PDA

View Full Version : Stats problem... need help!!!



LoveStruck16
11-18-2009, 05:08 PM
I have this stats question and I'm confused on what to do, any help would be greatly appreciated.

- Determine whether an observed sample mean of 84 based upon a sample of n=25 is significantly different from a (null hypothesized) population mean of 81 with a standard deviation of 10.
a) state your hypothesis
b) state your alpha value and the critical values
c) show your calculations
d) state your decision regarding the null hypothesis

tkhunny
11-19-2009, 11:01 PM
a1) H_{0}: Population Mean = 81
a2) H_{a}: Population Mean NOT= 81

Okay, you do part b).

LoveStruck16
11-22-2009, 08:47 PM
So the alpha/critical values would be 0.05 and 0.10?

tkhunny
11-22-2009, 10:13 PM
You pick it. It is also very important to decide what kind of test you will be performing. 1-tail? Maybe not. It goes to the nature of the question and the definition of the hypotheses. They way I defined them, it appears to be a 2-tail test. Perhaps you defined the hypotheses some other way.

LoveStruck16
11-23-2009, 12:58 AM
K so if i do a two tail test then the critical values would be..

0.10

Thus, 1-0.10 = 0.90
0.90/2 = 0.45

Then I look at the Z-table chart in my textbook and look for .10 and the value is then 1.812 and I divide that by 2 (1.812/2=0.906)

Therefore the critical value would be 0.906? ..

I hope the way i explained it wasn't confusing...

tkhunny
11-23-2009, 08:05 PM
1) That seems a little odd. Normally, one would work with the tail area. Certainly it would depend on how your table is designed.
2) Are you sure a Sample Size of 25 is big enough to use a z-score? Maybe a t-score would be more appropriate?

What say you?