Can a paired sample T-test in statistics have a directional alternative hypothesis?

I'm wondering if the alternative hypothesis in a paired t-test can be directional, or if it just has to be Ud[doesnotequal] 0. The directional hypothesis would be u1%26gt;u2 or u1%26lt;u2 and would be a one-tailed P-value. If not, how do you state the conclusion of the test if the null hypothesis is rejected?|||ANSWER: Yes!





T-test for paired data must have either "equal" or a "directional" (less than, greater than) alternative hypothesis. This is necessary because when the "test" is made, this determines when there is a “one-tailed” or a "two-tailed" test.