Ling-Chieh Kung Department of Information Management National Taiwan University 10 1 25×0.6×0.4 1. (a) Var(X 1 ) = 20 = 2 and Var(X 2 ) = 5 = 65 . It is then clear that Var(X 1 ) < Var(X 2 ). (b) As X 1 and X 2 are random variables with overlapping sets of possible outcomes, we cannot determine whether one is larger than the other one. 2. Let X be the sample mean and Z be a standard normal random variable. (a) The distribution of the sample mean is X ∼ ND(120, √40n ). (b) When n = 16, X ∼ ND(120, 10). The desired probability is
! −6 6 1 − Pr X ∈ [114, 126] = 1 − Pr Z ∈ √ , √ ≤ 0.01, 40/ n 40/ n 6√ i.e., Pr(Z > 40/ n ) ≤ 0.005. This requires 6 √ 40 √ ≥ 2.576 ⇔ n ≥ 2.576 × ≈ 17.17 ⇔ n ≥ 294.88. 40/ n 6 The smallest sample size allowed is thus 295. X 3. (a) Let X be the number of women in the sample that have volunteering experiences and p̂ = 150 be the sample proportion. According to the central limit theorem, we have r 0.25 × 0.75 p̂ ∼ ND 0.25, ∼ ND(0.25, 0.035). 150 It then follows that
0.23 − 0.25 Pr(X ≥ 34.5) ≈ Pr(p̂ ≥ 0.23) ≈ Pr Z ≥ ≈ Pr(Z ≥ −0.57) ≈ 0.714, 0.035 where Z is a standard normal random variable. (b) We need to first find the population proportion, which is the probability that a randomly selected person has volunteering experiences. This probability is 0.48×0.25+0.52×0.2 = 0.224. X Let X be the number of people in the sample that have volunteering experiences and p̂ = 300 be the sample proportion. According to the central limit theorem, we have r 0.224 × 0.776 p̂ ∼ ND 0.224, ∼ ND(0.224, 0.024). 300 It then follows that
0.2 − 0.224 0.25 − 0.224 Pr(0.2 ≤ p̂ ≤ 0.25) ≈ Pr ≤Z≤ 0.024 0.024 ≈ Pr(−0.997 ≤ Z ≤ 1.08) ≈ 0.701. where Z is a standard normal random variable.