This Exam P sample reference tests Exponential Distribution. An exponential observation exceeds its own mean with probability exp(−1). The number of exceedances in ten independent observations is binomial, and its nine-or-ten tail equals 0.000825499, matching choice B.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis reverses the coefficients that arise after factoring the two binomial tail terms. The factor must be 10-9/e, not 9-10/e.
CThis uses the nonexceedance probability 1-exp(−1) as though it were the exceedance probability and also applies an incompatible extra scaling.
DThis is the probability of exactly nine exceedances. It omits the additional qualifying outcome in which all ten observations exceed the threshold.
EThis raises the nonexceedance probability to the tenth power and attaches a factor of nine; it is neither an exact binomial mass nor the requested tail.
Original practice · fully worked
Original variant: long-duration sensor outages
Outage durations for a remote sensor are independent and exponentially distributed with mean 4 hours. During six outages selected at random, calculate the probability that at least five last longer than 4 hours.
A 0.002479
B 0.004259
C 0.006738
D 0.025555
E 0.028034
Variant answer in brief
Each outage exceeds four hours with probability exp(−1). Adding the binomial masses for five and six exceedances gives 0.0280339, so choice E is correct.
Setup
Setup
The threshold equals the exponential mean, so the individual survival probability has a simple form.
p=Pr(T>4)=e−4/4=e−1
Model
Model
Let K count long outages among the six independent observations.
K∼Bin(6,e−1)
Pr(K≥5)=Pr(K=5)+Pr(K=6)
Compute
Compute
Add the exactly-five and all-six probabilities.
Pr(K≥5)=6e−5(1−e−1)+e−6
Pr(K≥5)=0.0280339211…
Answer
Answer
The probability that at least five outages are long is approximately 0.028034.
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.