Condition on a row total
In a group of $120$ students, $50$ study physics, and $18$ of those physics students also study art. Find $P(\text{art}\mid\text{physics})$.
The denominator is $50$, not $120$, because physics is the condition.
Use conditional probability and Bayes’ theorem while keeping the direction of conditioning clear.
Why this matters. New evidence changes the relevant sample space. Most errors come from using the right numbers in the wrong direction.
$P(A\mid B)$ means the probability of $A$ among outcomes where $B$ has occurred.
Independence means the condition does not change the probability: $P(A\mid B)=P(A)$, equivalently $P(A\cap B)=P(A)P(B)$.
Checkpoint
Correct. The event after the conditioning bar is known to have happened.
Not quite. Read the notation as “probability of A given D.” The given event is D.
The event after the conditioning bar is known to have happened.
Example 1
Condition on a row total
In a group of $120$ students, $50$ study physics, and $18$ of those physics students also study art. Find $P(\text{art}\mid\text{physics})$.
The denominator is $50$, not $120$, because physics is the condition.
For the machine tree above, the total defective probability is
Checkpoint
Correct. Bayes starts with the joint path that satisfies both B and defective.
Not quite. A conditional rate on one branch is not yet the joint probability needed in the numerator.
Bayes starts with the joint path that satisfies both B and defective.
Example 2
Which machine made it?
Using the tree, find the probability that a defective item came from machine B.
A higher defect rate makes B more represented among defective items, even though B produces fewer items overall.
$0.05=P(D\mid B)$ is the forward defect rate within B. The question asks the reverse probability $P(B\mid D)$.
Compare a joint probability with the product of marginals, or compare a conditional probability with its corresponding marginal.
Checkpoint
Correct. $P(A)P(B)=0.4(0.5)=0.2=P(A\cap B)$.
Not quite. Mutual exclusivity would require a zero intersection when both events have positive probability.
$P(A)P(B)=0.4(0.5)=0.2=P(A\cap B)$.
Example 3
A screening result
A condition affects $3\%$ of a population. A test is positive for $92\%$ of affected people and $6\%$ of unaffected people. Find $P(C\mid +)$.
The positive predictive probability is not the same as the sensitivity $0.92$.
Five questions to check the main decisions from this lesson.
Your score is not saved.
The event after the bar is the condition.
The event after the bar is the condition.
Rearrange the conditional probability formula.
Rearrange the conditional probability formula.
This product rule characterizes independence.
This product rule characterizes independence.
A path describes an intersection of successive events.
A path describes an intersection of successive events.
A small false-positive rate applied to a large group may outweigh true positives from a small group.
A small false-positive rate applied to a large group may outweigh true positives from a small group.
Want to leave a comment?
Log in