This Exam P sample reference tests Discrete Random Variables. The four relevant count probabilities are proportional to 60, 20, 5, and 1. Their total weight 86 represents probability 0.95, while the first three weights total 85, giving 0.95(85/86)=0.938953 and choice D.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis is much too small because it effectively retains only the dominant low-count mass and discards material probability from the other allowed counts.
BThis can result from reversing or truncating one of the probability-ratio links. Every ratio must be propagated from count three before normalization.
COmitting the zero-count category gives weights 20, 5, and 1; normalizing those to 0.95 and then excluding count three produces about 0.91.
EA probability through count two cannot exceed the given probability through count three. This choice violates nesting of cumulative events.
Original practice · fully worked
Original variant: even-ranked search result
A retrieval system returns a result at rank R in {1,2,3,4}. Before normalization, rank r receives probability weight r(5-r). An analyst is told that the returned rank is even. Calculate the conditional expected rank.
A 0.400
B 0.600
C 2.500
D 2.800
E 4.000
Variant answer in brief
The four rank weights are 4, 6, 6, and 4. Conditioning on an even rank leaves weights 6 and 4 at ranks two and four, so the conditional mean is 2(0.6)+4(0.4)=2.8 and choice D.
Setup
Setup
Evaluate the unnormalized weight at each possible rank.
(w1,w2,w3,w4)=(1(4),2(3),3(2),4(1))=(4,6,6,4)
Model
Model
After the even-rank condition, only ranks two and four remain. Normalize their two weights.
Pr(R=2∣R even)=6+46=0.6
Pr(R=4∣R even)=6+44=0.4
Compute
Compute
Weight the two remaining ranks by their conditional probabilities.
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.