Manipulating salience: Causal assessment in natural settings

Suggests a procedure for identifying causal effects when it is impossible to manipulate the independent variable of interest or to control confounded variables statistically. If a variable has a causal influence, this influence should be more pronounced when the variable is salient than when it is not. The applicability of this salience approach is illustrated in a discussion of a study by the authors and their associates (1979) on the effects of the German job-ban policy on the political behaviors of 32 university students. (5 ref) (PsycINFO Database Record (c) 2007 APA, all rights reserved)