a*****n
发帖数: 40 | 1 Suppose a planning agency needs to know the proportion of households in an
area which are renters. It wants a sample estimate which is accurate to ±4
percent. How large of a sample must it have in order to be sure of this
level of accuracy at the 95 percent (±2 a) level of confidence?
I know it must be very simple, but I am toooo :( | c*********n
发帖数: 128 | 2 Wierd problem which I don't quite understand.
If you only asked "How large of a sample must it have in order to be sure of
this level of accuracy at the 95 percent (±2 a) level of confidence? "
then it is the very common error-estimation problem in statistics, which is
probably discussed more often in Statistics board.
Say, the real proportion is p, your sample size is N, then the number of
renters out of N is of binomial distribution. Thus, you can estimate the
error, which is discussed in any
【在 a*****n 的大作中提到】
: Suppose a planning agency needs to know the proportion of households in an
: area which are renters. It wants a sample estimate which is accurate to ±4
: percent. How large of a sample must it have in order to be sure of this
: level of accuracy at the 95 percent (±2 a) level of confidence?
: I know it must be very simple, but I am toooo :(
|
|
