Affiliations: [a] Schneider Institute for Health Policy, Brandeis
University, Waltham, MA, USA | [b] Substance Abuse and Mental Health Services
Administration, Rockville, MD, USA | [c] Statistical Group, WESTAT, Rockville, MD, USA
Correspondence:
[*]
Corresponding author: Grant Ritter, Schneider
Institute for Health Policy, MS035, Brandeis University, 415 South Street,
Waltham, MA 02454, USA. Tel.: +1 781 736 3872; Fax: +1 781 736 3905; E-mail:
[email protected]
Abstract: In the past, incentive payments have proven successful in increasing survey
response rates. In particular, incentives have been shown to increase the yield
rate for harder-to-interview respondents. Despite this success, incentives are
not without controversy. Opponents note that incentives may make the survey
appear less important, destroy civic responsibility, and create an unnecessary
expectation. Over time researchers may have to pay sizable incentives to achieve
the response rates obtained 20 years ago before the use of incentives was
common. Incentives could also have a detrimental impact on sample composition.
The Alcohol and Drug Services Study (ADSS) included an incentive payment
sub-study as part of its client follow-up phase. The sample for this sub-study
randomized clients into four groups. One group was paid $25, the amount received
by other responding clients in the ADSS main study. Respondents in the other
three groups were paid an alternative amount (i.e., $0, $10, or $35). In
examining the impact of varying incentives to sampling benchmarks, such as
response rate and sample composition, the ADSS incentive sub-study found that
the fears of opponents concerning the use of incentives were substantially
unfounded, at least among the substance abuse treatment clients in the follow-up
phase of ADSS.
