Abstract:

Refugee and immigrant communities’ histories and cultural differences lend themselves to unique types of vulnerabilities and challenges for achieving community resilience in response to environmental stressors and disasters. This mixed methods, community- based participatory research project engaged Cambodian and Laotian families living along coastal Alabama to determine how individual, family, and community level strengths and vulnerabilities contributed to community health and well-being and how community members utilized social networks and formal community services to respond to stressors. Important outcomes included the development of culturally responsive interventions and strategies for increasing community capacity and resilience when communities are faced with challenges associated with environmental stressors and disasters. Qualitative: Participants were recruited via purposive, snowball sampling of the key community partners working with the researchers. Specifically, there are four cultural advisory board members embedded in the community who worked closely with the researchers throughout participant recruitment and data collection (two Cambodian community members and two Lao community members). In total, 124 interviews were completed with 183 participants between March 2018 and July 2019. Of participants, 55 were men and 128 were women. 68 Lao and 113 Cambodian participants were interviewed, as well as 2 participants of other ethnicities. Participants 15 years and older were invited to participate. For individuals ages 15-18, both they and their parent or guardian completed the informed consent. Of participants, 20 were aged 25 or younger, 35 participants were ages 25-39, 82 participants were ages 40-60, and 46 participants were 61 or older. Additionally, of participants, 20 immigrated to the US less than 5 years prior to the time of the interview, 12 participants immigrated within 5-10 years, 35 immigrated within 10-20 years, 37 immigrated 20-30 years prior, 55 participants immigrated over 30 years prior to the interview, and 24 participants’ time since immigration was unknown. Participants received a $25 gift card as an incentive for participating in an interview (either ethnographic or photovoice). Participants came to community meetings and were offered a $10-retail gift card for participation in the survey data collection. Data were also collected in homes, businesses and recreational locations. Most of the data were collected in the Mobile County (73 percent, n=326) and the remaining responses came from a combination of Mississippi (n=58), Florida (n=34) and Louisiana (n=27) in October of 2017 through March of 2018. Responses were recruited through purposive snowball sampling of the key community partners working with the researchers. Specifically, there are four cultural advisory board members embedded in the community working closely with the researchers for data collection (two Cambodian community members and two Lao community members). Nearly all responses were collected face to face on iPad devices. The survey was available in English, Khmer and Lao. Institutional review board requirement and human subject approval were obtained by the (University of Georgia) and subjects indicated their consent to participate in the project. Including demographic information, the survey contained 67 questions and took approximately 15 min to complete. This dataset supports the publications: DeYoung, S. E., Lewis, D. C., Seponski, D. M., Augustine, D., & Phal, M. 2019. Disaster preparedness among Cambodian- and Laotian- Americans. Disaster Prevention and Management. Kelley, A., Seponski, D., & Lewis, D. (2022). ‘Swallow medicine, eat rice, pray about health’: Health, health care and health-seeking experiences of South-East Asian older refugees. Ageing and Society, 1-18. doi:10.1017/S0144686X22000290

Suggested Citation:

Lewis, Denise C., Desiree M. Seponski, and Sarah DeYoung. 2022. Cambodian-Laotian community strengths and resilience study conducted in coastal Alabama, Mississippi, Florida, and Louisiana between March 2018 and July 2019. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/W2CJ10FX

Publications:

Kelley, A. N., Seponski, D. M., & Lewis, D. C. (2022). ‘Swallow medicine, eat rice, pray about health’: health, health care and health-seeking experiences of South-East Asian older refugees. Ageing and Society, 1–18. https://doi.org/10.1017/s0144686x22000290

DeYoung, S. E., Lewis, D. C., Seponski, D. M., Augustine, D. A., & Phal, M. (2019). Disaster preparedness and well-being among Cambodian– and Laotian–Americans. Disaster Prevention and Management: An International Journal, 29(4), 425–443. https://doi.org/10.1108/dpm-01-2019-0034

Purpose:

This project contributed to three main sociocultural objectives of human dynamics tightly linked to resilience: (1) Engagement of the refugee community through collaboration on, development, testing, and implementation of empirically-informed, community-driven, and community-based programs that produced immediately useable information to prepare for, adapt to, and mitigate effects of environmental challenges; (2) Training community members in research, grant writing, and hazardous risk management skills to facilitate sustainability of developed programs/processes; (3) Creation of models of community-based risk reduction and household production of health including individual, family, and community intersections of sociocultural factors associated with strengths and vulnerabilities. This project demonstrated linkages across and enhanced, explicated, and coordinated existing informal networks and formal systems to better support individuals’, family members’, and community’s decisions surrounding access to and/or utilization of crisis response services and other formal systems (i.e., health care and legal services).

Data Parameters and Units:

Participant Selection Participants were recruited via purposive, snowball sampling of the key community partners working with the researchers. Specifically, there are four cultural advisory board members embedded in the community who worked closely with the researchers throughout participant recruitment and data collection (two Cambodian community members and two Lao community members). In total, 124 interviews were completed with 183 participants between March 2018 and July 2019. Of participants, 55 were men and 128 were women. 68 Lao and 113 Cambodian participants were interviewed, as well as 2 participants of other ethnicities. Participants 15 years and older were invited to participate. For individuals ages 15-18, both they and their parent or guardian completed the informed consent. Of participants, 20 were aged 25 or younger, 35 participants were ages 25-39, 82 participants were ages 40-60, and 46 participants were 61 or older. Additionally, of participants, 20 immigrated to the US less than 5 years prior to the time of the interview, 12 participants immigrated within 5-10 years, 35 immigrated within 10-20 years, 37 immigrated 20-30 years prior, 55 participants immigrated over 30 years prior to the interview, and 24 participants’ time since immigration was unknown. Participants received a $25 gift card as an incentive for participating in an interview (either ethnographic or photovoice).

Methods:

Interview Methodology Ethnographic Interviews In-depth interviews served to gain thick descriptions of the actions, beliefs, and behaviors that influence the lived experiences of community members. Interviews were conducted in the primary language of the participant(s). Interviewers either spoke the primary language of the participant, or a multilingual translator (either a member of the research team or key community member) was employed. Interviews were conducted at the homes of participants, the local temples, or the office of the research team, decided by the participants based on their convenience and comfort. Participants also elected to engage in interviews either individually or as family, resulting in in-depth interviews with one to four participants. In total, 92 ethnographic interviews were completed with 142 participants. Interviewers employed a structured interview guide that assessed family and community strengths and leadership, disaster recovery and preparedness, health and health-seeking behaviors, and resilience. Interviews lasted approximately 30 to 90 minutes. Photovoice Interviews Photovoice participants were recruited by community partners. 72 digital cameras were provided to 72 participants. Community partners instructed participants to take 10-12 photos of their home and community that illustrated why the community was a good place to live and what the community did well, and 10-12 photos that showed what the participants perceived as community hardships. Participants were asked to take the photos themselves. Due to extenuating circumstances, 32 photovoice interviews were completed with 41 participants, nearly one year later. Prior to the interview, two copies of the photos were printed. The participants kept one copy of the photos, while the research team kept the second copy. Interviewers and participants looked through the hard copies of the photos together during the interview, with participants narrating why they chose to take their photos and what they represented to them. Participants returned the camera SIM card with the photos to the research team, which were stored in the research office. Participants kept the cameras following the interviews and were provided with new, blank SIM cards as incentives for participating. Transcription Transcriptionists began by listening to the audio recorded interview at least once, all the way through to become familiarized with the subject matter, accents, and process. Transcriptionists paid attention to segments of data, which were defined as beginning with a question/comment by the Interviewer and concluding with the answer/comment by the Participant. Transcriptionists then engaged in a preliminary, rough transcription. Following initial transcription, transcriptions replayed the audio file while rereading the transcript, editing and revising to create a full and accurate transcription. The final transcription was read three times by the transcriptionist while listening to the recording, allowing for further revisions to be made and to increase accuracy. At the start of transcription, transcripts were labeled, including date, time, and location of the initial interview and the transcription. Interviewers and participants were identified via initials (“I” for interviewer, “P” for participant). Multiple participants within a single interview were differentiated via an assigned number, given to participants in order of beginning speech (e.g., The first participant to speak is P1, the second to speak is P2, etc. See below for a thorough description of participant and interviewer de/identification). Time markers were included in five minute intervals as well as at the end of the document. Recordings were transcribed verbatim (i.e., recorded word for word, exactly as said), including any nonverbal or background sounds (e.g., laughter, sighs, coughs, claps, snaps fingers, pen clicking, and car horn). All transcripts were written in English, with common Cambodian and Lao words, phrases, and proper nouns (e.g., names of holidays) retained for increased accuracy and salience within the community. Nonverbal sounds were typed in parentheses, (e.g., “short sharp laugh”). If interviewers or interviewees mispronounced words, these words were transcribed as the individual said them. The transcripts were not “cleaned up” by removing foul language, slang, grammatical errors, or misuse of words or concepts. Filler words (e.g., “hm”, “huh”, “mm”) or other similar sounds were transcribed. Word or phrase repetitions were also transcribed. If a word was cut off or truncated, a hyphen was employed at the end of the last letter or audible sound to indicate the point of cutoff. If an incorrect or unexpected pronunciation made a segment of data difficult to comprehend, the correct word was placed in square brackets. If speech overlapped within the interview, the phrase “crosstalk” indicated the indistinguishable segment of overlapping audio. A forward slash immediately behind the open square bracket and another in front of the closed square bracket denoted this corrected phrase (e.g., “I thought that was pretty pacific [/specific/]”). Inaudible information was noted with the phrase “inaudible segment”. Brief pauses and trailing off sentences were noted with ellipses, while long pauses were indicated with the phrase “long pause” in parentheses. If the transcriber was unsure of the accuracy of a statement made by a speaker, the statement was placed inside parentheses and a question mark in front of the open parenthesis and behind the close parenthesis (e.g., “I wanted to switch to ?(the hospital)? if they have a job available”). For audio files of English interviews or interviews with a multilingual translator, a primary English speaker or multilingual transcriptionist with high English proficiency transcribed the audio file. For audio files of interviews with Lao or Khmer speakers, multilingual transcriptionists engaged in simultaneous translation and transcription, transcribing interviews into English. Transcriptions were completed in InqScribe, and then formatted in Microsoft Word, at which point they were saved to the secure cloud-based database. Organization of Data Once interviews were completed, audio files were uploaded to a secure cloud-based database by interviewers. These files were then assigned to transcriptionists, whose transcription of the audio files were uploaded to the same cloud database in individual interview folders alongside their associated audio files. Individual interview transcripts were assigned a number according to the order in which they were transcribed and collected, followed by unique identifiers. Interviews are numbered one through 142. Eighteen numbers were not used, resulting from duplication and clerical error. Photovoice interviews are named “(PVC)” immediately following their interview number, while ethnographic interviews are not specifically indicated. De-identification of Interview Data For the data repository, we de-identified each interview transcript. Transcripts were de-identified through four rounds of review by two research assistants and the lab manager (twice), in order to ensure all items indicating participant identity were removed. The following criteria was used to de-identify transcripts: Participant Summary Each transcript is headed with a categorical summary describing each participant in general terms, as they were known to the interviewer and other researchers. These include: · Participant ID: how the participant is identified within the transcript. Typically, participants are identified with the letter “p”, sometimes followed by a number. · Age (category) · Gender (Male or Female) · Ethnicity (National-cultural identification. Typically Cambodian/Khmer, Laotian, or other Asian(-American)) · Level of Education (generalized) · Birthplace (Country of birth) · Length of Time Since Immigration to the US (category) Personal Names All personal names were removed from the data. Research staff names were replaced by a code used consistently across transcripts.Participant names were replaced by “P,” as well as a number in the case of multiple participants. Family member names were replaced by an indication of what relation they were to the speaker (e.g. “father,” “cousin,” “granddaughter” etc.). The names of community acquaintances were replaced by “[name]” or “[name removed],” with the exception of six community leaders, whose names were replaced by codes consistent across all transcripts. These codes appear as “[CCL1]” through [CCL4] (Cambodian Community Leaders 1-4) and “[LCL1]” through “[LCL2]” (Laotian Community Leaders 1 and 2) Place Names When used to indicate public or business institutions, or used in a general or community-wide narrative, place names were left as spoken by the participant and/or clarified by the transcript proofreader. Place names used by research staff members’ to describe themselves were typically left unedited as well. When used as a detail in a participant or a community interviewer’s personal narrative, place names were replaced by a more-general indication according to the following method: · Locations outside Mobile County, Alabama were replaced by a regional descriptor, such as “[state in the southeastern U.S.]” or “[city in another U.S. gulf state]” in accordance with generally accepted U.S. regional categories. · Large municipalities in Mobile County were typically left unedited. · Small municipalities, neighborhood, and street names were replaced by a more general indication, such as “[St. Elmo-Theodore area]” or “[Inland Mobile County]” as appropriate. · Detailed place descriptions or workplace names were removed only if they inferred a personal identity onto the participant or other community members. Age Across the entirety of the transcript, all specific age indicators were removed and generalized according to five age categories: Age under 15, 15 to 24, 25 to 39, 40 to 60, and 61+ Gender Across the entirety of the transcript, indications of gender were typically left unedited. Exceptions to this rule occur when some participants told potentially identificatory narratives where research staff felt it was safer to generalize. An example of this occurs when one participant indicates her mother just died within the past few months from complications of alcohol abuse. In this case, and other similar cases, “mom” or “mother” would be changed to “parent,” and gendered pronouns were changed to “him/her” or “s/he”. Duration of Residential Tenure Across the entirety of the transcript, all time periods indicated by participants to describe how long they had lived in a particular place (including south Alabama) were generalized according to five categories: <5 years, 5 to 10 years, 10 to 20 years, 30 to 40 years, and 40 to 50 years. Personal Information Health information was removed in part or in its entirety and generalized when research staff felt it could disclose identity. These include cases when health information was too specific (e.g. the location of cancer, or a particular limb injured), or when it referred to an unusual disease. Health care providers were removed and/or generalized if they referred to individual healthcare providers, which also might disclose participant identity. In each of these cases, efforts were made to generalize specific details while still retaining as much information as possible. Other personal information was removed in part or in its entirety, depending on its reference to personal information, such as a family member in jail or involvement in illegal or potentially shameful activities. Edited portions are marked with an indication, such as “[identifying information removed]”. Other Miscellaneous Identifying References Through our team’s work with the community and through a review of the collective body of interview data, research staff gained an acquaintance with particular cases where a particular element of a narrative might identify the participant. We removed or generalized the transcript in these cases. Edited portions are marked with an indication, such as “[identifying information removed]”. Quantitative Survey As described in our article (italics are verbatim), https://www.emerald.com/insight/content/doi/10.1108/DPM-01-2019-0034/full/html Participants came to community meetings and were offered a $10-retail gift card for participation in the survey data collection. Data were also collected in homes, businesses and recreational locations. Most of the data were collected in the Mobile County (73 percent, n=326) and the remaining responses came from a combination of Mississippi (n=58), Florida (n=34) and Louisiana (n=27) in October of 2017 through March of 2018. Responses were recruited through purposive snowball sampling of the key community partners working with the researchers. Specifically, there are four cultural advisory board members embedded in the community working closely with the researchers for data collection (two Cambodian community members and two Lao community members). Nearly all responses were collected face to face on iPad devices. The survey was available in English, Khmer and Lao. Institutional review board requirement and human subject approval were obtained by the (University of Georgia) and subjects indicated their consent to participate in the project. Including demographic information, the survey contained 67 questions and took approximately 15 min to complete.

Instruments:

Well-being Questions for well-being were adapted from the Subjective Well-Being Assessment Scale (Veronese et al., 2017) and the Personal Well-being Index (Sulaiman-Hill and Thompson, 2010). Sense of community Sense of community was assessed using the theoretical components developed by McMillan and Chavis (1986) and validated by Peterson et al. (2008). Disaster preparedness The disaster preparedness scale was adapted from preparedness research by DeYoung and Peters (2016), whose scales were adapted from Mishra and Suar (2012). Risk perception The risk perception scale was adapted from the scale also used in the DeYoung and Peters (2016) study on preparedness. Additional items The instrument also asked several “yes/no” questions including: ability to cope with a sudden financial crisis, confidence in the local governments’ capacity to respond to a disaster, confidence about one’s ability to engage in hazard preparedness and perceptions about major hazards.

Provenance and Historical References:

DeYoung & Peters: Disaster Preparedness. 250. International Journal of Mass Emergencies and Disasters. August 2016, Vol. 34, No. 2, pp. 250-282. McMillan, D. W., & Chavis, D. M. 1986. Sense of community: A definition and theory. Journal of Community Psychology, 14(1), 6–23. https://doi.org/10.1002/1520-6629(198601)14:1<6::AID-JCOP2290140103>3.0.CO;2-I Mishra, S., & Suar, D. 2012. Effects of anxiety, disaster education, and resources on disaster preparedness behavior. Journal of Applied Social Psychology, 42(5), 1069–1087. DOI: 10.1111/j.1559-1816.2011.00853.x Peterson, C., Park, N. and Sweeney, P.J. 2008. Group Well-Being: Morale from a Positive Psychology Perspective. Applied Psychology, 57: 19-36. https://doi.org/10.1111/j.1464-0597.2008.00352.x Sulaiman-Hill, Cheryl M.R. and Thompson, Sandra C. 2010. Selecting instruments for assessing psychological wellbeing in Afghan and Kurdish refugee groups. BMC Research Notes. 3 (237): pp. 1-9. Veronese N, Facchini S, Stubbs B, Luchini C, Solmi M, Manzato E, Sergi G, Maggi S, Cosco T, Fontana L. Weight loss is associated with improvements in cognitive function among overweight and obese people: A systematic review and meta-analysis. Neurosci Biobehav Rev. 2017 Jan;72:87-94. doi: 10.1016/j.neubiorev.2016.11.017. Epub 2016 Nov 24. PMID: 27890688.

Individual Files: