This Exam P sample reference tests Conditional Probability. The three prior-by-suburban weights are 0.1100, 0.1800, and 0.0525. Normalizing the under-25 weight gives 0.1100/0.3425=44/137=0.32117, so choice C is correct.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis stops at the joint numerator P(U and S)=0.55(0.20)=0.11 and omits division by the total suburban probability.
BThis assumes the three age groups have equal priors and normalizes only the suburban likelihoods: 0.20/(0.20+0.60+0.35)=0.1739.
DThis repeats the prior P(U)=0.55 and ignores the information that the observed location was suburban.
EThis reverses the Bayes likelihood ratio. Using prior odds 0.55/0.45 times P(S|not U)/P(S|U) gives posterior probability 0.7595; the likelihood ratio must be P(S|U)/P(S|not U).
Original practice · fully worked
Original variant: recover a supplier share from defect evidence
A component comes from supplier A or supplier B. Components from A are defective with probability 0.08, while those from B are defective with probability 0.02. Among all defective components, half came from supplier A. Calculate the overall proportion of components supplied by A.
A 0.02
B 0.08
C 0.20
D 0.50
E 0.80
Variant answer in brief
If p is A's prior share, the two defective joint weights are 0.08p and 0.02(1-p). A posterior share of one half makes them equal, so p=0.20 and choice C is correct.
Setup
Setup
Let p denote the unknown proportion supplied by A and let D denote a defective component.
Pr(A)=p,Pr(B)=1−p
Pr(D∣A)=0.08,Pr(D∣B)=0.02
Model
Model
Express the supplied posterior probability with Bayes' formula.
Pr(A∣D)=0.08p+0.02(1−p)0.08p=21
Compute
Compute
A one-half posterior means the defective weights from the two suppliers are equal.
The 3108-page Probability Proof Manual reorganizes 718 verified Exam P solutions by syllabus skill and adds formula proofs, error patterns, and original worked practice.