Memorandum United States Department of Education
Institute of Education Sciences
National Center for Education Statistics
DATE: June 28, 2017
TO: Robert Sivinski, OMB
THROUGH: Kashka Kubzdela, OMB Liaison, NCES
FROM: Carolyn Fidelman, MGLS:2017 Project Officer, NCES
This memorandum provides NCES responses to OMB passback received on June 23, 2017.
How many of the 1,236 initial schools sampled would be eligible for the additional $200 incentive? While in Table 1 of Part B (page 14) the goal seems to be to get 128 schools with a high prevalence of students with disabilities, the incentive boost is offered to all schools in the district where at least one school has a high prevalence of students with disabilities (Part B, page 28). So how many additional schools (with low prevalence) would you expect to have to pay the additional $200 to?
NCES Response: Of the 1,236 initially sampled schools, 1,017 are public schools and public schools identify IEP students (the remaining 219 sampled schools are private schools and counts of students with the focal disabilities are unavailable for private schools). The 1,017 public schools come from 839 public school districts. In the initial school sample, we sampled 176 high prevalence schools and they reside in 144 districts. There are an additional 56 low prevalence schools in those districts for a total of 232 sampled schools in 144 districts with at least one high prevalence school (out of the total of 1,017 sampled public schools in 839 districts). However, of the 144 districts with at least one sampled high prevalence school, 30 have given approval for us to contact the schools in their districts as of June 23, 2017, leaving 114 districts with 194 sampled schools (142 high prevalence schools and 52 low prevalence schools).
Given that we planned for a ~72.8% school participation rate (we started with this rate for our initial release to achieve 900 participating schools from the 1,236 initial sample of schools), we could extrapolate for the applicable schools to be ~ 194*0.728=141 participating schools (103 high prevalence and 38 low prevalence) to receive the incentive.
Of those 114 districts, 84 have a single school (a single high prevalence school) and 30 districts have more than one school. In fact: 84 districts have 1 school; 11 have 2 schools; 7 have 3 schools; 5 have 4 schools; 4 have 5 schools; 1 has 7 schools; 1 has 8 schools; and 1 has 12 schools.
The above information is based on the original sample and does not include counts of schools for the reserve sample. It is possible that we offer the additional incentive for a district (which agrees to participate) and later need to release schools in the same district – which could result in additional schools being offered the higher ($600) incentive.
To provide more information in the Supporting Statement of this request, we have added a portion of the first paragraph of this response to page 28 of the Supporting Statement Part B.
How did NCES decide on a boost of $200 for these targeted schools (Part B, page 28)? Why not $100?
NCES Response: A limited number of comparisons is available for determining the optimal incentive amount for schools in this case. One of the factors we considered is our experience with the school-incentive experiment in the MGLS:2017 OFT, in which $400 was more effective than $200. In addition, the NCES international assessment studies (e.g., PISA, PIRLS, TIMSS) begin recruitment with an offer of $200 to schools, and then an offer of an additional $600 (for a total of $800) to refusing schools. Most recently, this strategy was met with success on PIRLS 2016 when 11 of the 19 initially refusing schools agreed to participate once offered the $800 incentive. That said, we recognize that the circumstances for the international studies are different and should not be directly applied to MGLS:2017.
Refusals at the district level prohibit the study from contacting the school or schools in their jurisdictions to discuss participation in the study; thus a single district refusal can affect participation of multiple schools. This is even more damaging to the study when the refusing district contains high prevalence schools, meaning schools where the total number of students with primary IEP designation of AUT or EMN exceeds 17. Although the $200 boost will be provided to schools, because the district is a gatekeeper and was a significant factor in nonresponse during the OFT, we propose to offer the additional $200 in districts that contain high prevalence schools to maximize their approval of research and improve school participation rates. We will let each refusing districts know about the higher incentive offer as part of our district refusal conversion efforts. This is intended to encourage the district to open the door for us to contact schools directly, where we have greater likelihood of gaining cooperation. Upon OMB approval of this plan, we will identify all pending refusal districts containing high prevalence schools on an ongoing basis. We will offer all schools in those districts, including low prevalence schools, an additional $200 monetary incentive or $200 monetary equivalent in goods or services – for a total of $600 for the targeted schools – to achieve district participation. The purpose of offering consistent levels of incentives across schools in a district is to avoid treating schools in the same district differentially.
A financial incentive does not secure the assurance of participation, as districts and schools are increasingly protective of their instructional time and are reluctant to participate in voluntary studies. However, our recent experience demonstrates that many schools facing budget reductions find that an increase in the incentive is attractive and that it encourages their participation in the study. The proposed increase of $200 is a moderate increase compared with other studies, and we are concerned that a lesser offer may not be sufficient to offset the districts’ reasons for declining to participate in the study.
To further explain the incentive boost, we added the 2nd and 3rd paragraph above to page 28 of the Supporting Statement Part B.
| File Type | application/vnd.openxmlformats-officedocument.wordprocessingml.document |
| File Modified | 0000-00-00 |
| File Created | 0000-00-00 |