ExcelStatWorld
 

  You don't need to be an 'investor' to invest in Singletrack: 6 days left: 95% of target - Find out more

[Closed] ExcelStatWorld

3 Posts
3 Users
0 Reactions
44 Views
Posts: 5787
Full Member
Topic starter
 

V quick and probably dumb Excel question: I already have means and standard deviations for two samples, and want to run a quick t-test within Excel using those values. The options I've found all assume I have the raw observation data, rather than the means and SDs.

So... how do I go about just entering the means, SD and sample size and getting the t-test output?

ETA: I want to do it in Excel as I have a stack of these tests to run, and the means, sample sizes and SDs are all already in Excel


 
Posted : 18/04/2018 4:21 pm
Posts: 36
Free Member
 

from the means, sds and sample sizes you can get back to the sum of teh squares for each distribution and then run them back through the t-test formula I think.

None of that is excel forumlae though, just algebra.


 
Posted : 18/04/2018 4:40 pm
 poly
Posts: 8699
Free Member
 

I’m not sure a t-test can really be considered valid if you don’t have the raw data*?  I’m rusty but is that not when a z-test might be more useful?

*i presume it assumes a normal distribution.


 
Posted : 18/04/2018 6:19 pm
Posts: 5787
Full Member
Topic starter
 

Thank you, it's helpful. I've gone back to the raw data to bring that into Excel; it's a bit more hassle, but pushes the more cerebrally challenging part (the stats testing) onto Excel.


 
Posted : 18/04/2018 7:00 pm

6 DAYS LEFT
We are currently at 95% of our target!