An event is happening n times, and there are 2 hypothesis: H(0)=the probablity of success for each event is p, H(1)=the probablity of success for each event is q lets assume there were h success. I proved that $P(\Lambda(h) > a) = P(h > b)$ for some positive constants a, I got $b=\frac{log(a)-n*log(1-p/1-q))}{log(p/q)-log(1-p/1-q)}$
now I need to find the b, s.t the false negative probablity is 0.2. But I can't understand how to choose the value of a, for this to happen. (if calculating it for any n,p,q is hard just pick n=10,p=0.2,q=0.8 or any numbers you want)