Abstract
This study sought to examine problems in assessing field experiments with conventional survey-based data gathering methods. Two phenomena are examined: (a) increases in reactiveness, and (b) the recalibration of survey scales. The theoretical dimensions of such changes are discussed, including such processes as consistency, saliency, priming, and raised expectations. Methods of identifying these sources of contamination within two-group experimental designs are presented and illustrated with data from a recent field experiment involving 78 female sewing machine operators. In this instance, survey-data gathering processes increased "reactiveness," suggesting a general increase in the salience of experimental variables to subjects; and, scale recalibration appears to have occurred with respect to a primanV intervention variable. Methods of reducing or accommodating these possible sources of contamination are discussed, together with their implications for the protection of human subjects.
Get full access to this article
View all access options for this article.
