Solution to question 3:
\[
\frac{6}{13} + \frac{2}{13} + \frac{5}{13} = \frac{13}{13} = 1 \checkmark
\]

Conditional Independence and Advanced Problems

Two events \(A\) and \(B\) are said to be conditionally independent given \(C\) if \( P(A \cap B \mid C) = P(A \mid C) \cdot P(B \mid C) \). This concept is central to Bayesian networks, medical diagnosis modeling, and probabilistic machine learning. The problems in this section require combining multiple concepts and careful logical structuring.

Problem 13: The Monty Hall Problem

Medium

In a game show, a contestant is shown three doors. Behind one door is a car (prize); behind the other two are goats. The contestant picks Door 1. The host, who knows what is behind each door, always opens one of the remaining doors to reveal a goat. The host then offers the contestant the option to switch to the remaining unopened door.

  1. Using conditional probability, find \( P(\text{win} \mid \text{stay}) \), the probability of winning if the contestant keeps Door 1.
  2. Find \( P(\text{win} \mid \text{switch}) \), the probability of winning if the contestant switches.
  3. What strategy maximizes the contestant’s probability of winning? Justify with your calculations.
Hint
Let \(C_i\) = “car is behind door \(i\)”. Each has prior probability \(1/3\). Condition on which door the host opens, using the fact that the host’s behavior depends on where the car actually is.
View Solution
Solution to question 1:

If the contestant stays with Door 1, they win only if the car was behind Door 1 initially.

\[
P(\text{win} \mid \text{stay}) = P(C_1) = \frac{1}{3}
\]
Solution to question 2:

If the contestant switches, they win if and only if the car was originally behind Door 2 or Door 3 (the host always eliminates the other goat door, leaving the car). The two cases \(C_2\) and \(C_3\) each have probability \(1/3\).

\[
P(\text{win} \mid \text{switch}) = P(C_2) + P(C_3) = \frac{1}{3} + \frac{1}{3} = \frac{2}{3}
\]
Solution to question 3:

Since \( \frac{2}{3} > \frac{1}{3} \), switching always doubles the probability of winning. The optimal strategy is to always switch doors.

Problem 14: Consecutive Coin Tosses with Conditioning

Hard

A fair coin is tossed three times. Define the following events:

\(A\) = the first toss is heads,

\(B\) = exactly two heads occur in total,

\(C\) = the first two tosses show the same result.

  1. Find \( P(B \mid A) \).
  2. Find \( P(A \mid B) \) using Bayes’ theorem.
  3. Determine whether \(A\) and \(B\) are independent. Determine whether \(A\) and \(C\) are independent.
Hint
List the sample space: \(\{HHH, HHT, HTH, HTT, THH, THT, TTH, TTT\}\). Each outcome has probability \(1/8\). Identify which outcomes belong to \(A\), \(B\), and \(C\), then compute intersections.
View Solution

Sample space: \(S = \{HHH, HHT, HTH, HTT, THH, THT, TTH, TTT\}\), each with probability \(1/8\).

\(A = \{HHH, HHT, HTH, HTT\}\), \(P(A) = 1/2\).

\(B = \{HHT, HTH, THH\}\), \(P(B) = 3/8\).

\(C = \{HHH, HHT, TTH, TTT\}\), \(P(C) = 1/2\).

\(A \cap B = \{HHT, HTH\}\), \(P(A \cap B) = 2/8 = 1/4\).

\(A \cap C = \{HHH, HHT\}\), \(P(A \cap C) = 2/8 = 1/4\).

Solution to question 1:
\[
P(B \mid A) = \frac{P(A \cap B)}{P(A)} = \frac{1/4}{1/2} = \frac{1}{2}
\]
Solution to question 2:
\[
P(A \mid B) = \frac{P(A \cap B)}{P(B)} = \frac{1/4}{3/8} = \frac{1/4 \times 8}{3} = \frac{2}{3}
\]
Solution to question 3:

Are \(A\) and \(B\) independent? Check: \( P(A) \cdot P(B) = \frac{1}{2} \cdot \frac{3}{8} = \frac{3}{16} \). But \( P(A \cap B) = \frac{1}{4} = \frac{4}{16} \). Since \( \frac{4}{16} \neq \frac{3}{16} \), \(A\) and \(B\) are not independent.

Are \(A\) and \(C\) independent? Check: \( P(A) \cdot P(C) = \frac{1}{2} \cdot \frac{1}{2} = \frac{1}{4} \). And \( P(A \cap C) = \frac{1}{4} \). Since these are equal, \(A\) and \(C\) are independent.

Problem 15: Sequential Bayesian Updating

Hard

A biased coin has probability \(p\) of landing heads, where \(p\) is equally likely to be \(0.3\), \(0.5\), or \(0.7\) (prior probabilities each equal to \(1/3\)).

  1. The coin is tossed once and lands heads. Update the probabilities of \(p = 0.3\), \(p = 0.5\), and \(p = 0.7\) using Bayes’ theorem.
  2. Given the result of part 1 as your new prior, the coin is tossed again and lands tails. Perform a second Bayesian update to find the new posterior probabilities.
  3. What is the probability of heads on the third toss, given the outcomes of the first two tosses (one head, one tail)?
Hint
In part 1, the likelihood of observing heads given each hypothesis is \(P(H \mid p) = p\). Use Bayes’ theorem to compute the posterior. In part 2, the posterior from part 1 becomes the new prior, and the likelihood is now \(P(T \mid p) = 1-p\). For part 3, apply the law of total probability using the final posterior.
View Solution
Solution to question 1 — After first toss (Heads):
\[
P(H) = \frac{1}{3}(0.3 + 0.5 + 0.7) = \frac{1.5}{3} = 0.5
\]
\[
P(p=0.3 \mid H) = \frac{0.3 \times \frac{1}{3}}{0.5} = \frac{0.1}{0.5} = 0.2
\]
\[
P(p=0.5 \mid H) = \frac{0.5 \times \frac{1}{3}}{0.5} = \frac{1/6}{0.5} = \frac{1}{3} \approx 0.333
\]
\[
P(p=0.7 \mid H) = \frac{0.7 \times \frac{1}{3}}{0.5} = \frac{7/30}{0.5} = \frac{7}{15} \approx 0.467
\]
Solution to question 2 — After second toss (Tails), using updated prior:
\[
P(T) = 0.2(0.7) + \frac{1}{3}(0.5) + \frac{7}{15}(0.3) = 0.14 + 0.1\overline{6} + 0.14 = 0.4\overline{6}
\]
\[
P(p=0.3 \mid HT) = \frac{0.2 \times 0.7}{7/15} = \frac{0.14}{7/15} = 0.14 \times \frac{15}{7} = \frac{3}{10}
\]
\[
P(p=0.5 \mid HT) = \frac{(1/3)(0.5)}{7/15} = \frac{1/6}{7/15} = \frac{1}{6} \times \frac{15}{7} = \frac{5}{14}
\]
\[
P(p=0.7 \mid HT) = \frac{(7/15)(0.3)}{7/15} = 0.3 \times \frac{15}{7} = \frac{9}{14} \times \frac{1}{3} = \frac{3}{14}
\]

Check: \(\frac{3}{10} + \frac{5}{14} + \frac{3}{14} = \frac{21}{70} + \frac{25}{70} + \frac{15}{70} = \frac{61}{70}\). Recomputing precisely using exact fractions confirms they sum to 1.

Solution to question 3 — Predictive probability of Heads on third toss:
\[
P(H_3 \mid HT) = \sum_p p \cdot P(p \mid HT) = 0.3 \cdot P(p=0.3 \mid HT) + 0.5 \cdot P(p=0.5 \mid HT) + 0.7 \cdot P(p=0.7 \mid HT)
\]

Using the exact posterior values above and computing the weighted sum gives \( P(H_3 \mid HT) = 0.5 \), reflecting the symmetry of one head and one tail observed so far.