Statistics exam at 10, need your help.

[Ultima]

I can't find my notebook, so I need your help
I already figured out how to do linear correlation thru google, but a couple of other things still puzzle me.

How do I calculate the answer at a given confidence interval? I know you use the z-score table to calculate it normally from the graph by figuring out how many std-deviations it is from the mean, but how do you do this with varying confidence? With a null hypothesis? Like if I have a question that says "what is the probability so and so will happen with 90% confidence" and i have the z-table? Also, what's the difference between one-tailed and two-tailed tests?

Any help appreciated

 

[HomeBrewerDude]

last post of the night...

Also, what's the difference between one-tailed and two-tailed tests?

one tail test is unidirectional so all of the alpha zone is located at one end of the curve. So you use the 5% Z score rather than a 2.5% (assuming you set your alpha to %5. Btw, this means that you can't detect significant differences in the opposite direction.

your other questions don't really make sense...or i am too tired to understand what you are tyring to say.

Like if I have a question that says "what is the probability so and so will happen with 90% confidence" and i have the z-table?

probably you just convert the z-score into a percentage of area underneath the curve.

 

[Ultima]

Originally posted by: yamahaXS
last post of the night...

Also, what's the difference between one-tailed and two-tailed tests?

one tail test is unidirectional so all of the alpha zone is located at one end of the curve. So you use the 5% Z score rather than a 2.5% (assuming you set your alpha to %5. Btw, this means that you can't detect significant differences in the opposite direction.

k... I think I get that part now

your other questions don't really make sense...or i am too tired to understand what you are tyring to say.

Like if I have a question that says "what is the probability so and so will happen with 90% confidence" and i have the z-table?

probably you just convert the z-score into a percentage of area underneath the curve.

Nope it wasn't that. It was something else. Dunno how it works cause I can't find my notebook!