Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
one sided t test
Message posted by Amy on November 2, 1999 at 12:00 AM (ET)
is there a difference between the t test and the one sided t test? If so, what? and could you please provide me with some examples.
Thanks, Amy
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: one sided t test
Message posted by Vern Myers on November 2, 1999 at 12:00 AM (ET)
You are looking at two applications of the same test. The one-sided t-test is looking at the critical region in ONE tail of the bell curve; it is used when evaluating the hypotheses that the population mean is either equal to, or else greater (or less than) a specified value. The two-sided t-test is looking at the critical regions in BOTH tails of the bell curve; used evaluating the hypotheses that the population mean is either equal to, or else not equal to a specified value. The previous example is for the t-test for inference for a single mean. There are other applications, such as inference for two means, and inference for the mean of a paired difference. The calculations are a little more involved in the latter cases, but the principle is the same. Note that the t-test is for a small sample (n<30), for larger samples, the z-test is usually used.
Re: one sided t test
Message posted by JG on November 3, 1999 at 12:00 AM (ET)
A two sided t-test can also be performed as a one sided F-test.
Re: one sided t test
Message posted by Jack Tomsky on November 3, 1999 at 12:00 AM (ET)
This is because the square of a t with m degrees of freedom is an F with 1 and m degrees of freedom.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.