Generating and interpreting collinearity diagnostics when performing logistic regression in SPSS
Вставка
- Опубліковано 6 сер 2024
- This video provides a work-around for generating collinearity diagnostics when performing logistic regression through the SPSS menus. Additionally, a provide some general interpretation of discussion of the diagnostics generated in the context of the example provided.
Download a copy of the dataset here: drive.google.com/file/d/1zoLL...
Download Powerpoint referenced in video here: drive.google.com/file/d/1WKTe...
This is extremely helpful! Thanks a lot for doing what you are doing!
Interesting i have learnt plenty of things... keep in touch Miko...!!!
That's really helpful. Thank you for sharing
Thank you!
Hi, thank you for the video. Do we perform collinearity diagnostics in the same way if my predictor variables are categorical?
Dear Mike,
Thanks for this nice explanation.
I would like to repeat the stepps but within my indepedent variables there are 4 dichotomous variables and 1 continuus variable.
So my question is now: do i have to take this into account before i can run my binary regression analyses for multicollinearity?
Kind regards.
Hi Mike, Excellent info here. Pls would you be able to a video on how to split data in SPSS in training and validation sets while performing regression(binary or MLR). Thanks.
Do you have a video on using SPSS to examine a linear relationship between any continuous
independent variables and the logit transformation of the dependent variable or how to perform and use the results of a Box-Tidwell Test? (Thank you so much for your amazingly helpful videos by the way! :-) )
Hi miko i have looked this video , i would like to say thank u. Now i have question that can this method of collinearity diagnostic used for categorical variables in cox regression?
how would I reproduce this test for a logistic regression with both nominal and continuous scale predictors?
Thank you so much. This was really needed. Could you make a video on testing the assumptions of multilevel models (Linear mixed effects and generalized logistic mixed models?
Hi there. Thank you for your comment! I'll see what I can do on these topics in the future. It is certainly on my 'to-do' list. Best wishes!
@@mikecrowson2462 I want to thank you from the bottom of my heart for all the time and effort you put into creating these videos for novices like us! Stay safe and healthy!
Great video. Thank you! I note that your dependent variable of pass/fail is binary. I would therefore have thought that binary logistic regression is best suited to testing this. The assumptions of linear regression, in my (humble!) understanding, are not met when the dependent variable is binary prohibiting its use. Are you able to comment please?
Hi Kadhim, the issue of collinearity pertains to the relations among the predictors only. Unfortunately, SPSS does not have the option of generating the collinearity diagnostics via the logistic regression route. However, you can 'trick' the program into giving them to you by way of the regression option in SPSS. The collinearity diagnostics would be the same if the option was offered via the logistic regression menu.
Hope that helps.
Cheers!
what if variables are categorical. Then should we check collinearity assumption?
I hope u will come a gain with solution for multicollinearity
SPSS does not allow string variables in the dependent section in linear regression. How do you do it?
Hi jack, try just converting the dependent variable to Scale.
He is just talking too much about unnecessary stuff..bad explanation..he got me more confused