I am conducting an analysis of customer satisfaction survey data. Based upon my knowledge of this field, I know that some of the variables in my data are not simply statistically associated with each other, but some influence others. My question: is there a way to express the strength of this influence in numerical terms? I know that the width of an arrow indicates strength of influence, but my clients are accustomed to strength of influence being expressed numerically.
All of my variables are categorical, so one way that occurs to me is to compare the observed distribution of the dependent variable to its distribution as predicted by Genie. How do I obtain the distribution as predicted by Genie, using all of the cases in the data to perform the prediction?
The front end.
1 post • Page 1 of 1