Hi,
I have ran some correlation analysis and obtained correlation coefficents against series Y:
series X 0.82
series M 0.89
I would like to test that correlation for series M is statistically signifciantly different from the correlation for series X at the 95% confidence level.
Does anyone know what statistical function to use excel to do this and what parameters I need to enter?
Thanks,
Hamond
Hello Hamond,
hope I'm not completely off on this, one test for independence I know the Chi-Square test. There is a worksheet function in Excel for that.
Search in Excel help for CHI, and it'll display the results (CHITEST, CHIINV, CHIDIST).
HTH,
Isabella
Addendum: Or rather, FTEST? I'm not too good at statistics.
Hi Isabella,
I don't think it is either of these, defo not chi sqaure as this is a non parametric test.
Could be f-test or t-test or something else!
PLS DO NOT PM; OPEN A THREAD INSTEAD!!!
1) Posting Code
[CODE]PasteYourCodeHere[/CODE]
(or paste your code, select it, click # button)
2) Uploading File(s)
Go Advanced / Attachments - Manage Attachments / Add Files / Select Files / Select the file(s) (multiple files can be selected while holding Ctrl key) / Upload Files / Done
Replace company specific / sensitive / confidential data. Include so many rows and sheets etc in the uploaded workbook to enable the helpers visualize the data and table structure. Helpers do not need the entire workbook.
3) Testing the Codes
always back up your files before testing the codes.
4) Marking the Thread as Solved
from Thread Tools (on the top right corner, above the first message)
Posting Permissions
Reply With Quote
