statistical inference

[ravian]

hello,

during the testing of hypothesis, i cant figure out when to use a one tailed test and when to use a two tailed one.please help.

[duck]

Hi sayantan.. :)

When alternate hypothesis is a negation of null then we use Two tailed test.

Suppose, Null hypothesis: true population mean = u_0
Alternate hypothesis : true population mean is not equal to u_0


And when we have a case like this>
Null hypothesis: true population mean = u_0
Alternate hypothesis : true population mean is greater than/less than u_0
then, we use one tailed test.

[ravian]

yes, that is true. I want to know what test to use when the alternate hyp. is not specified(i.e. no info is given on the alternate hyp). Should i use a 2 tailed or a 1 tailed test??

On Mon, Jun 20, 2011 at 8:33 PM, duck [via Discussion forum] <[hidden email]> wrote:

Hi sayantan.. :)

When alternate hypothesis is a negation of null then we use Two tailed test.

Suppose, Null hypothesis: true population mean = u_0
Alternate hypothesis : true population mean is not equal to u_0


And when we have a case like this>
Null hypothesis: true population mean = u_0
Alternate hypothesis : true population mean is greater than/less than u_0
then, we use one tailed test.

---

If you reply to this email, your message will be added to the discussion below:

http://discussion-forum.2150183.n2.nabble.com/statistical-inference-tp6495543p6496117.html

To unsubscribe from statistical inference, click here.

[Amit Goyal]

Use common sense.