Alternate Title: Hurricane Katrina Study
Principal Investigator(s): Kessler, Ronald C., Harvard Medical School
One or more files in this study are not available for download due to special restrictions; consult the restrictions note to learn more. You can apply online for access to the data. Authentication is required to apply for access.
Due to the sensitive nature of the data, users interested in obtaining these data must complete an Agreement for the Use of Confidential Data, specify the reasons for the request, and obtain IRB approval or notice of exemption for their research. Apply for access to these data through the ICPSR Restricted Data Contract Portal, which can be accessed via the study home page.
Kessler, Ronald C. Hurricane Katrina Community Advisory Group Study [United States]. ICPSR22325-v2. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-06-10. doi:10.3886/ICPSR22325.v2
Persistent URL: http://doi.org/10.3886/ICPSR22325.v2
This survey was funded by:
- United States Department of Health and Human Services. National Institutes of Health (R01 MH070884-01A2)
Scope of Study
Subject Terms: American Red Cross, disaster relief, disasters, emotional states, Federal Emergency Management Agency, floods, health insurance, health status, household composition, housing conditions, hurricanes, insurance, life satisfaction, living arrangements, mental health, physical condition, property insurance, race, stress, suicide
Date of Collection:
Unit of Observation: individual
Universe: English-speaking adults aged 18 and older who lived in counties or parishes defined by the Federal Emergency Management Agency (FEMA) as eligible for individual assistance after Hurricane Katrina.
Data Types: survey data
Study Purpose: The purpose of this study is to inform policy makers of the impact of Hurricane Katrina on survivors' physical and mental health and barriers to treatment, as well as assist in future natural disaster planning efforts.
A multiple-frame sample design was used to recruit the Advisory Group. A multiple-frame design is one in which more than one frame is used to select respondents and all respondents sampled from any of the frames are asked to provide information that allows the researcher to estimate the probability of selecting the respondent from any of the frames. In this study, five frames were used to select the sample:
1. Safe lists. The American Red Cross, MSNBC, and a number of other organizations created 'safe lists' after the hurricane to allow survivors and their relatives to get in touch with each other by providing information about their whereabouts. These lists were downloaded from the Internet and the information in all of the lists was consolidated. From this consolidated list, a probability sample was selected for inclusion in the Advisory Group.
2. American Red Cross relief list. More than 1,300,000 families contacted the American Red Cross for relief after Hurricane Katrina. A probability sample from the American Red Cross relief list was also included in the Advisory Group.
3. Hotels. At the time the Advisory Group was recruited, in mid-January, 2006, the vast majority of people displaced by Hurricane Katrina had long since left hotels and shelters. However, a residual group of approximately 30,000 still resided in hotels. As a result, a probability sample of hotels in the communities reported by FEMA to be housing Katrina refugees was selected and hotel rooms were screened to obtain a probability sample of these individuals.
4. Random-digit dialing in the affected areas. A conventional random-digit dial (RDD) sample of randomly generated phone numbers in the affected counties and parishes was also used. This was based on telephone banks in these areas that had at least one listed number prior to the hurricane. The usefulness of this sample was increased dramatically by the fact that the Bell South telephone company maintained the land line phone numbers of people who were displaced by the hurricane and provided forwarding numbers for all individuals who had given such numbers to the telephone company. This made it possible to contact a large proportion of the people who were displaced by the hurricane simply by calling their old phone numbers and being transferred directly to them in California, Illinois, Pennsylvania, or elsewhere throughout the country.
5. Random-digit dialing outside the affected areas. RDD was also used to screen for people displaced by the hurricane who lived outside the affected area. This was done by calling a random sample of phone numbers based on the same sampling scheme as in the affected areas, but this time including numbers throughout the United States. In order to increase the efficiency of this screening effort, a digitally recorded message was used to call thousands of phone numbers each day throughout the recruitment period asking for people living in households with eligible people to indicate this fact either with a recorded voice response or a touch-tone phone response that indicated that a live interviewer should contact the household.
You can find more information via the sample characteristics utility:
Weighting was used to adjust for differential probability of selection depending on the number of sampling frames in which each person in the population was represented.
An additional weight was used to adjust for the fact that only one respondent in each sample household was selected to participate in the Advisory Group regardless of the number of eligible people living in that household. In special cases of group quarters -- for example, where several different families were living together -- we allowed two people to be selected for Advisory Group participation in order to truncate the within-household probability of selection weight.
A final post-stratification weight was also used to adjust for residual discrepancies between the sample and data obtained from the United States Bureau of the Census on the characteristics of people who lived in the affected areas at the time of the 2000 Census. All the information on the Census long form was aggregated to the census tract level and Advisory Group members were weighted to approximate the distribution of the cross-classification of these variables at that level of geographic aggregation.
Mode of Data Collection: telephone interview
Response Rates: The screening response rate was 64.9 percent. The participation rate among screening survey respondents was 41.9 percent.
- Created variable labels and/or value labels.
- Checked for undocumented or out-of-range codes.
Original ICPSR Release: 2009-07-22
- 2010-06-10 Value labels were added to variables AGE_CAT, BK_POV4, RACE_NEW, PRE_LOCATION, and POST_LOC_STATUS. Variable labels were added/changed for variables QR2 to QR9, QC23_4_L, QC24_C, QH3A, QH3B, NEW_WEIGHT_FINAL, and BK_POV4.
Related Publications (?)
- Citations exports are provided above.
Export Study-level metadata (does not include variable-level metadata)