Hello! I could use a bit of everyone's expertise!
I have a non-normal DV (heavy positive skew due to the type of scale used: starting from zero and moving upward; it's interval data), and three continuous predictors mean-centered.
I'm thinking the appropriate analyses would be to use a Poisson link function, but using the 'Generalized Linear Models' option gives me a 'numerical problems' error due to negative values (of which I'm assuming are coming from the centered predictors).
How do I get around this?
Thank you! I expect clarifying questions, and please let me know what, if any, images would help.