Abstract

Indisputably, sample surveys hold a legitimate, perhaps primary, role in the reduction of uncertainty and support of marketing decision making. Technological advances and competent research administration have enhanced the efficiency and professionalism of survey research. Recent methodologies for improved response quality have advanced self-report accuracy. To help marketers avoid similar pitfalls, the article has reviewed the scope of response errors. Moreover, it has outlined several approaches for coping with self-report bias. Successful management of response error, in fact, requires a systems view of total design error-sampling and all types of nonsampling error. Chief concepts of the systems view are cost and error trade-offs in survey and observational designs. Since the calculation of error terms from different designs is complex, a small computer is extremely valuable. But the computer model is only as good as its input. Creative thinking is required to escape the survey tradition and identify a set of potentially feasible designs.