This Exam P sample reference tests Binomial Distribution. This is a binomial upper-tail identification problem. A non-defective item has probability 1-p, so the probability of at least n non-defective items sums the binomial masses from n through 50 with factors (1-p) raised to k and p raised to 50-k; this is choice A.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
BThis sums the correct non-defective-count masses only from zero through n. It represents a lower tail, not the event that the count is at least n.
CThis uses p as the success probability and starts at n+1. It therefore describes more than n defective items rather than at least n non-defective items.
DThis is the lower tail for the number of defective items: it both swaps p with 1-p and sums in the wrong direction.
EThis has the requested upper-tail limits but assigns p to each non-defective item. The non-defective probability is 1-p, so the two powers are reversed.
Original practice · fully worked
Original variant: expected overflow from a compute rack
A compute rack contains ten independent processing nodes. Each node is online at a scheduled checkpoint with probability 0.80. A local controller can serve seven online nodes, and every additional online node must be rerouted. Let R be the number of nodes rerouted at the checkpoint. Calculate E[R].
A 0.1074
B 0.6778
C 1.0000
D 1.1610
E 5.9056
Variant answer in brief
If X is the online-node count, then X is binomial and R is the positive part of X-7. Weighting the masses at 8, 9, and 10 by overflow counts 1, 2, and 3 gives E[R]=1.1609833472, so choice D is correct.
Setup
Setup
Count the online nodes and express rerouting as the amount by which that count exceeds local capacity.
X∼Binomial(10,0.80),R=(X−7)+
Model
Model
Only counts eight through ten create overflow, and each mass is weighted by its number of rerouted nodes.
E[R]=k=8∑10(k−7)(k10)(0.8)k(0.2)10−k
Compute
Compute
Evaluate the three binomial masses and their overflow weights.
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.