h*********o
发帖数: 151 | 1 Look at the following two hypotheses:
H0: p1>p2. Given 90% confidence, critical value is 1.28.
H0: p1=p2. Given 90% confidence, critical values are 1.64 & -1.64.
Now suppose I conduct both hypotheses using the same sample. The test
statistic = 1.35 (should be the same for both tests). So that means p1 is
statistically higher then p2 BUT p1 and p2 are not significantly different?
I am thinking if “higher”, then should be “different”. How do you think? | s*r
发帖数: 2757 | | n**m
发帖数: 156 | 3 if test p1>p2, critical value should be -1.28, reject the null hypothesis if
test smaller than -1.28, right? |
|
