a. can use normal approximation if np and n(1-p) > 10. Here n = 50, p = 0.16. np = 8, so not appropriate to use normal approximation
b. n = 90 np = 14.4 so can use normal approximation.
z = (phat - p)/sqrt(p*(1-p)/n)
z = (0.2 - 0.16)/sqrt(0.16*0.84/90) = 1.04
P(p > 20) = P(z > 1.04) = 1 - P(z < 1.04) = 1 - 0.8508 = 0.1492
c. P( 0.15 < p < 0.18)
convert endpoints to z-score:
z = (0.15-0.16)/sqrt(0.16*0.84/90) = -0.26
z= (0.18-0.16)/sqrt(0.16*0.84/90) = 0.52
P(0.15 < p < 0.18) = P(-0.26 < z < 0.52) = P(z < 0.52) - P(z < -0.26) = 0.6985 - 0.3974 - 0.3011
d. P(p < 0.25)
z= (0.25 - 0.16)/sqrt(0.16*0.84/90) = 2.33
P(p < 0.25) = P(z < 2.33) = 0.9901
e. P(p > 0.25) = 1 - 0.9901 = 0.0099, so that would be very