Justification for Non-substantive Change

Notification of Change to the Sampling and Contact Methods for U.S. households in the National Smoke and CO Alarm Survey (SCOA Survey)

The administration of the SCOA survey, in its first two fielding locations, has experienced lower response rates relative to expectations established by the study’s approved methodology (OMB Control No. 3041-0180). The forecasted response rate for the SCOA survey was set at 25%, on average. Performance in the field, thus far, has produced a fraction of this response. For example, in Charlotte, NC and New Bern, NC where the study was conducted in February-March 2019 the response rate for each of these locations equaled less than 1%. Specifically, in New Bern, where OMB granted a 12% response rate allowance in January based on New Bern’s FEMA natural disaster designation, even with this 12% adjustment and much larger sample of contacts, the total response there equaled less than 1%. The contractor, EurekaFacts, followed all established protocols for administering the survey – as outlined in OMB No. 3041-0180 – using mail notifications, mail follow up letters and phone calls for recruitment and participant screening for interest and eligibility.

Following a review of the total attempted contacts for recruitment of participants, the sample dispositions for each contact, and the high level of effort made by interviewers (measured in hours worked), EurekaFacts determined that a 25% response rate for these specific metro areas and the remaining metro areas is an unattainable expectation. The contractor demonstrated that a change in the overall expected response rate for these and all other metro areas was justified.

To correct for this challenge and complete the number of homes surveyed as outlined in the contract with EurekaFacts, EurekaFacts proposed to CPSC a revised sampling method and contact protocol for participant recruitment. The CPSC project team has reviewed the plan with EurekaFacts and determined that this plan is an acceptable approach to reach the desired number of participants and to complete the study within the contracted timeline.

CPSC recommends the following revisions to the methodology:

• Increase sample of contacted households by reducing the expected response rate. Reduce the expected response rate for the study to 1% and in turn, purchase a sample of households to contact and recruit in each metro area that corresponds to a 1% response rate. This process will follow the same proportional sampling rules established by the original survey methodology. In each location a similar plan to increase the share of sampled housing units will be used to achieve the total number of homes surveyed as defined by the survey’s overall OMB approved methodology.

• Change the initial contact method. For the survey administration protocol, the recruitment and screening of participants will begin initially with phone calls to households from the address-based sample, instead of just the initial invitation letter. This change will drastically allow the contractor to stay within schedule of the contact.

• Change the order of notification mailings. In place of initial contact by mail (invitation letter), the households who are contacted, screened and scheduled for in-home interviews by telephone will then receive a confirmation notice by mail to confirm their participation. This reminder notification has a dual purpose; to increase household response rates and to provide legitimacy to the study.