Coffeehouse for desis
Would you like to react to this message? Create an account in a few clicks or log in to continue.


Statistics 101 question

2 posters

Coffeehouse for desis :: Coffeehouse :: Coffeehouse Main

Go down

Statistics 101 question Empty Statistics 101 question

Post by MulaiAzhagi Wed Feb 22, 2012 2:31 pm

For a two-tailed test, we can compute the confidence interval from a sample statistic and then check whether the population parameter specified in the null hypothesis is within the CI inteval for "failing to reject" the null hypothesis. If the CI interval does not contain the population parameter value, we can reject the null hypothesis.

For a one-tailed test, can we use the the confidence interval to "reject" or "fail to reject" the null hypothesis?


Looks like the CI method works only for a two-tailed test.


Why is it so?

MulaiAzhagi

Posts : 1254
Join date : 2011-12-20

Back to top Go down

Statistics 101 question Empty Re: Statistics 101 question

Post by truthbetold Sun Feb 26, 2012 6:35 am

MulaiAzhagi wrote:For a two-tailed test, we can compute the confidence interval from a sample statistic and then check whether the population parameter specified in the null hypothesis is within the CI inteval for "failing to reject" the null hypothesis. If the CI interval does not contain the population parameter value, we can reject the null hypothesis.

For a one-tailed test, can we use the the confidence interval to "reject" or "fail to reject" the null hypothesis?


Looks like the CI method works only for a two-tailed test.


Why is it so?

Sara,
both one tailed test and two tailed use an interval estimate, confidence level and similar logic reject or fail to reject. In one tailed(sided) test alternate hypothesis selected predetermined only side of estimate is of interest.

truthbetold

Posts : 6799
Join date : 2011-06-07

Back to top Go down

Statistics 101 question Empty Re: Statistics 101 question

Post by truthbetold Sun Feb 26, 2012 6:42 am

Sara,

You could search wikipedia with one and two tailed(sided) test and confidence interval for initial understanding. You can search for t test and z test to get a next level of clarity.

truthbetold

Posts : 6799
Join date : 2011-06-07

Back to top Go down

Statistics 101 question Empty Re: Statistics 101 question

Post by Sponsored content


Sponsored content


Back to top Go down

Back to top

- Similar topics

Coffeehouse for desis :: Coffeehouse :: Coffeehouse Main

 
Permissions in this forum:
You cannot reply to topics in this forum