Broom and Dozier (1990) compare evaluation research to a field experiment (discussed in Chapter 5). The public relations campaign is similar to an experimental treatment, and the target public is similar to the subjects in the experiment. If possible, public relations researchers should try to construct control groups to isolate campaign effects from other spurious factors. The public relations researcher takes before-and-after measures and determines if any significant differences exist that can be attributed to the campaign. However, Broom and Dozier call attention to the fact that public relations campaigns occur in dynamic settings, and as with most field experiments, it is difficult to control extraneous variables. As a result, it may not be scientifically possible to prove the program caused the results. However, from a management standpoint, systematic evaluation research may still represent the best available evidence of program effectiveness.