SUPPORTING STATEMENT
Part B
Cognitive Testing of the Medical Expenditure Panel Survey Health Insurance Verification Module
Agency for Healthcare Research and Quality (AHRQ)
1. Potential Respondent Universe and Sample Selection Method 3
2. Information Collection Procedures 4
3. Methods to Maximize Response Rate 4
5. Statistical Consultation and Independent Review 4
The potential respondent universe will be all residents of the metropolitan Washington, D.C. area and surrounding areas who fulfil the criteria required to test the questioning included in the MEPS health insurance verification module.
Ideally, the stimuli to which cognitive interview respondents are exposed mirrors as closely as possible that to which respondents in the main study are exposed. For this pretesting effort we face several challenges in that regard. One is that we cannot screen respondents for cognitive interview with the knowledge that they will misreport health insurance status for themselves or other household members when answering the main MEPS interview questions and therefore be administered the health insurance verification module. Another is that the vast majority of people do in fact, report their household health insurance status accurately. So even if we could screen on the likelihood of misreporting, that screening effort would be massive indeed. We also want to avoid priming respondents during screening, to the extent possible, to ensure a reliable test of the health insurance verification module items. Our design addresses these challenges while maintaining the integrity of the pretesting methodology and ensuring useful, high quality results. For screening, we propose identifying households with characteristics that we know from past research are prone to health insurance misreporting.
A pre-screener will be administered to all those who respond to the recruitment advertisement. A purposive sample of 44 English-speaking adults aged 18 or over will be recruited. This will be through: 1) follow-up with previous MEPS respondents who have agreed to take part in future research of this nature; 2) Westat’s database of individuals who have expressed interest in participating in qualitative research activities; 3) advertisements on Craigs List; 4) online ads using social media, or 5) contacts at local groups, including churches and other community organizations.
To adequately test all of the questions, a mix of respondents with the following characteristics will be recruited at each round of cognitive testing. The focus will be on recruiting larger households containing indirect family relationships, for example someone who lives with an elderly relative or adult sibling, in order to be able to test adequately the proxy reporting on other members of the RU/household. Table 1 lists the criteria for selection. Interviewees may fulfill one or more of the criteria listed.
Table 1. Sample selection criteria for each round of cognitive testing
Household Characteristic |
At least four household members |
At least one indirect family relationship within household |
At least one child under the age of 18 |
As least one household member aged 65 or over |
At least one household member with a disability age <65 |
At least one household member currently unemployed |
At least one household member who is employed (non-military) |
At least one self-employed household member |
At least one household member who is active duty or retired military |
At least one household member on Medicaid |
At least one household member with coverage through the Marketplace |
Household income $30,000 or less |
Household income $90,000 or more |
Respondent Characteristic |
High school diploma or less |
Lived in the U.S. for less than 5 years |
In addition to the mix shown in Table 1, we will also aim to achieve a spread of respondent characteristics related to age, sex and race/ethnicity.
We plan to conduct in-person interviews for all three rounds of cognitive testing to further refine the survey’s items and composites/dimensions. The MEPS health verification module questionnaire will be administered by the cognitive interviewer thereby replicating the approach used in the main MEPS survey. We plan to use valid cognitive interviewing techniques, which include scripted non-biasing retrospective cognitive probing, card sorting and respondent narrative, to identify any issues related to question comprehension and response processes. All interviews will be audio recorded for quality assurance and analysis purposes.
The questionnaire for this study is actually very short, although respondents will be asked to report on all members of their RU/household, very few questions are asked about any one household member.
We are using a cash remuneration as noted in #9 in Supporting Statement Part A.
The procedures for this specific project have not been subjected to testing. However, the contractor, Westat, has conducted many similar projects and is using well established research methods, that comply with the OMG standards and guidance for cognitive testing, with this project.
The following individuals have been consulted on the design of the pretest questionnaire:
Jessica Vistnes, Agency of
Healthcare research and Quality
Brandy Lipton, Agency of
Healthcare research and Quality
Wendy Hicks, Westat, Inc.
Angie Kistler, Westat, Inc.
Martha Stapleton, Westat, Inc.
Amanda Wilmot, Westat, Inc.
Karen Stein, Westat, Inc.
Heather Brotsos, Summit
Oswaldo Urdapilleta,Summit
| File Type | application/vnd.openxmlformats-officedocument.wordprocessingml.document |
| Author | Suzanne Streagle |
| File Modified | 0000-00-00 |
| File Created | 2021-01-22 |