This Exam P sample reference tests Poisson Distribution. This problem uses independence to compute the mean and variance of a weighted Poisson sum. Solving the coefficient-of-variation equation gives λ=0.4420, which selects choice A.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
BAt λ=0.82, the stated model gives CV=√(72,500λ)/(450λ)=0.6608, not 0.90. The dependence on λ is 1/√(λ), which is lost if the square root is handled inconsistently.
CTreating the weighted total itself as Poisson gives λ approximately 1/0.90²=1.23. Unequal event weights prevent that shortcut.
DSubstituting λ=1.50 gives CV=0.4886, so this candidate fails the defining equation; the squared weights must produce variance 72,500λ.
ESubstituting λ=2.25 gives CV=0.3989, so this candidate also fails the defining equation.
Original practice · fully worked
Original variant: infer a two-speed mixture from its variability
A document-processing task takes 2 minutes in normal mode and 8 minutes in long mode. Let p be the probability of long mode. The processing time has coefficient of variation 0.50, and its mean exceeds 5 minutes. Calculate p.
A 0.0352
B 0.2500
C 0.5000
D 0.6315
E 0.7500
Variant answer in brief
The two-point processing time has mean 2+6p and variance 36p(1-p). The coefficient-of-variation equation has two roots; the condition that the mean exceeds 5 selects p=(5+2sqrt(5))/15=0.6315 and choice D.
Setup
Setup
Let T be the processing time. Compute its first two moments from the two possible durations.
E[T]=2(1−p)+8p=2+6p
E[T2]=4(1−p)+64p=4+60p
Model
Model
Center the second moment to obtain the variance.
Var(T)=4+60p−(2+6p)2
Var(T)=36p(1−p)
Compute
Compute
Square the coefficient-of-variation equation and solve the resulting quadratic.
(2+6p)236p(1−p)=0.502
45p2−30p+1=0
p=155±25
Answer
Answer
The smaller root gives a mean below 5. The larger root is therefore required.
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.