Let "m" denote the margin of error, then "m" and it is given by,
"m=Z_{\\alpha\/2}*(\\sigma\/\\sqrt{n})"
In order to determine the number of subjects required, make n subject of the formula,
Therefore,
"m*\\sqrt{n}=Z_{\\alpha\/2}*\\sigma"
"\\sqrt{n}=(Z_{\\alpha\/2}*\\sigma)\/m"
Hence
"n=((Z_{\\alpha\/2}*\\sigma)\/m)^2"
Values given are,
"m=1"
"\\sigma=3.4"
"Z_{\\alpha\/2}=Z_{0.1\/2}=Z_{0.05}=1.645"
"n" therefore is,
"n=((1.645*3.4)\/1)^2=(5.593)^2=31.281649\\approxeq32subjects"
Therefore "n=32" subjects would be required to ensure with 90% confidence
that the generated estimate is within 1 minute of the true mean time.
Comments
Leave a comment