This Exam P sample reference tests Bayes' Theorem. This is Bayes' theorem with category-specific exponential survival likelihoods. The two prior-weighted survivor masses are 0.4 exp(−0.25) and 0.6 exp(−0.5); normalizing the first mass gives 0.4612, so choice C.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThe value 0.08 results if the stated exponential means 4 and 2 are misread as rate parameters: normalizing 0.4 exp(−4) and 0.6 exp(−2) gives about 0.083.
BThe value 0.27 is the posterior home share among policies canceled during the first year, using 0.4(1-exp(−0.25)) over the two cancellation contributions. It conditions on the opposite event.
DThe value 0.56 is approximately exp(−0.25)/[exp(−0.25)+exp(−0.5)]. It normalizes the two survival likelihoods without the category priors.
EThe value 0.66 is approximately 4/(4+2). It treats the two mean lifetimes as posterior weights instead of using exponential survival likelihoods and the category priors.
Original practice · fully worked
Original variant: device type after a second-hour failure
A device is type A with probability 0.37 and type B with probability 0.63. Conditional on type, its lifetime is exponential. The respective probabilities of surviving one hour are 0.50 and 0.80. A device is observed to fail after hour one but no later than hour two. Calculate the conditional probability that it is type A.
A 0.1600
B 0.2500
C 0.3700
D 0.4785
E 0.5215
Variant answer in brief
Exponential survival implies S(2)=S(1)², so the second-hour failure likelihoods are 0.50-0.25=0.25 for A and 0.80-0.64=0.16 for B. The prior-weighted masses are 0.0925 and 0.1008, giving posterior probability 0.478531 and choice D.
Setup
Setup
Let F be the event that failure occurs during the second hour. Exponential survival over two hours is the square of one-hour survival.
SA(2)=0.502=0.25,SB(2)=0.802=0.64
Model
Model
Subtract two-hour survival from one-hour survival to obtain each interval likelihood.
Pr(F∣A)=0.50−0.25=0.25
Pr(F∣B)=0.80−0.64=0.16
Compute
Compute
Apply Bayes' theorem to the two prior-weighted interval probabilities.
Pr(A∣F)=0.37(0.25)+0.63(0.16)0.37(0.25)
=0.19330.0925=1933925=0.4785307812
Answer
Answer
The posterior probability of type A rounds to 0.4785.
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.