This Exam P sample reference tests Conditional Distributions. Normalize the three joint masses in the specified column to obtain conditional probabilities 0.56, 0.32, and 0.12. The conditional second moment is 0.80 and the conditional mean is 0.56, so the variance is 0.4864 and choice D is correct.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AUsing the unnormalized column masses as though they formed a probability distribution gives 0.20-(0.14)²=0.1804, which rounds to 0.18.
BSubtracting the mean itself rather than its square from the second moment gives 0.80-0.56=0.24.
CIgnoring the conditioning column and using the marginal distribution of T gives an unconditional variance of 0.45.
EThis is the conditional second moment 0.80; it omits subtraction of the squared conditional mean.
Original practice · fully worked
Original variant: satellite diagnostic update
A satellite records zero, one, or two anomaly pings in a cycle with prior probabilities 0.50, 0.30, and 0.20. A diagnostic lamp activates with probability 0.10 after zero pings, 0.40 after one ping, and 0.80 after two pings. Given that the lamp activated, find the variance of the ping count.
A 0.330
B 0.525
C 0.610
D 1.333
E 2.303
Variant answer in brief
Bayes' rule turns the three prior-by-likelihood products into posterior masses 5/33, 12/33, and 16/33. Their variance is 52/99, or 0.525253, so choice B is correct.
Setup
Setup
Let N be the ping count and A the event that the diagnostic lamp activates. Record the prior probabilities and the count-specific activation likelihoods.
(Pr(N=0),Pr(N=1),Pr(N=2))=(0.50,0.30,0.20)
(Pr(A∣N=0),Pr(A∣N=1),Pr(A∣N=2))=(0.10,0.40,0.80)
Model
Model
Multiply each prior by its activation likelihood. Their sum is the evidence used to normalize the posterior distribution.
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.