Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# st: RE: creating combined correlation of dummy (ordered multilevel)

 From Nick Cox To "'statalist@hsphsun2.harvard.edu'" Subject st: RE: creating combined correlation of dummy (ordered multilevel) Date Thu, 9 Jun 2011 17:01:34 +0100

```A correlation coefficient is only likely to have meaning if the combined variable in question is at least on an ordered scale. Even to calculate a meaningless correlation for a six-category scale you would need to have such a combined variable.

Nick
n.j.cox@durham.ac.uk

Stefan Nijssen

I am struggling to find the correlation of various variables to a dummy variable (6level ordered). The problem is that since the dummies appear in the table as six variables, I cannot read a single correlation between the combined dummy and any other variable (see below). Does anyone know how to do this?

Stefan Nijssen

. correlate AAA AA A BBB BB B wc5 MAR COV
(obs=383)

|      AAA       AA        A      BBB       BB        B      wc5      MAR      COV
-------------+---------------------------------------------------------------------------------
AAA |   1.0000
AA |  -0.0303   1.0000
A |  -0.0767  -0.1249   1.0000
BBB |  -0.1187  -0.1932  -0.4892   1.0000
BB |  -0.0559  -0.0909  -0.2302  -0.3563   1.0000
B |  -0.0504  -0.0820  -0.2077  -0.3214  -0.1513   1.0000
wc5 |   0.3377   0.1016   0.1805  -0.1683  -0.0769  -0.1032   1.0000
MAR |   0.0030   0.0101   0.0146   0.0449  -0.0136  -0.0807   0.0176   1.0000
COV |   0.3052   0.1377   0.2432  -0.0967  -0.1235  -0.2545   0.2121   0.1220   1.0000

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```