JCC Help me with a Stat problem

So you can help Social Science Education majors pass their Stats class?

 

Or something, yeah.

 

I do stats because I'm not smart enough to do silly nonsense maths. Having said that, Conan > Russell.

Sent from my GT-I9505 using Tapatalk

 

I read this thread title as "Help me with a STAT problem." As in, help me with an immediate, it's an emergency, problem.
Then upon opening the thread I realized it was stat as in statistics and and had flashbacks to the summer I took statistics in college. *shutters*.

 

Well, honestly, is there any mathematician or philosopher anyone could rank as greater than Conan? No.

 

Ramza dp4m Jabba-wocky

So sorry guys this is the last actual problem I have but I'm stuck.

Problem is as such:

Construct a 95% confidence interval for mu1 - mu2. Two samples are randomly selected from normal populations. The sample statistics are given below.

n1=8
n2=7
x-bar1-4.1
x-bar2=5.5
s1=0.76
s2=2.5

I tried using the Confidence Interval for the Difference of Two Means....and am not getting the right answers. I then tried it in my TI-83 with a 2-Sample Z-Interval and get the same thing as I get on paper. What I get is (-3.333, .53254). What my test is saying I should get is (-3.813, 1.013)

Your help would be appreciated, sorry that I'm cutting into your Sunday.

 

darthcaedus1138

I think I've got it.

IIRC the formula is


Then you use 6 as your estimate for degrees of freedom. When I cross-referenced this table http://en.wikipedia.org/wiki/Student's_t-distribution#Table_of_selected_values it gave 2.447, I was getting precisely their answer at about 2.455, so I think that's about the whole of it.

C.R. http://www.stat.purdue.edu/~tlzhang/stat511/chapter9_1.pdf, http://www.stat.purdue.edu/~tlzhang/stat511/chapter9_2.pdf, though they're jerks and don't provide a t-table.

 

I've got a t-table in my book, looking for the 2.455 now. Thanks so much for your help.

Where did you get the degree of freedom btw?

Replicated very close with 2.447, that would practically be all I really needed, if it's multiple choice I only need to get close.

 

The second Purdue link gives the formula

(Technically they use m and n, whatever). That gives 6.9yadda, and you have to round down, hence 6.

 

Thanks very much, I went through the Purdue links but the different terminology they use gets confusing.

Once more a million thank yous.

Yup I got 6.96 so we're now on the same page.

It appears I was putting some of the information into my calculator wrong. The answer doesn't come out right exactly, but it's much closer to the answer. It also tells me my df would be 6.96, so that definitely would help.

 

I'm trying to decide if I'm disappointed I didn't get tagged in this thread or not. I mean, I only taught statistics at a university, after all.

 

I can't say anything bad, Inty, I'm still waiting on your Comic Draft vote...

 

You need to make more of a name for yourself as the go to math guy.

Right now Ramza has that title.

 

Indeed. Namedrops don't come free, you know.

 

I'll keep you in mind from now on DarthIntegral