hypotheses testing

• Oct 24th 2010, 07:40 PM
DINOCALC09
hypotheses testing
In a study, 900 subjects were treated with a drug compared with 100 in a placebo group. 60 of the subjects in the drug group and 4 in the placebo group complained of headaches. Use the binomial exact test at the 5% level to test whether headaches are more frequent among subjects treated with th drug. ( Do not use \pi = 0.5)

H_0 = ?
H_a = ?
\alpha =? I know its \alpha = 0.05
P-value = ?
What can be concluded?

************************************************** ************************************************** ***********
Does this problem have something to do with N-n in which you're supposed to reverse the "successes" and "failures? I read somewhere that in order to obtain the full power of th binomial exact test, n should be no more than 5% of N. It seems that 5% of 1,000 is 50 so what should I do.
• Oct 24th 2010, 09:27 PM
downthesun01
Well, since the problem clearly states to use the binomial exact test, you should use it
• Oct 25th 2010, 02:16 PM
DINOCALC09
My question was not whether or not to use the binomial exact test. I can read. Thanks. A little more insight would be appreciated.