ED Responses to Passback

Memorandum United States Department of Education

Institute of Education Sciences

National Center for Education Statistics

DATE: September 28, 2015

TO: Shelly Martinez, OMB

THROUGH: Kashka Kubzdela, NCES

FROM: Isaiah O’Rear, NCES

SUBJECT: ED School Climate Surveys National Benchmarking Collection 2016 Responses to OMB Passback (OMB # 1850-XXXX v.XXX)

Provided below are responses to ED School Climate Surveys National Benchmarking Collection 2016 Responses to OMB Passback (OMB # 1850-XXXX v.XXX)

1. Why are the schools administered by the Bureau of Indian Education excluded from the sample frame?

NCES has revised the sampling plan to include a stratum for BIE schools, and text stating that BIE school will be excluded has been removed from the Supporting Statement Part B. To ensure BIE representation, NCES plans to include four BIE schools in BIE strata. The Bureau of Indian Education has not provided data to the Common Core of Data since 2010-11. Therefore we will have to use the 2010-11 enrollment data to inform the sampling procedures.

2. OMB recommends cognitive testing for the sexual assault items which were not part of the initial cognitive testing.

NCES plans to administer small scale cognitive testing of the sexual assault items, and a statement to that effect was added to Supporting Statement Part A. For the testing NCES plans to recruit 10 middle grade students, 10 high school students, 10 teachers, and 10 instructional staff.

3. OMB has asked for justification for the exclusion of sexual assault items from middle grade student survey.

NCES plans to include middle grade students in the cognitive testing of the items to determine the appropriateness of the questions for middle grade students.

4. Part A, section 16 needs to provide more information about how the results of the benchmarking study will be published.

A description of how benchmark results will be integrated in the 2016 public release of the EDSCLS platform has been added to A.16:

“The EDSCLS national benchmark data will be made publically available through the second release of the EDSCLS platform in fall 2016. These national data will be included as the context for school scores. For example, figure 2 shows one possible way of showing tertiles based on the national data along with a school’s score. For more information, please see appendix E. We also plan to provide respondent-level subgroup reporting by grade (student only), sex, and race/ethnicity. If the number of respondents in some racial groups do not meet reporting standards, they will be collapsed into an “other” category.”

A draft of a Reporting Position Paper has been provided as Appendix E, showing how the benchmark results will be displayed.

5. Take a look at the school climate survey being proposed by REL in New Mexico. Decide whether it has anything that should be used in EDSCLS.

The items from the New Mexico REL survey can be divided into three groups. The first group of items is about academic motivation, a topic which is outside the scope of the EDSCLS. The second group is about individual behaviors and experiences. To preserve student privacy, the EDSCLS does not ask about individual behaviors and experiences. The third group of questions is very similar to questions currently on EDSCLS. There are no items which would be appropriate to add to EDSCLS.

6. Why are Department of Defense schools excluded from the sample?

DOD has not provided data to the Common Core of Data (CCD) since 2008-2009, making it difficult to integrate DOD schools into the sampling frame. Resource constraints also prevent NCES from administering the survey to overseas DOD schools.

7. Did NCES consider asking about “drug use” as a two-part question (marijuana and other drugs) given the differences in legal status and public opinion on the first?

In EDSCLS, adult respondents are asked, “At this school, how much of a problem is student drug use?” This item is taken from the Maryland Safe and Supportive Schools survey, in which the item has performed well. There were not existing marijuana-specific items.

For student respondents, we looked at candidate items from Louisiana Safe and Supportive Schools, which focused on marijuana use, "How do you feel about someone your age doing the following? Trying marijuana or hashish once or twice."

NCES discussed using these question, but ultimately decided that the changing legal status of marijuana wasn’t relevant to understanding school climate related to drug use. NCES thought the drug-specific questions weren’t well-suited for school climate measurement and opted instead to use the following item: “Students at this school think it is okay to try drugs.”

8. Please fix the race question instructions on all of the questionnaires to indicate “check all.”

Appendix B has been revised for all questionnaires to read: “What is your race? You may mark one or more races.”