Abstract
The U.S. Census Bureau conducted a series of experiments to evaluate alternative contact strategies. We hoped to identify effective mail strategies that increase timeliness of response and reduce the number of cases receiving more-costly follow-ups. These experiments were incorporated into the collection of several annual business surveys and one quarterly survey. This article summarizes results from five experiments whose designs are based on business survey decision-making and response processes. We obtained clear, positive results from tests involving manipulation of mail sequences, but only modest results from tests in which we varied messages and envelope appearance, whose effects were seen only in specific industry subgroups.