Incentive Memo

DATE:	July 17, 2008
TO:	Shelly Martinez, OMB
THROUGH:	James Griffith
FROM:	Linda Zimbler; Tracy Hunt-White, BPS:04/09 Project Officer
	Jennifer Wine, BPS:04/09 Project Director
SUBJECT:	Additional Information for Request for Approval to Increase BPS:04/09 Field Test Respondent Incentive

This memorandum provides additional information to OMB on the request to increase the incentive amount being offered current nonrespondents to the field test implementation of the 2004/09 Beginning Postsecondary Students Longitudinal Study (BPS:04/09). BPS:04/09 is being conducted for the National Center for Education Statistics (NCES) by RTI International (RTI) [OMB No.: 1850-0631]. The original memorandum, dated July 10, 2008, requested a $20 increase in the level of incentives offered for completed interviews for the remainder of the BPS:04/09 field test.

As discussed during the conference call with OMB, the BPS:04/09 field test sample is an atypical BPS longitudinal field test sample. It was first identified during the NPSAS:04 field test, the base year data collection for the BPS longitudinal study series. An insufficient number of students were determined to be eligible for the BPS cohort, defined as first time beginning postsecondary students. As a result, prior to the first follow-up data collection, a supplemental sample of potential first time beginners was selected from enrollment lists that had been collected for the NPSAS field test but not used.

A total of 2,121 additional sample members were added to the BPS first follow-up field test. Because the supplemental sample members were not part of the NPSAS field test sample, they were not interviewed during the base year nor was there any locating of the sample conducted between the NPSAS and the BPS first follow-up interviews. Only 784 (37 percent) of the supplemental sample was interviewed during the BPS:04/06 field test. All of the responding members of the supplemental sample were retained for the second follow-up interview, BPS:04/09.

The remaining supplemental sample continues to be a difficult group to interview despite their participation in the BPS:04/06 interview. To date, 36.2 percent of the supplemental sample has yet to participate in BPS:04/09 compared to 24.5 percent of the original BPS sample who participated in BPS:04/06 (Z=-3.54; p<.01). Even if all of the original BPS sample members remaining as nonrespondents were interviewed, we would still be unable to reach our goal of 865 interviews. Interviewing the most difficult cases – members of the supplemental sample and BPS:04/06 nonrespondents (of the 80 retained, 57.5% have not been interviewed) – is critical to our success.

Following OMB’s suggestion, we RTI investigated the feasibility of using certified mail to reach the remaining BPS nonrespondents, rather than Federal Express. To send a package certified mail, the cost is $4.22 per piece and all information on the return card would need to be written by hand. Additionally, since the Postal Service will require a signature, sample members would have to make a trip to their local post office if not available at the time of delivery. In contrast, Federal Express labels could be prepared electronically via computer and packages can be left at the sample member’s residence. Although somewhat more expensive ($5.35 per piece), Federal Express appears to be the more economical and feasible alternative. If the additional incentive request is approved by OMB, therefore, we propose to send notification of the higher incentive amount to nonrespondents via Federal Express.

The BPS:04 full-scale sample does not contain a supplemental sample as the field test did and, therefore, we do not expect a similar problem with our interview response rates for the second follow-up in 2009. For the first follow-up in 2006, 81 percent of eligible base-year respondents and 39 percent of eligible base-year nonrespondents completed the interview. Of course, we cannot be certain that our normal data collection efforts will be able to obtain the desired response rates in the full-scale study. By increasing the incentive for a small group of non-respondents during the field test, we can empirically see if offering a higher incentive will make the difference. Hopefully, a higher incentive amount will give us the yield that we are looking for in the field test and will be a back up plan for the full-scale study, if the need should arise. We would not implement the higher incentive amount in the full-scale study unless we were having difficulty meeting our response rate goals and unless we requested and obtained OMB approval to do so during the nonresponse phase of the full-scale study.

If we or RTI can provide any additional information, please let us know. We appreciate your consideration of our request.