Stat: 怎能知道reject or accept H0

http://en.wikipedia.org/wiki/Statistical_hypothesis_testing

Basically in hypothesis testing,
the test statistics under the null hypothesis will follow a certain
distribution (in you case is the t-distribution)
And then you check whether the observed test statistics
lie in the critical region or not.
The critical region is the tail part of the distribution
which determined by the significance level of your test and
your alternative hypothesis is whether two-sided or one-sided