I am running Generalized Linear Model and I have one continuous dependent variable, two categorical fixed factors and 22 continuous independent variables as covariates. When I run the model, I get quite a few significant factors, but I am unable to form the equation for categorical variables.

I have re-coded the project size into 6 categories (>3 million = 1; 3-10 million =2 and so on) same with the project complexity (Most complex=1, moderate =2 and least complex =3) The covariates are taken from the raw data without any re-coding. For eg: the parameter estimate output is as follows: My question is how can I frame the equation if Project size, project complexity and covariates C,D were the significant factors? Linear Regression analysis gives us one intercept for the categorical variables whereas GLM gives it for each category in the categorical variable. Kindly help me with the same. Parameter EstimatesModel Effects

  • $\begingroup$ what is A B C D and are you looking for results on the basis of complexity as a variable ? $\endgroup$ – Subhash C. Davar May 16 '16 at 9:43
  • $\begingroup$ A B C D are just the names of the continuous variables. I am looking forward to frame an equation such as Y = a +b1x1 +b2x2 +.... Not sure how to add the intercepts for categorical variables. The equation I am trying to frame is based on significant factors (90% confidence interval) so in this case complexity and C variable $\endgroup$ – Vijay May 17 '16 at 14:54
  • $\begingroup$ I think you need re-scoring for project complexity - 3, 2, 1 . what is your dependent variable. To proceed with model building, you need to identify key exlanatory variables and then implement regression model. May be you have not entered values for categorical variables properly. After scoring these variables become continuous variables.pl.check. $\endgroup$ – Subhash C. Davar May 17 '16 at 15:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.