I am re-posting because my first mail of a week ago has not yet attracted any replies.
Dear List =
I shall be obliged for answers to a few queries arising from the need to combine some 10 pairs of independent variables before running Multiple Regression (in SPSS15). All of these IVs are binary predictor variables. My questions are these:
1. To be suitable for combining do the variables in each pair need to have a similar impact on the dependent variable ? Some years ago I read that this was the case, but I cannot now find the reference.
2. Can dichotomous IVs be combined in SPSS15 (using the `Compute' command or otherwise)? Do such binary variables qualify as the numeric arguments which the process appears to require?
With advance thanks =
Owen
|