Historically, the National Survey of Recent College Graduates (NSRCG) has noted differences in survey response rate by sample members’ race/ethnicity, U.S. residency status and academic major. In 2006, for example, the unweighted response rates for minority sample members averaged about 62 percent, compared to 75 percent for non-Hispanic white sample members. Also in that same year, only 51 percent

of sample members with foreign addresses completed the survey, compared to 70 percent of those with U.S. addresses. Differences of five or more percentage points were also found across academic majors.

The NSRCG is a multimode survey of approximately 18,000 randomly selected recent college graduates. In 2008 NSRCG, a web survey option was made available for the entire duration of the survey for the first time. Also in 2008, NSRCG embedded an experiment that offered variations in timing and amount of incentives and a web response option. Sample members were randomly assigned to various incentive and web option groups. This paper examines impact of incentives and web response option on increasing

the survey completion rates among the sample groups with historically low survey responses. It takes advantage of an incentive experiment built into the study and compares rate of response among these subgroups by incentives offered and use of the web option.