The same tools apply except that the mean of a population is approximated by the proportion of "successes" in the sample data.

the variance is estimated by p*(1-p).

Thus the equivalent of sample distribution SEM is the sqrt((p*(1-p))/n)

In this example, according to the *Governor's Council for a drug-free
workplace Report* (Spring/Summer 1995), a large number of companies
in New Jersey were sent a questionnaire about whether employees' performance
was affected by the use of alcohol or other drugs.

The results from those who responded are given in the file **druguse.txt**

Is the observed ratio far enough away from 0 and 1.0 that we can use
the normal approximation to the binomial distribution?

Use a 95% confidence interval to estimate the proportion of New Jersey businesses with these substance abuse problems.

solution **phatdrug.xls**