I was wondering if there are discussions, examples of studies, or book chapters which model randomly varying slopes while disregarding intercepts. I am facing a situation where the dv (overall job performance) needs to be standardized within organization due to variations in scaling (e.g,. 3-, 5-, and 7-point rating scales). As a result, all lines pass through the same point on the y-axis thus eliminating between-unit variance in the outcome (i.e., ICC of zero).
I have variables originating at three levels (individual, task, organization) and theoretical predictions on when the relationship between
particular individual differences and overall job performance should vary in strength. In other words, I am only interested in cross-level
interactions. At this point, I am wondering if it would make more sense to use ordinary least squares regression or, perhaps, sift through the date for comparable scaling to model randomly varying intercepts. Thanks!