This Exam P sample reference tests Independence. The probability of no accidents is 0.2916. Exactly one accident contributes 0.3888 from a high-risk employee and 0.0648 from a low-risk employee. Adding these disjoint cases gives 0.7452 and choice E.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis is the probability of exactly one accident. It omits the allowed case in which nobody has an accident.
BThis is the probability that at least one high-risk employee has an accident while both low-risk employees avoid accidents. It does not describe every at-most-one case.
CThis includes no accidents and exactly one high-risk accident but omits the cases in which exactly one low-risk employee has an accident.
DThis is the complement of the no-accident probability, or the probability of at least one accident. It also includes cases with two or more accidents.
Original practice · fully worked
Original variant: parity of four sensor outages
Four independent sensors have outage probabilities 0.15, 0.25, 0.35, and 0.45 during a test window. Calculate the probability that an even number of sensors are offline.
A 0.0105
B 0.2279
C 0.4948
D 0.5053
E 0.7721
Variant answer in brief
For independent Bernoulli indicators, the even-parity probability is one-half of one plus the product of one minus twice each outage probability. The product is 0.0105, giving 0.50525 and choice D.
Setup
Setup
Encode outage-count parity through the expectation of minus one raised to the count.
Pr(N even)−Pr(N odd)=E[(−1)N]
Model
Model
Independence makes the parity expectation a product of four Bernoulli factors.
E[(−1)N]=j=1∏4(1−2pj)
Compute
Compute
Evaluate the product and combine it with the probabilities summing to one.
(0.70)(0.50)(0.30)(0.10)=0.0105
Pr(N even)=21+0.0105=0.50525
Answer
Answer
An even number of sensors are offline with probability approximately 0.5053.
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.