# Need help finding mu

• May 25th 2011, 01:10 PM
lancelot854
Need help finding mu
Ok I have to do a 2 sample t-test so i have 2 different sets of data that are similar to this.
alpha = .05

If i take the average of each set, I can find the x-bar but i need to find the mu in order to complete the test statistic. Any help is appreciated, thanks
• May 25th 2011, 02:39 PM
pickslides
You you need to establish if there is any difference between the two sets then the null hypothesis is

$\displaystyle H_0 : \mu_1 - \mu_2 = 0$

and

$\displaystyle H_A : \mu_1 - \mu_2 \neq 0$

The $\displaystyle \mu_0$ you require is the hypothesised difference so make $\displaystyle \mu_0=0$ in your test statistic.
• May 25th 2011, 03:38 PM
lancelot854
The formula includes mu0 and mu1 though, so what would i put in that place?
Also, wouldn't the null hypothesis be mu0-mu1= c, where c = some constant
• May 25th 2011, 06:13 PM
matheagle
I thought Moo was online.
• May 25th 2011, 06:42 PM
lancelot854
Quote:

Originally Posted by matheagle
I thought Moo was online.

Am I missing something? Sorry if I don't get the reference. o.0
• May 25th 2011, 06:49 PM
matheagle
• May 25th 2011, 07:22 PM
lancelot854
Quote:

Originally Posted by matheagle

Oh okay. Although, that doesn't really help xD. I just want to confirm: the mu0 and mu1 in the formula are both set equal to 0? Also, if the sample sizes are the same, the degrees of freedom is n1 + n2 - 2, right? Thanks for the help
• May 25th 2011, 09:27 PM
matheagle
well someone needs to win some joke points away from Mr Fantasy
First of all, what are you trying to prove?
we need to know, what the alternative hypoethesis is.
Next we need to knw if this is a paired difference experiment, since we
have the same number of boys and girls.
But I doubt that it is a paired diff experiment.
Next we need to know if the population variances are equal or not.
They seem to be, if thats the case.
The test stat would be

${(\bar X_1-\bar X_2)-(\mu_1-\mu_2)\over s_p\sqrt{{1\over n_1}+{1\over n_2}}$

where we are pooling the sample variances.
Then we need to know if this is a one-sided or two-sided test.
So, you need to tell me what they are asking you to prove
and the distribution under the null hypothesis would be a t with df
equal to the sum of the two sample sizes minus two.