A Convenient Truth: University Employees as Heterogeneous and Inexpensive Experimental Samples*
Published online on October 26, 2016
Abstract
Objectives
Building on work by Kam, Wilking, and Zechmeister, we argue that academic researchers ought to make greater use of campus employees for survey experiments. For applications where a representative sample is unattainable and student samples provide too little variation on key characteristics, staff members provide an inexpensive alternative that offers greater statistical leverage and the ability to detect conditional treatment effects
Methods
Using an e‐mail survey experiment conducted on employees at a large public university, we explore how two modest incentives for participation affect the quantity and quality of responses.
Results
While overall differences among conditions are modest, a cash lottery generates somewhat higher response rates with little effect on sociodemographic skew or level of satisficing. The condition offering a charitable contribution fared worse on both counts.
Conclusion
University employees provide a convenient, heterogeneous, and inexpensive population for experimental studies.