I have been thrown out into the wilderness of data analysis and statistics while working on a paper. Based on a survey of employees in a hospital I have built a multiple regression model using the forward selection method. The model uses a binary dependent variable of submittedIdea (0 if no, 1 if yes). The IVs include 3 measured on a likert scale from 1-4. In addition there are 3 categorical control variables (position, sex and research y/n).
The model produces the attached residuals plot. The plot looks nothing like the residual plots I have seen anywhere in class or in explanation videos online. What does this plot tell me? Heteroskedasticity, yes, but why do I get these funky straight downward sloping lines?
