Re: To accept the Null Hypotheses

On 8 Ago, 23:07, "Luis A. Afonso" <lic...@xxxxxxxxxxx> wrote:
To accept the Null Hypotheses

__1__Suppose that the Test Statistics falls in the so-called * Acceptance Interval *.

__The Test is a ONE-TAIL: ACCEPT H0
__TWO-TAILED: no sufficient evidence to reject H0.
*****
__2__On the contrary if the TEST falls in the * Rejection Interval *, then Reject H0 whatever the test.
********
I think that the no distinction of what the test kind (one or two tailed) had originate a lot of misunderstands like: the test statistics aims to check an evident impossibility such the Null Hypotheses be true, or that’s possible to infer that the Null Hypotheses is true whatever the test. This later nonsense is explicitly backed by the IMBECILE DUO: Jack Tomsky / John Smith, namely when H0: m=c, Ha: m=/c, c=constant.
_________
Open to discussion.

Luis Amaral Afonso [The moderator destroyer]

.