=+successes occur three times as frequently in population 1 as they do in population 2. Alternatively, if the ’s refer to success proportions for two different treatments, then a ratio of 3 implies that the first treatment is three times as likely to result in a success as is the second treatment. Consider independent random samples of sizes n1 and n2 from the two different populations, which result in sample proportions p1 and p2, respectively. Also let u 5 number of successes in the first sample and v 5 number of successes in the second sample. When the n’s are both large, the statistic ln(p1 yp2)has approximately a normal sampling distribution with approximate mean value and standard deviation