Dear community,
I ran a logistic regression with continious IV in SPSS. In the table "variables in the equation" one variable is missing (despite using entry method) and without any message from SPSS . When browsing through the web I understood that this might happen due to collinearity. However Collinearity diagnostics did not return a clear sign for it. Highest VIF-values are 6.1 and the highest Conditionindex is 21.1.
So my question are:
1. Is my regression model still valid despite SPSS dropping one variable?
2. Are there other reasons than collinearity why the IV is missing in the model.
Thanks Ilka
Hello Ilka,
Three possibilities come to mind (and, SPSS is frequently uninformative or under-informative when it comes to issues with data):
1. The omitted IV has no variance.
2. The omitted IV does, in fact, represent some linear combination of other IVs, and is therefore redundant. One common example would be the case wherein a total score and its constituent subscores are all included as IVs in a model.
3. A sufficient number of missing data points on IV(s) were imputed, and by so doing, generated a set of incoherent relationships (covariances or correlations); a Heywood case is one manifestation of the problem of incoherent relationships (though not always due to missing data imputation).
Is the model valid? Well, it's accurate for your data set, given the variables that ended up being included. Is it useful? That's for you and your audience to decide.
Good luck with your work.
In Regression model, you have dependent and independent variables, marked. If independent variable is not contributing linearly to the model, it will not appear in the model generated by the soft ware. Say, by adding a variable X, there is not much change in R^2 (significant enough), then the model will not select that variable rather drop it from the list. Another reason could be the collinearity which you have already stated.
Take a look at this paper and it's references and you will not want to use SPSS for logistic regression again. David Booth
Try to look at the descriptive statistics followed by univariate regression to see if you can spot the issue from there.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS