This Exam P sample reference tests Variance. Marginalizing the joint table gives a positive-test probability of 0.09. A Bernoulli variable with that success probability has standard deviation equal to the square root of 0.09 × 0.91, so its coefficient of variation is approximately 3.18 and choice D.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis reverses the coefficient, dividing the mean by the standard deviation, which gives approximately 0.31.
BThis divides the variance by the mean. For a Bernoulli variable that ratio is 0.91, not the coefficient of variation.
CThis is the marginal probability that the underlying condition indicator is zero. It uses the wrong variable and is not a variability measure.
EThis is the coefficient of variation of the underlying condition indicator, whose positive probability is 0.05, rather than of the test indicator.
Original practice · fully worked
Original variant: upload attempts until acceptance
A data terminal repeats independent upload attempts until the first accepted transmission. Each attempt is accepted with probability 0.36. Let N be the total number of attempts, including the accepted attempt. Calculate the coefficient of variation of N.
A 0.36
B 0.64
C 0.80
D 2.22
E 2.78
Variant answer in brief
The attempt count is geometric with mean 25/9 and standard deviation 20/9. Their ratio is 0.80, so choice C.
Setup
Setup
Use the geometric distribution that counts the successful attempt.
p=0.36,q=1−p=0.64
Model
Model
Write the geometric mean and variance.
E[N]=p1=925
Var(N)=p2q=81400
Compute
Compute
Take the square root of the variance and divide by the mean.
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.