Dear List,
In CAR models, I have obtained statistically significant
interaction/modification effects between the continuous dependent and
continuous independent variables for levels of a dichotomized factor variable.
The factor is defined on the basis of the spatial units (high or low social
disadvantge). These spatial units are also used to define the spatial
connectivity matrix by adjacency of boundaries.
Is it reasonable to use standard procedures to assess the statistical
significance of each coefficient's difference from zero when separately
considering the two levels of the dichotomized factor?
Thank you in anticipation,
Dora
Dora Pearce
PhD student
School of Science & Engineering
University of Ballarat
PO Box 663, Ballarat, Victoria, 3353
Email: dpearce@students.ballarat.edu.au
Ph: (03) 5327 9264
|