Hey guys, in my statistic course the null hypothesis is always equal the value to be tested. For one tail test: Ho: x=a and Ha: x>a or Ha: x

Ho: x=a is not really a one-tailed test because if the value is greater or less then you reject. I suppose in your set-up (Ho: x=a and Ha: x>a or Ha: x

You get me wrong, what i’ve told is about one tail test. In the CFA text: Ho: x>( or equal) a Ha: x

use the cfai way. that is the way you will be tested. h >= a is null, h < a is alternate for a 1-tailed test h = a is null, h <> a is alternate for a 2-tailed test.