REU-Questions from John Kraemer

August 20, 2007

SRI’s response to questions from John Kraemer regarding the OMB clearance request for

RESEARCH EXPERIENCES FOR UNDERGRADUATES (REU) IN THE DIRECTORATE OF ENGINEERING (ENG).

1. Why not send the incentive up front to all participants?

The primary reason not to send incentives up front to all participants is the cost. NSF has indicated that there are no additional funds available. Also, as OMB has noted, we have previously obtained very satisfactory response rates (76% and 80%) using incentives provided after completion of the questionnaire. As discussed in our response to question 2, below, it appears that our student population size is smaller than anticipated in our proposal, thereby lowering the total estimated cost of incentives. Briefly, our proposal estimated a 4, 589 student participants; we now estimate approximately 3,700 students. The funds budgeted may, therefore, allow us to provide incentives up front to the reduced number of student participants. However, the majority of research regarding incentive use has been done on mail and telephone surveys rather than on internet surveys. We suggest using this study to test the relative effectiveness of providing the incentive up front vs offering the incentive after completion of the online questionnaire, with random assignment of half of the participants to each group.

2) Why not do a sample instead of a census?

Participant information we have obtained so far indicates that the actual numbers of PIs, faculty mentors, and students in our population are considerably fewer than those estimated for purposes of our proposal (Table 1). In addition, because of the likely transience of our student sample, we suspect that, even with considerable effort, we will be unable to locate many of the earlier year participants. Given these two factors, we are concerned that selecting a sample of participants will seriously limit our ability to conduct analyses requested by NSF to identify differences by year and type of award and by NSF division. We therefore prefer to survey the census, as proposed.

Table 1 Estimated Number of ENG REU Participants for 2003-2006 and Actual Numbers Received to Date
Participant Type	Estimated	Actual	Difference
PIs	682	413	61%
Faculty Mentors	4,039	1,955	48%
Students	4,590	3,375	74%

2. Add “to the extent permitted by law” to the confidentiality sentence.

We will be happy to add this clause to the confidentiality statement.