## Two Regression questions -- one about Logistic Regression

### Two Regression questions -- one about Logistic Regression

Hi everyone,

I'm having a problem running my regression analyses.  Basically, I have 5
Personality Profiles that I have created using Cluster Analysis.  I am
trying to use these personality profiles to predict career success
(extrinsic and intrinsic).  I have two measures for each type of success --
I combined the intrinsic measure linearly, and I created a 9-box matrix of
categories for Extrinsic Success where there are three groups for each of
the two variables (salary and no. of promotions) so you have Hi:Lo Hi:Avg,
Lo:Hi etc..

First question and problem:
For my Intrinsic composite I entered my 7 control variables into a Linear
regression equation - so far so good.  I dummy coded the 5 personality
profiles (1=belongs to profile; 0=does not belong to profile).  I then
entered the 5 dummy coded variables into the regression, and it keeps
leaving one of my personality profiles out of the equation due to the
Tolerance Statistic.  How do I get around this?  Is it possible?  Am I
better off NOT dummy coding my personality profile grouping variable and
entering that into the equation and then using some other method to
determine which of the 5 personality profiles is the significant predictor?

Second question and problem:
For my extrinsic variable -- its a categorical variable as described above
(i actually have the data cut different ways -- 9-box grid, and 4-box grid
(only Hi and Lo categories for each variable).  Interested group being those
that are high on both variables -- High Salary and High No. of Promotions.
So, I'm looking at a Multinomial Logistic Regression in this case.  How do
deal with control variables here?  I entered them into the regression and it
said it couldn't run it because several cells were empty.  How do I get
around this?

In addtion, when i only enter the personality profiles variables, the
parameter estimates only show for cells 1-8 and not for cell 9 which is my
High:High group that I'm interested in.  I'm new to Logistic Regression and
I've read up on interpreting the Output, but I guess I'm still a little
confused.  Any help is greatly appreciated.

Thanks!
Claudia

SAS-L,

I've come across an inconsistency in the results output from LOGISTIC and
GENMOD for a logistic regression.  The inconsistency appears with
categorical predictor variables, the coefficient and standard error
estimates from GENMOD are exactly double the LOGISTIC estimates for
dichotomous variables, and for the x5 (values 1,2,3) variable they don't
seem to have an exact relationship.  This occurs for the dichotomous
variables when they are included in the CLASS statement.

The estimates for the continuous variables agree regardless of what other
variables are in the model.  I believe that the LOGISTIC and GENMOD code
should be providing the same models.  Can someone help explain this?  I hope
there is just something simple I'm missing. [Example dataset and code below]

I've got v8.2  TS02M0 on Win98.

Thanks!
Jennifer

data test;
input y x1 x2 x3 x4 x5;
cards;
1 1 0 2.45 16.12 1
1 0 1 3.45 13.18 2
1 1 1 2.34 14.27 3
1 1 1 3.12 24.23 3
1 1 1 2.34 16.56 3
1 1 0 3.89 14.34 2
1 0 0 1.34 20.56 2
0 0 0 1.56 18.45 1
0 1 0 1.34 15.45 1
0 0 1 2.14 20.34 2
0 0 1 2.56 19.53 3
0 0 0 2.32 18.45 3
0 1 0 1.89 19.98 2
0 0 0 2.68 18.45 2
0 0 0 2.98 16.12 1
0 0 0 2.57 12.34 2
;
run;

/*MODEL 1*/
proc logistic data=test descending;
class x1;
model y=x1;
run;
proc genmod data=test descending;
class x1;
run;

/*MODEL 2*/
proc logistic data=test descending;
class x1 x2;
model y=x1 x2;
run;
proc genmod data=test descending;
class x1 x2;
run;

/*MODEL 3*/
proc logistic data=test descending;
class x1;
model y=x1 x3;
run;
proc genmod data=test descending;
class x1;
run;

/*MODEL 4*/
proc logistic data=test descending;
class x1 x2;
model y=x1 x2 x3;
run;
proc genmod data=test descending;
class x1 x2;
model y=x1 x2 x3 /dist=bin link=logit;
run;

/*MODEL 5*/
proc logistic data=test descending;
model y=x3 x4;
run;
proc genmod data=test descending;
run;

/*MODEL 6*/
proc logistic data=test descending;
class x1 x5;
model y=x1 x5;
run;
proc genmod data=test descending;
class x1 x5;