Ordinal Regression - test of parallel lines problem

[Katina Dimoulias]

Hello,

I have run a number of ordinal regression analyses with different sets of
predictor variables and I continue to get the below results for the Test of
Parallel Lines.  My predictor variables have 3 ordered levels and DV has 2
ordered levels.  Can someone please give advice as to what I may be doing
wrong and how I can fix it or if these are legitimate results (from what I
have read and understood they seem not to be) that I am getting?

Test of Parallel Lines(a)

Model-2 Log LikelihoodChi-SquaredfSig.
Null Hypothesis20.693
General20.693.0000.
The null hypothesis states that the location parameters (slope
coefficients) are the same across response categories.
a  Link function: Complementary Log-log.

Your advice will be greatly appreciated,

Katina

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Henrik Lolle]

Hi Katina,

Actually, it's your dependant variable that should be ordinal scaled in
this type of regression, not the independant. The dependant variable in
this analysis of yours is a dummy, so you will get the same results in
ordinary logistic regression and it gives no meaning to test for
parallel lines because there is only one.

Best,
Henrik

Quoting Katina Dimoulias <[hidden email]>:

> Hello,
>
> I have run a number of ordinal regression analyses with different sets of
> predictor variables and I continue to get the below results for the Test of
> Parallel Lines.  My predictor variables have 3 ordered levels and DV has 2
> ordered levels.  Can someone please give advice as to what I may be doing
> wrong and how I can fix it or if these are legitimate results (from what I
> have read and understood they seem not to be) that I am getting?
>
> Test of Parallel Lines(a)
>
> Model-2 Log LikelihoodChi-SquaredfSig.
> Null Hypothesis20.693
> General20.693.0000.
> The null hypothesis states that the location parameters (slope
> coefficients) are the same across response categories.
> a  Link function: Complementary Log-log.
>
> Your advice will be greatly appreciated,
>
> Katina
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>

************************************************************
Henrik Lolle
Department of Economics, Politics and Public Administration
Aalborg University
Fibigerstraede 1
9200 Aalborg
Phone: [phone]
************************************************************

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD