Request for change

Change Request

Proposed Changes to the National Intimate Partner and Sexual Violence Survey (NISVS)

(OMB No. 0920-0822 Exp. Date 3/31/2023)

May 4, 2021

Background and Justification

The National Intimate Partner and Sexual Violence Survey (NISVS) is an ongoing, nationally representative survey of U.S. adults and their experiences of sexual violence, intimate partner violence, and stalking. The NISVS program underwent experimentation and feasibility testing in 2020 for the purpose of redesigning the methodology for NISVS, which has traditionally used random-digit-dial (RDD) telephone survey methodology.

This document serves as a change request for the currently approved NISVS (OMB# 0920-0822, expiration 3/31/2023) for the pilot testing phase of the NISVS Redesign Study. This change request is the result of the previously approved feasibility study (approved 3/20/2020) and is needed to conduct the pilot testing that was also approved in the previous ICR revision 2019. As a reminder, the change request was discussed in a previous call with OMB (11/15/2019); a plan to submit a non-substantive change request for pilot testing was described in the previously approved OMB package. The goal of the pilot study is to field test the survey using the features anticipated for the full-scale NISVS collection (to be completed at a future time under a future OMB submission). The plan is to complete 200 surveys using the recommended design. The pilot study is not intended to generate prevalence estimates. Rather, this study will primarily be an operational test to assess whether the recommended design works as anticipated from the feasibility study. These include the impact of the additional screener and extended survey mailing, the call-in CATI process using the web-survey instrument rather than the original RDD CATI instrument, and how receptive respondents are to the alternative mode (push to call-in CATI) in the absence of the paper survey choice.

The feasibility study tested alternative designs (RDD and ABS) and modes (web, CATI, paper) and considered the implications for data quality, representativeness, non-response bias, effects on prevalence, and concerns about privacy, confidentiality and minimizing harm. The following recommendations from the feasibility study are guiding the implementation of the pilot study and this change request (see Attachment N):

1. Address-based sampling (ABS) frame with push-to-web methodology

Rationale: Feasibility testing showed that the ABS had a higher response rate (33.1%) and cooperation rate (94.1%) compared to RDD (response rate = 10.8%; cooperation rate = 25.9%) and lower cost. The ABS approach is also less burdensome in terms of survey completion time. For respondents with no victimization, the median time to complete the survey was 12 minutes for ABS/web vs. 30 minutes for RDD. For those with 3 or more different types of victimizations, the completion time was 25 minutes for ABS/web vs. 50 minutes for RDD.

2. Additional mailings (add 1 to the screener phase and 1 to the extended/full survey phase)

Rationale: Because the feasibility study data collection time period was limited, the standard number (four) of attempted contacts was not implemented. Additional contact attempts will bring the number of screener contact attempts to four and the extended/full survey contact attempts to three. Note that the extra mailing will be not be implemented in the pilot study given the compressed timeline.

3. Web/Computer Assisted Telephone Interview (CATI) optional group instead of web/paper

Rationale: Using call-in CATI as an option for those who cannot (or prefer not to) complete the survey by web allows for: a complete dataset (the paper version was a shorter, less detailed version of the NISVS survey); less item-missing data compared to the paper version; more guidance to respondents who use the web or call-in CATI vs. the paper version; inclusion of those without internet access (which paper would also do but at the cost of less data). Additionally, despite the paper survey being shorter than the web and CATI versions, more paper respondents stated that the survey was burdensome and too long compared to those who used the web or call-in CATI.

4. Probability method of respondent selection

Rationale: Feasibility testing results showed negligible differences between probability and non-probability selection modes.

5. Items on the NISVS questionnaire that allow for assessment of representation and bias

Rationale: Inclusion would be for benchmarking purposes (e.g., American Community Survey (ACS), National Health Interview Study (NHIS)). We would include questions from another national survey (e.g., ACS or NHIS) in the NISVS survey to compare for consistency. An example is to include the ACS question in the screening instrument about whether the household has internet access.

6. One additional item to measure the attention of respondents on the web survey

Rationale: An additional item will help assess whether the respondent is carefully reading the questions. During feasibility testing we included one item at the beginning and another at the end of the survey. It was recommended to add an attention-related item to the middle of the survey. For example, a sample item instructs the respondent to select a specific answer choice among the response options available (e.g., “If you are paying attention, please choose Silver below.”).

We are requesting approval to incorporate, into the pilot design, the procedures recommended at the conclusion of the feasibility study (see the six recommendations above). See Attachment O for the detailed methods of the pilot study. Specifically, we request approval to do the following:

1. Remove RDD sampling and use only address-based sampling (ABS).

2. Remove non-probability respondent selection and use only probability-based selection.

3. Use a web instrument as the primary data collection instrument and remove the CATI instrument.

4. Remove the optional full paper survey instrument. Use call-in CATI as an option for those who cannot (or prefer not to) complete the survey by web. Note that the web instrument will be used by the interviewer.

5. Remove the non-response follow-up (NRFU) phase. There is not a need to pilot test the NRFU phase. There is no reason to expect that the NRFU would perform very differently in the pilot as it did in the feasibility test. Also, the small sample size and short data collection period would not allow for many additional gains in completed interviews.

Add 1 survey question designed to assess how carefully the respondent is reading the survey questions along with minor survey revisions to update the instrument for the pilot study (see Attachment K). For example, we have updated the programming instructions in the screening tool, made revisions to aid in respondent selection, corrected typos discovered during feasibility testing, replaced benchmarking questions with newer versions, dropped survey questions that did not perform well in feasibility testing, and dropped some debriefing questions that did not require further testing.
Effect of Proposed Changes on Currently Approved Instruments

Revisions were made to the survey programming instructions to reflect the change in sampling procedures and respondent selection. Other screener and survey changes included: corrections of minor typos (e.g., inconsistencies between survey versions), updated incentive amounts, updated benchmarking questions, removed questions that do not require pilot testing, and replaced attentiveness questions with improved ones. The revised web survey is presented in Attachment F.1, and survey question revisions are shown in track changes. A survey crosswalk is presented in Attachment K. Other materials were updated to reflect the pilot study status. See Table 1 for a listing of the attachments that underwent revisions as a result of this change request.

IRB Approval

CDC’s IRB has deferred to the contractor’s IRB. The IRB amendment obtained through the study contractor is presented in Attachment E. As approved in the study protocol, CDC will not have contact with study participants, nor will CDC have access to PII.

Table 1 below shows the attachments that were revised as a result of this change request. Previously approved attachments that are not shown in this table did not change (Attachments A, B, C.1, C.2, D, G, H, L.1, L.2, M).

Table 1. Change to Previously Approved Attachment Listing

Previous Attachment Number	Previous Title	Change Request Attachment Number	Change Request Title	Type of Change
E	Institutional Review Board (IRB) Approval	E	Institutional Review Board (IRB) Approval	Revision approval
F.1	Survey - National Intimate Partner and Sexual Violence Survey (NISVS), CATI	--	--	Removed
F.2	Survey - National Intimate Partner and Sexual Violence Survey (NISVS), web	F.1	NISVS Web Screener and Survey (both in English and Spanish)	Updated for pilot procedures and question revisions
F.3	Survey - National Intimate Partner and Sexual Violence Survey (NISVS), paper	--	--	Removed
F.4a-F.4b	Screeners for paper survey	F.2	NISVS Paper Screener	Updated for pilot procedures
F.5	FAQs for ABS	--	--	Combined with advanced letter
F.6	FAQs for Extended Paper Survey	--	--	Removed
I.1	ABS Advance Letter for Screener	I.1	Advance Letter for Screener	Updated and combined with FAQs
I.2	ABS Advance Letter for Survey	I.2	Advance Letter for Survey	Updated and combined with FAQs
I.3	ABS NRFU Letter for Screener	I.3	Screener Follow-up Letter	Updated for pilot procedures
I.4	ABS NRFU Letter for Survey	I.4	Survey Follow-up Letter	Updated for pilot procedures
I.5	ABS Reminder Postcard	I.5	Reminder Postcard	Updated for pilot procedures
I.6	RDD Advance Letter	--	--	Removed
J	Thank You Incentive Letter	J	Screener Thank You Letter	Updated for pilot procedures
K	Crosswalk of Survey Revisions	K	Crosswalk of Survey Revisions	Replaced with new revisions
--	--	N	Westat Recommendations for NISVS	New document
--	--	O	Pilot Sample Design Methodology and Data Collection Procedure	New document

Previously Approved Burden and Costs

In March 2020, OMB approved the NISVS data collection plans for the NISVS Redesign Study. At that time, 113burden hours and a cost of $2,988for pilot testing was approved.

Current Request

The contractor will collect complete pilot survey data from a total of 200 respondents in June-August 2021. The survey completion time is estimated to be 21-49 minutes, on average, including screening and verbal informed consent. Consistent with the feasibility study methodology, screening and full survey participants will receive a monetary incentive at the same levels as used in the feasibility study (see Table 2). Given the reduced sample size, the total respondent burden will be lower than previously approved estimates. Tables 3-4 describe the respondent burden for the current data collection for a one-year period. For the current request, 334 respondents will complete the 3-minute screener and 200 respondents will complete a 25 to 40-minute survey (depending on the mode), resulting in 17 burden hours for screening and 84 burden hours for survey completion (101 total burden hours); see Table 3.

For pilot testing, the annualized cost was derived by using 334 as the expected number of households screened and 200 completed interviews. This results in costs of $441 for screening and $2,235 for survey completion, for a total cost of $2,676 for pilot testing; see Table 4.

Table 2. Use of Incentives by Sample Size

Incentive groups within ABS sample frame	Sample Size Offered Incentive
Pre-paid incentive
$5 for initial request	815
Promised incentive
To complete household screener
$10 to complete by web	278
$5 to complete by paper	56
To complete NISVS
$15 to complete by web	195
$15 to complete by call-in	5
*Number of eligible sampled persons who are offered the incentive. Estimate contingent on response rates assumed in the sample design.

Table 3. Estimated Annualized Burden Hours for 2021 Data Collection

Type of Respondent	Mode	Number of Respondents	Number of Responses per Respondent	Average Burden per Response (in hours)	Total Burden (in hours)
U.S. General Population, adults (18 and older)	Screener Respondents
	ABS, web	278	1	3/60	14
	ABS, paper	56	1	3/60	3
	Full Survey Respondents
	ABS, web	195	1	25/60	81
	ABS, in-bound telephone	5	1	40/60	3
	Total Annualized Burden Hours				101

Table 4. Estimated Annualized Burden Costs for 2021 Data Collection

Type of Respondent	Mode	Number of Respondents	Number of Responses per Respondent	Average Burden per Response (in hours)	Average Hourly Wage	Total Cost
U.S. General Population, adults (18 and older)	Screener Respondents
	ABS, web	278	1	3/60	$26.42	$367
	ABS, paper	56	1	3/60	$26.42	$74
	Full Survey Respondents
	ABS, web	195	1	25/60	$26.42	$2,147
	ABS, in-bound telephone	5	1	40/60	$26.42	$88
	Total Annualized Burden Cost					$2,676

6