This Exam P sample reference tests Uniform Distribution. For a continuous uniform interval beginning at zero, the standard deviation is the upper endpoint divided by √(12). The first standard deviation fixes c=24; the second interval therefore has width 54 and standard deviation 9√(3)≈15.6, so choice D is correct.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis comes from the invalid operation |3(4√3) − 18| = 2.784…, which applies an endpoint shift directly to a standard deviation. Constants inside an interval endpoint change its width before the uniform formula is used.
BThis treats the recovered width 54 as though it were a variance and reports √54 = 7.348…. A uniform interval of width 54 has variance 54²/12, not 54.
CThis uses the incorrect rule SD = w/(3√3), producing 54/(3√3) = 6√3 ≈ 10.4. The continuous-uniform denominator is √12 = 2√3.
EThis ignores the 18-unit reduction in the second interval width. Using width 3c = 72 gives 72/√12 = 12√3 ≈ 20.8.
Original practice · fully worked
Original variant: calibrate a transformed sensor reading
A sensor reading X is uniformly distributed on an unknown interval. Its mean is 30, and its 90th percentile is 42. A processed reading is defined by Y=2X-5. Calculate the variance of Y.
A 60
B 75
C 150
D 300
E 900
Variant answer in brief
The 90th percentile is 0.4 interval widths above the midpoint, so the width is 30. Thus Var(X)=30²⁄¹²=75, and the scale factor 2 multiplies variance by 4, giving 300 and choice D.
Setup
Setup
Write the unknown support as an interval with midpoint 30 and width w.
X∼Uniform(30−2w,30+2w)
Model
Model
A 90th percentile lies 90% of the width above the lower endpoint, hence 40% of the width above the midpoint.
42=30−2w+0.90w=30+0.40w
Compute
Compute
Recover the width, compute the uniform variance, and apply the squared scale factor in the affine transformation.
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.