621 Investigating Mobile Home Residents, Tornadic Weather, and Vulnerability: A Review of Survey Data Focused on Social and Physical Resources and Responses to Tornadic Information

Tuesday, 8 January 2019
Hall 4 (Phoenix Convention Center - West and North Buildings)
Michael Egnoto, Walter Reed Army Institute of Research, Silver Spring, MD; and K. D. Ash, S. M. Strader, W. S. Ashley, D. Rouche, C. Edwards, and K. E. Klockow-McClain


Residents of mobile homes (MHR) present a unique challenge when it comes to fostering tornado resilience. Recently MHR have been identified by scholars as disproportionately affected by the consequences of severe weather. Competing lines of thought have speculated whether the current state of MHR is a function of the construction of mobile homes (MH) and the mechanisms with which they are secured, or if individuals who live in MH have other social confounds that complicate their ability to receive and act upon information that might help mitigate negative tornadic outcomes. Therefore, we seek to assess MHR vulnerability and capacities to address tornado warnings by using survey results of 248 MHR throughout the Southeast U.S. to improve scientific understanding of how people living in mobile interact with tornado situations given their physical and social vulnerability.

This work evaluates MHR perceptions of quality for their specific MH relate to fear of tornadoes, what associations exist between protective action taking and perceived MH quality, and how resources (cognitive and emotional) impact protective action taking in tornadic contexts.

Data and Method

A total of 248 MH subjects from the Southeast completed an online survey via Qualtrics.com. Respondents identified as female (75.8%), with 0.4 percent not reporting a binary gender identity and 22.2% identifying as male. Most respondents lived in Alabama (65.7%) with the remaining subjects living in Mississippi. Approximately two-thirds of respondents owned their MHs (65.9%) with the rest identifying as renters. Most residents racially identified as white (73.8%) or black (20.6%), with the remaining subjects identifying as Latino/Hispanic (2.0%), Asian (1.2%), or other. Roughly half of the respondents had completed high school (46.0%) with smaller percentages having an associate’s (16.5%), bachelor’s (17.3%), or advanced (8.5%) degrees. The remaining MH subjects did not complete high school (8.1%) or have a technical profession. Additionally about two-thirds of respondents have a live-in partner (62.5%) with approximately half of all respondents reporting that they had children living with them in unit (47.6%). In total, 39.7% of respondents indicated that they worked full-time, with 13.4% indicating part-time employment, 23.9% indicating they were currently unemployed, and 22.3% indicating they had an employment situation that could not be described as unemployed, part-time, or full-time. Investigation of open-ended responses on employment status indicated that the 22.3% that described their status as other were largely retired, disabled, or acting home-makers, with some identifying as students. Finally, respondents averaged 42.96 years old (SD = 16.13) and ranged from 18 to 81 years.


Perceived Housing Quality and Fear of Tornadoes were assessed with multi-item Likert type measures. Housing quality (α=.913) and fear of tornadoes (α=.907) both indicated one-factor solutions and good reliability, and were used as composite variables averaging the scores of their individual items. Bi-variate correlation indicated no association between perceived housing quality and fear of tornadoes. Results were further investigated by cutting housing quality into three groups with the low perceived quality group being members more than one standard deviation below the mean, and the high perceived quality group being more than one standard deviation above the mean, with the median group being within one standard deviation of the mean. Results further confirmed no association between perceived housing quality and fear of tornadoes (F(2,238)= .812, p = NS).

Perceived Housing Quality and Protective Action Taking were investigated to determine how people react when they are in a hypothetical situation with only 15 minutes to prepare before being impacted by an impending tornado while they are home. Respondents were presented with 13 possible actions developed from prior research that people commonly engage in to prepare for tornadoes. Responses were categorized as desirable when people placed a priority on moving themselves and individuals immediately around them to shelter, where as other actions were considered less desirable (i.e. moving vehicles out from under trees, moving lawn furniture indoors, etc.). ANOVA indicated a significant relation such that people who were more likely to collect individuals immediately and seek shelter had greater confidence in their shelters than those less likely to immediately seek shelter (F(2,240) = 3.181, p = .043). Subsequent analyses indicated that although those with more confidence in their homes were more likely to immediately seek shelter, they also had significantly more fear of tornadoes (F(2,242)= 9.356, p < .001). No association between education and employment could be found with desirable action-taking behavior.

Personal Resource Access and Protective Action Taking were evaluated with the same action taking construction as before, with resources being assessed with self-reported items like, “I can usually think of a way to get myself out of a jam” (cognitive) or, “I am looking forward to the future ahead of me” ( emotional). ANOVA results were consistent such that better protective actions were associated with higher self-reported cognitive (F(2,244)=3.834, p = .023) and emotional (F(2,237)=4.503, p=.012) resources.


The lack of association between perceived quality of housing and fear of tornadoes is promising. This result may indicate that MHR are not over-estimating the ability of their residences to withstand severe weather. Secondly, the relation between prioritizing life-saving actions and higher self-reported housing quality ratings and greater fear may be indicative of higher knowledge of tornado consequences. Further, non-financial resources in-terms of cognitive and emotional capability were associated with better protective action taking. These results are indicative of a potential way forward to improve outcomes for individuals with limited fiduciary means by targeting non-financial resources as a way forward to improve behaviors during tornadic weather.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner