Supppose we have a binomial experiment in which success is defined to be a particular quality or attribute that interest us.
"\\text{for a binomial distribution}"
"\\text{mean \u03bc of a binomial }"
"\\mu=np"
"\\text{standard deviation of a binomial}"
"\\sigma=\\sqrt{np(1-p}"
"\\text{For a normal distribution, \u03bc should be 3 standard deviations away from 0 and n}"
"\\mu- 3\\sqrt{np(1-p)} > 0 \\text{ or}"
"np>9(1-p);(1)"
"\\mu+3\\sqrt{np(1-p)}<n \\text{or}"
"n(1-p)>9p;nq>9p(2)"
"np = 33*0.2=6.6"
"np <9(1-0.2)=7.2"
"nq =33*(1-0.2)=26.4"
"nq>9p=1.8"
"\\text{\u0441ondition(1) not met}"
"\\text{it is not possible to approximate } \\hat{p}"
"\\text{by a normal distribution}"
Answer:it is not possible to approximate a binomial experiment to a normal distribution
np = 6.6 nq =26.4
Comments
Leave a comment