An Early Assessment of Hurricane Harvey’s Impact on Vulnerable Texans in the Gulf Coast Region: Their Voices and Priorities to Inform Rebuilding Efforts

Appendix A: Survey and Focus Group Methodology

Survey Methodology

The Kaiser Family Foundation/Episcopal Health Foundation Post-Harvey Survey was conducted by telephone October 17 – November 20, 2017 among a random representative sample of 1,635 adults age 18 and older living in 24 counties along the Texas Gulf Coast. The counties were chosen based on a mapping analysis of Harvey property damage developed by FEMA, in an effort to examine a contiguous area of counties that suffered the largest share of property damage (see Figure 1). The region was further divided into four groupings of counties: 1) Harris County; 2) Counties surrounding Harris (Liberty, Chambers, Galveston, Brazoria, Matagorda, Wharton, Colorado, Austin, Waller, Fort Bend, Montgomery, and Walker counties); 3) Golden Triangle (Jefferson, Hardin, and Orange counties); and 4) Coastal counties (Nueces, San Patricio, Refugio, Aransas, Calhoun, Victoria, Jackson, and Lavaca counties).

Interviews were administered in English and Spanish, combining random samples of both cellular and landline telephones (note: persons without a telephone could not be included in the random selection process). Sampling, data collection, weighting and tabulation were managed by SSRS in close collaboration with Kaiser Family Foundation and Episcopal Health Foundation researchers. Episcopal Health Foundation paid for the costs of the survey fieldwork, and Kaiser Family Foundation contributed the time of its research staff. Both partners worked together to design the survey and analyze the results.

The sampling procedures were designed to reach set numbers of respondents in each of the four county-groups and to oversample particular vulnerable subpopulations who were likely to require assistance in the aftermath of the hurricane, namely: people who experienced property damage as a result of the hurricane, those with household incomes near or under poverty level, Hispanic residents (in particular, non-native Hispanics), and Black residents. Some respondents were reached by oversampling cellular and landline numbers matching directory-listings in areas where data from the Federal Emergency Management Agency (FEMA) indicated large amounts of property damage due to Harvey. The sampling and screening procedures included an oversample component designed to increase the number of low-income respondents, specifically low-income Hispanic and Black respondents. This included 104 respondents who were reached by calling back respondents in the affected areas who had previously completed an interview on the SSRS Omnibus poll and indicated they fit one of the oversample criteria (based on income and race).

The dual frame cellular and landline phone sample was generated by Survey Sampling International (SSI) using random digit dial (RDD) procedures. All respondents were screened to verify that they resided in one of the 24 counties covered by this study at the time Harvey hit Texas. For the landline sample, respondents were selected by asking for the youngest adult male or female currently at home based on a random rotation. If no one of that gender was available, interviewers asked to speak with the youngest adult of the opposite gender. For the cell phone sample, interviews were conducted with the adult who answered the phone.

A multi-stage weighting design was applied to ensure an accurate representation of the population of each county-group. The first stage of weighting involved corrections for sample design, including accounting for oversampling of the most-affected areas, as well as non-response for the callback sample. In the second weighting stage, demographic adjustments were applied to account for systematic non-response along known population parameters, within each county-group. Population parameters included gender, age, race, Hispanic origin (broken down by nativity), educational attainment, and phone status (cell phone only or reachable by landline). This stage excluded the low-income oversample component. Based on this second stage of weighting, estimates were derived for the share of low-income respondents (Black, Hispanic and other) in the population. The third stage of weighting included all respondents in each county-group and included income-status (low or high) by race/ethnicity based on the previous stage’s outcomes. In the last stage each county-group was weighted to accurately represent its adult-population share within the 24-county region. Weighting parameters were provided by SSI based on estimates from the U.S. Census Bureau’s 2016 American Community Survey (ACS) for Harris County and 5-year (2011-2015) cumulative data from the ACS for other county groups.

The margin of sampling error, including the design effect for the full sample, is plus or minus 3 percentage points. Numbers of respondents and margins of sampling error for key subgroups are shown in the table below. For results based on other subgroups, the margin of sampling error may be higher. Sample sizes and margins of sampling error for other subgroups are available by request. Note that sampling error is only one of many potential sources of error in this or any other public opinion poll. Kaiser Family Foundation public opinion and survey research is a charter member of the Transparency Initiative of the American Association for Public Opinion Research.

Focus Group Methodology

As part of this project, the Kaiser Family Foundation and the Episcopal Health Foundation conducted focus groups to gather qualitative data from vulnerable populations affected by Hurricane Harvey. Topics covered in the focus groups were similar to those covered in the survey. The groups were held in Houston on November 8-9, 2017 and Beaumont on November 9, 2017. In total, five two-hour groups were conducted with eight participants each.

To qualify for each group, individuals must have said: 1) that their total family income from all sources in 2016, before taxes, was less than $60,000; and 2) that they experienced damage to their home or vehicle or lost job-related income as a result of Hurricane Harvey, and that the loss was at least a minor problem for them and their family. In addition, participants in each session were recruited with several additional aims: That at least half said their loss of income or property was a major problem for them and their family, and at least half said they their home had major damage or was destroyed. Recruitment also aimed to target a mix of owners and renters.

In Houston, two groups were conducted with Hispanics in Spanish – one of which consisted of participants who self-identified as undocumented immigrants. A third Houston group, consisting of mixed races/ethnicities, was conducted in English. In Beaumont, two groups were conducted in English, one with Black residents and the other with participants of mixed races/ethnicities. Participants in four of the groups were given an incentive of $100, and participants in one Houston group were given $125 since the group was conducted in the morning, a more difficult time to recruit for.

ConneXion Research recruited and provided moderator services for each group. The Houston groups were held at ConneXion’s research facility, and the Beaumont groups were held at a health service organization. The screener questionnaire and discussion guide was developed by researchers at the Kaiser Family Foundation and the Episcopal Health Foundation. Groups were audio and video recorded with participants’ permission. Focus group costs were paid for by the Episcopal Health Foundation.

Appendix B: Additional Resources

In Episcopal Health Foundation’s (EHF) initial effort to study the scope of property damage caused by Harvey, staff created a user-friendly, interactive map by using a geographic information system (GIS) mapping file from the FEMA website. The analysis was conducted using parcel data on buildings and homes, as well as flood inundation data showing the depth of water levels in different areas. EHF staff also overlaid the CDC-developed social vulnerability index data on these maps, identifying communities in need of additional support to prepare or recover from a disaster. Staff further identified and analyzed FEMA application data at the zip code level across the 41 counties, showing tremendous diversity of needs in the region.