Question 997052

n=1000; N is much larger so can use normality, don't need finite population correction factor, and can use z-test
Ho: p>=0.38
Ha:p<0.38
alpha=0.05
Test statistic is 1-sample proportion test
critical value is z< -1.645
z=phat-p/SE
=0.38-0.42/0.0156
z= -2.5641; probability (p-value) of this occurrence by chance is 0.0052
std error is  sqrt {(0.42)(0.58)/1000}=0.0156