1 EFFECTS OF THE GULF OIL SPILL IN ESCAMBIA COUNTY, FLORIDA By KELCEY RAY KILLINGSWORTH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Kelcey Ray Killingsworth
3 T o my girls
4 ACKNOWLEDGMENTS I thank my entire family for supporting me through this process including my parents, in laws, and extended family I espec ially thank my grandmother, Jean Ray, for a list of things so long it would constitute another paper. I love you Ma. I thank my husband, Cliff, for making it possible for me to quit my job and devote my time to this endeavor. I also thank him for taking my children on trips at strategic crazy writing method. I thank Cameron and Gillian for gracing my life and for understanding when momma could not go with them to do fun things because I had to work on my I thank my Chair, Peter Sherrard, for sticking with me even when it may have seemed I would never finish. I truly appreciate your kindness and patience through the years. I also appreciate my committee memb ers Drs. Clark, Dawson, and Puig, for being willing to stick with me through the dissertation process which took a lot longer than anyone planned. My eternal gratitude goes to Dr. Elizabeth Pearman for walking me through this process and keeping me motiva ted. I thank Drs. Jaime Jasser, Teresa Leibforth, and Heather Rask for inspiring and supporting me at critical times The Fab Four finally finished! My thanks go to the entire Glagola family who treated me as one of their own and gave me love and support throughout my academic career. When I was missing my family, they were there for me with a meal, a couch, and company. I love you guys. I need to thank David and Brian Marshall, who may no longer be here in body but are definitely with me in spirit. Wh
5 to see me finish. Finally I would like to thank the people of Escambia County, Florida for participating in my sur vey and making it possible for me to do the research to complete my dissertation. I love our community and hope that we appreciate what a truly wonderful place it is we get to call home.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF TERMS ................................ ................................ ................................ ........... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Scope of the Study ................................ ................................ ................................ 15 Theoretical Framework ................................ ................................ ........................... 17 Statement of the Problem ................................ ................................ ....................... 18 Need for the Study ................................ ................................ ................................ .. 19 Purpose of the Study ................................ ................................ .............................. 20 Rationale for the Methodology ................................ ................................ ................ 20 Research Questions ................................ ................................ ............................... 22 Overvie w of the Study ................................ ................................ ............................. 22 2 REVIEW OF LITERATURE ................................ ................................ .................... 23 Conservation of Resources Theory ................................ ................................ ........ 24 Disasters ................................ ................................ ................................ ................. 26 Natural Disasters ................................ ................................ .............................. 27 Technical/Man made Disasters ................................ ................................ ........ 30 The Gulf Oil Spill ................................ ................................ ................................ ..... 33 Resilience ................................ ................................ ................................ ............... 36 Psychological Stress ................................ ................................ ............................... 38 Summary ................................ ................................ ................................ ................ 40 3 METHODOLOGY ................................ ................................ ................................ ... 41 Research Method and Design ................................ ................................ ................ 41 Research Design ................................ ................................ .............................. 42 Population and Sampling ................................ ................................ .................. 44 Informed Consent ................................ ................................ ............................. 45 Confidentiality ................................ ................................ ................................ ... 45 Instrumentation ................................ ................................ ................................ ....... 46 Resource Change ................................ ................................ ............................ 46 21 Item Depression Anxiety Stress Scales ................................ ....................... 47 14 Item Resilience Scale ................................ ................................ .................. 48 Demographic Questionnaire ................................ ................................ ............. 48
7 Open ended Question ................................ ................................ ...................... 49 Data Collection ................................ ................................ ................................ ....... 49 Data Analysis ................................ ................................ ................................ .......... 50 Research Question 1 ................................ ................................ ........................ 50 Research Questions 2 and 3 ................................ ................................ ............ 52 Research Questions 4, 5, and 6 ................................ ................................ ....... 55 Methodological Limitations ................................ ................................ ...................... 56 Study Validity ................................ ................................ ................................ .......... 56 Summary ................................ ................................ ................................ ................ 58 4 RESULTS ................................ ................................ ................................ ............... 60 Sample Demographics ................................ ................................ ............................ 62 Reliability and Validity of Instruments ................................ ................................ ..... 64 Results ................................ ................................ ................................ .................... 65 Research Question 1 ................................ ................................ ........................ 65 Research Question 2 ................................ ................................ ........................ 65 Research Question 3 ................................ ................................ ........................ 68 Research Question 4 ................................ ................................ ........................ 71 Research Question 5 ................................ ................................ ........................ 72 Research Question 6 ................................ ................................ ........................ 74 Open Ended Question ................................ ................................ ...................... 75 Summary ................................ ................................ ................................ ................ 77 5 DISCUSSION ................................ ................................ ................................ ......... 89 Overview of the Chapter ................................ ................................ ......................... 89 Discussion of the Descript ive Data ................................ ................................ ... 90 Discussion of Instrumentation ................................ ................................ ................. 91 Discussion of Hypotheses ................................ ................................ ....................... 91 Discussion of Hypothesis 1 ................................ ................................ .............. 93 Discussion of Hypothesis 2 ................................ ................................ .............. 94 Discussion of Hypothesis 3 ................................ ................................ .............. 95 Discussion of Hypothesis 4 ................................ ................................ .............. 95 Discussion of Hypothesis 5 ................................ ................................ .............. 96 Discussion of Hypothesis 6 ................................ ................................ .............. 96 Discussion of Open ended Question ................................ ................................ 97 Clinical Implications ................................ ................................ ................................ 97 Theor etical Implications ................................ ................................ .......................... 99 Limitations ................................ ................................ ................................ ............... 99 Recommendations for Future Studies ................................ ................................ ... 100 Conclusion ................................ ................................ ................................ ............ 101 APPENDIX A INFORMED CONSENT ................................ ................................ ........................ 102
8 B BUSINESS OWNER SURVEY ................................ ................................ ............. 104 C RESIDENT/WORKER SURVEY ................................ ................................ ........... 111 LIST OF REFERENCES ................................ ................................ ............................. 118 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 125
9 LIST OF TABLES Table page 4 1 Study Participants Mean Age by Group ................................ .............................. 78 4 2 Study Participants by Gender ................................ ................................ ............. 78 4 3 Study Participants by Ethnic Group ................................ ................................ .... 78 4 4 Study Participants by Education Level ................................ ............................... 79 4 5 Study Participants by Employment Status ................................ .......................... 80 4 6 Study Participants by Household Income ................................ ........................... 81 4 7 Study Particip ants by Hours Worked per Week ................................ .................. 82 4 8 Reliability Coefficients for Study Scales ................................ ............................. 82 4 9 Correlation Matrix ................................ ................................ ............................... 83 4 10 Model Summary for the DASS 21 Depression Subscale ................................ .... 83 4 11 Coefficients for DASS 21 Depression Subscale ................................ ................. 83 4 12 Model Summary for the DASS 21 Anxiety Subscale ................................ .......... 83 4 13 Coefficients for DASS 21 Anxiety Subscale ................................ ....................... 83 4 14 Model Summary for the DASS 21 Stress Subscale ................................ ............ 83 4 15 Coefficients for DASS 21 Stress Subscale ................................ ......................... 84 4 16 Model Summary fo r the DASS 21 Total Score ................................ ................... 84 4 17 Coefficients for DASS 21 Total Score ................................ ................................ 84 4 18 Summary of Regression Analysis for DASS 21 ................................ .................. 84 4 19 Model Summary for the RS 14 Self Reliance Subscale ................................ .. 84 4 20 Coefficients for RS 14 Self Reliance Subscale ................................ ................ 84 4 21 Model Summa ry for the RS 14 Meaning Subscale ................................ .......... 85 4 22 Coeffi cients for RS 14 Meaning Subscale ................................ ....................... 85 4 23 Model Summary for the RS 14 Equanimity Subscale ................................ ...... 85
10 4 24 Coefficients for RS 14 Equanimity Subscale ................................ ................... 85 4 25 Model Summary for the RS 14 Perseverance Subscale ................................ 85 4 26 Coefficients for RS 14 Perseverance Subscale ................................ ............... 85 4 2 7 Model Summary for the RS 14 Existential Aloneness Subscale ...................... 85 4 28 Coefficients for RS 14 Existential Aloneness Subscale ................................ ... 85 4 29 Model Summary for the Total RS 14 ................................ ................................ .. 86 4 30 Coefficients for Total RS 14 ................................ ................................ ................ 86 4 31 Summary of Regression Analysis for Resilie nce Scale ................................ ...... 86 4 32 Means and Standard Deviations for Stress and Resource Change .................... 86 4 33 Summary Analysis Question 5 Significant R esults ................................ ............. 86 4 34 Means and Standa rd Deviations for Significant RS 14 Total and Subscales ..... 87 4 35 Means and Standard Deviations for Resource Change ................................ ...... 88 4 36 ANOVA Results for Question 6 Significant Results ................................ ............ 88 4 37 Mean and Standard Deviation for Significant Fi ndings by Claim/No Claim ........ 88
11 LIST OF TERMS A NXIETY : Negative affective state characterized by fear and somatic responses to fear (Lovibond & Lovibond, 1995). C LAIM M ONEY : Money individuals and businesses c ould apply to receive to compensate for financial loss due to the Gulf Oil Spill (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011). C ONSERVATION OF R ESOURCES T HEORY : theory that individuals seek gain and a void loss, and that if an individual loses (or is threatened with loss of) resources, that individual will experience psychological stress (Hobfoll 1988, 1989, 2012). D EPRESSION : Negative affective state characterized by loss of self esteem and incentive (Lovibond & Lovibond, 1995). E CONOMIC S TRESSORS : Negative financial and economic events, including job loss, 1989, 2012). E SCAMBIA C OUN TY B USINESS O WNERS : Those respondents who self identified as owning a business in Escambia County, Florida. E SCAMBIA C OUNTY R ESIDENTS : Those respondents who self identified as living in Escambia County, Florida. E SCAMBIA C OUNTY W ORKERS : Those respo ndents who self identified as working in Escambia County, Florida. G ULF O IL S PILL (GOS): The oil spill beginning in the Gulf of Mexico with an explosion of the Deepwater Horizon oil drilling rig on April 20, 2010 and lasting until July 15, 2010 when the oil well was capped (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011). N EGATIVE PSYCHOLOGICA L RESPONSE : A reaction to events that is characterized by a range of responses, such as adjustment disorders, depression, anxiety, stress, Post traumatic Stress Disorder, etc. (Lovibond & Lovibond, 1995). R ESILIENCE : The positive adaptation to negative life events or stressors (Wagnild & Young, 1993).
12 R ESILIENCY T HEORY : The theory that living a purposeful life, persev ering, having a balanced view of life, being self reliant, and being comfortable with yourself lead to positive reactions to negative life events or traumas (Wagnild & Young, 1993). R ESILIENT R ESPONSE : Responding to adversity in an adaptive, positive man ner with limited disruption in functioning and that leads to personal growth (Wagnild & Young, 1993). R ESOURCES : Objects, conditions, personal characteristics, and energies that individuals seek to obtain, retain, and avoid losing (Hobfoll 1988, 1989, 20 12). S TRESS : A persistent negative affective state characterized by persistent arousal and low tolerance of frustration (Lovibond & Lovibond, 1995). T ECHNOLOGICAL D ISASTER : A disaster caused by the actions of man, also referred to as a man made disaster (Palinkas, 2012).
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF THE GULF OIL SPILL IN ESCAMBIA COUNTY, FLORIDA By Kelcey Ray Killingsworth December 2012 Chair: Peter Sherrard Major: Mental Health Counseling The purpose of this study was to assess the impact of the British Petroleum Gulf Oil Spill on resource change, psychological stress, and resilience for business owners, residents, and workers in Escambia County, Florida. (1988, 1989) Conservation of Resources theory. All business owners, residents, and workers over the age of 18 in Escambia County, Florida were elig ible to participate in the online survey. A total of 146 participants completed a survey using the 21 Item Depression Anxiety Stress Scale s (DASS 21) the 14 Item Resilience Scale (RS 14) resource change scale, and demographic questionnaire. Utilizing co rrelation, multiple regression, and analysis of variance ( ANOVA ) the findings indicated there were significant models predicting stress and resilience and significant differences between respondents by income group, claim status, and respondent type (busi ness, resident, or worker ). Results indicated resource change was a predictor for the DASS 21 total and subscale scores, income group predicted resilience, and claim status resulted in higher stress for those with a claim. Results of the study are present ed, limitations addressed, and the implications for theory, practice, and future research are discussed.
14 CHAPTER 1 INTRODUCTION The Gulf Oil Spill (GOS) began on April 20, 2010 when the Deepwater Horizon oil drilling rig exploded in the Gulf of Mexico app roximately 49 miles south of Louisiana, instantly killing 11 men and starting a chain reaction of events. The initial explosion and the resultant oil spill led to environmental and economic devastation along the entire Gulf Coast and had national and inte rnational ramifications (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011). The Deepwater Horizon rig burned for two days after the initial explosion and then sank to the floor of the Gulf of Mexico nearly a mile below. The sinking caused damage to the underwater well, allowing oil to flow at an unknown rate into the Gulf of Mexico for 87 days until engineers were able to cap the well. The GOS is now called Institute of Medicine, 2010, p.ix). The Gulf Coast, the United States, and the entire world watched as oil spread and began to come ashore. The actual amount of oil spilled from this one disaster may never be known. Estimates are 57,000 barrels per day, or a total of 5,000,000 barrels of oil spilled between April 20 and July 15, 2010 when the well was finally capped. It was not officially declared dead until September 19, 2010 (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drill ing, 2011). In addition to the oil, the well also released approximately 100 million standard cubic feet of natural gas every day, Immediately following the GOS, the process of assigning blame for the incident began. British Petroleum (BP), Transocean (the company operating the oil rig), and
15 Halliburton were implicated through their systemic failure of risk management (National Commission on the BP Deepwater Horizon Oil Spil l and Offshore Drilling, 2011). Almost immediately, calls for BP to pay for the damage they caused to the environment and economies of the Gulf Coast communities began. BP set up a compensation fund of $20 billion and the claims process began in June of 2010 (Sole, 2011). Federal fines were also scheduled to be levied under the Resources and Ecosystems Sustainability, Tourism Opportunities and Revived Economies (RESTORE) Act (King & Berry, 2012), equating to millions of dollars set aside for local areas impacted by the spill. In the past there have been other technical disasters including oil spills; however, man made disasters of this scale are rare. The GOS is different because of its magnitude, duration, and the complexity involved in assessing the hu man and environmental impact (Institute of Medicine, 2010). More than two years after the disaster began, Gulf Coast papers continue to report submerged oil mats, BP claims 1; Gulf Coast Ecosystem Restoration Task Force, 2011; Jansen, 2010). The psychological impact of a technological disaster of this scale on the people of the Gulf Coast is unknown and has not been investigated. Scope of the Study A disaster on the scale of the GOS has potential to impact the global community, but it is the local people of the Gulf Coast region bearing the brunt of the impact. The Gulf Coast of the United States (U.S.) border s the Gulf of Mexico and consists of Florida, Alabama, Mississippi Louisiana, and Texas. Th e Southern U.S. coastline also includes bays and wetlands adding up to approximately 16,000 miles of coast (U.S. Environmental Protection Agency, 2010). Florida alone accounts for approximately
16 5,095 miles of this Gulf Coast tid al shoreline ( Gulf Coast Ecosystem Restoration Task Force, 2011 ). The Gulf Coast produces more than one harvest, produces most of the domestic gas and oil for the country, and supports the tourist industry providing so me 800,000 jobs (Gulf Coast Ecosystem Restoration Task Force, 2011). The entire United States is impacted by what happens in and around the Gulf of Mexico, but none more so than the people who choose to call the Gulf Coastal region home. Escambia County, Florida is the western edge of the State of Florida bordered by Alabama to the west and north, and by the Gulf of Mexico to the south. This area of Florida is known for its beautiful sugar white sand beaches driv ing the local economy through tourism ( e.g hotels, restaurants, and tourist attractions), charter and commercial fishing, and real estate industries. Other industries in the area including service industries are dependent on tourism, fishing and real estate to stay in business. Escambia County has a current unemployment rate of approximately 10%, higher than the current national unemployment rate of 8.3% (U.S. Bureau of Labor Statistics, 2012). Unfortunately the unemployment rate does not capture the underemployed or those no longer looking for gainful employment (U.S. Bureau of Labor and Statistics, 2012). Escambia County is unique due to its geographical location, natural environment, and population. The county has a diversified economy; however, a large part of the economy is driven by tour ism specifically related to the Gulf of Mexico. The economy of Escambia County has been weakened in recent years with hurricane losses (Hurricane Ivan in 2004) followed by a weak real estate market.
17 Florida does not allow offshore drilling, nor does it b enefit from the monetary resources offshore oil drilling brings, and yet the economy of this county was significantly impacted by the GOS drilling disaster. The first tar balls from the GOS reached Escambia County (Santa Rosa Island) on June 4, 2010 47 days after the explosion of the Deepwater Horizon (Flemming, 2010) Until this point the residents of Northwest Florida had been living with the uncertainty of when the tar and oil would reach their shore, when their water ways would be closed to swimmers and boat traffic, and when the local fish would no longer be safe to eat. Th e environment al impact of the GOS was immediately apparent. However, t he impact of the GOS on the lives of the residents of Escambia County is unknown even tw o year s after the oil flow was finally stopped. Local b usinesses have closed. Families have had to move. Unemploym ent rates are still high (U.S. Bureau of Labor and Statistics, 2012 ). The local economy is s till struggling. The local coastal environmen t appears to be clean, but there is still uncertainty about how much oil is still hidden under the water of the Gulf of Mexico Theoretical Framework This study on the effects of the GOS on the residents of Escambia County, Florida was developed using two theories, the Conservation of Resources Theory (COR) (Hobfoll, 1988, 19 8 9) and the theory of resilience (Wagnild & Young, 2003). The COR Theory developed by Hobfoll addresses stress. It is based on the assumption people try to obtain, keep, and protect t hose things, termed resources, they value (Hobfoll, 1988, 1989; Hobfoll & Lilly, 1993). The COR theory posits if individuals strive to obtain, retain, and protect their resources, psychological stress will occur if: 1) an
18 lost, and 3) if an individual fails to gain resources after investing other resources (Hobfoll & Lilly, 1993). Resilience has been defined as a characteristic some people have helping them to moder ate the negative effects of stress and allowing for change a nd growth (Wagnild & Young, 1993 ). Others have defined resilience as not just an absence of psychopathology in the face of negative events but as the ability to adapt and grow from these events ( Bonanno, 2004 & 2005). Wagnild and Young (199 3 ) identified five characteristics of resilience, forming the foundation of their Resilience Scale: a purposeful life, perseverance, equanimity, self reliance, and existential aloneness. Resilience can be thou Depression, anxiety and stress are negative affective conditions and have been studied for their theoretical and clinical importance (Lovibond & Lovibond, 1995). These negative affective conditions can have an impact life stressors and on society as a whole (Monroe, 2008). Depression, anxiety and stress often have similar symptoms or features; however, they are each a distinct state awford & Henry, 2003). Statement of the Problem The psycho social impact of an event as complicated and immense as the GOS is unknown; however, trauma research on other disasters has indicated there will be an impact on those living in Gulf Coastal commun ities ranging from mild adjustment disorders to major psychopathological disorders (Becker, 1997; Flynn & Norwood, 2004; Palinkas, 2012). Psychopathology or maladjustment to an event can lead to physical health problems, such as high blood pressure, heart disease and gastrointestinal issues (Bisgaier & Rhodes, 2011). Physical health problems resulting from exposure to the oil
19 can also lead to mental health problems including depression, anxiety, and stress (National Center for Disaster Preparedness, 2010) It is important to begin to understand the mental and physical health ramifications for coastal Florida residents and how the GOS has impacted their lives. Need for the Study The impact of the GOS on the environment was apparent from the birds covered w ith oil to the tar balls washing up on the beaches to the oil stained Gulf waters. The impact on the residents of the Gulf Coast was also apparent in lost livelihoods and damage to businesses and individuals. Shortly after the spill began there was specu lation about the effects this disaster would have on the mental health of Gulf Coast residents (Kunzelman, 2010; London, 2010; Mulvihill, 2010; Navarro, 2010). Workshops were held to determine what research might be conducted and how to proceed. A meetin g was convened for June 22 23, 2010 in New Orleans by the the long term healt h effects on GOS cleanup workers, residents of Baldwin County, Alabama, and other local areas (National Institute of Environmental Heal th Science, 2010; Grattan, Roberts, Mahan, McLaughlin, Otwell, & Morris 2011). However, researchers have neglected the Northwest Florida area in the mental health impact studies completed or proposed to this point. While the impact of the GOS was apparent for these residents, workers and business owners, the effects of the GOS hav e not been explored in Escambia County, Flo rida The impact on this area of Florida was as profound as any other area and affected residents in different ways ensuring this study will add to the literature on the GOS.
20 Purpose of the Study The purpose of this study was to assess the effect of the G ulf Oil Spill on business owners, residents and workers in Escambia County, Florida. The study assessed differences in outcomes for business owners, residents and workers as well as identified predictors of resilience, psychological stress, and resources. All residents of the county were eligible to complete an online survey designed to measure resource loss, resilience, and psychological stress. The information gathered may be useful in deciding what services may serve community members to decrease nega tive mental and physical health outcomes associated with technical disasters. The information can also be used as a tool to assist in allocating RESTORE Act ( King & Berry, 2012) funds to projects that may help local residents the most. Rationale for the M ethodology This study investigated the effects of the GOS on business owners, residents, and workers in Escambia County, Florida. The study used an online survey and employed quantitative data gathering and analysis consisting of a resource loss scale, th e 21 Item Depression Anxiety Stress Scale (DASS 21) (Lovibond & Lovibond, 1995), the 14 Item Resiliency Scale (RS 14) ( Wagnild, 200 9 ) and demographic information. Because there has been no research conducted in the Escambia County, Florida area on the ef fect of the GOS, one open ended question was also used to allow business owners, residents, and workers to express anything they wanted to on the subject of the GOS, resulting in qualitative analysis of the responses to that question. Previous research on natural disasters, technological disasters, and other oil spills has addressed different areas. Some research has attempted to determine the impact on mental health of disasters using models and theories to assist communities
21 ( Arata, Picou, Johnson, & McN ally 2 000; Becker, 1997; Freedy Shaw, Jarrell, & Masters, 1992). Previous disaster research has focused on those directly impacted by the disaster such as the fishermen in the Exxon Valdez oil spill or those losing their homes due to Hurricane Katrina ( Arata et al., 2000; Picou & G ill, 1996; University of New Hampshire, 2011). While being directly impacted by a disaster is important to study, it is also important to research the larger community to assess how widespread the impact really is and how a di saster impacted the lives of more community members (Littleton, Kumpula, & Orcutt, 2011). Thus it was important for this study to include a broader spectrum of community members to assess how the GOS affected them. Three groups of Escambia County, Florida residents were selected for this study to assess the impact of the GOS on economic, physical, or mental well being: business owners, residents, and people who work in the county (workers). Residing in an area makes a person a stakeholder in the community People tend to take pride in the place they call home and feel a sense of community. Working or being employed in an area also makes an individual a stakeholder in the community even if they live elsewhere. While a worker may not actually reside in th e county, they none the less spend a considerable amount of time in the area earning their living. Business owners are invested in the community and may have selected the location of their business based on environmental and economic factors. The communi ty in which a business owner operates provides the resources for the business venture ( Byers, Kist, & Sutton, 1997). Despite the type of business, business owners are stakeholders in the community and depend upon the community to survive and stay in busin ess. Each of these groups of stakeholders may have been affected in some way by the GOS disaster. The study
22 sought to determine how a disaster such as the GOS affected these dif ferent groups and in what ways. The research questions presented below repres ent an exploration of the effects of the GOS on Escambia County, Florida business owners, residents and workers. Research Questions The following research questions were examined in the study: RQ 1 : Is there a relationship among resource change, psychologic al stress, and resilience ? RQ 2 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change predict scores on the DASS 21 subscales and total ? RQ 3 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age and resource change predict scores on the RS 14 subscales and total ? RQ 4 : Are there differences in the responses of business owners, res idents, or workers on resource change DASS 21 subscales, DASS 21 total RS 14 subscales, and RS 14 total ? RQ 5 : Are there differences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales and RS 14 total for Escambia County? RQ 6 : Are there differences based on claim status for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales and RS 14 total for Escambia County? Overview of the Study The remainder of the study is organized into four chapters. Chapter 2 provide s a review of the relevant literature. Chapter 3 provides an overview of the methods used in the study. Chapter 4 provides the research results of the data analysis. The final chapter include s a discussion of the major findings, limitations, and implica tions of the results.
23 CHAPTER 2 REVIEW OF LITERATURE The literature reviewed for this study addresses the Conservation of Resources Theory (COR) of psychological stress as it relates to the Gulf Oil Spill (GOS) and its effects on business owners, resident s and workers of Escambia County, Florida. Literature describing the COR Theory is reviewed, as well as literature about other theories of psychological stress. Disasters, including natural and technical or man made, the effects of disasters, and researc h on effects of disasters are also examined. The GOS and research completed or proposed is presented, as well as the claims process and its history. In addition, literature about psychological stress and resilience is presented and reviewed and related t o disasters. The GOS was a complex technical disaster impacting a large geographical area encompassing five states and the entire Gulf of Mexico. While there were similarities in the way regions were affected by the GOS, there were also differences. Five different states with different localities, geography and environments were directly impacted by the GOS. It was necessary to have different ways to help local communities, maximize gains, and minimize negative effects from the GOS. A large sum of money has and will be distributed to state and local governments impacted by the spill. If no one as ks about who was impacted by the GOS there is the possibility of missing the opportunity to help people who may have never accessed mental health care before, an d they may not know how to access the help available There are millions of dollars at stake and one chance to help those in need. Difficult economic times make it necessary to turn this disaster into an opportunity to learn how to help those in physical, mental, and financial need. This study examined the effects of the GOS on business owners,
24 residents, and workers in Escam bia County, Florida and how the GOS impacted their lives, earning a living, and quality of life. Conservation of Resources Theory Co nservation of Resources (COR) theory (Hobfoll, 1988; 1989) is a theory of stress integrating environmental factors and internal processes (Hobfoll, 2001). COR theory addresses the concepts of human nature and the human condition in relationship to psychol ogical stress (Hobfoll, 2001; Quick & Gavin, 2001). The basic concept of COR is individuals are motivated to obtain, retain, and protect resources in their lives (Hobfoll, 1988; 1989; 1998; 2001). Resources are divided into four categories: objects, cond itions, personal characteristics, and energies (Hobfoll, 1989). Individuals experience psychological stress when their resources are threatened, lost, or they have failed to gain resources after an initial resource investment (Hobfoll, 1989; 2011; Hobfoll & Lilly, 1993). 2001). Primacy of resource loss means resource loss outweighs resource gain and has a greater impact than resource gain (Hobfoll, 2001; 2012). The secon d principle of COR theory is people must make a resource investment to gain resources, protect themselves from resource loss, or to recover from a resource loss (Hobfoll, 2012). Those with more resources are less vulnerable to resource loss because they c an use other resources to try and gain resources, and those with fewer resources are more vulnerable to resource loss because they have fewer resources to use to try and gain resources (Hobfoll, 2012). Hobfoll developed the Conservation of Resources Evalu ation (COR E; Hobfoll, 2001) to assess the loss of resources and the impact of loss of resources. In the
25 original form, the COR E is 74 items. Freedy, Shaw, Jarrell, and Masters (1992) and Littleton, Kumpula, and Orcutt (2011) successfully adapted the CO R E to measure resource loss for their own research purposes. The modification was necessary because the original scale was long and the researchers were looking to measure a specific kind of resource loss rather than all four categories of resource loss. Originally applied to individuals, Hobfoll has also applied COR theory at the community level because communities can objectively identify and share threats to resources (Hobfoll, 2012). COR theory has been used at the individual and community levels to understand stress, trauma, disaster, burnout, and other psychological stressors better (Arata et al. 2000; Bonanno, Galea, Bucciarelli, & Vlahov, 2007; Curan, Totenhagen, & Serido, 2010; Freedy et al. ,1992; Picou & Gill, 1996; Picou, Marshall, & Gill, 20 04; Sattler, Preston, Kaiser, Olivera, Valdez, & Schlueter, 2002; Sumer, Karanci, Berument, & Gunes, 2005). The successful application of COR theory in different research studies across different fields adds to its credibility and usefulness. Theories of stress such as COR theory, are not new to the literature. Schwarzer (2001) delineated the response based, stimulus based, and cognitive transactional paradigms of stress research. The cognitive relational theory of stress developed by Lazarus (1966) fal ls into the cognitive transactional paradigm and emphasized cognitive appraisal of stressful events (Schwarzer, 2001). There are similarities between COR theory and cognitive relational theory, such as addressing resource loss as a cause of psychological stress. However, cognitive relational theory views objective resources as antecedents and subjective resources as direct cause of stress (Lazarus, 1966; Schwarzer, 2001). COR theory places more emphasis on the objective resource status
26 and change of reso urce status. Negative change leads to stress. COR theory takes resources and context and builds a robust theory of stress allowing for the further development of coping modes (Quick & Gavin 2001; Schwarzer, 2001). It is for these reasons the COR theory was selected over cognitive relational theory for this study. Disasters There are different types of disasters. A disaster is considered a large, negative event resulting in loss of life, damage, or severe hardship, all of which can be considered resourc es as defined by Hobfoll (1988). Disasters either derive from nature, man, or a combination of both. Natural disasters include hurricanes, tornados, earthquakes, and tsunamis and derive from nature. Technical or man made disasters include terrorist atta cks, mass shootings, the GOS or the Exxon Valdez oil spill. A natural disaster may also be complicated by man made problems such as the failure of the levy system during Hurricane Katrina in New Orleans. A technical disaster can have an environmental impa ct but they are different in cause and often in scope. Disasters can have devastating effects on people and entire communities and regions. Death is a possibility in any natural disaster and sometimes in technical disasters, too. Depending on many diff erent factors, disasters can create a wide range of problems with varying degrees of impact and time frames. For example, an initial earthquake may be followed by aftershocks, leading to further destruction, injuries, or even deaths in a community. A dis aster can have a limited impact in a compact geographical area such as a sinkhole damaging a single house or a disaster can have a wide impact over a vast geographical area such as the tsunami in 2005 displacing or affecting 4.35 million people and killing 176,685 people in Southeast Asia (Carbello, Heal, & Horbaty, 2006). Research studies have been conducted on natural and
27 technical disasters over time. A discussion of some disaster research is addressed in the following sections. Natural Disasters Carb allo, Heal, and Horbay (2006) researched the tsunami that struck Southeast Asia on December 26, 2005. This tsunami was a large scale, multi national disaster and is considered to be the most serious natural disaster in recorded history. The researchers f ound individuals impacted by the disaster were resilient, able to cope under extreme duress, and attempted to seek normalcy quickly after the tsunami. The researchers also noted there was little monitoring of the situation and tsunami data may not be repr esentative of individuals affected by the tsunami due to the large geographical area and diverse populations The authors suggested disaster response planning needed to be made more situation and culture specific, take into account what type of support or help people may need based on culture, and who should provide this support. The researchers also identified what they considered to be vulnerable populations, including children, the elderly, the disabled, and women. Health workers and other relief/aid workers were also identified as needing support and training to cope with burnout associated with giving aid post disaster. Sumer, Karanci, Berument, and Gunes (2005) examined the predictive power of personal resources, earthquake experience, coping, and self efficacy on distress, intrusion, and avoidance symptoms of survivors of the 1999 Marmara earthquake in Turkey. This earthquake was considered devastating registering 7.4 on the Richter scale and killing an estimated 18,000 people. Of the survivors, many were displaced and forced to live in tent cities for extended periods of time because of the damage to infrastructure. The r esearch found those people experiencing more material and human
28 loss, those threatened by the earthquake and women were more vulnerable to general distress and symptoms of posttraumatic stress disorder (PTSD). The coastal zones of the southern U. S. are very familiar with hurricanes. Hurricanes bring the potential for high winds, tornados, and flooding leading to the destructio n of property and possible loss of life. Hurricane season is six months long and often results in multiple storms of varying strengths making landfall along the Gulf and Atlantic coasts. Freedy, Shaw, Jarrell, and Masters (1992) researched the short term adjustment following Hurricane Hugo in South Carolina using the Conservation of Resources ( COR ) theory. The authors found the following: resource loss was positively related to psychological distress, resource loss was a better predictor of psychological distress than personal characteristics and coping behavior, and resource loss was a risk factor for developing clinically significant psychological distress. Sattler, Preston, Kaiser, Olivera, Valdez, and Schlueter (2002) conducted a cross national study examining preparedness, resource loss, and psychological distress in the United States Virgin Islands, Puerto Rico, Dominican Republic, and the United States (U.S.) after Hurricane Georges. They found in each location Acute Stress Disorder symptoms were a ssociated with personal characteristic resource loss and low social support. The researchers also found object resource loss in the U.S. Virgin Islands and in the U. S. accounted for a significant portion of Acute Stress Disorder symptom variance, as did energy resource loss in Puerto Rico, basic resource loss in the Dominican Republic, and condition resource loss in the Dominican Republic and the U. S. The findings supported the COR theory of stress as well as previous research, and lend credibility to t he idea of resource spirals. Resource spirals are described as
29 the continued loss of resources after a disaster and can be exacerbated by secondary stressors. Secondary stressors, such as additional stressful events, strains, and complications, can happe n after disasters. Secondary stressors can deplete personal characteristic, energy, and condition resources, which can lead to further psychological stress. On September 16, 2004, Hurricane Ivan made landfall in the vicinity of the Alabama Florida state l ine as a Category 3 hurricane. The destruction caused by the storm was concentrated in Escambia, Santa Rosa, and Okaloosa counties in Florida (Bureau of Beaches and Coastal Systems of the Florida Department of Environmental Protection, 2004). Ivan claime d 14 lives in Florida and disrupted thousands of resident s lives by destroying homes, businesses and infrastructure, such as the Interstate 10 Bridge over Escambia Bay. It is estimated Ivan caused over 13 billion dollars in damages in the United States ( NOAA Technical Memorandum NWS NHC 6, 2011) It took years for the Northwest Florida area to heal from the physical damage of Ivan and some residents still suffer negative psychological effects from the storm ( NOAA Technical Memorandum NWS NHC 6, 2011; Rug giero, Amstadter, Acierno, Kilpatrick, Resnick, Tracy & Galea, 2009 ). Ruggiero, Amstadter, Acierno, Kilpatrick, Resnick, Tracy, and Galea (2009) researched self rated health in relation to disaster characteristics, social resources, and post disaster out comes in adults experiencing the 2004 Florida Hurricanes of Charley, Frances, Ivan, and Jeanne. The research was conducted in the 33 counties in the path of the hurricanes. The study found poor self rated health was associated with older age, extreme fea r during the storm, low social support, and depression. Social support
30 and depression are variables possibly mitigated by targeted interventions after disasters. Poor self rated health is of concern because self rated health status has been related to mo rbidity, mortality, and impairment in social and occupational functioning. The authors suggested targeting these modifiable variables post disaster may allow for improved access to community resources and may reduce the long term economic burden for the i ndividual and society. Technical/Man made Disasters Littleton, Kumpula, and Orcutt (2011) used the COR theory to examine whether psychological resource loss predicted posttraumatic stress disorder (PTSD) symptomology in college women following a campus sh ooting at Northern Illinois University (NIU). The researchers gathered data immediately following the shooting and then again 8 months later. The results supported resource loss as a predictor of PTSD symptomology in the immediate aftermath of the shooti ng and 8 months later. Of note in this study was the finding the whole community was affected by the trauma, not just those who witnessed the shooting or were fired upon. The researchers also found even after 8 months, 12% of the sample was still reporti ng PTSD symptomology. This suggests many individuals may need services many months following a disaster or trauma. The Exxon Valdez oil spill (EVOS) in Alaska in March of 1989 released approximately 42 million liters of oil into Prince William Sound devas tating local fisheries and contaminating the coastal ecosystem (Picou & Gill, 1996). Until the Gulf Oil Spill, the EVOS had been considered the largest and most damaging spill in North American history (Picou, Marshall & Gill, 2004). The environmental and s ocial impact from the EVOS were immediate and long term. Several research studies were conducted on the
31 EVOS from different theoretical perspectives, including studies on the political context and management of the EVOS and social impact of the EVOS (Pi cou & Gill, 1996). Picou and Gill (1996) applied the COR model to the EVOS and attempted to evaluate the long term psychological impact of the EVOS. The researchers used an ex post facto design and compared data gathered from three comparable communiti es; two in the impact zone of the EVOS and one outsi de of the impact zone. Data were collected via face to face interviews, telephone surveys and mail surveys. The focus of the research was on indicators of chronic stress, community type and occupational role. The Impact of Events Scale (IES) was used to measure levels of psychological distress. Community, occupational, and demographic comparisons were made. Using the Mann Whitney U statistic to test whether the comparison groups were selected from the same population, the researchers found that there were negative psychological impacts of the EVOS and elevated stress levels in the communities impacted by the EVOS. Arata, Picou, Johnson, and McNally (2000) applied the COR model to the EVOS to measure mental health function six years post disaster. They surveyed 125 commercial fishers in Cordova, Alaska anonymously via mail and attempted to measure economic and social impacts of the spill, as well as coping and psychological functioning. Using Pearso n correlations and multiple regression, the researchers found although there was no loss of life or damage to personal possessions caused by the EVOS, there were higher levels of depression, anxiety and PTSD symptoms in those residents who were directly im pacted by the EVOS as compared with study participants from another similar location, with no EVOS exposure. They also found empirical support for the relationship between resource loss and chronic psychological symptoms.
32 This research suggests that the GOS the focus of this study, may have an effect on the residents and business owners of the Gulf Coast for many years to come. Picou, Marshall, and Gill (2004), building on past research using the COR model, used structural equation modeling (SEM) to lin k litigation to chronic psychological stress of the EVOS. They found being a plaintiff in civil litigation increased psychological stress, and could lead to a large negative c ommunity impact. The results of their research suggests fighting for damages an d payment from BP as a result of the GOS could possibly result in long term psychological distress for the residents and business owners of Northwest Florida. Disasters, whether natural or man made, are known to have a potential negative impact (physicall y, mentally, and financially) on individuals and communities. Individuals live and work in their communities. If individuals are suffering from symptoms of stress or PTSD, they are unable to perform socially or professionally to the best of their ability. The stress reactions have potential of spreading throughout families, businesses, and communities. According to COR theory, once resource loss starts it may continue into a spiral of loss leading to more and more loss and compounding the impact of loss The loss spiral can be on an individual level as well as a community level (Hobfoll, 1989, 2001) The literature suggests psychological reaction to a natural disaster or a technological man made disaster is different and technological disasters can have longer term mental health impact than natural disasters (Picou & Gill, 1996). Technological disasters are often characterized by prolonged uncertainty and litigation,
33 with chronic negative psychological impact on the communities affected by the disaster ( Arata et al. 2000; Marald, 2001 ; Palinkas, 2012 ). The Gulf Oil Spill (GOS) The GOS is the largest accidental oceanic oil spill and the largest man made disaster in American history (Bourne, 2010; Mulvihil, 2010; National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011 ). The GOS began April 20, 2010 and continued for 47 days when a cap was placed on the underwater well spewing an unknown amount of oil and gas into the Gulf of Mexico. Visible damage was apparent very early o n the disaster and continued even as a massive cleanup effort was made by BP and local communities along the Gulf Coast. Two years after the well was capped, most of the beach cleaning effort has stopped because visible signs of the oil were gone on most o f the beaches although not all. However, whenever a storm forms in the Gulf, oil and oil residue are churned up and brought ashore all along the Gulf Coast (Blair, 2012). Even when the well was capped, the oil stopped spilling, and the cleanup was mostly complete, there were still on going individual and class action lawsuits awaiting resolution in the court system (Sole, 2011). It is thought these lawsuits may be settled soon although what is considered soon varies from person to person. Billions of dol lars are at stake and most of these dollars are intended for communities directly impacted by the GOS (King & Berry, 2012). Escambia County, Florida is one of these communities. As the GOS began, the importance of mental health care for the local communit ies, especially fishermen and those in the tourist trade, were made (London, 2010). There was also an immediate need for research on mental health and some research was conducted toward the beginning of the spill. Grattan, Roberts, Mahan, Jr.,
34 McLaughlin Otwell, and Morris Jr. (2011) studied the early psychological impact of the GOS in a comparison study of Baldwin County, Alabama (direct exposure to oil) and Franklin County, Florida (no direct exposure to oil). The researchers attempted to determine th e level of distress, mechanisms of coping, and perceived risk in counties directly and indirectly impacted by the GOS and sought to identify the extent of economic resource loss to explain these factors. The study found there were no significant differenc es in community groups for distress, adjustment, neurocognition, or environmental worry. Both communities had residents with clinically significant depression and anxiety. The researchers felt the psychological impact of the GOS might be much broader tha n just those areas directly on the water. They also pointed out income loss related to the GOS may have a greater psychological impact than the presence of oil on the shoreline. The National Institute of Environ mental Health Sciences (NIEHS; 2011) is cond ucting the Gulf Long Term Follow Up Study for Oil Spill Clean Up Workers and Volunteers, or GuLF STUDY. The study is a long term physical and mental health study of oil spill clean up workers and volunteers helping with the clean up after the GOS. There w as simply not enough information about the physical and mental health ramifications of an oil spill of this size and the study is an effort to understand the health consequences faced by all exposed to the spill whether worker or resident. The GuLF STUDY began on February 28, 2011 and is expected to continue for ten years. The GOS had an immediate impact on some areas and a slower impact in others. In Escambia County, Florida, the oil spill occurred in the middle of the spring break season just before th e summer tourist season was about to begin. Businesses,
35 residents, and workers watched the news as the oil came closer and closer to Escambia County shores reaching them on June 4, 2010. Prior to the first tar balls arriving, hotels and rental agencies o n Pensacola Beach and Perdido Key were already receiving cancellations for summer vacation rentals. These cancellations continued throughout the summer and into the fall (Dixon, Wolfgram, Dehart, & Devonshire, 2010). Local waterways were blocked off with giant lengths of boom trying to prevent the oil and tar from entering the Intracoastal Waterway, Perdido Bay, Pensacola Bay, and Escambia Bay through the Pensacola Pass or the Perdido Pass. In spite of this effort, oil and tar did come in to the Pensacol a and Escambia Bay areas, as well as other areas. The closing of the waterways meant that boats could not go in and out. This was not an issue for most fishermen because most local waters were already closed to fishing. Early on in the disaster, BP set as ide money to be dispersed to individuals and businesses with direct loss claims due to the GOS. Claims offices were set up in different locations to make the process more accessible. Individuals and businesses were told to bring in documentation (taxes, pay stubs, etc.) to document the difference in their income in 2010 from previous years (Sole, 2012). Individuals and businesses also had to have some sort of direct relation to the beach or shoreline, tourist industry or service industry to qualify for a claim, but sometimes they did not (Weaver, 2012). There were complaints almost immediately about fraudulent claims being paid, large sums of money being paid to restaurant and bar industry workers with little or no proof of previous income. Legitimate cl aims were being denied by overwhelming bureaucracy and paperwork required to file a claim (Elliot, 2012). The claims process was fraught
36 with difficulties and appeared to be unfair (Elliot, 2012; University of New Hampshire, 2011). As mentioned before, th ere was great concern, too, about the physical and mental health of the local residents. Residents were told not to go into the water on certain days and not to touch the oil and tar balls washing up on the shore. Residents also complained of smelling pe troleum fumes when they were close to the water and dead and dying wildlife washed up on shore (Bourne, 2010; Dixon et al., 2010; Mulvihill, 2010). Residents were told not to eat local seafood and they were told it was safe to eat the seafood again. Ther e was a lot of uncertainty about what was safe and what was unsafe and very little reliable information available on which to base decisions (Casselman, 2011; Goldstein, Osofsky, & Lichtveld, 2011). Resilience There are many definitions of resilience altho ugh many of them are similar (Wagnild, 2009). In this study, resilience refers to the ability of people to maintain a healthy psychological and physical functioning after being faced with an adverse or traumatic event (Bonanno, 2004). Resilience is a hea lthy reaction to a disturbing or stressful event (Flynn & Norwood, 2004). Wagnild (2009) stated resilience is made up of five characteristics, the Resilience Core (RC). The RC consists of a purposeful life, perseverance, equanimity, self reliance, and exi stential aloneness. Meaning or a purposeful life is the realization a person has something to live for. Perseverance is the act of continuing to try in spite of adversity or discouragement. Equanimity is the ability to have a balanced perspective on lif e and experiences and to roll with the punches of life without having extreme responses to adversity. Self reliance is the ability to depend on oneself and recognize personal strengths and weaknesses. Existential aloneness is
37 the realization we are all l iving our own unique lives and only we can live our life. These five characteristics constitute the subscales of the Resilience Scale and the 14 Item Resilience Scale (Wagnild & Young, 1993). Bonanno (2005) reviewed research on resilience and found after a traumatic event, resilience is the most common outcome and there are multiple factors that may support a resilient outcome. The study also found there is a difference between adult resilience and childhood resilience. Adult resilience is usually in re sponse to a single event though not always (Bonanno, 2005). Bonanno (2004) also distinguished between recovery and resilience, stating recovery implies a departure from normal functioning and then a return to normal functioning and resilience indicates th e ability to maintain normal functioning after a traumatic event. Bonanno, Galea, Bucciarelli, and Vlahov (2007) studied the association between resilience and socio contextual factors (gender, age, race/ethnicity, education, level of trauma exposure, inc ome change, social support, frequency of chronic disease, and recent and past life stressors) in New York City after the September 11, 2001 terrorist attack. The study found resilience was a mix of various factors including person centered variables and s ocio contextual variables as well as risk and protective factors. Research on the concept of resilience is important because it can inform policy makers and responders about what the focus of interventions and assessments should be in order to support res ilient reactions to traumatic events in individuals and throughout communities. Research has also found widespread psychological treatment following a traumatic event is not helpful and can actually cause more stress and is
38 harmful to some individuals (Bo nanno et al., 2007). Assessments and treatment need to be tailored to the individual in order to support a resilient reaction. Psychological Stress Psychological stress is a negative reaction brought about by an event (or events) causing loss of resources threatening loss of resources, or causing no resource gain after investment of resources (Hobfoll, 1989). As previously stated, there are theories of stress focusing on perception, cognition, and appraisal, as well as those focusing on stimulus definiti ons of stress (Hobfoll, 1989; Monroe, 2008). Psychological stress is an important phenomenon to understand because it can lead to mental and physical health problems including: major depression, anxiety, high blood pressure, and heart disease (Amstadter, Broman Fulks, Zinzow, Ruggiero, & Cercone, 2009; Monroe & Reid 200 9 ). Specific negative affective states such as depression, anxiety, and stress have been associated with trauma and disasters (Becker, 1997; Freedy et al 19 92; Littleton et al. 2011). A psychological response to a disaster is normal; however, some people who experience a disaster may exhibit higher levels of distress (Flynn & Norwood, 2004). Lovibond and Lovibond (1995) developed the Depression Anxiety Stress Scales (DASS) to measure th e specific negative affective states of depression, anxiety, and stress which are related and yet different from each other. Depression is defined as a loss of self esteem and incentive, anxiety is defined as an enduring state of fear, and stress is defin ed as persistent tension and arousal (Lovibond & Lovibond, 1995; Lovibond, 1998). Without intervention, these negative affective states may worsen over time leading to clinical psychopathology such as major depression, generalized anxiety disorder, or eve n PTSD (Flynn & Norwood, 2004; McNally, Bryant, & Ehlers, 2003).
39 There has been discussion and debate over the assessment and treatment of psychological stress following a traumatic incident such as a disaster. Some researchers have suggested everyone exp osed to a disaster should be treated with critical incident stress debriefing (CISD) (Phipps & Byrne, 2003) while others have suggested this is not necessary and may actually be harmful (Bonanno et al., 2007). Research has been conducted on participatio n in trauma or disasters r esearch and whether or not participation in research causes added stress or trauma to the participants. Griffin, Resick, Waldrop, and Mechanic (2003) studied the impact of trauma research participation on trauma survivors. They examined participant reactions to trauma assessment in domestic violence, rape, and physical assault cases. The results indicated trauma survivors were able to tolerate participation in the trauma research and some trauma survivors found participation in the research a valuable experience. Ferrier Auerbach, Erbes, and Polusny (2009) also found participants in trauma related research tolerated trauma related questionnaires without being overwhelmed and stressed. They may even find they derive a benefit fro m participating in the research. These studies indicated participants in a survey about a stressful event such as the GOS should not experience further stress as a result of their participation in the research and they may derive a benefit. Even with all of the previous research on disasters, and specifically on oil spills, the GOS is such a different phenomenon because of how it began, how long it lasted and its scale. The possibility exists that there may be other variables of interest in the context o f the GOS to study, making the addition of a qualitative research question prudent (Wang, 2008). It is possible that the objective reality of the GOS for the people
40 of Escambia County, Florida is not discoverable, but in asking if they have anything else to say about the GOS, they are given the chance to voice their expe rience in their own words. The open ended question option may provide further information about variables of interest for future research and it may also provide a healthy outlet for the p eople of Escambia County, Florida to let their thoughts and feelings about the GOS be known. Summary The review of the literature yielded several pertinent ideas including the lack of a study of this nature has not been conducted in the Northwest Florida community since the GOS began on April 20, 2010. The literature review also found the COR model has been applied to the impact of the EVOS in Alaska and to research about the impact of hurricanes and other natural disasters in various locations. Previous ly there has not been an incident such as the GOS to study. Escambia County, Florida has a history of dealing with hurricanes. Relying on data gathered from previous research focusing on post natural disaster mental health is insufficient to address the n eeds of the people impacted by the GOS. The GOS is considered a technological disaster because the root cause of the disaster was man made ( Mabus, 2010; National Commission on the BP Deepwater Horizon Oil Sp ill and Offshore Drilling, 2011 ) There will be a prolonged uncertainty and litigation in the lives of residents, business owners, and workers related to the GOS that will last for years. Chapter 2 has presented the literature associated with and supporting the questions and hypotheses posed for this s tudy. Chapter 3 will present the methodology to address the questions and hypotheses.
41 CHAPTER 3 METHODOLOGY The purpose of this study was to examine the perceptions of Escambia County, Florida business owners, residents, and workers about the Gulf Oil Spi ll ( GOS ) More specifically, the study sought to identify how residents coped with resource change, psychological stress, and resilience. This chapter presents the methodology used in the study including the design of the study, study variables, study po pulation and sampling procedures, instrumentation, data collection and data analysis procedures, research hypotheses, and methodological limitations. Research Method and Design A quantitative research method was chosen over a qualitative or mixed method t o meet the needs of the study. When attempting to identify predictors or differences between groups, numerical data and a quantitative method is an appropriate choice over qualitative ( Cooper & Schindler, 2008 ). Cooper and Schindler identified a quantitat ive methodology as being beneficial when working with larger samples, removing potential researcher bias, and applying the results to larger populations. Qualitative methodology using interview or observations might have benefits onal stories, reactions, and interpretations of the GOS might be useful. However, bias is always a possibility when researchers are in direct contact with the research participants as is required in qualitative research (Cooper & Schindler, 2008). It might also have been difficult to obtain the breadth of responses to the GOS through a limited number of interviews with individuals. A mixed method study might have allowed for the use of the best of quantitative and qualitative research methodologies. Howev er, a mixed method study might not
42 allow variables to be measured accurately and findings of the study might not apply to other populations in disasters (Cooper & Schindler, 2008). A quantitative approach was deemed the most appropriate method to include a ll different groups of Escambia County residents involved in the GOS. Research Design The quantitative cross sectional survey design selected for the study was appropriate, as the variables are measurable (Ary, Jacobs, & Razavieh, & Sorenson, 2009). Quan titative survey methods are appropriate for gathering information from a large number of participants about the GOS. In quantitative research, the research questions or hypotheses are specific to gather measurable and clear d ata on variables (Creswell, 200 9 ). The study investigated responses about the effects of the GOS using a cross sectional survey design methodology utilizing electronic Internet technology. In cross sectional research, data are obtained at one point in time from respondents of differen t ages or in different stages of development in their personal and professional lives and in this study reflected how the GOS had affected businesses, residents, and workers. Cross sectional research is an alternative to longitudinal research or following participants for long periods of time. An advantage of cross sectional research is that sample attrition is not an issue, as the data are collected at one point in time (Gall, Borg, & Gall, 1996). The cross sectional approach was much less expensive given the short time span of the study (Salkind, 2003). The surveys were administered to a group of Escambia County, Florida residents. The study was also descriptive by exploring and describing the effects of the GOS on the lives of Escambia County residents. W hile descriptive studies are simple in
43 design and execution, they can yield important data and information for informing policy and the direction of future research (Gall et al., 1996). Quantitative survey data was collected to investigate relationships a mong resilience, psychological stress, resource change, and demographic variables. Survey or questionnaire collection of data from a wide variety of sources in a timely and concise manner is relatively easy (Dillman, 2007). Various methods of survey data c ollection include personal interviews, telephone interviews, mailed questionnaires, and directly administered questionnaires (Ary et al., 2009). Regardless of the method chosen, the six basic steps involved in conducting survey design research are planning defining the population, sampling, constructing the instrument, conducting the survey, and processing the data (Ary et al., 2009). Numerous innovations in survey design, data collection, and methodology have emerged, beginning with the use of the telepho ne for data collection (Dillman, 2007). Other recent innovations have included the use of computers and the Internet. Using the Internet to conduct surveys has changed survey methodology. E mail or Web surveys eliminate the costs associated with postage, p aper, mailing, and data entry. The Internet makes it possible to overcome international boundaries, increase sample size, and significantly shorten the time required to collect data (Dillman, 2007). Learning the software necessary to construct an Internet system for collecting data used to be time consuming and difficult (Dillman, 2007). Currently there are a number of websites designed to make conducting surveys and collecting data online easier. The www.surveymonkey.com website was used for this study a nd permitted the researcher to format a survey. The SurveyMonkey site provided the ability to collect both text and numerical data and return the data in a usable format. Respondents were
44 able to provide informed consent by clicking a box and since no name s were used, all responses were confidential (http://www.surveymonkey.com). Respondents were also able to complete demographic data, such as age, gender, and other information. Internet surveys also give the researcher control over which items and the numb er of items a respondent can see at any time (Dillman, 2007). Population and Sampling The study surveyed adult residents (over the age of 18) in Escambia County, Florida. Participants had to be living or working in Escambia County at the time of the GOS (April 20, 2010) or since. The U.S. Census Bureau (2012) estimated the population of the county to be 299,144 in 2011 and the survey was open to any resident of the county over the age of 18 years of age. Since there was no exact list of residents, it wa s necessary to inform residents of the study and solicit their participation. Several methods were used to solicit participants for the study. Fliers were taken to local restaurants, bars, churches, and businesses across the county advertising the survey a nd asking for participation. A website was ( www.gosproject.org ) developed with information about the study and had a direct link to the survey. There was a page on www.faceboo k.com advertising the survey and directing respondents to the survey link. The Independent News, a local online and print newspaper, wrote a blog post to generate interest in the survey. Information and links to the survey were published on multiple list servs in the community, including the Pensacola Fishing Forum and www.northescambia.com Repeated efforts were made by the researcher to solicit participation in the study. Cohen (1992) noted it would be necess ary to solicit a participant pool of 102 participants for a 7 predictor multiple regression study with an alpha=.05, a medium effect, and power = .80. The analysis of variance hypotheses
45 would require between 45 and 64 individuals per group with an alpha = .05, medium effect size, and power = .80. It was hoped approximately 300 individuals would volunteer to participate in the study. Informed Consent The study used a consent procedure to inform voluntary participants of the study. Participants were infor med of the purpose of the research, the time involved, assessment of minimal risk and benefits to participants, contact for questions about the research, and contact information for questions about their rights as a research participant. This information w as provided in the informed consent on the survey website and participants had to consent prior to being allowed to view the survey. When respondents linked to the electronic survey site, they were asked to check a box on the survey form indicating they ha ve been provided informed consent to participate in the study ( see Appendix A for a copy of the consent form ). All participants were informed that their participation was voluntary. Confidentiality Every research participant in a study has a right to pri vacy and the expectation the data will be kept confidential at all times. The right to privacy and confidentiality was disclosed to research participants prior to the start of a study. Research participants have a right to expect respect for autonomy, tru st, scientific integrity, and fidelity. Every research participant has the right to expect there will be no chance of being identified by name at any time, before, during, or after the study. No personally identifying information or data was collected and data was only reported in an aggregated format. Creswell (2009) suggested the fundamental role for ethical research is to do no harm, including physical, psychological, social, economic, or legal harm. Participants were
46 also informed they had the option n ot to complete the survey; however, their participation was appreciated and would make an addition to the study. Internet surveys do not collect personally identifying information and it was impossible to identify any individual since identification numb ers were randomly assigned as the surveys were completed through www.surveymonkey.com. The electronic surveys had no identifying information on them and were stored as an encrypted file. The encrypted file will be saved on a compact disc in a separate sto rage office. After a period of three years, the electronic disc containing the data will be shredded in a crosscut shredder. Instrumentation The survey consisted of the measure of resource change, the 21 Item Depression Anxiety Stress Scale s (DASS 21), the 14 Item Resiliency Scale (RS 14), a demographic questionnaire, and one open ended question. Each instrument used in the collection of data is presented separately. A complete copy of both types of the survey (business owner and resident/worker) can be found in Appendix B and C Resource Change (2001) COR E and adaptions made to the COR E used by other researchers (Freedy et al ., 1992 ; Littleton et al., 2011), and conversations with individuals about adverse economic impact. The measure of resource change was the same for residents, businesses, and workers since the events or situations remaine d the same for someone living or working in Escambia County, Florida. The wording for businesses, residents, or workers was consistent depending upon the point of view of the respondent. One
47 item on the business resource change ( I had to fire employee(s )) was omitted because it did not match the questions for residents and workers. The resource change items ranged from lost a job/business to defaulted on a loan, had to apply for government assistance/aid Items were both positive and negative or a measure was primary to resource gain, and resource gain still required the investment of resources, thus making the listing of resource change a measure of change. It is not a valuation of res ource loss or gain, whether objective or subjective. A resource change is an indicator a business owner, resident, or worker may have felt psychological stress according to COR theory. The resource change items were answered with a Yes (1) or No (0) and w ere summed to create a resource change score. 21 Item Depression Anxiety Stress Scale s The 21 Item Depression Anxiety Stress Scale s (DASS 21; Lovibond & Lovibond, 1995) consists of 3 subscales Depression (DASS D), Anxiety (DASS A) and Stress (DASS S) and a total score. The DASS 21 has 21 items utilizing a response scale of None of the time (0), some of the time (1), most of the time (2), and all of the time (3). None of the items are reverse scored and the Cronbach alpha for the total score has bee repor ted as a=.966. The Depression subscale measures depression as a loss of self esteem and incentive as well as the possibility of not attaining life goals and consists of 7 items. The reported Cronbach alpha is a=.947. The Anxiety subscale emphasizes an acut e response to fear and nervousness. There are 7 items in the Anxiety subscale with a Cronbach alpha of a=.879. The Stress subscale also has 7 items and purports to measure tension and a low threshold for being upset and frustrated. The reported Cronbach al pha was a=.933. The DASS 21 total score and
48 subscales were previously analyzed using factor analysis to lend validity to the scale and subscales (Lovibond & Lovibond, 1995). 14 Item Resilience Scale The 14 Item Resilience Scale (RS 14; Wagnild & Young, 19 93) was designed to assess the concept of resilience and the five characteristics of the Resilience Core. Resilience is thought to be a characteristic allowing individuals to overcome negative aspects of life such as stress. The characteristics or subscale s include: living a purposeful life, perseverance, equanimity, self reliance, and existential aloneness. Participants respond to the 14 statements using a 7 point Likert response scale of Strongly disagree (1) to Strongly agree (7). The RS 14 was develope d as a shorter form of the original 25 item Resilience Scale (RS) developed by Wagnild and Young (1993). The RS was originally tested on 782 middle aged and older adults and had strong internal consistency ( =0.91) as well as good concurrent validity. Th e RS 14 and the RS are strongly correlated and the internal consistency alpha for the RS 14 was 0.89 (Waginald, 2009). Reliability was not reported for the characteristic subscales. Demographic Questionnaire A demographic questionnaire designed to gather information regarding individual characteristics of the study participants was included. The questionnaire collected information regarding age, sex, ethnic/cultural identification, marital status, completed level of education, living situation, type of re sidence, employment status, household income group, and hours worked per week. The demographic items can be found in the business or resident / worker survey in Appendix B and C Respondents were also asked about their claim status for compensation from BP, including filing a claim, claim being denied, approved, joining a class action suit, satisfaction with amount from BP,
49 accepting payment, rejecting payment, involved in litigation, and no claim with BP. Each of these was answered as a check indicating they had some claim with BP. Respondents were also asked to rate their physical health, mental health, finances, and relationships on a 8 point scale of No Problem (0) to Severe Problem (8) for both before and after the GOS. Open ended Question An open ende d question was also included to allow respondents to make any comments about the GOS not addressed on the survey. The open ended question, Is there anything else you would like to tell us about the impact of the Gulf Oil Spill on your life? was included b ecause no one had asked the people of Escambia County, Florida what the impact of the GOS has been on their lives. It was also an opportunity to learn about other factors impacting the lives of Escambia County residents. Data Collection Solicitation of s tudy participants and data collection commenced after receiving permission to conduct the study from the University of Florida Internal Review Board. After permission was received, the researcher began contacting organizations, media, and set up a website to solicit study participants. There was no list of individuals to whom surveys could be sent necessitating using multiple means to solicit potential participants for the study. The process for participants involved hearing about the study, accessing a com puter, and linking to the survey website. Cards and flyers were prepared and handed out at various organizations and places of business to inform potential participants and to provide the website address. When participants connected to the survey website, it was necessary for them to read and agree to the informed consent prior to being taken to the survey. Participants were asked to identify
50 themselves as a business owner, resident, or worker in order to send them to the correct survey. After that selecti on, they were asked their zip code to ensure they lived or worked in Escambia County. At this point participants were taken to the business, resident, or worker survey. Upon completion of the study the participant was thanked and the data recorded. After repeated attempts to increase sample size the survey was closed and the data downloaded for analysis. Data Analysis The data analysis for the study occurred in a series of steps. The first step was to inspect the data using descriptive statistics (mean, median, mode, and frequency) to gain a general understanding of the data. Cronbach Alpha Reliability Coefficients were calculated for each total and subscale scores of the three instruments. The factor structure of each of the scales was assessed using a principal components factor analysis with a varimax rotation. These psychometric properties had been determined previously but the instruments used in this study had not been administered to this particular population previously and reliability and validit y can be person and situation specific (Ary et al., 2002). Once the preliminary data analysis had been completed and the subscales determined to be valid and reliable, a mean score for each viable subscale was calculated and used in further analysis. A sta tistically significant relationship was defined as p =.05 or less. Research Question 1 The firs t research question and null hypothesis were as follows: RQ 1 : Is there a relationship among resource change, psychological stress, and resilience?
51 Ho 1 : There w ill be no relationship among resource change, psychological stress, and resilience To address question 1, a Pearson Product Moment correlation was used to test for the strength of the relationships among the total resource change score, the total DASS 21 score and the total RS 14 score. A correlation is a measure of the strength of the relationship among variables and a correlation coefficient can range from 1.0 to 0 to +1.0. The closer to 1.0 the correlation coefficient is, the stronger the relationshi p. Correlations can be positive or negative In a positive correlation, as one variable goes up the second variable also goes up and in a negative correlation as one variable goes up the second variable goes down (Ary, et al,. 2009). Hopkins (2009) provid ed a useful means to evaluate the strength of a correlational relationship and it will be used in this study. The scaling of correlation coefficients is as follows: _____________________________________________________________________ Correlation Coefficie nt Descriptor _____________________________________________________________________ 0.0 0.1 trivial, very small insubstantial tiny, practically zero 0.1 0.3 small, low, minor 0.3 0.5 moderate, medium 0.5 0.7 large, high, major 0.7 0.9 very large, very high, huge 0.9 1.0 near, practically, or almost perfect, distinct, infinite _____________________________________________________________________
52 Research Questions 2 and 3 Multiple linear regression was used to test t he null hypotheses posed for research questions 2 and 3. The research questions and null hypotheses are as follows: RQ 2 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and re source change predict scores on the DASS 21 subscales and total ? Ho 2 : There will be no predictive relationship among the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change and the DASS 21 subscales and total RQ 3 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change predict scores on the RS 14 subscales and total ? Ho 3 : Ther e will be no predictive relationship among the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age and resource change and the RS 14 subscales and total Regression analysis is not caus al in nature and its purpose is the development of an equation for predicting values on a Dependent Variable (DV). Research Questions 1 and 2 and their null hypotheses are predictive in nature. Simple linear regression involves a single Independent Variab le (IV) and a single Dependent Variable (DV). The goal of simple regression is to create a linear equation that can predict the value of the DV if there is a value for the IV. In multiple regressions, a set of predictor Independent Variables are selected ( IVs) as potential predictors of a DV as is the case in this study. Multiple regressions are an extension of simple linear regression involving more than one predictor variable. It is used to predict the value of a single DV from a weighted linear combinati on of IVs. One problem with multiple regressions may be the existence of multi collinearity. Multi collinearity is a problem arising when there are moderate to high inter correlations among predictor variables. The problem lies with the possibility there may be two or
53 more variables that are measuring essentially the same information (Mertler & Vanatta, 2005). Nothing is gained by adding variables to a regression analysis measuring the same thing but multi collinearity can cause real problems with the anal ysis itself. Stevens (1992) pointed out three reasons why multi collinearity can cause problems. They include: (a) multi collinearity limits the size of the response since the IVs are going after much the same variability in the DV; (b) multi collinearity can cause difficulty because individual effects are confounded when there is overlapping information; and (c) multi collinearity tends to increase the variances of the regression coefficients resulting in unstable prediction equations. The simplest method of diagnosing multi collinearity is to investigate high inter correlations among the IV predictor variables. A second method is to inspect the Variance Inflation Factor (VIF), an indicator of the relationships among predictors (Stevens, 1992). Stevens also noted VIF values greater than 10 are generally cause for concern. The data for all regression analyses was checked to ensure multi collinearity did not present a problem in the analysis. If multi collinearity did exist, a variable may be deleted or variab les may be combined to create a single construct. The data for the regression analysis was also checked to ensure compliance with the assumptions of regression. The assumptions of regression include: (a) the independent variables are fixed (the same value s would be found if the study were replicated), (b) the IVs are measured without error, (c) the relationship among the IVs and the DV is co linear, (d) the mean of the residuals for each observation on the DV is zero, (e) errors on the DV are independent, (f) errors are not correlated with the IV, (g) variance across all values of the IV is constant, and (h) errors are normally distributed
54 (Mertler & Vanatta, 2005). The assumptions were inspected through examination of residual scatter plots, assessment of linearity, inspection of normality through skewness, kurtosis, and Kolmogorov for homosecdasticity (Mertler & Vanatta, 2005). Multiple regressions were the statistical analysis to be used for Research Qu estions 2 and 3 and were appropriate for use in predictive studies. A stepwise multiple regression method often referred to as statistical multiple regression, was used When there are multiple predictor variables, a statistical multiple regression may b e used to determine which specific IVs contribute to the model (Mertler & Vanatta, 2005). Forward, stepwise, and backward are methods of entering and keeping variables in the model. In using a stepwise selection method, at each step tests are performed to determine the significance of each IV already in the equation. If a variable were entered into the analysis and is measuring much the same construct as another, a reassessment of the variables may conclude the first variable is no longer contributing anyth ing to the analysis. In a stepwise selection procedure, the variable would then be dropped out of the analysis even though it might have been a good predictor at one time. The variable may not be found to provide a substantial contribution to the model (Me rtler & Vanatta, 2005). After ascertaining the data was appropriate for regression analysis and checking for multi collinearity, the multiple regression procedure was completed using a probability level of p =.05 as the level of significance. A stepwise sel ection multiple regression procedure was used to develop the model for these questions and analysis.
55 Research Questions 4, 5, and 6 Research question 4, 5, and 6 used analysis of variance (ANOVA) as the analytical technique and were as follows: RQ 4 : Are th ere differences in responses of business owners, residents, or workers on resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales and RS 14 total? Ho 4 : There are no differences in responses of business owners, residents, or workers on resource c hange, DASS 21 subscales, DASS 21 total, RS 14 subscales and RS 14 total. RQ 5 : Are there differences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales and RS 14 total for Escambia County ? Ho5 : There will be no di fferences in based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County RQ6 : Are there differences in responses based on claim status for resource change, DASS 21 subscales, DASS 21 to tal, RS 14 subscales and RS 14 total for Escambia County ? Ho 6 : There will be no differences in responses based on claim status for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County Analysis of varianc e (ANOVA) was used as the statistic to address hypotheses 4, 5, and 6. The data was assessed to ensure the data met the assumptions of ANOVA. The assumptions of ANOVA are: observations are independent of one another, a normal distribution occurs, and an e quality of variance exists. Generally, ANOVA is robust with respect to violations of the assumptions (Field, 2009). The dependent variables will be mean response to the resource change scale, the total and subscale scores of the DASS 21, and the total an d subscale scores of the RS 14 scale. A probability level of p = .05 was the criterion for failing to reject or rejecting the null hypotheses.
56 Methodological Limitations In the study, quantitative research methods were used to address the research questio ns posed for the study. Quantitative methods were effective in gathering a large amount of data without affecting efficiency (Johnson & Turner, 2002). A limitation of the study was the method of sampling. Soliciting participants through various media may h ave limited access to a broader population had names and emails of county residents been available. Generalizing to other locations or disasters may be difficult given the procedures used in this study. A second limitation was the use of volunteers in the study. Volunteers tend to be different that non volunteers (Ary, et al., 2009). Volunteers have a need for approval, may be highly educated, have an agenda they want to pursue, or may seek social approval (Ary et al. 2009 ). Care was taken in interpreting the findings of the study. The honesty of the study participants was also a limitation. The questions can always be asked as to whether survey respondents are as honest as possible or whether they replied with what they thought was a socially acceptable re sponse. Study Validity In descriptive survey research with no treatment and control or comparison group, validity and reliability of results remain important. Study validity can be internal or external. External validity is the generalizability of the st udy results to another group of participants, in another place, or at another time. To what other populations, settings, or measurement variables could the findings of the study be generalized? Ary, Jacobs, Sorenson, and Razavieh (2009) identified three types of external validity: population, ecological, and external validity of operations. Population validity concerns identifying other populations to which the findings of a study are generalizable, and depends upon
57 how subjects were selected for a study In the present study, the population was Escambia County, Florida residents, therefore generalizing to other populations in different areas would be difficult. No threat from interaction between subjects and treatment was present; however, using volunt eers presents a problem. Volunteers may have characteristics not typical of the population and their motivation to participate is unknown. Why non volunteers do not choose to participate and how they would have answered survey items is also unknown (Ary et al., 2009). Ecological validity involves generalizing the study results to other situations. Careful consideration of the environment in which the research is done is necessary. In the present study, threats to ecological validity (pre testing, nove lty effect of a new treatment, or attitudes developed over the course of the study) did not present problems. Neither pre testing nor treatments were conducted and the study was of a short enough duration not to affect the attitudes and perceptions of the participants. External validity of operations addresses how the study was conducted with specific operational definitions. Would the same results be expected with different investigators using different operational definitions or measurement procedures? The constructs for investigation are derived from the literature review and were pilot tested. Campbell and Stanley (1963) defined internal validity as the extent to which extraneous variables are controlled by the researcher. Extraneous variables are those variables possibly affecting the outcomes of a study. Eight factors affecting internal validity are of concern in the present study: history, maturation, testing, instrumentation, statistical regression, differential selection, experimental mortalit y, and selection maturation interaction. History might present a problem, as outside events might
58 influence the views of participants. The occurrence of outside organizational, world, or natural events is beyond the control of the researcher, but any occu rrence will be noted in the study if necessary. Maturation was assumed to have little influence due to the short time line of the study and the adult age of participants whose developmental sequence is not as rapid as in young children. Testing was not a problem as there was no pre testing. The two instruments used in the study have been used previously with established and acceptable validity and reliability. With no repeated measures, statistical regression to the mean w as not a problem nor was diffe rential selection. adults and not likely to change over the short time of the study. Experimental mortality (subjects dropping out of the study) and a low response rat e were problems in the present study. The open ended questio n included in the survey was treated as qualitative data and m odified content analysis was used as the analysis technique. Theme s present in the writing were iden tified and reported. Themes were n ot determ ined prior to analysis but emerge d from the responses of participants. Summary Chapter 3 has articulated the methodology used in addressing the research problem and questions stated for the study. The research method was described and why quantita tive methods were selected for the study. The research design was a survey, the population was described, as well as the method of accessing and soliciting participants for the study. The instruments used to measure psychological stress, resilience, resour ce change, and descriptive variables were articulated, as well as the procedure for collecting the data. Data analysis to address the questions and
59 hypotheses posed for the study were described. The chapter closed with methodological limitations and study validity. Chapter 4 presents the results of the data collection and analysis.
60 CHAPTER 4 RESULTS Chapter 4 presents the results of the analysis of the data collected for this study. The purpose of this study was to assess the consequences o f the British P etroleum oil spill in 2010 on businesses, residents, and workers in Escambia County Florida. The G ulf Oil Sp i l l (GOS) survey was administered to volunteer business owners, residents, and workers in the county. The GOS survey assessed resource change psyc hological distress, resilience, impact, and demographic variables. The participants in the survey are presented and described followed by the psychometric properties of the instruments used in this study and the analyses of the research questions posed f or this study. There were six researc h questions posed and they were as follows: RQ 1 : Is there a relationship among resource change, psychological stress, and resilience ? RQ 2 : Do the demographic variables of income group, employment status, hours worked pe r week, marital status, level of education, age, and resource change predict scores on the DASS 21 subscales and total? RQ 3 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age and resource change predict scores on the RS 14 subscales and total? RQ 4 : Are there differences in the responses of business owners, res idents, or workers on resource change, DASS 21 subscales, DASS 21 total RS 14 subscales, and RS 14 total ? RQ 5 : Are there d ifferences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County? RQ 6 : Are there differences based on claim status for resource change, DASS 21 subscales, DASS 21 total, RS 14 subs cales, and RS 14 total for Escambia County? Data for this study was collected electronically using the SurveyMonkey website ( www.surveymonkey.com ) The survey was open to any and all residents of Escambia
61 County Florida and was publicized through interviews with the media, local listservs, networking, and access to other media resources. Data collection took longer than expected and yielded fewer respondents than desired. The plan was to have at least 300 partic ipants if possible; however, after data collection started it became obvious recruiting survey participants was more difficult than expected. Lack of media coverage, newspapers not responding, and radio stations not being willing to make public service an nouncements made publicizing the survey difficult and limited the number of survey responses. A local weekly paper ran an online article about the survey and published a link to the website with the survey. Local businesses were approached with flyers a nd business cards from the Alabama line on Perdido Key (the farthest point in the county to the southwest) to Pensacola Beach. Most people were receptive to the information about the survey and expressed interest; however, some people were hostile to the survey. The researcher was asked to leave a yacht sales shop without them even hearing about the survey. Another issue encountered during data collection was that business owners, residents, and workers thought they had to have a loss from the GOS to part icipate in the survey. The researcher had to explain the purpose of the study was to measure the impact of the GOS and all responses were of interest and valuable. Other respondents expressed some paranoia about the survey process and how the data collec ted would be used and presented. Potential respondents thought the researcher was working for the government, BP, a law firm, or another entity involved in the oil spill to whom the information gathered from the survey would be presented.
62 I t was necessary to constantly publiciz e the study and survey during the data collection p rocess and w hen there were no more respondents the survey was closed and the data downloaded T here were a number of individuals connecting to the survey and completing the consent but not completing any of the survey items I ndividuals not completing any items other than the consent were deleted from the study. There were also a number of places where respondents did not complete each and every item. When respondents missed items the data cells were marked as missing and were not included in the data analysis. Missing was treated as missing data and no effort was made to impute values for the missing data as this was perceived to be putting words into the mouths of respondents. The re was a reason why items were not complete d although the reason is unknown. Thus, there may be a different number of responses for different scales, subscales, and variables. Sample Demographics A total of 151 people participated in the study. There wer e 23 business owners (15.8%), 86 residents (57.0%), and 37 workers (25.3%) in the group of respondents in Escambia County, Florida. Respondents not identifying whether they were a worker, resident, or business owner and not providing their zip code were n ot included in the analysis resulting in a total sample of 146 individuals P articipants ranged from 18 83 years of age (M= 46.06 SD= 13.06 ). The average age of b usiness o wners was 45.05 (SD=12.34, n=21), r esidents were on average 46.32 years of age (SD=1 3.88, n=76) and w orkers were 46.10 years of age (SD=11.73, n=30 ). Respondents consisted of males (n =55, 42.6%) and females (n=74, 57.4%). Table 4 1 presents the de scriptive data for age by group and Table 4 2 present gender by group.
63 The participants in this study were nearly all Caucasian. This was true across the groups of business resident, and worker. Table 4 3 illustrates the number and percentage for each group by ethnic group Most respondents had a high school diploma or GED as only 2 people ind icated common across business owners, residents and workers (n=42, 32.8%), followed by finally Do ctorate or Professional degrees (n=5, 3.9%). Table 4 4 illustrates the number and percentage for each group by education level Most people responding to the survey were employed in a full time job (n=85, 65.9%). Table 4 5 illustrates the number and per centage for each group by employment status. There was a wide range of household income represented by the survey ranged from $71,201 to $110,000 (n=34, 27.0%). Table 4 6 illustrates the number and per centage for each group by household income Most respondents work 40 or more hours per week (n=44, 34.6%) or 40 or less hours per week (n=39, 30.7%). Table 4 7 illustrates the number and per centage for each group by hours worked per week Responde nts were most frequently married or partnered (n=80, 62.0%), single (n=22, 17.1%) or divorced (n=19, 14.7%). Respondents described their living situation spouse/partner and children
64 most frequent residence listed across all groups (n=110, 85.3%), followed by condominium/townhouse (n=10, 7.8% (n=1, 0.8%). Reliability and Validity of Instruments Reliability can be person and situation specific (Ary et al., 2009) and it was important to assess the reliability of the scales used in this study for this particular group of respondents. The 21 Item Depression Anxiety Stress Scale s (DASS 21; Lovibond & Lovibond, 1995) and the 14 Item Resilience Scale (RS 14; Wagnild & Young, 1993) as well as the 18 item GOS measure of resource change scale were used i n this study. Cronbach alpha reliability coefficients were calculated for the total scale or subscales for each instrument used in this study. Reliability was not available for the RS 14 subscales and reliability was not available for the GOS measure of re source change or the Before and After GOS scale as these were new scales developed and used in this study. Inspection of the reliability coefficients presented in Table 4 8 indicate the study alpha for the RS 14 total was slightly higher than previous repo rts and comparable with previously reported reliability coefficients for the DASS 21 total and subscales. The reliability for the GOS measure of resource change and the Before and After GOS were moderately high and acceptable. The Before and After GOS scal es were also compared using a paired t test to determine if the respondents indicated there were differences in their Before and After GOS physical health, mental health, finances, and relationships. Results of the paired t test indicated there were signif icant differences between the Before (M=.8320, SD=.9966) and After (M=1.6536, SD=1.804) GOS scores, t (127) = 6.499, p=<.001.
65 Results Prior to commencing the statistical analysis of the data collected for this study, it was necessary to code the data an d create subscale scores. The instructions for the DASS 21 and the RS 14 were followed as articulated in Chapter 3. However, a change needed to be made to the measure of resource change. There were 19 items on the business owner version of the survey an d 18 matching items on the resident and worker surveys. The extra item ( I had to lay off an employee(s) on the business survey was deleted from the analysis to answer the research questions. The extra item deletion on the business owners survey resulted in 18 matching items for the business owners, residents, and workers. The items in the resource scale with a yes response were summed to create a score for each individual and used in further analysis. Research Question 1 The first research question and n ull hypothesis was as follows: RQ 1 : Is there a relationship among resource change, psychological stress, and resilience? Ho 1 : There will be no relationship among resource change, psychological stress, and resilience. To test this hypothesis a Pearson produ ct moment correlation was calculated for scores on the measure of resource change, the DASS 21, and the RS 14. Results indicate there was a significant negative relationship between the RS 14 and DASS 21; however, it was statistically very small, as was t he negative relationship between the RS 14 and resource change. The strongest relationship was between resource change and the DASS 21 (r=.513). Table 4 9 presents the results of this analysis. Research Question 2 Research question and null hypothesis 2 were as follows:
66 RQ 2 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change predict scores on the DASS 21 s ubscales and total ? Ho 2 : There will be no predictiv e relationship among the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change and the DASS 21subscales and total The DASS 21 consists of three subscales and a total s cale score and each subscale and the total scale served as the dependent variable or predicted variable. The subscales include depression, anxiety, and stress. The predictor or independent variables were income group, employment status, hours worked per we ek, marital status, level of education, age, and resource change score. Each of the subscales is addressed followed by the DASS 21 total score. A stepwise multiple regression was used to address the subscales and total DASS 21 scale null hypothesis 2. Mult i collinearity was assessed for each subscale and total scale and was not present in this analysis. The variance inflation factor (VIF) and tolerance were well within limits noted by Stevens (2009). The first subscale to be tested was the depression subsc ale of the DASS 21. The results of this analysis indicated there was a statistically significant 2 step model developed, R =.476 R 2 =.227, R 2 adj = .213, F (1, 117) = 6.748, p =<.011. Resource change and employment status accounted for 22.7% of the variance i n total resilience. Resource change accounted for 18.2% and employment status added 4.5% of the variance in the final model. The null hypothesis was rejected for resource change and employment status; however the study failed to reject the null hypothesis for income group, hours worked, marital status, level of education, and age. Table 4 10 presents the stepwise model summary and Table 4 11 presents the coefficients for this analysis.
67 The second DASS 21 subscale addressed was anxiety. Results of the step wise regression analysis indicated there was a statistically significant two step model developed for anxiety, R =.478 R 2 =.229, R 2 adj = .216, F (1, 117) = 10.865, p =.001. The model accounted for 22.9% of the variance in anxiety with resource change accounti ng for 15.7% and employment status adding 7.2% to the total variance. The null hypothesis was rejected for resource change and employment status ; however, the study failed to reject the null hypothesis for income group, hours worked, marital status, level of education, and age. Table 4 12 presents the step wise model summary and Table 4 13 presents the coefficients for this analysis. The third DASS 21 subscale addressed was stress. Results of the stepwise regression analysis indicated there was a statistic ally significant two step model developed for stress, R =.611 R 2 =.373, R 2 adj = .362, F (1, 117) =3.933, p =<.050. The model accounted for 37.3% of the variance in stress with resource change accounting for 35.2% of the variance and hours worked adding 2.1% o f the variance. The null hypothesis was rejected for r esource change and hours worked. The study failed to reject the null hypothesis for income group, employment status, marital status, level of education, and age. Table 4 14 presents the stepwise model summary and Table 4 15 presents the coefficients for this analysis. The DASS 21 total scale scores were also tested to identify statistically significant predictors of the total scale score. Results of this analysis indicated there was a statistically si gnificant two step model for predicting the DASS 21 total score, R =.551 R 2 =.303, R 2 adj = .291, F (1, 117) = 5.884, p =<.017. The two step model accounted for 30.3% of the variance in DASS 21 scores and resource change (26.8%)
68 and hours worked adding 3.5% of the variance in DASS 21 total scores. Table 4 16 presents the step wise model summary and Table 4 17 presents the coefficients for this analysis. The demographic variables of income group, employment status, hours worked per week, marital status, level o f education, age, and resource change were used to predict scores on the DASS 21 total and subscales of depression, anxiety, and stress. Statistically significant models were found for depression, anxiety, stress, and the total DASS 21 scale score. It is i nteresting to note resource change was a pr edictor in each model. Table 4 18 presents a summary of the predicted variable and the significant predictors for each model. Research Question 3 The third research question and null hypothesis was as follows: RQ 3 : Do the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change predict scores on the RS 14 subscales and total ? Ho 3 : There will be no predictive relationship among t he demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change and RS 14 subscales and total The dependent or predicted variables were self reliance, meaning, equanimity, p erseverance, existential aloneness, and RS 14 total score. The independent variables were income group, employment status, hours worked per week, marital status, level of education, age, and resource change. A stepwise multiple regression was used to addre ss the subscales and RS 14 total scale null hypothesis 2. Multi collinearity was assessed for each subscale and total scale and was not present in this analysis. The variance inflation factor (VIF) and tolerance were well within limits noted by Stevens (20 09).
69 The first subscale to be tested was the self reliance subscale of the RS 14. The results of this analysis indicated there was a statistically significant 3 step model developed, R =.370 R 2 =.137, R 2 adj = .114, F (1, 116) = 3.931, p =.050. Income, age, a nd marital status accounted for 13.7% of the variance in self reliance. Income accounted for 6.4%, age added 4.4%, and marital status added 2.9% of the variance in the final model. The null hypothesis was rejected for income, age, and marital status ; howev er, the study failed to reject the null hypothesis for income group, hours worked, employment status, level of education, and resource change. Table 4 19 presents the step wise model summary and Table 4 20 presents the coefficients for this analysis. The se cond subscale to be tested was the meaning subscale of the RS 14. The results of this analysis indicated there was a statistically significant 1 step model developed, R =.326, R 2 =.106, R 2 adj = .099, F (1, 118) = 14.022, p =<.001. Income accounted for 9.9% of the variance in meaning. The null hyp othesis was rejected for income; however, the study failed to reject the null hypothesis for age, marital status, hours worked, employment status, level of education and resource change. Table 4 21 presents the step wi se model summary and Table 4 22 presents the coefficients for this analysis. The third subscale to be tested was the equanimity subscale of the RS 14. The results of this analysis indicated there was a statistically significant 1 step model developed, R =.2 39, R 2 =.057, R 2 adj = .049, F (1, 118) = 7.119, p =.009. Income accounted for 5.7% of the variance in equanimity. The null hypothesis was rejected for income ; however, the study failed to reject the null hypothesis for age, marital status, hours worked, empl oyment status, level of education and resource change. Table 4 23
70 presents the step wise model summary and Table 4 24 presents the coefficients for this analysis. The fourth subscale to be tested was the perseverance subscale of the RS 14. The results of t his analysis indicated there was a statistically significant 1 step model developed, R =.345, R 2 =.119, R 2 adj = .112, F (1, 118) = 15.970, p =<.001. Income accounted for 11.9% of the variance in perseverance. The null hypothesis was rejected for income ; howev er the study failed to reject the null hypothesis for age, marital status, hours worked, employment status, level of education, and resource cha nge. Table 4 25 presents the stepwise model summary and Table 4 26 presents the coefficients for this analysis. The fifth subscale to be tested was the existential aloneness subscale of the RS 14. The results of this analysis indicated there was a statistically significant 1 step model developed, R = .336, R 2 =.113, R 2 adj = .105, F (1, 118) = 15.006, p =<.001. Income a ccounted for 11.3% of the variance in existential aloneness. The null hypothesis was rejected for income ; however, the study failed to reject the null hypothesis for age, marital status, hours worked, employment status, level of education and resource cha nge. Table 4 27 presents the step wise model summary and Table 4 28 presents the coefficients for this analysis. The sixth analysis was to test for predictors of the RS 14 total score. The results of this analysis indicated there was a statistically signifi cant 1 step model developed, R = .320, R 2 =..102, R 2 adj = .095, F (1, 118) = 13.443, p =<.001. Income accounted for 10.2% of the variance in total RS 14. The null hypothesis was rejected for income ; however, the study failed to reject the null hypothesis for age, marital status, hours worked,
71 employment status, level of education, and resource change. Table 4 29 presents the step wise model summary and Table 4 30 presents the coefficients for this analysis. The demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change were used to predict scores on the RS 14 total and subscales of self reliance, meaning, equanimity, perseverance, and existential aloneness. Statistically significa nt models were found for self reliance, meaning, equanimity, perseverance, and existential aloneness, and the total RS 14 score. It is interesting to note income was a predictor in each mode l. Table 4 31 presents a summary of the predicted variable and the significant predictors for each model. Research Question 4 Research question and null hypothesis 4 were as follows: RQ 4 : Are there differences in responses of business owners, residents, or workers on resource change, DASS 21 subscales DASS 21 total R S 14 subscales and RS 14 total ? Ho 4 : There are no differences in responses of business owners, residents, or workers on resource change, DASS 21 subscales DASS 21 total and RS 14 subscales and RS 14 total Research question four was addressed through the use of a univariate ANOVA. The independent variables were business owners, residents, or workers and the dependent variables were the three subscale scores of the DASS 21 (Depression, Anxiety, and Stress), the DASS 21 total score, the five subscale score s of the RS 14 (Self reliance, Meaning, Equanimity, Perseverance, and Existential Aloneness), the RS 14 total score, and resource change. Results of the ANOVA analysis of the total scal e and subscales of the R S 14 DASS 21 and resource change indicated th ere were statistically significant differences
72 aming business owners residents, and workers for the stress subscale of the DASS 21 scale, F (2, 128) = 4.117, p=.018 and resource change, F ( 2, 134) = 10.611, p=<.000 Business owners had a hig her mean scor e on stress (M=6.64, SD=5.45 ) than did residents (M=3. 76, SD=3.94) or workers (M=5.03 SD= 4.15 ). The Bonferroni post hoc test for stress indicated business owners differed significantly from residents (p=.019) but residents did not differ from workers (p=. 472), and workers did not differ from business owners (p=.524) For resource change, the Bonferroni post hoc test indicated business owners differed significantly from residents (p=<.000) and workers differed significantly from business owners (p=.025), b ut residents did not differ from workers (p=.227). The null hypothesis was rejected for DASS 21 stress and resource change ; however, the study failed to reject the null hypothesis for DASS 21 total, depressi on, and anxiety. Table 4 32 presents the means and standard deviations for the statistically significant findings. Research Question 5 Research question 5 and null hypothesis 5 were as follows: RQ 5 : Are there differences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 1 4 subscales, and RS 14 total for Escambia County? Ho 5 : There will be no differences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County. Research question 5 asked about differenc es in Escambia County respondents by income group on the scales and subscales used in the study. The independent variable was income group and the dependent variables were depression, anxiety, stress, DASS 21 total score, self reliance, meaning, equanimity perseve rance, existential aloneness, RS 14 total score, and resource change. Data collected for income was
73 grouped into 8 groups from $0.00 to 230,001 or more dollars in income. There were too few individuals in the higher income groups, $150,000 190,000 ( n =7), $190,001 230,000 ( n =3), and $230,001 ( n =4), and these 3 groups were collapsed into one group of $150,000 and higher ( n =14) for this analysis resulting in 6 income groups. Results of the ANOVA analysis of the scales and subscales of the DASS 21, R S 14, and resource change indicated there were statistically significant differences by income group for 4 of the subscales of the RS 14 (meaning, equanimity, perseverance, and existential aloneness), the RS 14 total score, resource change scale and the nu ll hypothesis was rejected. There were no statistically significant differences for the self reliance subscale, the DASS 21 total score, the depression, anxiety, or stress subscales therefore the study failed to reject the null hypot hesis for these variab les Table 4 33 summarizes the results of this analysis. Income group 5 ($110,001 150,000) had a higher mean score on meaning ( M =18.77, SD =3.34) and income group 6 ($150,001 and higher) had a higher mean score on meaning ( M =19.36, SD =1.60) than did incom e group 1 ($0 $22,757; M =14.30, SD =3.34). Income groups 5 and 6 also had higher mean scores on equanimity, perseverance, and RS 14 total than did income group 1. Income groups 4, 5, and 6 had higher mean scores on existential aloneness than did income gr oup 1. Table 4 34 presents the means and standard deviations for the statistically significant findings. Table 4 35 illustrates the mean resource change for respondents by income group. It is interesting to note the lower income groups experienced more r esource change as opposed to those with higher incomes. This finding indicates those with
74 lower incomes indicated they had more loss, los s of job, and loss of income than did those with higher incomes. The Bonferroni post hoc test for meaning indicated t hat income group 5 (p=.027) and income group 6 (p=.005) differed significantly from income group 1, but there were no other significant differences among the income groups. For equanimity, the Bonferroni post hoc test indicated income group 5 (p=.038) and income group 6 (p=.027) differed significantly from income group 1. There were no other significant differences among income groups. For perseverance, income group 5 (p=.040) and income group 6 (p=.032) differed significantly from income group 1 with no other significant differences among income groups. For existential aloneness, income group 4 (p=.029), income group 5 (p=.002), and income group 6 (p=.001) differed significantly from income group 1 with no other significant differences among income grou ps. For the RS 14 total, income group 5 (p=.039) and income group 6 (p=.018) differed significantly from income group 1 with no other significant differences among groups. Research Question 6 Research question 6 and the null hypothesis were as follows: RQ 6 : Are there differences based on claim status for resource change, DASS 21 su bscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County? Ho 6 : There will be no differences based on claim status for resource change, DASS 21 subscales, DA SS 21 total, RS 14 subscales, and RS 14 total for Escambia County. Research question six asked if there were differences in the responses based on those making a claim and those not making a claim. There were 122 individuals making no claim (80.8%) and 29 individuals participating in this survey making a claim (19.2%). The data was analyzed using a univariate ANOVA. The independent variable was claim
75 or no claim and the dependent variables were the three subscale scores of the DASS 21 (depression, anxiety, and stress), the total DASS 21 score, the five subscale scores of the RS 14 (self reliance, meaning, equanimity, perseverance, and existential aloneness), the total RS 14 score, and resource change. Results of the ANOVA analysis of the scales and subscale s of the DASS 21, RS 14, and resource change indicated there were statistically significant differences for those with and without a claim for the stress subscale, the DASS 21 total score, and resource change. The null hypothesis was rejected for DASS 21 stress subscale, DASS 21 total, and resource change. The study failed to reject the null hypothesis for all of the RS 14 subscales, RS 14 total, and the DASS 21 subscales of d epression and anxiety. Table 4 36 summarizes these results. Those submitting a claim had a higher mean score on the stress subscale of the DASS 21 ( M =7.10, SD =4.49) than those with no claim ( M =3.84, SD =4.08). Those with a claim also had a higher mean score for the DASS 21 total ( M =14.48, SD =10.99) and resource change ( M =4.10, SD =2. 77) than those indicating no claim. Table 4 37 presents the mean and standard deviations for the statistically significant findings. Open E nded Question The open ended question was included because the measure of resource change, the DASS 21, and the RS 1 4 were very specific measures that may or may not have yielded information about the impact of the GOS on respondents, and the open ended question allowed all respondents the opportunity to tell their story about the impact of the GOS on their lives. Sixt y four respondents chose to answer this question. Sixty two (42.4% of the total respondents) respondents expressed concerns, fears, and frustrations about the oil spill, the clean up process, the claims process, environmental
76 damage, health effects, and e ffects on children. Themes identified in the analysis are presented below. One theme repeated throughout the responses to the open ended question was future, seafood, and the water of the Gulf of Mexico. the responses. A nger was expressed over the damage done to the environment, loss of life savings, and the claims process. There were a number of complaints mentioned in the responses of the 62 participants. There were physical health complaints (self and others; cancer; asthma; flu symptoms in children), environmental complaints (smelling fumes; seeing oiled wild life; use of oil dispersants), and loss of recreation complaints, such as being un able to swim, boat, or walk on the beaches. There were a number of complaints about the claims process. One respondent the GO S did not just impact the beach, but impacted the entire area. on the beach or who worked on the water (fishermen, charter boats Beach properties and rental management companies, etc.) it effect ed all businesses, no matter where they were located in Escambia or Santa Rosa County supplying businesses dependent Frustration at the process and seemingly unfair payouts to certain people were also mentioned. t appears individuals, such as
77 waiters and employees of impacted businesses had a much easier time of getting In spite of the resource change measure being included in the survey, econo mic complaints were also mentioned repeatedly. These complaints included losing customers, not making enough money, homes losing value, not being able to fish and eat seafood, losing life savings, and losing business opportunities. The clean up process al so brought about concerns. Some respondents stated they were hired as clean Others thought clean up workers were not actually working and clean up workers were bei ng told not to clean up hidden oil. Government and big business complaints were also prevalent including: trusting the government, losing faith in the government, and not trusting BP. Summary In this chapter, the results of a survey to assess the conseque nces o f the B P oil spill in 2010 on businesses, residents, and workers in Escambia County Florida were presented The participants in the survey were presented and described, as were the psychometric properties of the instruments used in this study. The s questions were answered by providing a detailed explanation of the results of the data analysis. In Chapter 5 the results will be discussed as well as the theoretical implications, practice implications, and study limitations. Additionall y, recommendations for future research will be presented.
78 Table 4 1. Study Participants Mean Age by Group Business Resident Worker Total M SD M SD M SD M SD Age 45.05 12.33 46.32 13.87 46.10 11.72 46.06 13.06 Table 4 2 Study Partici pants by Gender Business Resident Worker Total N % N % N % N % Male 18 81.8 27 35.5 10 32.3 55 42.6 Female 4 18.2 49 64.5 21 67.7 74 57.4 Table 4 3 Study Participants by Ethnic Group Business Resident Worker Total N % N % N % N % Black 0 0.0 2 2.6 1 3.2 3 2.3 Asian 0 0.0 3 3.9 1 3.2 4 3.1 White 21 95.5 68 89.5 29 93.5 118 91.5 Hispanic 1 4.5 2 2.6 0 0.0 3 2.3 Other 0 0.0 1 1.3 0 0.0 1 .8
79 Table 4 4 Study Participants by Education Level Business Resident W orker Total N % N % N % N % Less than High school 0 0.0 1 1.3 1 3.2 2 1.6 High School/GED 6 28.6 17 22.4 5 16.1 28 21.9 Degree 4 19.0 14 18.4 4 12.9 22 17.2 Degree 7 33.3 23 30.3 12 38.7 42 32.8 Degree 3 14.3 20 26.3 6 19.4 29 22.7 Doctorate/ Professional Degree 1 4.8 1 1.3 3 9.7 5 3.9
80 Table 4 5 Study Participants by Employment Status Business Resident Worker Total N % N % N % N % Full Time Job 21 95.5 40 52.6 24 77.4 85 65.9 Part Time Job 1 4.5 8 10.5 2 6.5 11 8.5 Multiple Part Time Jobs 0 0.0 2 2.6 1 3.2 3 2.3 Retired 0 0.0 8 10.5 1 3.2 9 7.0 Homemaker 0 0.0 8 10.5 0 0.0 8 6.2 Unemployed 0 0.0 6 7.9 3 9.7 9 7.0 Disabled 0 0.0 4 5.3 0 0.0 4 3.1
81 Table 4 6 Study Part icipants by Household Income Business Resident Worker Total N % N % N % N % 0 22,757 0 0.0 5 6.8 5 16.7 10 7.9 22,758 44,500 4 18.2 19 25.7 4 13.3 27 21.4 44,501 71,200 5 22.7 19 25.7 4 13.3 28 22.2 71,201 110,000 6 27.3 19 25.7 9 30. 0 34 27.0 110,001 150,000 3 13.6 5 6.8 5 16.7 13 10.3 150,001 190,000 2 9.1 4 5.4 1 3.3 7 5.6 190,001 230,000 0 0.0 2 2.7 1 3.3 3 2.4 230,001 and above 2 9.1 1 1.4 1 3.3 4 3.2
82 Table 4 7 Study Participants by Hours Worked per Week Business Re sident Worker Total N % N % N % N % 20 hours or less 1 4.5 9 12.2 2 6.5 12 9.4 30 hours or less 1 4.5 7 9.5 1 3.2 9 7.1 40 hours or less 6 27.3 21 28.4 12 38.7 39 30.7 More than 40 hours 14 63.6 17 23.0 13 41.9 44 34.6 Not Applicable 0 0.0 20 27.0 3 9.7 23 18.1 Table 4 8. Reliability Coefficients for Study Scales No. items Previous Study Study Mean Study SD RS 14 Total 14 .89 .955 81.17 14.746 Self reliance 5 N/A .880 29.78 5.299 Meaning 3 N/A .859 12.11 3.480 Equanimity 2 N/ A .717 11.00 2.430 Perseverance 2 N/A .777 11.37 2.510 Existential Aloneness 2 N/A .825 11.39 2.592 DASS 21 Total 21 .966 .961 10.19 11.214 Anxiety 7 .879 .885 2.25 3.346 Depression 7 .947 .930 3.26 4.269 Stress 7 .933 .916 4.63 4.379 GOS Scale to tal 18 N/A .780 1.19 2.012 Before GOS Scale 4 N/A .756 3.33 3.986 After GOS Scale 4 N/A .857 6.62 7.201
83 Table 4 9 Correlation Matrix Variables DASS 21 Total RS 14 Total Resource Change DASS 21 Total 1.00 RS 14 Total Resource Change .181* .5 1 3 ** 1.00 .115 1.00 p<.05 (two tailed), ** p<.0 1 (two tailed). Table 4 10 Model Summary for the DASS 21 Depression Subscale Step R R 2 R 2adj 2 F chg Df P 1 .343 .182 .175 .182 26.256 1, 118 <.001 2 .476 .227 .213 .045 6.748 1, 117 .011 Tab le 4 11 Coefficients for DASS 21 Depression Subscale B t p Bivariate r Partial r Resource Change .668 .408 4.995 <.001 .427 .419 Employment status .493 .212 2.598 .011 .248 .234 Table 4 12 Model Summary for the DASS 21 Anxiety Sub scale Step R R 2 R 2adj 2 F chg Df P 1 .397 .157 .158 .157 22.034 1,118 <.001 2 .478 .229 .216 .072 10.885 1,117 .001 Table 4 13 Coefficients for DASS 21 Anxiety Subscale B t p Bivariate r Partial r Resource Change .478 .379 4.574 <.001 .397 .389 Employment status .489 .269 3.296 .001 .392 .292 Table 4 14 Model Summary for the DASS 21 Stress Subscale Step R R 2 R 2adj 2 F chg Df P 1 .593 .352 .346 .352 34.002 1, 118 <.001 2 .611 .373 .362 .021 3.933 1, 117 .050
84 Table 4 15 Coefficie nts for DASS 21 Stress Subscale B t p Bivariate r Partial r Resource Change .985 .584 7.967 <.001 .593 .583 Hours Worked .562 .145 1.983 .050 .180 .145 Table 4 16 Model Summary for the DASS 21 Total Score Step R R 2 R 2adj 2 F chg Df P 1 .518 .268 .262 .268 43.227 1,118 <.001 2 .551 .303 .291 .035 5.884 1,117 .017 Table 4 17 Coefficients for DASS 21 Total Score B t p Bivariate r Partial r Resource Change 2.136 .501 6.467 <.001 .518 .513 Hours Worked 1.137 .188 2.426 .017 .233 .219 Table 4 18 Summary of Regression Analysis for DASS 21 Scale Predictor Predictor Depression Resource Change Employment Status Anxiety Resource Change Employment Status Stress Resource Change Hours Worked DASS 21 Total Resource Change Hours Worked Table 4 19 Model Summary for the RS 14 Self Reliance Subscale Step R R 2 R 2adj R 2 F chg Df P 1 .253 .064 .056 .064 8.059 1, 118 .005 2 .328 .108 .092 .044 5.716 1, 117 .018 3 .370 .137 .114 .029 3.931 1, 116 .050 Table 4 20 Coefficients for RS 14 Self Reliance Subscale B t p Bivariate r Partial r Income 1.142 .333 3.69 2 <.001 .253 .324 Age .115 2.78 2.941 .004 .137 .254 Marital status 1.121 .181 1.983 .050 .081 .181
85 Table 4 21 Model Summa ry for the RS 14 Meaning Subscale Step R R 2 R 2adj R 2 F chg Df P 1 .326 .106 .099 .106 14.022 1, 118 <.001 Table 4 22 Coefficients for RS 14 Meaning Subscale B t p Bivariate r Partial r Income .725 .326 3.745 <..001 .326 .326 Table 4 23 Model Summary for the RS 14 Equanimity Subscale Step R R 2 R 2adj R 2 F chg Df P 1 .239 .057 .049 .057 7.119 1, 118 .00 9 Table 4 24 Coefficients for RS 14 Equanimity Subscale B t p Bivariate r Partial r Income .379 .239 2.668 .009 .239 .239 Table 4 25 Model Summary for the RS 14 Perseverance Subscale Step R R 2 R 2adj R 2 F chg Df P 1 .345 .119 .112 .119 15.970 1,118 <.001 Table 4 26 Coefficients for RS 14 Perseverance Subscale B t p Bivariate r Partial r Income .569 .345 3.996 <.001 .345 .345 Table 4 27 Model Summary for the RS 14 Existential Aloneness Subscale Step R R 2 R 2adj R 2 F chg Df P 1 .336 .113 .105 .113 15.006 1,118 <.001 Table 4 28 Coefficients for RS 14 Existential Aloneness Subscale B t p Bivariate r Partial r Income .556 .336 3.874 <.001 .336 .336
86 Table 4 29 Model Summary for the Total RS 14 Step R R 2 R 2adj R 2 F chg Df P 1 .320 .102 .095 .102 13.443 1,118 <.001 Table 4 30 Coefficients for Total RS 14 B t p Bivariate r Partial r Income 3.096 .320 3.666 <.001 .320 .320 Table 4 31 Summary of Regression Analysis for RS 14 Scale Predictor 1 Predictor 2 Predictor 3 Self Reliance Income Age Marital status Meaning Income Equanimity Income Perseverance Income Existential Aloneness Income RS 14 Total Income Table 4 32 Means and Standard Deviations for Stress and Resource C hange Mean Std Deviation Stress Business Owners 6.64 5.45 Residents 3.76 3.94 Workers 5.03 4.15 Resource Change Business Owners 3.96 3.08 Residents 1.35 2.13 Worker 2.23 2.50 Table 4 33 Summary Analysis Question 5 Significant Results Variable df F p Meaning 5, 120 3.506 .005 Equanimity 5, 120 3.041 .013 Perseverance 5, 120 4.014 .002 Existential Aloneness 5, 120 4.883 .000 RS 14 Total 5, 120 3.543 .005 Resource Change 5, 234 2.307 .049
87 Table 4 34 Means and Standar d Deviati ons for Significant RS 14 Total and Subscales Income group N Mean Std Deviation Meaning Grp 1 ($0 22,757) 10 14.30 3.36 Grp 2 ($22,758 44,500) 27 17.03 3.01 Grp 3 ($44,501 71,200) 28 16.78 4.49 Grp 4 ($71,201 110,000) 34 17.82 3.21 Grp 5 ($110,0 01 150,000) 13 18.77 2.49 Grp 6 ($150,001 & up) 14 19.36 1.60 Equanimity Grp 1 ($0 22,757) 10 9.00 3.09 Grp 2 ($22,758 44,500) 27 10.88 1.78 Grp 3 ($44,501 71,200) 28 10.42 2.97 Grp 4 ($71,201 110,000) 34 10.58 2.51 Grp 5 ($110,001 150,000) 13 1 2.15 1.99 Grp 6 ($150,001 & up) 14 1.21 1.84 Perseverance Grp 1 ($0 22,757) 10 9.50 2.87 Grp 2 ($22,758 44,500) 27 10.51 2.42 Grp 3 ($44,501 71,200) 28 10.60 2.78 Grp 4 ($71,201 110,000) 34 11.76 2.60 Grp 5 ($110,001 150,000) 13 12.69 1.70 Grp 6 ($150,001 & up) 14 12.71 1.72 Existential Aloneness Grp 1 ($0 22,757) 10 8.80 3.25 Grp 2 ($22,758 44,500) 27 11.33 2.25 Grp 3 ($44,501 71,200) 28 10.53 2.99 Grp 4 ($71,201 110,000) 34 11.55 2.36 Grp 5 ($110,001 150,000) 13 12.76 1.36 Grp 6 ($ 150,001 & up) 14 12.85 1.40 RS 14 Total Grp 1 ($0 22,757) 10 69.60 14.63 Grp 2 ($22,758 44,500) 27 78.14 11.97 Grp 3 ($44,501 71,200) 28 76.67 19.61 Grp 4 ($71,201 110,000) 34 81.61 14.74. Grp 5 ($110,001 150,000) 13 88.38 9.30 Grp 6 ($150,001 & up) 14 89.50 8.81
88 Table 4 35 Means and Standard Deviations for Resource Change Group N M SD Grp 1 ($0 22,757) 10 4.10 3.41 Grp2 ($22,758 44,500) 27 2.52 3.30 Grp3 ($44,501 71,200) 28 2.32 2.80 Grp4 ($71,201 110,000) 33 1.45 1.75 Grp 5 ($110,001 1 50,000) 13 1.38 1.98 Grp 6 ($150,001 & up) 13 1.23 1.64 Table 4 36 ANOVA Results for Question 6 Significant Results Variable df F p Stress 1, 129 13.772 <.000 DASS 21 Total 1, 130 6.829 .010 Resource Change 5, 135 29.384 <.000 Table 4 37 Mean an d Standard Deviation for Significant Findings by Claim/No Claim Claim Status Mean Std Deviation Stress Claim 7.10 4.49 No Claim 3.84 4.08 DASS 21 Total Claim 14.48 10.99 No Claim 8.58 10.67 Resource Change Claim 4.10 2.77 No Claim 1.45 2.21
89 C HAPTER 5 DISCUSSION Technical disasters are nothing new; however the scale, geographical area impacted, and damage caused by the G ulf O il S pill (GOS) makes it the worst technical disaster in American history (National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling, 2011) Research conducted on previous technical disasters, including oil spills such as the Exxon Valdez in Alaska, identified Conservation of Resources theory as useful for understanding the psychological stress felt by people impacted (Picou et al., 2004; Picou & Gill, 1996). This study examined the relationship s among resource loss, psychological stress, resilience, and demographic characteristics of residents of Escambia County, Florida after the GOS. The study was de signed to help identify who was impacted by the GOS in Escambia County, Florida in order to help those impacted seek and access tangible goods and services. The results of the study offer information about business owners, residents, and workers in Escamb ia County and may provide baseline data to understand the far reaching effects of the GOS. The desired outcome of the study is for the information gathered to be used to inform public service initiatives, delivery of services, and policy making in the loc al area in order to better serve the popula tion of Escambia County, FL Overview of the Chapter C hapter five and hypothes e s along with the implications of each. Finally, chapter 5 des cribes the limitations of the study and makes recommendations for future research.
90 Discussion of the Descriptive Data The study used a convenience sample to collect data by gathering information from business owners, residents, and workers in Escambia Coun ty, Florida. The demographic data collected was not entirely consistent with the demographic data describing Escambia County, Florida. The first major difference in Escambia County demographic data and the demographic data collected from the survey was t he majority of respondents to the study were Caucasian (91.5%) whereas in the county, that number is 69.9% (US Census Bureau, 2012). Of particular note is the low number of Black survey respondents (2.3% of the total respondents) versus the 22.2% of Black people in the population of Escambia County. Females represented 57.4% of the study sample whereas they represent 50.5% of the population. Respondents had more education than the general population of the county, with 59.4% of respondents having a bache (US Census Bureau, 2012) Also of note is that respondents had a full time job (65.9%), made upwards of $44,501 per year (70.7%) and were an average age of 46 years old. The study relied on self reported data and there is no way to verify all of the demographic data provided by respondents. It is important to note that study participants were volunteers V olunteers may be different from non volunteers in level of education their inclination to add to a topic of study, they may be less authoritarian, and are less conforming (Gall et al., 2006). The current convenience sample of Escambia County, Florida may have elicited the participation of those members of the community who felt they had something to say about the GOS and was not representative of the entire county. The convenience
91 sample may also have been limited to those with access to computers and the I nternet, as well as those who heard about the study. Every attempt was made to spread word about the study through the use of social media, internet news sources, listservs, flyers, and business cards. Discussion of Instrumentation The study assessed resource change, psychological stress, and resilience in the aftermath of the GOS in Escambia County, Florida. The three instruments used were designed to gather responses pertinent to the research questions and hypotheses. The (2001) COR E, and w as used to determine number of resources changed for respondents. The second instrument, the 21 Item Depression Anxiety Stress Scale s (DASS 21 ; Lovibond & Lovibond, 1995 ), was used to determine levels of psychological stress in business owners, residents, and workers. The 14 Item Resilience Scale (RS 14 ; Wagnild & Young, 1993 ) was used to determine levels of resilience in the sample. The reliability of these instruments was moderately high and acceptable to this study, even though the measure of resource change was new and had never been tested before. These instruments represented the best assessments for measuring the constructs of interest for this study. Discussion of Hypotheses The six null hypotheses were designed to determine if there was a relati onship among resource change, psychological stress, and resilience, if there was a predictive relationship among demographic variables ( income group, employment status, hours worked per week, marital status, level of education, and age) and resource change DASS 21 scale and subscale scores, and RS 14 scale and subscale scores, if there
92 was a difference among business owners, residents, and workers on the resource change measure, DASS 21 scale and subscales, and RS 14 scale and subscales, if there was a dif ference based on income group or claim status on the measure of resource change, the DASS 21 scale and subscales, and the RS 14 scale and subscales. The first hypothesis sought to determine if there was a relationship among the measure of resource change, the DASS 21, and the RS 14. Hypothesis 2 used multiple regression to determine if there was a predictive relationship among the demographic variables of income group, employment status, hours worked per week, marital status, level of education, age, and r esource change and the DASS 21 total and DASS 21 subscales. Hypothesis 3 used multiple regression to determine if there was a predictive relationship among the demographic variables of income group, employment status, hours worked per week, marital status level of education, age, resource change, total RS 14, and RS 14 subscales. Hypothesis 4 utilized a univariate ANOVA to determine if there was a difference in responses of business owners, residents, or workers on the resource change, DASS 21 subscales, DASS 21 total, and RS 14 subscales and RS 14 total. Hypothesis 5 utilized an ANOVA to determine if there were differences based on income group for the DASS 21 total scale and subscales, RS 14 total scale and subscales, and resource change scale for Esca mbia County. Hypothesis 6 utilized a univariate ANOVA to determine if there were differences based on claim status for the DASS 21 total scale and subscales, RS 14 total scale and subscales, and resource change scale for Escambia County.
93 Discussion of Hyp othesis 1 Ho 1 : There will be no relationship among resource change, psychological stress, and resilience. As seen in Table 4 9 there was a significant negative association between psychological stress as measured by the DASS 21 total score and resilience as measured by the RS 14 total score and there was a significant positive association between psychological stress as measured by the DASS 21 and resource change. There was a non significant negative association between resilience as measured by the RS 14 total and resource change. It makes sense that there would be a negative association between DASS 21 scores and RS 14 scores since the higher the rating of depression, anxiety, and stress a respondent was feeling, the lower their resilience would be. Li kewise, the higher the resilience scores, the lower the scores would be on the measure of psychological stress. Study finding s contention that as RS 14 scores increased, reported depression decreased and overall health status incr eased. Notably, when there was more resource change, there was more psychological stress (depression, anxiety, and stress) a s measured by the DASS 21. These finding s correspond to COR theory in that the more resource change (or resource loss, specifically ), the more likely a person would feel psychological stress. It was interesting to find that there was not a significant negative association between resilience and resource change as one might think that the more resource change, the lower resilience may be However, the results could be explained by the overall rates of resilience for the population or the fact those who responding to the survey were resilient.
94 Discussion of Hypothesis 2 Ho 2 : There will be no predictive relationship among the demograph ic variables of income group, employment status, hours worked per week, marital status, level of education, age, and resource change and the DASS 21 subscales and total Resource change was a predictor for scores on the DASS 21 total and each of the subsca les (depression, anxiety, and stress). Results indicated the relationship between resource change (and resource loss) and psychological stress. The more resource change, the more psychological stress. Employment status was a predictor for depression, an xiety, and DASS 21 total. This finding is imp ortant because it reaffirms financial circumstances, such as having a job, or not having one, impacts psychological stress (Bisgaier & Rhodes, 2011). Hours worked was a predictor only for stress. It i s intere sting the amount a person works, whether it is many hours or few, impacts their level of psychological stress. What may be a more important factor is the culture of the workplace, as t he culture stress (Hobf oll, 2011). Income group, marital status, level of education and age were not predictors of psychological stress. Age has been identified in disaster literature as a demographic variable predicting coping ability; however age w as not a significant predict or of coping in th e study (Flynn & Norwood, 2004). Higher levels of income or having more resources w as identified in the COR literature as being a protective factor against psychological stress (Hobfoll, 2001). Study findings do not appear to confirm th ese ideas possibly due to the variety of individuals participating in the study Marital status, particularly having a partner, is thought to be a protective factor against depression, anxiety, and stress (Flynn & Norwood, 2004). Again, that does not see m to be the case
95 with the data on the GOS. Level of education, whether high or low, does not offer any predictive ability in relation to psychological stress in the GOS. Discussion of Hypothesis 3 Ho3: There will be no predictive relationship among the de mographic variables of income group, employment status, hours worked per week, marital status, level of education, age, resource change, and RS 14 subscales and total Income was a predictor of the RS 14 total and all five subscales of self reliance, meani ng, equanimity, perseverance, and existential aloneness. Age and marital status were also predictors of self reliance. Wagnild (2009) found age was a predictor of resilience. There were no other predictive relationships among income group, employment st atus, hours worked per week, level of education, or resource change and the RS 14 or its subscales. Income appears to be an important factor in resilience. Income as an important factor in resilience is supported by research finding those experiencing in come decline being less resilient than those who did not (Bonanno et al., 2007). Discussion of Hypothesis 4 Ho 4 : There are no differences in responses of business owners, residents, or workers on resource change, DASS 21 subscales, DASS 21 total, and RS 14 subscales and RS 14 total. Business owners had a higher stress subscale score of the DASS 21 than residents and workers. B usiness owners may be feeling more stress, correspond ing with their responsibilities of running a business and having people depend on them in a social context (Byers et al., 1997) Business owners also had higher scores on the resource change measure than did residents or workers. Results indicated business owners had more resource change and potentially lost more when compared to r esidents and workers. There were no significant differences among business owners,
96 residents, and workers on the DASS 21 total, the depression subscale, or the anxiety subscale. Discussion of Hypothesis 5 Ho 5 : There will be no differences based on income group for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County. In the analysis, income did make a significant difference in the subscales of meaning, equanimity, perseverance, and existential aloneness, t he RS 14 total, and resource change. Specifically, those making $110,001 and above scored higher on meaning, equanimity, perseverance, existential aloneness, the RS 14 total, and resource change than did those making $0 to $22,757. On existential alonene ss, those making $71,201 and above scored higher on existential aloneness than those in the lowest income group rais ing several interesting points. First, the analysi s indicated those with the lowest incomes reported the highest resource change (or loss) than those with higher incomes. It also appears that compared with the lowest income group, the higher income groups have statistically significant higher levels of resilience as measured by the RS 14. These results corresponded with some of the literatu re about resilience and income (Bonanno et al., 2007). Income did not make a significant difference in the self reliance subscale of the RS 14, the DASS 21 total, or the DASS 21 subscales of depression, anxiety, and stress. Discussion of Hypothesis 6 Ho 6 : There will be no differences based on claim status for resource change, DASS 21 subscales, DASS 21 total, RS 14 subscales, and RS 14 total for Escambia County. Those making a claim against BP had higher scores for stress, higher scores on the DASS 21 tota l, and higher resource change scores than those who made no claim.
97 A claim has to b e justified with documentation asserting resource loss attributable to the GOS. The finding of claimants having higher scores for stress and the overall DASS 21 coincides with previous research about litigation leading to corrosive community (Picou et al., 2004). There was no difference based on claim status for the depression and anxiety subscales of the DASS 21, or the RS 14 total and subscales. Discussion of Open ended Question The final question of the survey was open was included because the measure of resource change, the DASS 21, and t he RS 14 were very specific measures that may or may not have yielded information about the impact of the GOS on respondents, and the open ended question allowed all respondents the opportunity to tell their story about the impact of the GOS on their lives Sixty four respondents chose to answer the open ended question. Sixty two of these respondents expressed concerns, fears, and frustrations about the oil spill, the clean up process, the claim process, environmental damage, health effects, and effects o n children. Respondents were very open in their discussion about how the GOS impacted their lives. The information gathered from the open ended question may be used to explore other aspects of the GOS that have not yet been researched. Clinical Implicati ons S tudy results may be useful to mental health practitioners, public policy makers, and government officials in a position to help people dealing with the impacts of the GOS and other disasters. There was psychological stress experienced across all inc ome groups participating in the study. It is important for there to be mental health outreach to all of the individuals in the community when there is a disaster. For mental
98 health counselors, it is important for them to make the association between reso urce change and psychological stress discussed in disaster research and also found in this study ( Flynn & Norwood, 2004; Palinkas, 2012). The resondents to the GOS study indicated resource change increased symptoms of depression, anxiety and stress. Coun selors may need to focus their work with impacted populations on adapting and adjusting to the changes a disaster has brought, as well as other clinical needs in the population. The results of the study indicated business owners experienced more resource change and more stress than residents and workers. Mental health counselors might customize their services to business owners and help them manage their stress associated with the GOS. Mental health counselors may also help business owners through case m anagement activities, such as being aware of available community resources and how to access those resources. Customizing services to the unique needs of business owners may help engage them in seeking services and make it easier for them to seek services In turn, seeking and engaging in mental health servic es to reduce stress may reduce the physical manifestations of stress, such as heart disease and high blood pressure. T he number of people cho osing to respond to the open ended question at the end of the survey was of interest Many of these people did not have measureable resource change and were not impact ed directly by the GOS ; however, many of the respondents expressed worry, fear, and anger about the spill and its aftermath. Comments suggest ed t he entir e community and not just those with beach or tourism
99 businesses or interests w ere impacted. It is important to understand who is in need of help in a disaster such as the GOS and not to make assumptions. Theoretical Implications The Conservation of Resources (COR) theory was the prim ary theory guiding the study COR did predict an increase in psychological stress when resources were threatened or changed. Resource change was a predictor of depression, anxiety, and stress, as well as the overall 21 Item Depression Anxiety Stress Scales ( DASS 21 ; Lovibond & Lovibond, 1995) score. These findings add to the literature about COR. Income was a predictor for resilience as measured by the 14 Item Resilience Scale ( RS 14 ; Wagnild & Young, 1993) and its s ubscales. Income was an important study variable serving as a predictor of overall resilience and all five of the RS 14 subscales. The importance of income in dealing with a disaster might be addressed through economic initiatives proposed by policy make rs to help the Gulf Coast region recover from the GOS. Economic recovery may increase resilience in this area. Business owners experienced the most resource change and had higher levels of psychological stress compared to residents and workers in Escambi a County. Business owners were distributed across all income groups in the study. T hey were not necessarily in the highest income group, nor were they in the lowest income group. COR theory may be especially applicable to the business owners Limitation s Study p articipants in this study were drawn from Escambia County, Florida through a convenience sample. The sample size was small compared to the popu lation of Escambia County. The small sample size may be due to the fact respondents did not believe th eir information would remain anonymous, they did not trust the origin of the
100 survey, or they simply did not hear about the survey. The survey was only available online possibly limiting responses. The survey was also conducted nearl y two years after the GOS began and i nterest in a survey may have waned. Whatever the case, because of the small sample size it is difficult to generalize the results of the survey to the population of Escambia County, Florida or anywhere else. It is questionable whether all o f the variables measured were due to the GOS or some other factor such as the general economic climate in the area or the collapse of the real estate market. Another limitation of the survey was the high representation of Caucasian s in the group of respo ndents The homogeneity of respondents may make generalization to other racial and ethnic groups difficult. Additionally the resource change scale used in the s tud y was developed by the researcher and validity ha s not yet been established For these re asons, generalizations beyond the results of this study should be done with caution. Recommendations for Future Studies Future research should include expanding the sample size to allow for broader inquiry into the effects of the GOS on local populations a long the entire Gulf Coas t For example, a study with a larger sample size and includ ing more ethnic minorities in Escambia County, as well as the seven other Florida counties in Florida affected by the oil spill would add to the literature. A study focu sing on what resources were lost by wha t employment sectors would add dimension to the research agenda. A q ualitative stud y would add to the data on how people experience disasters as well as psychological stress, resilience and resource change. The situ ation in Escambia County, FL may actually be worse than the data indicated based on the sample size
101 and results. F urther study of the impact of disasters on people and their lives is needed. Conclusion The goal of this study was to determine what effects, if any, the GOS had in Escambia County, Florida through the use of the Conservation of Resources theory of psychological stress proposed by Hobfoll (1988, 1989) Results indicated there were effects experienced in Escambia County, including resource loss and psychological stress. The study also found there was a predictive relationship between resource loss and psychological stress and income was a predictor of resilience Also of note was the number of respondents who choosing to answer the final ope n ended question asking for anything else they would like to have known about their experience with the GOS. Forty four percent of the respondents answered the last question, and of those, 97% expressed fear, anger, distrust, and worry about the future. T he GOS has been a disaster unlike any other in its scope, breadth, and damage and continues to inflict damage on the Gulf Coast even two years later It will take many years to fully understand all of th e environmental, physical and psy chosocial impacts of the GOS disaster. The time to make a plan for understanding the impact of a disaster is now, when there is funding available to help those affected. The plan needs to take into account all of those impacted and attempt to provide the greatest help to the most people.
102 APPENDIX A INFORMED CONSENT Protocol Title: Effects of the Gulf Oil Spill in Escambia County, Florida Please read this consent document carefully before you decide to participate in this study. Inclusion criteria to participate in the study: You must be over the age of 18 and you must have been a business owner, resident or an employee who worked in Escambia County, Florida on or after April 20, 2010 when the Gulf Oil Spill began. Purpose of the research study: The purpose of this stud y is to assess resource loss, psychological stress and resiliency in Escambia County, Florida after the Gulf Oil Spill. What you will be asked to do in the study: Your participation in this research includes choosing whether to answer the survey as a busi ness owner, a resident, or an employee who works in Escambia County, Florida. You will then be asked to complete 3 questionnaires and fill out a brief questionnaire that provides background information The final question allows you to tell us anything el se you would like us to know about the impact of the Gulf Oil Spill on your life. You do not have to answer any items that you do not wish to answer. Time required: Approximately 15 minutes Risks and Benefits: There are no anticipated risks If you f eel you need counseling or assistance due to the Gulf Oil Spill, please contact these local resources: Lakeview Center, Inc. 432 1222 United Ministries 433 2333 Catholic Charities 436 6425 A potential b enefit for participating in this study is helpi ng to identify unmet needs in this community so that they may be addressed. Compensation: No monetary compensation will be given as a result of participation in this study. Confidentiality: Your identity will be unknown to us Y ou will not be asked to pr ovide your name o n any of the questionnaires There is minimal risk that security of any online data may be breached, but since (1) no indentifying information will be collected, (2) the online host (surveymonkey.com) uses encryption and firewalls, and (3 ) your data will be removed
103 from the server soon after you complete the study, it is highly unlikely that a security breach of the online data will result in any adverse consequence for you. If you have further questions or concerns, please refer to http://www.surveymonkey.com/mp/policy/security/ Voluntary participation: Your participation in this study is completely voluntary. There is no penalty for not p articipating. Right to withdraw from the study: You have the right to withdraw from the study at anytime without consequence. Legal Consideration: Neither the researcher nor the supervisor have any interest in the outcome of lawsuits being persued by ind ividuals or in class action lawsuits against British Petroleum (BP) or any of its subsidiaries. The information gathered from this research is for academic and informational purposes only. Whom to contact if you have questions about the study: Kelcey Kil lingsworth, Ed.S., Doctoral Candidate 1216 Norman Hall, PO Box 117046, Gainesville FL 32611 kelcey@GOSproject.org (850) 206 8802 Peter Sherrard, Supervisor 1216 Norman Hall, PO Box 117046, Gainesville FL 32611 email@example.com (352) 273 4339 Whom to contact about your rights as a research participant in the study: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; (352) 392 0433. If you consent to participate in this research study, agree to the terms above, and meet the pa page for your records and/or bookmark it for future reference.
104 APPENDIX B BUSINESS OWNER SURVE Y Effects of the Gulf Oil Spill in Escambia County, Florida The purpose of t his survey is to find out how the Gulf Oil Spill of April 2010 has affected or impacted you and your life in Escambia County, Florida. Please answer each question your honesty will be greatly appreciated. While the general economy has been bad, too, ple ase answer each statement only about the impact of the GULF OIL SPILL. Thank you! Please choose one of the following categories to apply to you. More than one category may apply but choose only one. Did you own a business, live, or work in Escambia Coun ty, Florida on or after April 20, 2010? Check only one. BUSINESS RESIDENCE WORK
105 PLEASE ENTER THE ZIP CODE FOR YOUR BUSINESS _________________________ AFTER OR BECAUSE OF THE GULF OIL SPILL, I: YES NO 1 Lost my business 2 Started a new business 3 Had to lay off an employee (s) 4 Worked fewer hours in my business 5 Had to work more hours in my business 6 Considered seeking government assistance/aid 7 Had to apply for government assistance/aid 8 Had to seek charitable assistance/aid 9 Made less mo ney in my business 10 Had to move 11 Had to sell my house 12 Had to find a cheaper place to rent 13 Had/have trouble paying my bills 14 Defaulted on a loan 15 Borrowed from family/friends 16 Declared bankruptcy 17 A m under financial stress 18 Got a new loan 19 Made more money in my business
106 Please read each statement and mark the box which indicates how much the statement applied to you over the past week. There are no right or wrong a nswers. Do not spend too much time on any statement. None of the time Some of the time Most of the time All of the time 20 I found it hard to wind down 21 I was aware of dryness in my mouth 22 I couldn't seem to experience any posit ive feeling at all 23 I experienced breathing difficulty (eg, excessively rapid breathing, breathlessness in the absence of physical exertion) 24 I found it difficult to work up the initiative to do things 25 I tended to over rea ct to situations 26 I experienced trembling (eg, in the hands) 27 I felt that I was using a lot of nervous energy 28 I was worried about situations in which I might panic and make a fool of myself 29 I felt that I had n othing to look forward to 30 I found myself getting agitated 31 I found it difficult to relax 32 I felt down hearted and blue 33 I was intolerant of anything that kept me from getting on with what I was doing 3 4 I felt I was close to panic 35 I was unable to become enthusiastic about anything 36 37 I felt that I was rather touchy 38 I was aware of the action of my heart in the absence of physical exertion (eg, sense of heart rate increase, heart missing a beat) 39 I felt scared without any good reason 40 I felt that life was meaningless
107 Please read each statement and select the box to the right of each statemen t that best indicates your feelings about the statement. Respond to all statements. Strongly Disagree Strongly Agree 41 I usually mange one way or another. 42 I feel proud that I have accomplished things in life. 43 I usually take things in stride. 44 I am friends with myself. 45 I feel that I can handle many things at a time. 46 I am determined. 47 I can get through difficult times because fficulty before. 48 I have self discipline. 49 I keep interested in things. 50 I can usually find something to laugh about. 51 My belief in myself gets me through hard times. 52 In an emer can generally rely on. 53 My life has meaning. 54 usually find my way out of it.
108 Please share some information about yourself. Thank you. Di d you file a claim for compensation from BP funds (check all that apply) I filed a claim My claim was denied My claim was approved I joined a class action suit I was satisfied with the amount from BP I accepted payment I rejected payment I am involved in litigation with BP I had no claim BEFORE the Gulf Oil Spill did you have problems with: No Problem Severe Problem Your physical health Your mental health Your finances Your relationships AFTER the Gulf Oil Spill did you have problems with: No Problem Severe Problem Your physical health Your mental health Your finances Your relationships Your gender Male Female Your age in years: ___________________
109 What best describes your Cultural Identification Black/African American Asian/Pacific Islander White Hispanic American Indian Other M arital Status Single Divorced Married/partnered Widowed Separated Your highest level of completed education Less than high school High school/GED Doctorate/ Professional degree Your living situation Alone With Friends With Children With Relatives With Spouse/Partner Other _____________ With Spouse/Partner/Children Your residence House Rented Apartment Condominium/to wnhouse Assisted Living Retirement community Other_______________ Your employment status Full time job Homemaker Part time job Retired Multiple Part Time job Disabled Unemployed Your household income $0 22,757 $110,001 150,000 $22,758 44,500 $150,001 190,000 $44,501 71,200 $190,001 230,000 $71,201 110,000 $230,001 or more Hours worked per week 20 hours or less More than 40 hours 30 hours or less Not Appli cable 40 hours or less
110 How did you hear about this survey? Internet Church Newspaper Other Radio Is there anything else you would like to tell us about the impact of the Gulf Oil Spill on your life?
111 APPENDIX C RESIDENT/WORKER SURV EY Effects of the Gulf Oil Spill in Escambia County, Florida The purpose of this survey is to find out how the Gulf Oil Spill of April 2010 has affected or impacted you and your life in Escambia County, Florida. Please answer each question your honesty will be greatly appreciated. While the general economy has been bad, too, please answer each statement only about the impact of the GULF OIL SPILL. Thank you! Please choose one of the following categories to apply to y ou. More than one category may apply but choose only one. Did you own a business, live, or work in Escambia County, Florida on or after April 20, 2010? Check only one. BUSINESS RESIDENCE WORK
112 PLEASE ENTER THE ZIP CODE OF YOUR RESIDENCE/WO RK______________________ AFTER OR BECAUSE OF THE GULF OIL SPILL, I: YES NO 1 Lost my job 2 Found a new job 3 Worked fewer hours 4 Had to work more hours 5 Considered seeking government assistance/aid 6 Had to app ly for government assistance/aid 7 Had to seek charitable assistance/aid 8 Made less money 9 Had to move 10 Had to sell my house 11 Had to find a cheaper place to rent 12 Had/have trouble paying my bills 13 Defaulted on a loan 14 Borrowed from family/friends 15 Declared bankruptcy 16 Am under financial stress 17 Got a new loan 18 Made more money
113 Please read each statement and mark the box which indicates how mu ch the statement applied to you over the past week. There are no right or wrong answers. Do not spend too much time on any statement. None of the time Some of the time Most of the time All of the time 19 I found it hard to wind down 20 I wa s aware of dryness in my mouth 21 I couldn't seem to experience any positive feeling at all 22 I experienced breathing difficulty (eg, excessively rapid breathing, breathlessness in the absence of physical exertion) 23 I found it difficult to work up the initiative to do things 24 I tended to over react to situations 25 I experienced trembling (eg, in the hands) 26 I felt that I was using a lot of nervous energy 27 I was worried about situation s in which I might panic and make a fool of myself 28 I felt that I had nothing to look forward to 29 I found myself getting agitated 30 I found it difficult to relax 31 I felt down hearted and blue 32 I was in tolerant of anything that kept me from getting on with what I was doing 33 I felt I was close to panic 34 I was unable to become enthusiastic about anything 35 36 I felt that I was rather touchy 37 I was aware of the action of my heart in the absence of physical exertion (eg, sense of heart rate increase, heart missing a beat) 38 I felt scared without any good reason 39 I felt that life was meaningless
114 Please read each statement and select the box to the right of each statement that best indicates your feelings about the statement. Respond to all statements. Strongly Disagree Strongly Agree 40 I usually mange one way or another. 41 I feel proud that I have accomplished things in life. 42 I usually take things in stride. 43 I am friends with myself. 44 I feel that I can handle many things at a time. 45 I am determine d. 46 I can get through difficult times because 47 I have self discipline. 48 I keep interested in things. 49 I can usually find something to laugh about. 50 My belief in myself gets me through hard times. 51 can generally rely on. 52 My life has meaning. 53 usually find my way out of it.
115 Please share some information about yourself. Thank you. Did you file a claim for compensation from BP funds (check all that apply) I filed a claim My claim was denied My claim was approved I joined a class action suit I w as satisfied with the amount from BP I accepted payment I rejected payment I am involved in litigation with BP I had no claim BEFORE the Gulf Oil Spill did you have problems with: No Problem Severe Problem Your physical health Your mental health Your finances Your relationships AFTER the Gulf Oil Spill did you have problems with: No Problem Severe Problem Your physical health Your ment al health Your finances Your relationships Your gender Male Female Your age in years: ___________________
116 What best describes your Cultural Identification Black/African American Asian/Pacific Islander White Hispanic American Indian Other Marital Status Single Divorced Married/partnered Widowed Separated Your highest level of completed education Less than high school egree High school/GED Doctorate/Professional degree Your living situation Alone With Friends With Children With Relatives With Spouse/Partner Other _____________ With Spouse /Partner/Children Your residence House Rented Apartment Condominium/townhouse Assisted Living Retirement community Other_______________ Your employment status Full time job Homemaker Part time job Retired Mult iple Part Time job Disabled Unemployed Your household income $0 22,757 $110,001 150,000 $22,758 44,500 $150,001 190,000 $44,501 71,200 $190,001 230,000 $71,201 110,000 $230,001 or more Hours worked per week 20 hours or less More than 40 hours 30 hours or less Not Applicable 40 hours or less
117 How did you hear about this survey? Internet Church Newspaper Other Radio Is there anything else you would like to tell us about the impact of the Gulf Oil Spill on your life?
118 LIST OF REFERENCES Amstadter, A., Broman Fulks, J., Zinzow, H., Ruggiero, K., & Cercone, J. (2009). Internet based interventions for traumatic stress related mental h ealth problems: A review and suggestion for future research. Clinical Psychological Review, 29 410 420. Arata, C., Picou, J., Johnson, G., & McNally, T. (2000). Coping with technological disaster: An application of the Conservation Resources Model to the Exxon Valdez oil spill. Journal of Traumatic Stress, 13 23 39. Ary, D., Jacobs, L. & Razavieh, A. (2009). Introduction to Research in Education. River Grove, IL : Cengage Publishing. Becker, S. M. (1997). Psychological assistance after environmental accidents: A policy perspective. Environmental Health Perspectives, 105, 1557 1563. Bisgaier, J. & Rhodes K. V. (2011). Cumulative adverse financial circumstan ces: Associations with patient health status and behaviors. Health & Social Work, 36 129 137. Blair, K. (2012, September 16). Crews cleaning up tar balls on Pensacola area beaches. The Pensacola News Journal Retrieved from http://www.pnj.com Bona nno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist, 59 20 28. Bonanno, G. A. (2005). Resilience in the face of potential trauma. Current Directions in Psychological Science, 14 135 138. Bonanno, G. A., Galea, S., Bucciarelli, A., & Vlahov, D. (2007). What predicts psychological resilience after disaster? The role of demographics, resources and life stress. Journal of Counseling and Cl inical Psychology, 75, 671 682. Bourne, J. K. (2010). The gulf of oil. National Geographic, 218 28 75. Bureau of Beaches and Coastal Systems of the Florida Department of Environmental Protection (October 2004). Hurricane Ivan: Beach a nd dune erosio n and structural damage assessment and post storm recovery plan for t he Panhandle coast of Florida. Retrieved from http://bcs.dep.state.fl.us/reports/ivan.pdf Byers, T., Kist, H., & Sutton, R. I. (1997). Characteristics of the entrepreneur: Social creat ures, not solo heros. In R.C. Dorf (Ed.), The Handbook of Technology Management. Boca Raton, FL: CRC Press.
119 Campbell, D. & Stanley, J. (1963). Experimental and quasi experimental designs for research. Boston: Houghton Mifflin. Caraballo, M., Heal, B ., & Horbaty, G. (2006). Impact of the tsunami on psychosocial health and well being. International Review of Psychiatry, 18 (3), 217 223. Casselman, A. (2011). National Geographic. Retrieve d from http://news.nationalgeographic.com/news/2011/04/110420 gulf oil spill anniversary health mental science nation/ Cohen, J. (1992). A power primer. Psychological Bulletin, 112 (1), 155 159. Cooper, D. R., Schindler, P. S. (2008). Business Resea rch Methods (10 th ed). New York: McGraw Hill. Crawford, J. R. & Henry, J. D. (2003). The Depression Anxiety Stress Scales (DASS): Normative data and latent structure in a large non clinical sample. British Journal of Clinical Psychology, 42 111 131. Creswell, J. W. (2009). Research Design: Qualitative, quantitative, and mixed method approaches. Sage Publications Ltd. Curran, M., Totenhagen, C, & Serido, J. (2010). How resources (or lack thereof) influence advice seeking on psychological well be ing and marital risk: Testing pathways of the lack of financial stability, support, and strain. Journal of Adult Development, 17 44 76. Dillman, D. A. (2007 ). Mail and internet surveys: The Tailored Design Method, 2 nd ed. New York: John Wiley & Sons, Inc. Dixon, W.O., Wolfgram, Z., Dehart, J., & Devonshire, C. (2010). The slick effect. 850 Magazine 3, 30 33. Independent News, pp. 6 7. Elliott, D. (2012, February 24). On Gulf Co ast, frustration at BP claims process. National Public Radio. Retrieved from http://www.npr.org Ferrier Auerbach, A., Erbes, C., & Polusny, M. (2009). Does trauma survey research cause more distress than other types of survey research? Journal of Tr aumatic Stress, 22 320 323. Field, A. (2009). Discovering statistics using SPSS. Thousand Oaks, CA: Sage.
120 Flemming, P. (2010, June 5). Resources for Florida lag as oil comes ashore. Pensacola News Journal, pp. 1A & 8A. Flynn, B., & Norwood, A. (2 004). Defining normal psychological reactions to disaster. Psychiatric Annals, 34 597 603. Freedy, J. R., Shaw, D. L., Jarrell, M. P., & Masters, C. R. (1992) Towards an understanding of the psychological impact of natural disasters: An application of the Conservation of Resources Stress model. Journal of Traumatic Stress, 5 441 455. Gall, M. D., Borg, W. R., & Gall, J. P. (1996). Educational Research: An Introduction. White Plains, NY: Longman. Goldstein, B. D., Osofsky, H. J., & Lichtveld, M. Y. (2011). The Gulf oil spill. The New England Journal of Medicine, 364, 1334 1348. Grattan, L. M., Roberts, S., Mahan, W. T., McLaughlin, P. K., Otwell, W. S., & Morris, J. G. (2011). The early psychological impacts of the Deepwater Horizon oil s pill on Florida and Alabama communities. Environmental Health Perspectives, 119, 838 843. Griffin, M., Resick, P., Waldrop, A., & Mechanic, M. (2003). Participation in trauma research: Is there evidence of harm? Journal of Traumatic Stress, 16 221 22 7. Gulf Coast Ecosystem Restoration Task Force (2011, October 5). Gulf of Mexico Regional Ecosystem Restoration Strategy (Preliminary). Retrieved from www.restorethegulf.gov Hobfoll, S. E. (1988). The ecology of stress. Washington, D.C.: Hemisphere. Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44 513 524. Hobfoll, S. E. (1998). Stress, culture, and community: The psychology and philosophy of stress. New York: Plenum. Hobfoll, S. self in the stress process: Advancing conservation of resources theory. Applied Psychology: An International Review, 50 337 421. Hobfoll, S. E. (2011). Conservation of resource caravans an d engaged settings. Journal of Occupational and Organizational Psychology, 84, 116 122. Hobfoll, S. E. (2012). Conservation of resources and disaster in cultural context: The caravans and passageways for resources. Psychiatry, 75, 227 232.
121 Hobfoll, S E. & Lilly, R. S. (1993). Resource conservation as a strategy for community psychology. Journal of Community Psychology, 21 128 148. Hopkins, W. (2009). Scale of magnitude for effect statistics. Retrieved from http://www.sportsci.org/resource/st ats/effectmag.html Institute of Medicine (2010). Assessing the effects of the Gulf of Mexico oil spill on human health: A summary of the June 2010 workshop. Washington, DC: The National Academies Press. Jansen, B. (2010, June 16). Spill raises health concerns: Illnesses could last for years, officials tell U.S. lawmakers. Pensacola News Journal, p. 9A. Johnson, B. & Turner, L. A. (2002). Data collection strategies in mixed methods research. On A. TashaKKori & C. Teddlie (Eds.) Handbook of mixed m ethods in social and behavioral research (pp. 297 320). Thousand Oaks, CA: Sage. King, L. & Berry, D. B. (2012). Gulf Coast states prepare to spend billions in fine money. USAToday. Retrieved from http://www.usatoday.com/news/nation/story/2012 07 05/r estore act gulf coast/56043544/1 Kunzelman, M. (2010 June 25 ). Gulf coast residents feel increasingly powerless in face of oil spill. The Pensacola News Journal. Retrieved from http://www.pnj.com Lazarus, R. S. (1966). Psychological stress and the coping process. New York: McGraw Hill. Littleton, H., Kumpula, M., & Orcutt, H. (2011). Postraumatic symptoms following a campus shooting: The role of psychosocial resource loss. Violence and Victims, 26, 461 476. London, R. T. (2010). PTSD and the Gulf oil spill Clinical Psychiatry News. Retrieved January 20, 2011, from http://psych.imng.com/fileadmin/content_pdf/cpn/archive_pdf/vol38iss7/70280_m ain.pdf Lovibond, P. F. & Lovibond, S. H. (1995). The structure of negative emotional states: Comp arison of the Depression Anxiety Stress Scales ( DASS) with the Beck Depression and Anxiety Inventories. Behavioral Research and Therapy, 33 335 343. Lovibond, P. F. (1998). Long term stability of depression, anxiety, and stress syndromes. Journal of Abnormal Psychology, 107 52 526.
122 Marald, E. (2001). The BT Kemi scandal and the establishment of the environmental crime concept. Journal of Scandinavian Studies in Criminology and Crime Prevention, 2 149 170. Mabus, R. (September 2010). Americ the Deepwater Horizon oil spill. Retrieved from http://www.restorethegulf.gov/release/2010/09/28/america%E2%80%99s gulf coast long term recovery plan after deepwater horizon oil spill McNally, R., Bryant, R., & Ehlers, A. (2003). Does early psychological intervention promote recovery from posttraumatic stress? Psychological Science i n the Public Interest, 4 45 79. Mertler, C. A. & Vannatta, R. A. (2005). Advanced and multivariate statistical methods, 3 rd ed. Glendale, CA: Pyrzcak Publishing. Monroe, S.M. (2008). Modern approaches to conceptualizing and measuring human life stress. Annual Review of Clinical Psychology, 4 33 52. Monroe, S.M. & Reid, M.W. (2009). Life stress and major depression. Current Directions in Psychological Science, 18 68 72. Mulvihill, J. (2010). Concerns grow over long term health effects of oil spill Fox News. Retrieved from http://www.foxnews.com National Center for Disaster Preparedness (2010, August 3). Im pact on the children and families of the Deepwater Horizon oil spill: Preliminary findings of the Coastal Population Impact Study. New York: Columbia University Mailman School of Public Health. National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling Report to the President (January 2011). Deepwater: The Gulf oil disaster and the future of offshore drilling. Retrieved from http://www.oilspillcommission.gov National Institute of Environmental Health Sciences, (2011 July 7). Th inking globally to improve mental health. Retrieved from http://www.epa.gov/agingepa/press/othernews/2011/2011_0707_ons_1.htm New York Times. Retrieved from: http://www.nytime s.com/2010/06/17/us/17human.html?_r=0 NOAA Technical Memorandum NWS NHC 6 (2011, August 10). The deadliest, costliest and most intense United States tropical cyclones from 1851 to 2010 (and other frequently requested hurricane facts) Retrieved from htt p://www.nhc.noaa.gov/pdf/nws nhc 6.pdf
123 Palinkas, L. A. (2012). A conceptual framework for understanding the mental health impacts of oil spills: Lessons from the Exxon Valdez oil spill. Psychiatry, 75, 203 222. Phipps, A., & Byrne, M. (2003). Brief interventions for secondary trauma: review and recommendations. Stress and Health, 19 139 147. Picou, J. S., & Gill, D. A. (1996). The Exxon Valdez oil spill and chronic psychological stress. American Fisheries Society Symposium, 18, 879 893. Pic ou, J. S., Marshall, B., & Gill, D. A. (2004). Disaster, litigation, and the corrosive community. Social Forces, 82 1493 1522. Quick, J. C. & Gavin, J. H. (2001). Four perspectives on Conservation of Resources Theory: A commentary. Applied Psycho logy: An International Review, 50, 392 400. Ruggiero, K. J., Amstadter, A. B., Acierno, R., Kilpatrick, D. G., Resnick, H. S., Tracy, M., & Galea, S. (2009). Social and psychological resources associated with health status in a representative sample of adults affected by the 2004 hurricanes. Psychiatry, 72, 195 210. Salkind, N. J. (2003). Exploring research, 5 th ed. Upper Saddle River, NJ: Prentice Hall. Sattler, D. N., Preston, A. J., Kaiser, C. F., Olivera, V. E., Valdez, J., & Schlueter, S. (20 02). Hurricane Georges: A cross national study examining preparedness, resource loss, and psychological distress in the U. S. Virgin Islands, Puerto Rico, Dominican Republic, and the United States. Journal of Traumatic Stress, 15 339 350. Schwarzer, R. (2001). Stress, resources, and proactive coping. Applied Psychology: An International Review, 50 400 407 Journal of Corporation Law, 37 245 263. Stevens, J. (1992) Applied multivariate statistics for the social sciences. Hillsdale, NJ: Lawrence Erlbaum Associates. Sumer, N., Karanci, A. N., Berument, S. K., & Gunes, H. (2005). Personal resources, coping self efficacy, and quake exposure as predictors of psych ological distress following the 1999 earthquake in Turkey. Journal of Traumatic Stress, 18 331 342.
124 U. S. Bureau of Labor and Statistics (2011). Economy at a glance: Pensacola Ferry Pass Brent. Retrieved from http://www.bls.gov/eag/eag.fl_pensacola_m sa.htm U. S. Census Bureau (2012 ). State and county quickfacts: Escambia county. Retrieved from http://quickfacts.census.gov/qfd/states/12/12033.html U. S. Environmental Protection Agency (2010, May 17). General facts about the Gulf of Mexico. Retrie ved from h ttp://www.epa.gov/gmpo/about/facts.html University of New Hampshire (2011, April 12). Louisiana, Florida residents differ on views of long term effects of oil spill. ScienceDaily. Retrieved from http://www.sciencedaily.com/releases/2011/04/11 0412121245.htm l Weaver, J. (2012, August 19). Fraud pollutes BP oil spill compensation fund for Gulf Coast victims from Florida to Louisiana. The Miami Herald. Retrieved from http://www.miamihearld.com Wagnild, G. (2009). s Guide for the English version of the Resilience Scale and the 14 Item Resilience Scale (RS 14). Wordon, MT: The Resilience Center. Wagnild, G. & Young, H. (1993). Development and psychometri c evaluation of the Resilience Scale. Journal of Nursing M easurement, 1 165 178. Wang, Y. (2008). Qualitative Research. In P. Heppner, B. Wampold, & D. Kivlighan, Jr. (Eds.), Research design in counseling 3 rd ed., 256 295. Woods Hole Oceanographic Institution (2011, September 6). Study sharpens picture o f how much oil and gas flowed in Deepwater Horizon spill. ScienceDaily Retrieved from http://www.sciencedaily.com/releases/2011/09/110905160912.htm
125 BIOGRAPHICAL SKETCH Kelcey Ray Killingsworth was born in May 1973 to Ronnie and Judy Ray in Pensacola, Florida. She attended local schools and graduated from the International Baccalaureate Program at Pensacola High School in 1991. She attended Birmingham Southern College for two years before transferring to the University of Florida in 1993. She ea rned her Bachelor of Arts in political s cience in 1994 and her Master of Arts in public a dministration in 1997. She worked for a few years and then returned to the U niversity of F lorida to study mental health counseling, earning her Master of Education an d Sp ecialist in Education degrees in mental health counseling in 2004 She married Cliff Killingsworth in 2005 and had two daughters, Cameron (2006) and Gillian (2007). She received her Doctor of Philosophy in mental health counseling in December 2012 After graduation she will work in private practice in the Pensacola area.