This Exam P sample reference tests Normal Linear Combination. The average of the three independent normal variables is normal with mean 100 and variance (13.5+27+40.5)/9=9. The cutoff is two standard deviations above the mean, giving the upper tail 1-Φ(2) and choice D.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis finds the correct standardized cutoff 2 but uses the lower tail Φ(2) instead of the upper tail.
BThis incorrectly assigns the sum variance 81 to the average, obtains z=2/3, and then uses the lower tail.
CThis uses the upper tail but still assigns the sum variance 81 to the average, producing the wrong z-value 2/3.
EThis divides the sum variance by 3 instead of 3², giving variance 27 for the average and the incorrect cutoff 2/√(3).
Original practice · fully worked
Original variant: weighted calibration score
Three independent calibration readings A, B, and C are normal. Their means are 2, -1, and 0, and their variances are 4, 9, and 16, respectively. A diagnostic score is W=A-2B+0.5C. Calculate the 90th percentile of W.
A -4.501
B 4.000
C 10.901
D 12.501
E 16.816
Variant answer in brief
The score is normal with mean 4 and variance 4+4(9)+0.25(16)=44. Its 90th percentile is 4+1.2815516sqrt(44)=12.5009, so choice D is correct.
Setup
Setup
Apply the score coefficients to the three component means.
E[W]=2−2(−1)+0.5(0)=4
Model
Model
Independence eliminates covariance terms, while each variance is multiplied by the square of its coefficient.
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.