This Exam P sample reference tests Conditional Distributions. This is a conditional-variance calculation for a binary random variable. Within the specified row, the conditional probability of one is 4/5, so the Bernoulli variance is (4/5)(1/5)=4/25 and choice A is correct.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
BThe fraction 15/64 is (5/8)(3/8), the variance of the indicator that X equals one. It uses the conditioning-row probability as a Bernoulli parameter instead of forming Y's conditional distribution.
CUsing the unnormalized joint mass 1/2 as though it were the conditional success probability gives (1/2)(1/2)=1/4.
DThe value 3/4 is the complement of the separate joint cell with mass 1/4; taking that complement does not condition Y on X equal to one.
EThe fraction 4/5 is the conditional probability and conditional mean of Y, not its conditional variance.
Original practice · fully worked
Original variant: warning-conditioned calibration spread
A sensor line operates in fast mode with probability 0.60 and precision mode with probability 0.40. A warning appears with probability 0.20 in fast mode and 0.50 in precision mode. Given the mode, the warning is independent of calibration adjustment Y. In fast mode, Y has mean 0 and variance 1; in precision mode, it has mean 3 and variance 2. Given that a warning appeared, calculate Var(Y).
A 0.3200
B 1.6250
C 1.8750
D 2.1094
E 3.7344
Variant answer in brief
Bayes' rule gives posterior mode weights 3/8 for fast and 5/8 for precision after a warning. The conditional variance is the 1.625 within-mode component plus the 2.109375 between-mode component, totaling 3.734375 and selecting choice E.
Setup
Setup
Let W denote a warning. Compute its probability and update the operating-mode weights using Bayes' rule.
Pr(W)=0.60(0.20)+0.40(0.50)=0.32
Pr(F∣W)=0.320.60(0.20)=83,Pr(P∣W)=85
Model
Model
Apply the conditional law of total variance across the two posterior modes.
Var(Y∣W)=E[Var(Y∣M,W)∣W]+Var(E[Y∣M,W]∣W)
Compute
Compute
Conditional independence allows the supplied mode-specific moments to be used after observing the warning.
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.