Thread: Two Sample T-Test

I have a t-test question that I need help with.


This is the information that has been given to me.


Male Female
m=41 n=17

x(bar) =69.2 y(bar)= 66.7

S(x)=2.66 S(y)= 2.34

Test to see whether the difference between mean male and female is 5.5 inches.
I got the test statistic which was 3.57, and the pooled estimate variance to be 2.57


The problem is i cant get the p value, i dont understand where i am meant to look in the tables.

In my book it says the forumula to get the P-Value is: P=2*P(T>|t|) ,where t is observed value.

Any help please

Thomas

Originally Posted by thomasboateng

I have a t-test question that I need help with.


This is the information that has been given to me.


Male Female
m=41 n=17

x(bar) =69.2 y(bar)= 66.7

S(x)=2.66 S(y)= 2.34

Test to see whether the difference between mean male and female is 5.5 inches.
I got the test statistic which was 3.57, and the pooled estimate variance to be 2.57


The problem is i cant get the p value, i dont understand where i am meant to look in the tables.

In my book it says the forumula to get the P-Value is: P=2*P(T>|t|) ,where t is observed value.

Any help please

Thomas

What is your test statistic? What are the number of degrees of freedom?

See this site for more information.

CB

I still dont get it,

my test statistic is 3.37. not sure what my degrees of freedom are?

Originally Posted by thomasboateng

I still dont get it,

my test statistic is 3.37. not sure what my degrees of freedom are?

Read the material on the link I gave you.

CB

yes the degrees of freedom is 56 for me

but in the example its 66 and its cv is

If using 2-tailed test

Alpha.05, tcv= 1.997

it just tells me the cv is 1.997 is it always going to be 1.997 ?

Thomas

Thomas,

Yes you are correct it is always going to be 1.997 for two sample t tests.

Luke

Originally Posted by thomasboateng

yes the degrees of freedom is 56 for me

but in the example its 66 and its cv is

If using 2-tailed test

Alpha.05, tcv= 1.997

it just tells me the cv is 1.997 is it always going to be 1.997 ?

Thomas

No you need to look up the cumulative t-distribution with 56 degrees of freedom for a critical value of 0.975, which will give t_cv=2.004 rather than the t_cv=1.997 for 66 degrees of freedom.

(You may have a table of critical values for t-tailed tests then you would look up the critical value that corresponds to 0.95 probability of the t-score being between +/- t_cv)

CB

Originally Posted by lukeheselden

Thomas,

Yes you are correct it is always going to be 1.997 for two sample t tests.

Luke

This is wrong, see my other post.

CB