ExcelBanter

ExcelBanter (https://www.excelbanter.com/)
-   Excel Worksheet Functions (https://www.excelbanter.com/excel-worksheet-functions/)
-   -   Multiple Correlations and Multicollinearity (https://www.excelbanter.com/excel-worksheet-functions/103817-multiple-correlations-multicollinearity.html)

montreal1775

Multiple Correlations and Multicollinearity
 

I was wondering if there is a way to get excel to run correlation test
between every combination of 37 or 38 independent variable. I need to
find the best way to reduce multicollinearity, and finding every
possible correlation would give me the best idea of which direction to
take. I know the actual computation would take quite a while, but the
computer that would run the tests is very fast, so even 38! runs
shouldn't take very long.

I was also wondering if there was a way to find correlation between one
independent variable and every possible combination of the others,
although this would take much more computing time and might not even
be necessary.

Any guidance is appreciated!


--
montreal1775
------------------------------------------------------------------------
montreal1775's Profile: http://www.excelforum.com/member.php...o&userid=36994
View this thread: http://www.excelforum.com/showthread...hreadid=569518


Dave F

Multiple Correlations and Multicollinearity
 
That sounds like something you need a statistics package for.

"montreal1775" wrote:


I was wondering if there is a way to get excel to run correlation test
between every combination of 37 or 38 independent variable. I need to
find the best way to reduce multicollinearity, and finding every
possible correlation would give me the best idea of which direction to
take. I know the actual computation would take quite a while, but the
computer that would run the tests is very fast, so even 38! runs
shouldn't take very long.

I was also wondering if there was a way to find correlation between one
independent variable and every possible combination of the others,
although this would take much more computing time and might not even
be necessary.

Any guidance is appreciated!


--
montreal1775
------------------------------------------------------------------------
montreal1775's Profile: http://www.excelforum.com/member.php...o&userid=36994
View this thread: http://www.excelforum.com/showthread...hreadid=569518



montreal1775

Multiple Correlations and Multicollinearity
 

Which are the best packages? Do you think they can run multiple
correlation tests over several columns, or would I need to find a macro
to do that?

Thanks for your response.


--
montreal1775
------------------------------------------------------------------------
montreal1775's Profile: http://www.excelforum.com/member.php...o&userid=36994
View this thread: http://www.excelforum.com/showthread...hreadid=569518


montreal1775

Multiple Correlations and Multicollinearity
 

Also, I know Excel has a correlation function, so the only difficulty is
automating it to run and rerun the test over a range. If you think there
is a way to do so with a macro or otherwise, please let me know.


--
montreal1775
------------------------------------------------------------------------
montreal1775's Profile: http://www.excelforum.com/member.php...o&userid=36994
View this thread: http://www.excelforum.com/showthread...hreadid=569518


Mike Middleton

Multiple Correlations and Multicollinearity
 
montreal1775 -

Tools | Data Analysis | Correlation

If Data Analysis doesn't appear on your Tools menu, check the Analysis
ToolPak box at Tools | Add-Ins. If it's not listed there, add it from your
Office/Excel installation CD.

Please don't post in multiple forums, and if you must, please state that you
are doing so.

- Mike
http://www.mikemiddleton.com

"montreal1775"
wrote in message
news:montreal1775.2c8dui_1155056711.9456@excelforu m-nospam.com...

I was wondering if there is a way to get excel to run correlation test
between every combination of 37 or 38 independent variable. I need to
find the best way to reduce multicollinearity, and finding every
possible correlation would give me the best idea of which direction to
take. I know the actual computation would take quite a while, but the
computer that would run the tests is very fast, so even 38! runs
shouldn't take very long.

I was also wondering if there was a way to find correlation between one
independent variable and every possible combination of the others,
although this would take much more computing time and might not even
be necessary.

Any guidance is appreciated!


--
montreal1775
------------------------------------------------------------------------
montreal1775's Profile:
http://www.excelforum.com/member.php...o&userid=36994
View this thread: http://www.excelforum.com/showthread...hreadid=569518





All times are GMT +1. The time now is 06:06 AM.

Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
ExcelBanter.com