sample size calculation with a specific alternative

[Peter_Y]

Hello:

  I need to calculation sample size for comparing 2 proportions p1 and p2 where the null hypothesis is p1=p2 and the alternative hypothesis is |p1 - p2| > 0.2. That is, the 2 proportions have to differ at least 02. Does anyone know how to calculate power given different sample sizes and true proportions e.g for n1=50 n2=50 p1=0.5 p2 = 0.2

Thanks,

Peter