Determining the importance of an independent variable: A path analytic solution

Various methods have been proposed for evaluating the importance of an independent variable in determining a given dependent variable. This paper suggests that the analysis of direct and indirect effects in a recursive causal model can provide a more satisfactory evaluation than the ordinary correlation and regression techniques commonly used. The interpretation of an independent variable's impact derived from a path model is generally more adequate because it is able to take into account the operation of indirect effects and spuriousness. To demonstrate this contention, data gathered by the author to explain organizational innovation are presented.Peer Reviewedhttp://deepblue.lib.umich.edu/bitstream/2027.42/22338/1/0000783.pd