Interviewing in the Face of Disaster: Conducting a Survey of Hurricane Katrina Evacuees May 21, 2006

1
Interviewing in the Face of DisasterConducting
a Survey of Hurricane Katrina EvacueesMay 21,
2006AAPOR Annual Conference, Montreal,
QuebecMollyann Brodie, The Henry J. Kaiser
Family FoundationClaudia Deane, The Washington
PostMelissa J. Herrmann, ICRErin Weltzien,
The Henry J. Kaiser Family FoundationLori
Robbins, ICR
2

• A survey of Hurricane Katrina evacuees who were
  housed in Houston shelters was conducted by The
  Washington Post, The Henry J. Kaiser Family
  Foundation, and The Harvard School of Public
  Health.

The goal of the research was to represent
Hurricane Katrina evacuees in Houston.
3

• The general issue with properly capturing this
  group of people was their unexpected mobility and
  the constant fluidity of the situation.

The solution was found through strong
communication and extreme adaptability.
4
Timeline

• Monday, August 29th, 2005 Katrina hit the
  northern Gulf Coast

• Thursday, September 1st, 2005 Evacuees arrive in
  Houston

• Tuesday, September 6th, 2005 Contacted to
  perform a study
• Erin Weltzien (KFF) arrives in
  Houston

• Tuesday, September 6th -Thursday September 8th,
  2005
• Developed plan for the study hired local
  Houston in-person interviewers

• Wednesday, September 7th, 2005
• Lori Robbins (ICR) arrives in Houston

• Friday, September 9th, 2005
• Questionnaire was final and Briefed/trained
  interviewers

• Saturday, September 10th through Monday,
  September 12th, 2005 Field period    

• Tuesday, September 13th, 2005
• Data is delivered and Erin and Lori depart from
  Houston

• Friday, September 16th, 2005 Article Published
  in The Washington Post

5
General Planning

• Needed to hire field interviewers that could make
  it into the area without any trouble
• Hired a higher proportion of African-American
  interviewers
• Decided against offering incentives, for the
  protection of the interviewers and respondents
• Needed IRB approval in order to interview these
  respondents

6
General Planning

• Worked with 30 interviewers
• Set ground rules for interviewing
• Always ask permission
• No pressure or recontact if they refuse
• Assure respondents that their answers are
  anonymous
• Try to minimize disruption of the lives of
  individuals in the shelter or their schedules
• Respect privacy

7
There were three main steps in our process

• Gaining access to the large shelters
• Sampling of the smaller shelters
• Interviewer placement and sampling

8
Gaining Access to the Large Shelters
9
Larger Shelters

• The large shelters were comprised of

• The Reliant Complex encompassing the Reliant
  Stadium, Reliant Center, Reliant Astrodome and
  Reliant Arena.

• The George Brown Convention Center

10
The Plan

• We identified the large shelters and received
  permission from the head of the Red Cross before
  arriving in Houston.

11
What we encountered

• Permission from top officials did not gain us
  access at the Reliant Complex or George Brown
  Convention Center.

12
The Process

• Everyone worked on getting access

• Rich Morin (WP) called Red Cross contacts trying
  to reach the officials running these large
  shelters

• Erin Weltzien (KFF) followed up in-person with
  officials running the large shelters

• A field reporter from the Washington Post spoke
  with media officials on site

• Lori Robbins (ICR) spoke with supervisors at each
  shelter

• Bob Blendon (HSPH) put us in contact with the
  Harvard physician who was running the Medical
  Center

13
The Result

• We received access through the media department
  at the Reliant Complex and through the Red Cross
  and local county officials at the George Brown
  Convention Center.

14
Sampling Smaller Shelters
15
The Plan

• A list of 17 shelters was generated based on a
  25-30 mile radius of downtown Houston. The list
  was narrowed down to ten shelters for the field,
  based on the estimated population of residents.

• Permission from the head of Red Cross would help
  ease the burden of gaining access to smaller
  shelters.

16
What we encountered

• Once we arrived in Houston, the reported number
  of residents at these shelters was changing
  rapidly. Evacuees were moving into more stable
  housing arrangements once they received FEMA
  money and hotel vouchers.

• Permission from top officials did not guarantee
  access at the shelters, which were churches and
  other centers with their own leadership (i.e. not
  Red Cross people). Additionally, there were
  security concerns and problems locating shelters.
  

17
The Process

• We made substitutions to our original list of 17
  to replace the shelters that were vacant or that
  we were unable to locate.

• Lori (ICR) cajoled shelter supervisors to let us
  in in some cases, we simply showed up
  unannounced and convinced the shelter leader to
  participate.

18
The Result

• We were able to complete interviews at 5 small
  shelters within the Houston Area.

19
Interviewer Placement and Sampling
20
The Plan

• To ensure that we would provide as good a
  probability as possible to interview evacuees
  throughout the entire shelter facility

• Interviewers would review the floor plan of the
  evacuee areas. This included areas where cots
  were located as well as any other rooms,
  hallways, corridors, and walkways.

• Supervisors would divide the site into quadrants
  to provide adequate coverage to the entire
  facility and avoid clustering of interviewers.

• The majority of the interviewers would be placed
  where the audience would be most stable cot
  area.

21
What we encountered

• Once inside the shelters, interviewing
  arrangements at the Reliant Complex and the
  George Brown Convention Center differed

• On the first day of interviewing inside the
  Reliant Complex, escorts were required. Many
  times these escorts were only available for an
  hour or two. When an escort had to leave, the
  interviewer had to drop off and gather a new
  escort, thus wasting time out in the field.

• Escorts were not needed at the George Brown
  Convention Center or any of the 5 smaller
  shelters.

22
What we encountered

• We discovered that the cot areas were largely
  empty during the daytime as people were out
  accessing services.

• The majority of the evacuees that were in this
  area consistently were elderly people or people
  with limited mobility.

23
The Process

• Because of difficulties gaining access to sites,
  the first day of interviewing took place outside
  the Reliant Complex both on the street and in
  the nearby shopping areas- and outside of the
  George Brown Convention Center.

• Interview procedures were similar to our intent

• Interviewers were stationed in various locations
  and they counted off people who passed, or groups
  of individuals sitting on grassy areas, in
  parking lots, etc.

24
The Result

• The escort system at the Reliant Complex was
  phased out as the authorities grew more
  comfortable with our project and were assured
  that we were not pressuring residents to
  participate.

25
The Result

• In large shelters, interviewer placement was
  constantly changing to adapt to the flow of
  residents.

• During the day, most residents were active.
  Interviewers were stationed in hallways,
  entrances and exits, and outside.

• Cot areas were largely vacant during the day
  except for elderly or immobile residents fewer
  interviewers were stationed there.

• At meal times, interviewers shifted to the
  cafeteria areas.

• Towards the evening and after school, the cot
  area would populate and interviewers shifted
  accordingly.

26
Lessons Learned

• The cell phone is the greatest thing to happen to
  in-person research!

• Communication is paramount in situations like
  this.

• The interviewers were often the best sources of
  information about the daily flow of residents,
  and often had suggestions about where to station
  themselves.

• The key to success is extreme adaptability!

27
For more information

• Please contact
• Melissa J. Herrmann
• ICR
• 484-840-4404
• mherrmann_at_icrsurvey.com
• Link to the full survey results Post story
• http//www.kff.org/newsmedia/7401.cfm