I am doing multiple regression with PET data with 7 different conditions. I have modeled a linear and an exponential function separately and plotted the functions. When they are the sole covariates the plots are linear and exponential functions. However when I include both types of functions are included as covariates the plots are no longer linear or exponential. This happens when I examine the fitted responses for the corresponding contrast's adjusted response plotted against scan or time. They look very different from when I included modeled them as the sole covariates. When I examine the design -> explore -> covariates the graphs look like the linear and exponential functions, not the graphs I get when plotting. I think this may have something to do with orthogonalization, but am not entirely certain. I have looked in the list archive, but I have not found anything that helped me track down exactly how to work around this problem. I would appreciate any suggestions. Thank you.
Paul Corbitt