THE ASSOCIATION BETWEEN MARITAL FUNCTIONING, FAMILY 
CLOSENESS, AND TSUNAMI RELATED HEALTH:  
MODERATION BY RELIGIOSITY 
 
Except where reference is made to the work of others, the work described in this thesis is 
my own or was done in collaboration with my advisory committee. This thesis does not 
include proprietary or classified information. 
 
 
 
_______________________________________ 
Alyssa J. Banford 
 
 
 
 
 
Certificate of Approval: 
 
        
_____________________________                     _____________________________                    
Margaret K. Keiley       Thulitha Wickrama, Chair 
Professor        Assistant Professor 
Human Development and Family                                 Human Development and Family  
Studies         Studies 
 
 
 
 
 
_____________________________                           _____________________________                    
Scott A. Ketring     George T. Flowers 
Associate Professor     Dean 
Human Development and Family                               Graduate School  
Studies  
 
 
THE ASSOCIATION BETWEEN MARITAL FUNCTIONING, FAMILY 
CLOSENESS, AND TSUNAMI RELATED HEALTH:  
MODERATION BY RELIGIOSITY 
 
Alyssa J. Banford 
 
 
 
A thesis  
submitted to 
the Graduate Faculty of 
Auburn University 
in Partial Fulfillment of the  
Requirements for the 
Degree of  
Master of Science 
 
 
Auburn, Alabama 
August 10, 2009 
 
 
 
iii 
 
The Association Between Marital Functioning, Family  
Closeness, and Tsunami Related Health:  
Moderation by Religiosity 
 
Alyssa J. Banford 
 
Permission is granted to Auburn University to make copies of this thesis at its discretion, 
upon request of individuals or institutions and at their expense. The author reserves all 
publication rights. 
 
 
 
______________________________ 
       Signature of Author 
______________________________ 
       Date of Graduation 
 
 
 
 
 
 
iv 
 
 
VITA 
 Alyssa Jane Banford, daughter of Michael Banford and Patricia Banford, was 
born on October 21, 1985, in Salt Lake City, Utah. She attended Brigham Young 
University for three years and graduated in April 2007 with a Bachelor of Science degree 
in Marriage, Family, Human Development. She then entered graduate school in August 
2007 at Auburn University to pursue her Master of Science degree in Marriage and 
Family Therapy. She plans on pursuing her doctoral degree in Marriage and Family 
Therapy at Texas Tech University upon the completion of the requirements for her 
master?s degree. 
 
 
 
 
 
 
 
 
 
 
 
v 
 
 
 
THESIS ABSTRACT 
THE ASSOCIATION BETWEEN MARITAL FUNCTIONING, FAMILY 
CLOSENESS, AND TSUNAMI RELATED HEALTH:  
MODERATION BY RELIGIOSITY 
 
Alyssa J. Banford 
 
Master of Science, August 10, 2009 
(B.S., Brigham Young University, 2007) 
 
 
142 typed pages 
 
Directed by Thulitha Wickrama 
 
 To date, marital outcomes and family closeness following natural disasters have 
not been extensively investigated. Previous research indicates that disaster experiences 
relate to adverse mental health outcomes, and there is evidence suggesting that these 
effects persist over time. While the rise of family problems is documented in disaster  
vi 
 
research, marital satisfaction remains unreported on in disaster studies. In response to the 
adverse outcomes related to disaster, researchers have investigated what factors may be 
protective of mental health for disaster survivors. Religious coping is a prevalent topic in 
studies of trauma, stress, and disaster. The quality of coping employed tends to influence 
the outcomes realized. The purpose of the present study is to examine the relationships 
among physical health persisting form the 2004 Asian Tsunami, mental health outcomes, 
family closeness and marital satisfaction in Sri Lankan mothers. Secondarily, the 
moderating influence of religiosity on these relationships was also investigated. Each of 
these constructs was measured using self report instruments data were entered into a path 
analysis, where the data were examined. 
 
vii 
 
ACKNOWLEDGEMENTS 
 I would like to express sincere gratitude to my advisor Dr. Thulitha Wickrama for 
his patience with my endless emails and genuine sincerity and helpfulness. I would also 
like to thank Dr. Scott Ketring for his advisement on this project, and appreciate the 
friendship and counsel he has extended to me over the course of this program. To Dr. 
Margaret Keiley, I am grateful for encouragement and nurturing of my interests in 
therapy and research, and for her diligent help and statistical consultation on this project. 
I wish to express gratitude to my family for their support, prayers, and love. I express 
gratitude to all of the important people here in Auburn and at home for their confidence 
in me. Finally, and most of all, I want to thank my Heavenly Father for opportunities I?ve 
been given educationally, and the way I have been supported to accomplish my goals.  
 
 
 
 
 
 
 
 
 
viii 
 
Style manual used: Publication Manual of the American Psychological Association, Fifth 
Edition 
Computer software used: Microsoft Office Word, 2007, SPSS, Mplus Version 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ix 
 
TABLE OF CONTENTS 
LIST OF TABLES x
LIST OF FIGURES xi
INTRODUCTION 1
REVIEW OF LITERATURE 8
METHOD 70
RESULTS 76
DISCUSSION 90
REFERENCES 109
APPENDIX A 123
APPENDIX B 124
APPENDIX C 125
APPENDIX D 126
APPENDIX E 128
APPENDIX F 130
APPENDIX G 131
 
 
 
 
 
x 
 
LIST OF TABLES 
Table 1: Estimated Cronbach Alphas for Sample 76
Table 2: Univariate Statistics for Measures 77
Table 3: Pearson Correlation Coefficients for Measures 78
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
xi 
 
LIST OF FIGURES 
Figure 1: Hypothesized Path Model 75
Figure 2: Path Diagram for Model 1 81
Figure 3: Path Diagram for Model 2 83
Figure 4: Path Diagram for Model 3 85
Figure 5: Path Diagram for Model 4 87
Figure 6: Path Diagram for Model 5 89
 
 
 
 
 
 
 
 
 
1 
 
INTRODUCTION 
  
 A variety of events fit within the classification of a natural disaster. This term 
encompasses mass destroyers such as supercylclones (Suar, Mandal & Khuntia, 2002), 
floods (Smith, Pargament, Brant, & Oliver, 2000), hurricanes (Caldera, Palma, Penayo, & 
Kullgren, 2001), tornadoes (Menzel Baker, Hunt & Rittenburg, 2007), earthquakes 
(Proctor et al. 2007), and tsunamis (Bronisch et al. 2006).  
 One such disaster began from a large scale under water earthquake off the coast of 
Eastern Asia.  Tsunami waves were spawned, massive enough to ravage a large portion 
of the continent including multiple countries. Hundreds of thousands were killed, maimed 
or lost within the writhing waves (Anderson, 2007). Accounts of mothers and fathers 
literally having their children ripped from their arms by the relentless tides, and others 
searching helplessly for missing loved ones amidst the debris and wreckage (Bronisch et 
al., 2006) are some examples of the traumatic events experienced by survivors. Many 
were compelled to search among decaying bodies in hopes of identifying the remains of 
friends and family members (Bronisch et al., 2006). The spread of carnage was estimated 
to have claimed more than 126,000 lives within the first minutes (Anderson, 2007). The 
effects of this disaster were devastating, and we need more information regarding how 
the results of the 2004 tsunami are still influencing survivors today.       
 
2 
 
 Disasters, these incredibly destructive stressors, have captured the interest of 
researchers for decades. The ensuing research has focused largely on the mental health 
outcomes of individuals relatively soon after the disaster (Assanangkornchai, 
Tangboonngam, & Edwards, 2004; Catapano et al., 2001; Suar et al., 2002). The 
incidence of posttraumatic stress disorder (PTSD) and depressive symptoms are among 
the most common mental health constructs examined in disaster research. Studies 
investigating what personal characteristics or disaster circumstances predict variance in 
these symptoms is a prolific question within disaster research. Unexpectedness of the 
disaster is one such circumstance that relates significantly with the incidence of PTSD 
(Assanangkornchai et al., 2004). Furthermore, property destruction and loss of lives 
explained 10% of the variance in symptoms of both PTSD and depression in a study with 
a population of survivors from the 2004 tsunami (Wickrama & Kaspar, 2007). Other 
disaster experiences, like being injured have also been connected with a heightened 
likelihood of PTSD as well (McMillen, Smith & Fisher, 1997). This higher likelihood of 
PTSD also holds true for those who have experienced the injury or death of a family 
member (Caldera et al., 2001).  
 Personal characteristics such as a previous mental health problems (McMillen, 
Smith & Fisher, 1997, Caldera et al., 2001), trauma exposure (Weidmann, Fehm, & 
Fydrich, 2007), being female (McMillen, Smith & Fisher, 1997, Caldera et al., 2001), 
never having been married, or being illiterate (Caldera et al., 2001) have been identified 
as risk factors for the incidence of PTSD post disaster. Additionally, the intensity of 
resource loss explained 42% of the variance in depressive symptoms in one study of 
 
3 
 
earthquake damages in El Salvador (Sattler et al., 2006). It is clear that research has 
indicated that disasters relate to higher levels of mental health distress symptoms. These 
mental health outcomes may be immediate or long term (North, Hong, Suris, & 
Spitznagel, 2008). However, no studies have looked at the long term relational 
consequences for disaster survivors.  
 With regard to factors that may be protective against mental health problems post 
disaster, social support from the family, other trauma exposed peers and a sentiment of 
social acknowledgement of survivor trauma explained 23.3% of variance in lower 
indications of depression in one study (Weidmann, Fehm, & Fydrich, 2007). Other 
evidence of the importance of family members was found when adolescent survivors of 
the 2004 tsunami exhibited lower levels of depression and PTSD symptoms if they had a 
high quality relationship with their mothers (Wickrama & Kaspar, 2007). Family 
relationships are beginning to be addressed in the context of natural disaster within the 
literature. It is clear that positive and close relationships assist in coping with disaster. 
The relationship trauma survivors have with a higher power has also been investigated 
with regard to trauma.     
 Some aspects of religion and spirituality may facilitate recovery and even increase 
mental or emotional maturity known as posttraumatic growth (see Shaw, Joseph & 
Linley, 2005 for a review). The findings of researchers indicate that the use of religion in 
coping with disaster can yield both positive and negative outcomes. For instance, the use 
of positive religious coping predicted 33% of the variance in positive psychological 
outcomes after a flood in the American Midwest. On the other hand, negative religious 
 
4 
 
coping, including anger or distance from God or church and praying for miracles or 
bargaining with God, accounted for 21% of negative psychological outcomes (Smith et 
al., 2000). Religious coping in a negative way has also been found to predict a lower 
degree of hope as well (Ai, Peterson, & Huang, 2003). Other evidence indicates that 
those subscribing to a religious attribution style endorsing the concept of karma in 
response to trauma experienced more severe PTSD symptoms (Davidson, Connor & Lee, 
2005). Further examples of religious coping resulting in positive outcomes include a 
finding by Calhoun et al. (2000) suggesting that openness to religious change is 
associated with post-traumatic growth. Positive religious-spiritual coping was predictive 
of optimism in Bosnian war refugees. This positive coping style was enhanced by the 
degree of religiousness indicated by respondents (Ai, Peterson & Huang, 2003). One 
possible explanation for what makes the difference in positive and negative use of 
religious coping and subsequent outcomes is found by Chen (1997). He notes that 
societies that are avoidant of grief after loss forfeit the opportunity to experience spiritual 
growth (Chen, 1997). Overall however, research on religion and trauma presents a 
scattered picture of what characteristics of religious activity and belief serve to buffer 
stress or exacerbate it.  
 While family functioning in stress has not been neglected empirically, disaster 
researchers tend to focus on family process as a pre-existing factor predictive of various 
mental health outcomes (Proctor et al., 2007; Wickrama & Wickrama, 2007). This use of 
family variables as predictors rather than outcomes leaves a gap in the knowledge 
regarding how disaster actually affects the normative functioning of families. However, 
 
5 
 
pre-disaster family functioning has proved an important predictor of some post-disaster 
outcomes. For example, it appears that parental stress and negative mental health 
outcomes in response to disaster do predict poorer child mental health after a disaster 
(Proctor et al., 2007; Wickrama & Kaspar, 2007). However, the mechanisms responsible 
for this relationship are not as yet fully understood. Post disaster family outcomes which 
have been studied have been focused on increased negative family interactions or 
tensions (Catapano et al., 2001). Relationship quality and interactions post disaster seem 
sometimes ambiguous outcomes in the disaster research and deserve further attention. For 
example, Crabbs and Heffron (1981) note that loss and relocation associated with disaster 
lead to interpersonal problems, but they do not detail the process by which this occurs, or 
how this construct is defined. These interpersonal problems often stem from subsequent 
relocation or death which results in a loss of relationships (Crabbs & Heffron, 1981). 
Current tsunami research has begun to include constructs of family problems and how 
they interact with mental health outcomes and even religiosity (Wickrama & Kaspar, 
2007). Preliminary findings indicated that family problems may mediate the relationship 
between tsunami exposure and mothers? mental health. Also, a strong sense of family 
cohesion predicted lower depressive symptoms post-tsunami. Religious participation also 
predicted fewer depressive symptoms although religion was only measured with one item 
(Wickrama & Wickrama, 2007).  There still exists no solid understanding of how 
religious participation influences family cohesion or marital satisfaction, particularly in 
the context of a natural disaster. Studies looking at disaster outcomes appear to be of 
increasing salience because there is a large and growing number disasters being recorded 
 
6 
 
around the world (Centre for Research on the Epidemiology of Disasters (CRED), 2005) 
although all the reasons for the exponential increase in recorded disaster is unclear. More 
specifically, it may be that more disasters are reported as detection and communication of 
disasters becomes easier with technology. It is not ruled out however, that there may 
actually be more disasters occurring in the world now than before. 
 The premise of the current study is to understand how the persistent results of the 
2004 tsunami including lingering health problems as studied in survivors of Sri Lanka, 
influence family relationships including marriage four years after the occurrence of the 
tsunami. Secondly the objective is to understand whether religiosity, may moderate that 
influence. The role of Judeo-Christian faiths and other theistic orientations has been 
documented in coping with stress (Ai, Peterson & Huang, 2003; Smith et al., 2000). 
However, the body of literature systematically studying non-theistic religions including 
Buddhism to cope with disaster was until recently, non-existent (De Silva, 2006). 
Religious participation in a Buddhist population in Sri Lanka found religious activity to 
be protective against post disaster depression (Wickrama & Wickrama, 2007). However, 
religion was not measured in detail.  
 Previous research findings and theoretical suggestions surrounding stressful 
events prompt the following questions. How do persisting physical health problems 
resulting from the 2004 tsunami impact marital satisfaction after controlling for mental 
health status four years after the tsunami? How do persisting physical health problems 
resulting from the 2004 tsunami relate to family functioning after controlling for mental 
health status four years after the tsunami? Finally, does Buddhist religiosity moderate the 
 
7 
 
relationship between physical health problems from the tsunami and marital satisfaction 
and family functioning after controlling for mental health status four years after the 
tsunami? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8 
 
REVIEW OF LITERATURE 
 
 The purpose of this literature review is to examine the existing research 
concerning natural disasters, marital satisfaction, family functioning and religiosity. This 
review will specifically summarize research findings relating to the relationships between 
family functioning and varying measures of disaster exposure that prompt the questions 
examined in this study. Also it will review studies investigating the possible moderating 
role that religiosity plays between family functioning and disaster exposure as examined 
through the framework of the conservation of resources (Hobfoll, 1989) model.  
Theoretical Background 
 The phenomenon of family stress has been a construct of interest for theorists and 
researchers alike across a rich and long history of investigation (Boss, 1992; Hobfoll, 
1989; Hobfoll & Spielberger, 1992a; Hobfoll & Spielberger, 1992b; Kazak, 1992; Lavee, 
McCubbin & Patterson, 1985). The body of research literature does not always agree 
regarding how to define stress. This challenge has presented difficulty in empirically 
providing evidence that any one framework conceptualizing stress has validity over 
another (Hobfoll, 1989). The process by which events disturb resources and what 
attributes of remaining or replaced resources offset loss must be understood in order to 
face this ambiguity. The conservation of resources (COR) model (Hobfoll, 1989) is a 
progression of Hill?s seminal ABCX model (Hill, 1958) by examining families of world 
9 
 
war II returning soldiers (Hobfoll & Spielberger, 1992b). This model has been employed 
to examine the effects of natural disaster by Sattler et al. (2006) who discovered that a 
significant 58% of variance in acute stress disorder symptoms and 42% of the variance in 
depressive symptoms was accounted for by resource loss in a study of survivors of the El 
Salvador Earthquakes in 2001. 
 The basic tenets of the COR model suggest that change in and of itself is not 
stressful, and that in fact, what creates stress for families is loss, an aspect frequently 
present in changes. Conceptualizing stress in terms of loss puts great emphasis on the 
importance of resources. Hobfoll (1989) suggests that specific types of resources hold 
import for people; these include object resources, condition resources, personal 
characteristic, and energy resources. Some stressors can evoke a string of sequential 
resource loss, producing stress (Hobfoll, 1989). The 2004 tsunami represents such a 
stress because the disaster prompted immediate losses such as injury and death 
(Anderson, 2007), as well as losses which proved to provide continuing challenges such 
as extensive home damages (Wickrama & Wickrama, 2007).  
 Just as home damages present an obvious loss of resource, when physical and 
mental health is disturbed, the condition resource of health is lost. This loss may provoke 
other losses such as the ability to work to acquire other object resources. Injury, or loss of 
health from a natural disaster is a well documented variable in the research on disasters 
(Catapano et al., 2001; Proctor et al., 2007). However, current research lacks 
investigation of the connection between physical health problems that persist from 
 
10 
 
natural disasters and marriage and family functioning. Family variables such as increased 
conflicts were related to tsunami exposure in one study which also identified family 
cohesion as a protective factor (Wickrama & Wickrama, 2007). However, no direct 
studies of the influence of physical health problems from disasters on marital satisfaction 
and indicators of positive family functioning like closeness have been performed. Hobfoll 
(1989) through his COR model suggested that resource loss leads to stress which leads to 
varying negative outcomes or changes. Hobfoll (1989) then, would likely support 
hypotheses suggesting that the loss of physical health would produce stress. This stress in 
turn would be hypothesized to disturb family closeness and satisfaction felt within 
marriage. 
 In order to offset the loss provoked by stress, Hobfoll (1989) suggests that there 
are various ways to conserve, protect, or to manage loss of resources. These include a 
shifting of attention regarding resources in that there is reinterpretation or reframe of the 
stress as an opportunity for growth. In addition to reframing the stressor, resources can 
also be re-evaluated. This may entail de-valuing resources that are lost. Hobfoll (1989) 
suggests that this method of managing resource loss may be problematic in that it may 
require some to change their basic system of determining the value of resources. In fact 
posttraumatic growth is hampered by an avoidant attitude regarding loss according to 
Chen (1997) as minimizing loss precludes the opportunity to grieve and experience a 
transcendence of what is lost.   
 
11 
 
 The COR model provides a basis for understanding natural disasters. An 
exploration of what is lost as a result of disaster will likely help to explain why stress is 
experienced by survivors. Also, Hobfoll (1989) explains how the loss of resources can be 
quelled by a variety of mental and physical buffers. This theoretical model is applied to 
address the questions posed in the current study because of its attendance to resource loss 
and what tends to offset the negative effects of resource loss.  
Marital Satisfaction and Physical Health problems 
 The COR (Hobfoll, 1989) model would lead us to believe that the association 
between persistent physical health problems from natural disasters and marital 
satisfaction and family functioning would be negative. Hobfoll might explain this 
hypothesis in that health problems resulting from a natural disaster entail direct resource 
loss and potentially tax various resources within family relationships. There is a body of 
evidence in support of this hypothesis suggesting that persisting physical health issues 
relate to problems in marital and family relationships (Booth & Johnson, 1994; Murray, 
Murray, & Daniels, 2007; Trief et al., 2006). However, the relationship between family 
closeness and physical health problems has not been investigated in the context of natural 
disaster. Much of the natural disaster research has focused on an investigation of mental 
health outcomes (Assanangkornchai, Tangboonngam, & Edwards, 2004; Catapano et al. 
2001, Suar, Mandal, & Khuntia, 2002). Mental health disorders which occur after 
stressful experiences have been shown to relate negatively to marital outcomes (Dekel & 
Solomon, 2006, Renshaw, Rodrigues & Jones, 2008). However, mental health disorders 
 
12 
 
resulting from disaster have not been investigated thoroughly in relation to marital 
satisfaction and family closeness.  
 In a longitudinal study examining health status and marital quality, Booth and 
Johnson (1994) tracked both over a three year period in a national sample from a study of 
2,033 married individuals. All participants were under age 55. Participant households 
were chosen for the study using a clustered random-digit-dialing procedure, and 
information was attained through phone interviews. Exclusion criteria precluded those 
individuals were no longer married at the follow up data collection after 3 years and those 
whose health improved. This resulted in a final sample size of 1,298 persons. The 
researchers examined four questions: Does a decline in health affect the marital quality 
reported by the sick person? By the spouse? If so, by what mechanism do health declines 
affect the quality of the relationship? What factors moderate the effects of health declines 
on marital quality (Booth & Johnson, 1994)?  
In order to measure changes in health status, a one question indicator, ?In general, 
would you say your own health is excellent, good, fair, or poor?? was used at the first 
response and at follow up. Respondents were asked the same question regarding their 
spouse. Marital happiness was assessed using an 11-item scale which carries a reliability 
coefficient of .86. These items measured marital happiness, understanding received, 
agreement between spouses, information about the sexual relationship, whether the 
spouse was someone who would do things around the house, whether the spouse was 
someone they could do things with, and faithfulness. Divorce proneness, as an indication 
of marital stability was measured using a 13-item scale (?=.91) which indicated 
 
13 
 
frequency and timing of thoughts that the marriage was in trouble, talking to others about 
the possibility of divorce, separating from the spouse, or filing a petition. Four potential 
mediators including financial problems, household division of labor (indicating whether 
respondents thought the present division was fair), marital interaction (assessment of 
frequency of positive activities done together), and behavioral problems (easily angry, 
jealousness, dominance, criticism, moodiness, or not talking to the other) were assessed 
as well. The reliability coefficient for the marital interaction scale was marginal at ?=.63. 
The problem behavior scale asked about the self if the respondent was sick, and about the 
spouse if the respondent was married to a sick person. The scale demonstrated marginally 
adequate internal consistency (?=.61; Booth & Johnson, 1994).  
The researchers employed multiple regression equations to answer their questions 
and found that those who reported declines in health over the 3 year period also reported 
small declines in their marital happiness. Those respondents who were sick did not 
experience a significant increase in divorce proneness even if their health declined. 
However, respondents reporting about a sick spouse experienced a significant decline in 
marital happiness and increase in divorce proneness. This indicated that spouses 
experienced a somewhat greater impact on their marital happiness due to the sickness of a 
spouse than did the sick spouse themselves (Booth & Johnson, 1994).  
The researchers also tested for what mechanisms explained the relationship 
between health declines and marital happiness. Booth and Johnson (1994) found a 
mediating effect manifested in a nullification of the relationship between respondents? 
health and marital happiness when all the potential mediating factors were taken together. 
 
14 
 
The health-marital happiness coefficient represents the linear relationship of marital 
quality regressed on health, and was modified as the mediators were entered stepwise. 
Money problems, thinking that the division of labor was not fair, and behavior problems 
removed nearly an entire point from the coefficient whereas decline in income and 
increase in household chores didn?t appear to affect the coefficient. Booth and Johnson 
(1994) concluded that for those who experienced a decline in physical health and their 
spouses, the attending decline in marital happiness could be explained at least in part by 
the 4 mediating factors proposed. For the respondents reporting about a sick spouse, 
reduction in marital interaction and increase in behavior problems related to a reduction 
in the relationship between marital happiness and physical health. For those reporting 
health declines in their spouse, it appears that decreased interaction and increase in 
problem behaviors of the sick spouse had the greatest impact on marital happiness.  
The findings of Booth and Johnson (1994) show that physical health problems 
tend to influence happiness in marriage through the toll illness takes on positive 
interactions and an increase of problem behavior. Research literature needs to expand on 
this finding to include considerations of health problems relating to natural disasters and 
would also benefit from including consideration of possible protective factors of family 
relationships in the context of physical health problems. 
 Further evidence of the connection between physical health and marital 
functioning shows a relationship between diabetes outcomes and marital quality (Trief et 
al. 2006). A sample of 134 diabetes patients in upstate New York, were recruited as part 
of IDEATel, a randomized controlled trial funded by the Center for Medicare and 
 
15 
 
Medicaid services enrolling more than 1,500 elderly diabetes patients. Recruiting criteria 
included being a Medicare beneficiary, having diabetes, living in a medically 
underserved area, and being married or cohabiting for more than 1 year.  
 To measure marital quality, the Spanier Dyadic Adjustment Scale (DAS) (Spanier, 
1976) (? = .96), and the Perceived Marital Stress Scale (PMSS) (Pearlin & Schooler, 
1978) (? = .86) were used.  Demographic information including age gender, the length of 
respondent diabetes, education level, income, body mass index (BMI), and smoking 
history was gathered for the baseline data collection and at a one-year follow up. 
Additionally, blood pressure and blood and urine samples were collected. Blood glucose 
control was assessed by measuring glycated hemoglobin (HbA1c) levels which indicated 
blood glucose over the past 2-3 months. Depressive symptoms were assessed using the 
SHORT-CARE Depression Scale (?=.80) (Gurland, Golden, Teresi & Challop, 1984). 
Diabetes related distress was assessed using the Diabetes Distress Scale (?=.89) 
(Polonsky et al., 2005). Finally, comorbidity of other health problems was indicated and 
coded using the Charlson co-Morbitity Index (Charlson, Pompei, Ales, & MacKenzie, 
1987). The researchers found that greater marital satisfaction predicted improvement in 
1-year HvA1c in the group participating in the IDEATel project. This project is an 
intervention giving diabetes patients information on tracking blood sugar, nutrition and 
other medical topics through telephone contact and internet video conferencing with 
medical professionals. The researchers also found that while the best predictor of 
depression at follow up was depression at baseline, other significant predictors of 
variance in depression at follow up included greater marital stress (13%), and less marital 
 
16 
 
satisfaction (8.7%) at baseline. However, the indicators of marital satisfaction did not 
predict diabetes related distress at follow up. Finally, for those in the intervention group, 
it was found that greater marital stress at baseline predicted greater diabetes related 
distress (DDS) at follow up (p = .089) albeit at a marginally significant level. As the 
monitoring of blood pressure and blood glucose readings was a continuing process over 
the course of this study, it is evident that physical health problems influence daily life 
(Trief et al. 2006). It appears that physical health problems relate negatively to marital 
satisfaction. This idea is supported in the finding that martial stress predicted later 
physical health symptom distress. Depression was also predicted in part by marital 
problems suggesting that physical and mental health problems both take a toll on marital 
satisfaction. Moreover, problems with the marital relationship may even hamper one?s 
ability to cope with illness. However, receiving community support from program 
enrollment seemed to offset the negative outcomes on marriage from the ill-effects of 
diabetes. Physical health can influence marital satisfaction, and there are also some 
supports that tend to offset the impact of poor health on marital satisfaction. Further 
research uncovering other potential protective factors of marital satisfaction in the 
context of physical health problems is needful to understand what else may buffer the 
stress of a physical ailment.  
Family functioning and Physical Health problems 
 Physical health problems also relate to family functioning in a context broader 
than the marital relationship (Murray, Murray, & Daniels, 2007). Murray et al. (2007) 
 
17 
 
investigated two questions relating to family functioning and physical health. The 
researchers selected their sample for study among those suffering from fibromyalgia, a 
disease affecting the soft tissue in the joint and often resulting in chronic pain. The two 
questions investigated were, can the symptom severity of fibromyalgia be predicted by 
the level of differentiation of self, age-of-onset, and stressful events occurring within a 
year prior to the age-of-onset? Also, which predictor variables account for the greatest 
amount of variance in symptom severity of fibromyalgia?  
The researchers conducted the study with 201 respondents whose average age was 
47 years (SD=10.4 years). The sample was comprised of 169 females and 31 males and 
one transgendered person.  Within the sample were 196 Caucasian individuals, 2 
American Indian, and 2 bi-racial participants. The age of onset of fibromyalgia symptoms 
averaged at 31 with a range from birth to 62 years (SD=13.55 years). Participants were 
recruited through various internet contacts and comprised a non-randomized convenience 
sample. 
The Differentiation of self inventory-revised (DSI-R; Skowron & Schmitt, 2003), 
a Bowenian based scale designed to tap into differentiation of self, was used to indicate 
family relationship functioning (?=.92). The researchers assessed stressful events using 
the Perceived stress scale (PSS; Cohen, Kamarck & Mermelstein, 1983). The PSS was 
used to measure respondents? perception of stress in his/her life on a daily basis (?=.78-
86). Also, the Life Experiences Scale (LES; Sarason, Johnson, & Siegal, 1978) was used 
to identify positive and negative life events and their impact (?=.56-.88). Finally, the 
Fibromyalgia Impact Questionnaire (FIQ) is comprised of 20 items measuring the current 
 
18 
 
health status of fibromyalgia patients (Burckhardt, Clark, & Bennett, 1991) and was 
utilized to assess symptom severity (?=.79).  
The researchers found through a simultaneous multiple regression model with 
adequate fit (F
3,297
=16.80, p < .0001) that level of differentiation and stressful life events 
explained 19% of variance in symptom severity of fibromyalgia patients. Age of onset 
was not a significant predictor of variance in symptom severity. The investigators used 
stepwise multiple regression equations to discern which variables predicted the most 
variance in symptom severity. In the first regression model, perceived stress in daily life 
accounted for 27% of the variance in symptom severity. In the second model, perceived 
daily stress and negative life events in the year prior to symptom onset taken together 
predicted nearly 30% of the variance in symptom severity (Murray et al., 2007). In other 
words, physical health problems may be aggravated in their severity by the daily stress 
that accompanies them, and previous stressful life events.  
In another multiple regression model, the level of differentiation, age of onset, 
and stressful life events in the year prior to onset predicted 19% of the variance in 
symptom severity. The researchers noted that consistent with Bowen theory, those who 
had higher levels of differentiation were able to appraise life events and symptoms less 
negatively and seemed to have better ability to cope. The reverse is also possible, that 
those with lower symptom distress and fewer stressful life events seem to have higher 
levels of differentiation. It seems that one?s relationship with family relates to the 
experience, and perhaps coping of a physical ailment.  
 
19 
 
We need to know more about how family cohesion and other elements of 
functioning relate to physical health problems resulting from a variety of causes. If 
positive family functioning aids in managing the effects of physical health problems, 
interventions focused on the family level could aid those suffering from persisting 
physical health problems. Booth and Johnson (1994), Trief et al. (2006), and Murray et 
al. (2007) add to our understanding of how physical health problems relate to marriage 
and family. The same questions investigated by these researchers, and others, need to be 
examined in the unique context of natural disaster as disaster survivors may be unique in 
the physical health and relationship challenges they face. 
 Disasters and Mental Health 
 What researchers have focused more intently on in relation to natural disasters are 
mental health outcomes. Specifically, investigators of this topic have often attended to 
posttraumatic stress disorder (PTSD) and depression as main outcomes 
(Assanangkornchai et al., 2004; Catapano et al., 2001; Suar et al., 2002). For instance, 
Catapano et al., (2001) sought to assess the prevalence of PTSD in the community of 
Sarno, in Southern Italy after a landslide which occurred in May of 1998. One year after 
the landslide, a random sample of 272 individuals living in Sarno were contacted by mail 
and recruited to participate. The sample also included a control group selected from a 
nearby town, San Valentino Torio, which was not physically damaged in the disaster. 
 The researchers gathered information with three assessments. The 
sociodemographic form is a 34-item self-report index developed for the study. It gathered 
details regarding respondents? demographics, physical injuries of self or family members, 
 
20 
 
and material damages including home damage. It also asked about the need to move after 
the landslide, debts and increased expenses. It included information on family difficulties 
and work problems. This entailed tensions, unusual behaviors, drug abuse, school 
problems, loss or change of job, or hard work conditions. It also assessed sources of aid 
such as family, friends, public institutions or volunteer associations. Finally, it included 
whether victims had received information on the event, and behaviors to adopt in case of 
further emergencies. The General Health Questionnaire (Fontanesi, Gobetti, 
Zimmermann-Tansella & Tansella, 1985; Goldberg, 1972) was administered as a general 
screen for mental health problems. Finally, the Self-Rating Scale for Post-Traumatic 
Stress Disorder (Calier, Lambers, Van Uchelen, & Gersons, 1998) was used to assess for 
PTSD. Information on the reliability of these measures was not included. 
 Using t-tests, the researchers found that the only demographic difference between 
the two groups was that those recruited in the control group were more often employed. 
Of the affected group in Sarno, 39% reported home loss or damage. Sixty-five percent 
reported losing at least one family member or friend in the disaster, and 48% has 
resulting financial problems. Finally, 29% experienced occupational difficulties and 32% 
complained of family problems resulting from the landslide. Most of the respondents in 
Sarno (80%) reported never having received any information about the cause of the 
landslide, 79% received no information about the risk of subsequent disasters, 53% were 
not given any information about an evacuation plan, and 44% reported receiving no 
direction on how to behave in case of ?alarm?. Fifty-six percent allegedly received no 
support from public institutions or other sources post-disaster. 
 
21 
 
 The average GHQ score for the affected group in Sarno was 8.6 (SD=6.8). A 
score of 5-6 or above indicates a mental health problem. In the non-affected group in San 
Valentino Torio, the average GHQ score was 5.2 (SD=5.5). These averages were 
significantly higher in Sarno (F=15.6, df = 1, 342, P < .0001). In Sarno, 59% of 
respondents were considered probable cases of PTSD, and 35% of the respondents in 
Valentino Torio. The number of respondents who were victims of PTSD was 
significantly higher in the affected group (?
2
 = 13.3, df = 1, p = 0.0001). This suggests 
that close proximity to disaster, or intensity of exposure relates to mental health 
outcomes. 
 A GHQ score indicating a mental health problem was significantly associated 
with being female (?
2
 = 18.7, df = 1, p < 0.000), having a low education level (?
2
 = 8.4, 
df = 2, p < 0.015), and unemployment (?
2
 = 18.2, df = 2, p < 0.01). Also, a GHQ score 
above cut-off was reported significantly more frequently by those in Sarno who reported 
financial problems (?
2
 = 7.3, df = 1, p < 0.007), personal physical injuries (?
2
 = 5.6, df = 
1, p < 0.02), physical injuries of family members (?
2
 = 8.7, df = 1, p < 0.003), 
psychological problems suffered by family members (?
2
 = 12.6, df = 1, P < 0.0001) and 
leaving their home in the last 3 months (?
2
 = 8.7, df = 1, P < 0.003). Mental health 
problems relate negatively to physical injuries sustained in the disaster to the respondent 
and family members. However, we do not have information about physical health 
problems that persist long term after disasters and their influence on mental health 
outcomes, or family relationships. 
 
22 
 
 Factor analysis of the GHQ yielded distinct subscales including anxiety and sleep 
disturbance, subjective efficiency, depression, and social impairment. Respondents in 
Sarno had higher scores on these subscales than their San Valentino Torio counterparts. 
Those who were categorized as having PTSD also more often experienced personal 
physical injury (?
2 
= 9.6, df = 1, P < 0.01), injuries sustained by family members (?
2
 = 
8.1, df = 1, P < 0.01), family problems (?
2
 = 9.0, df = 1, P < 0.01), and having to leave 
home in the past 3 months (?
2
 = 6.2, df = 1, P < 0.05) (Catapano et al., 2001). This 
finding lends information about the co-existence of injuries, and family problems with 
mental health problems. Studies including family problems as an outcome of disaster 
effects may yield distinct and important findings even controlling for mental health 
status.  
 From these findings it is safe to assume that the intensity of natural disaster 
exposure is related to higher mental health distress levels. It is also clear that the 
incidence of mental health problems relates to the outcomes of disaster, namely, physical 
injury of self or family members, and family problems. It is evident that physical injury 
relates negatively to psychological outcomes. However, this study gathered information 
regarding the incidence of injury as opposed to persisting health problems resulting from 
the disaster. Also, differentiating between family problems and problems within the 
marital relationship may or may not have shown that disaster influences these 
relationships in distinct ways. 
 Suar, Mandal, and Khuntia (2002), expanded on these findings to further 
investigate the affects of disaster exposure, but went further to study if disaster exposure 
 
23 
 
was buffered by the mobilization of resources. They hypothesized that psychological 
outcomes following a supercyclone in Orissa, India could be explained relative to the 
magnitude of resource loss that survivors incurred. They predicted that those closer to the 
epicenter of the cyclone would experience greater psychological distress and exhibit a 
more external locus of control. They also hypothesized that distress would be magnified 
proportional to resource loss and conversely, distress would diminish with resource 
support. They formulated these hypotheses based on Hobfoll?s (1989) COR model. 
 Participants were recruited informally from survivors living in temporary camps. 
The sample of 65 persons excluded those under age 18 or those with a history of any 
psychiatric or other medical disorders. There were 45 males and 20 females in the 
sample, with an average age of 38 years (SD = 14.1). The education level averaged 9.2 
years (SD = 5.2) and all were from semi-urban areas. Another group of 65 participants, 
considered as an unaffected comparison group, lived 40 km away or more from the 
epicenter of the cyclone and reported no loss of property or life. In the unaffected group 
there were 32 males and 33 females whose average age was 35.5 years (SD = 11.9) and 
whose education level averaged near 10.6 years (SD = 5.0). The two groups were not 
significantly different in age or education level, however the affected population (within 
10 km of the actual devastation area) tended to be comprised of joint families more often 
than the unaffected group.  
 The researchers measured neuroticism with 20 items from the Maudsley 
Personality Inventory (Eysenck, 1959). The scale demonstrated a marginally adequate 
internal consistency coefficient of ? = .62. Internal versus external locus of control was 
 
24 
 
assessed using the Internal-External Locus of control Scale (Rotter, 1966). The alpha 
reliability for this scale was also relatively low at .61. Anxiety was assessed using 9 items 
from the Trait Anxiety Inventory (Speilberger, Gorsuch, & Lushene, 1970). Depression 
was measured using 15 items from the Beck Depression Inventory (Beck, 1967), and 
PTSD was assessed with a clinical interview developed on the basis of the DSM-IV 
(American Psychiatric Association, 1994).  
 Suar et al. (2002) also used some scales developed specifically for their study to 
assess total loss and social and economic support. Total loss was assessed by asking 
respondents to indicate a pre and post number of family members, amount of agricultural 
land, total rooms in their homes, rooms damaged, total livestock. Social and economic 
support ranged from 0-6 and was indicated by questions asking respondents if they had 
received social or economic support from various sources including family, relatives, 
community members, nongovernmental organizations, government or any other 
resources.  
 To assess for differences in mental health outcomes for those affected and not 
affected, the researchers dummy coded each respondent as either 1 = affected or 0 = 
unaffected. Multiple regression models revealed positive beta coefficients when status of 
being affected was regressed on psychological outcomes. The beta coefficients were 
significant for anxiety (? = .24 p < .01), depression (? = .21 p < .05) and PTSD (? = .77 p 
< .001), but not for external locus of control. When gender and neuroticism were 
controlled, the beta coefficients were all heightened, and the effect of external locus of 
control reached significance (? = .18 p < .05). The variables of total loss, as well as social 
 
25 
 
and economic support, were added to the regression analyses of the affected group as 
controls. Losses incurred by those in the affected group controlling for gender, 
neuroticism and support still explained 15% of the variance in locus of control, 35% of 
variance in anxiety, 31% of variance in depression, and 61% of PTSD symptoms (Suar et 
al., 2002). The researchers? hypotheses were supported in that exposure intensity was 
related to greater psychological distress just as in Catapano et al.?s 2001 study. 
Furthermore, in support of the COR model, greater social and economic support was 
negatively related to anxiety and depression. However, high levels of support had no 
protective effect on PSTD or locus of control (Suar et al., 2002).  
 This study is in clear support of the COR model in that loss seems to be what 
predicted adverse outcomes for survivors as opposed to the event itself. Injury of self and 
family members was related negatively to psychological outcomes, but ongoing health 
problems as a result of the disaster were not considered. The population was also limited 
to adults who were not studied in a relational context. Finally, the researchers considered 
some of the resources that allow survivors to increase, replace or manage the loss of other 
resources. However, a more complete consideration of buffer resources may have 
included religious activity and belief. 
 The influence of religion was considered by Assanangkornchai, Tanboonngam, 
and Edwards (2004) in a study of survivors of the Hat Yai flood which hit Thailand in 
2000. The subjects in this study were recruited from three urban areas in Hat Yai where 
water levels ranged from 2-2.5 m and flood duration lasted from 4.5-5.5 days, and one 
rural sub-district in the area. One hundred households were randomly selected and all 
 
26 
 
residents over 14 years of age were interviewed culminating in a total sample of 590 
individuals. The sample was comprised of 208 men and 382 women. Their ages ranged 
from 14-85 with the mean age being 41.9 years (SD = 16.5). Sixty-four percent of the 
sample was married, and 94% were Buddhists. Roughly one third of the sample had 
finished primary or high school while 11% had no formal education and twenty-one 
percent were recipients of college, university, or other higher education. Occupations 
within the sample varied with 10% laborers, 17% house-workers and the rest of the 
sample was comprised farmers, gardeners, government officers or private business 
officers.   
 A three part structured questionnaire was administered to respondents. The first 
section was designed to assess type and condition of property, household size, water level 
in and near property, duration of shortage of food, drinkable water and electricity as well 
as personal property damage and loss.  The next section asked about demographic 
information, intensity of exposure to stressors including injury, illness or seeing dead 
people or body parts. It also gathered information on loss of relatives, friends, and 
acquaintances as well as respondents? perception of the degree of their losses. This 
section gathered indication of how prepared respondents were for the flood and the level 
to which it was an unexpected event. Finally, the second section included items assessing 
activities during the flood, previous disaster exposure and subjective emotional and 
behavioral responses to the flood. The third section of the questionnaire included a Thai 
version of the General Health Questionnaire GHQ (Nilchaikovit, Sukying, & Silpakit, 
 
27 
 
1996) to assess mental health status and the Impact of Event Scale (Horowitz, Wilner, & 
Alvarez, 1979) to investigate PTSD symptoms in respondents. 
 There were several factors which increased the likelihood of poor mental health 
outcomes. Logistic regression analyses revealed that survivors who perceived that the 
flood had caused severe personal loss were twice as likely to have a mental health 
problem as those who reported little or no loss. Additionally, having a large household 
(3-6) meant that respondents were 1.95 times as likely to report a mental health problem 
(p = .01) as those with smaller households. Knowing someone killed in the flood resulted 
in a risk 1.98 (p = .014) times larger than those who did not for having a mental health 
problem. Interestingly, those able to collect almost all of their possessions were at 2.14 
times the risk of reporting a mental health problem than those who recovered none of 
their possessions (p = .018). Perhaps being able to collect part of one?s possessions was a 
reminder of the extent of the destruction and what was lost. 
 Other challenges resulting from the flood included 30% of the sample reporting 
unemployment as a negative life change 10 weeks after the disaster. Fifty-nine percent of 
the sample declared that they experienced an increased difficulty working. Eight percent 
of respondents indicated that they experienced more quarrels with family members 10 
weeks post-flood. The researchers noted that these conditions existed together in their 
sample, but no further analysis was performed to investigate how family interaction 
problems related to injury or mental health outcomes (Assanangkornchai et al., 2004). 
 Furthermore, there were also other factors were linked with the increased 
likelihood of experiencing PTSD symptoms. Those who perceived that their losses were 
 
28 
 
severe were 3.9 times as likely to have PTSD (p < 0.001). Religious affiliation also 
influenced the likelihood of PTSD. Those who were not Buddhist were a significant 3.41 
times as likely to have PTSD as those who were Buddhist (p =.013). This suggests that 
being part of a minority religion was related to an increase risk of an adverse emotional 
reaction to the disaster. Perhaps some were separated from support systems or a sense of 
community that came from being a part of the majority religion. These findings might 
also indicate that being religious or even specifically Buddhist provides a buffer against 
mental health problems from disaster. The literature needs to have a greater 
understanding of the mechanisms by which religion serves as a buffer or exacerbating 
force with regard to mental health outcomes and disasters.  
 It is clear that many of the circumstances that follow a natural disaster are linked 
to adverse mental health outcomes. While we know that family relationships tend to 
suffer in tandem with the occurrence of disaster, the literature doesn?t contain much detail 
regarding how family closeness may be influenced. Furthermore, if family relationships 
suffer, it will be important for the research literature to take into account the possible 
influences of disaster on marriage as well. 
Stressful Experiences and Marriage 
  Unfortunately, there is no existing research to draw upon to explain how adverse 
mental or physical health outcomes from natural disasters relate to marriage. However, 
mental health outcomes and marriage have been studied within the context of war (Dekel 
& Solomon, 2006; Renshaw et al., 2008). Wars and natural disasters have key differences 
that make them distinct experiences. Those affected by war may be embittered by a sense 
 
29 
 
of vengeance against the enemy (Ai, Peterson & Huang, 2003), while those affected by a 
disaster may blame a variety of influences and factors for their losses (Smith et al., 2000). 
One of the most traumatizing aspects of natural disaster is the unexpected nature of most 
of them (Assanangkornchai et al., 2004). Those traumatized by war experiences may not 
anticipate specific traumas, but soldiers are likely prepared to some degree for the danger 
they face. Natural disasters and war are also similar in that they both are comprised of 
varying traumatic events, and usually affect a large number of people simultaneously 
(Menzel Baker, Hunt & Rittenburg, 2007). Researchers studying war have queried about 
whether poor marital outcomes are due to traumas or mental health outcomes due to 
trauma itself (Dekel & Solomon, 2006). 
 Marital relations in a study by Dekel and Solomon (2006) were conceptualized by 
marital adjustment, spousal aggression, and sexual satisfaction. The variables were 
compared across three groups. The comparison groups comprised those who were 
prisoners of war (POWs) with PTSD, POWs without PTSD, and veterans who were not 
POWs. General findings indicated that the marital difficulties experienced by POWs were 
indeed related to PTSD symptoms. Also, POW?s number of PTSD symptoms associated 
significantly with marital adjustment (r = -.43, p = <.001), verbal aggression (r = .41, p = 
<.001), physical aggression (r = .23, p = <.05), and sexual satisfaction (r = -.27, p = 
<.001). Regarding whether marital difficulty was due to PTSD symptoms or the trauma 
experience of being a POW, there was some evidence to suggest that the experience itself 
controlling for PTSD symptoms had a significant relation to lower quality marital 
adjustment. This was noted in the finding that the POWs without PTSD were two times 
 
30 
 
as likely to score beneath the clinical cutoff for the Dyadic Adjustment scale (DAS; 
Spanier, 1976) than the control group (Dekel & Solomon, 2006).  
Dekel and Solomon?s study (2006) clarified that PTSD symptoms, trauma 
experiences and adverse marital outcomes are related. Further, the researchers noted that 
beyond the adverse influence of mental health problems, other aspects of the traumatic 
experience appear to relate negatively to marital adjustment. Renshaw, Rodrigues and 
Jones (2008) presented findings which further support that negative mental health status, 
following a traumatic event or events, associates negatively with indicators of marital 
functioning. Using the Relationship Assessment Scale (RAS; Hendrick, 1988), a 7 item 
questionnaire with Likert-type scale indicators, Renshaw et al. (2008) investigated 
marital satisfaction post national guard duty in Iraq. They also included measures of 
PTSD, depressive symptoms and then asked each respondent to give the same 
information about their spouse. They found that when spouses perceived that their 
husband experienced of high levels of combat exposure (i.e. controlling for level of 
combat exposure), the combat exposure reported by the soldier didn?t relate significantly 
to the spouse?s marital satisfaction (r = .-07, p = .59). On the other hand, when the spouse 
perceived that their husband experienced low levels of combat exposure, the soldier?s 
report of his combat exposure was significantly, negatively related to spouse?s marital 
satisfaction (r = .-.44, p = .001). This suggests a possible moderation of spouse?s 
perception of exposure on the relationship between self-reported exposure and spouse 
marital satisfaction. Links between combat exposure and marital satisfaction in spouses 
have not been found in previous research. Furthermore, when spouse?s perceived low 
 
31 
 
levels of combat exposure, soldier self-report of symptoms was negatively related to 
spouse marital satisfaction (r = .-46, p < .001 for PTSD symptoms and r = -.48, p = <.001 
for depressive symptoms). This suggests that spouse?s perception of combat exposure 
levels also may moderate the relationship between self-reported symptoms and spouse 
marital satisfaction.  
A negative relationship appears to exist between persisting physical health 
problems and family relationships (Booth & Johnson, 1994; Murray et al., 2007; Trief et 
al., 2006). Persisting physical health problems resulting from natural disasters have not 
been studied in a family context, however. Adverse mental health outcomes such as 
PTSD and depression are more often studied as disaster outcomes (Assanangkornchai et 
al., 2004; Catapano et al., 2001; Suar et al., 2002). Some researchers have indicated an 
increase in family problems or conflict after disaster (Catapano et al., 2001; Suar et al., 
2002). However, we still do not know how or if these problems persist along with other 
lingering effects of disaster, including mental health problems. Researchers have 
documented findings suggesting that other traumatic experiences such as combat 
exposure in war, relate to both adverse mental health outcomes, and marital functioning 
problems (Dekel & Solomon, 2006; Renshaw et al., 2008). The distinct mental health and 
physical health consequences of disaster will also need to be considered in their effect on 
marriage if we are to obtain a comprehensive understanding of trauma, mental health, and 
family relationships. These findings are in general support of the COR model (Hobfoll, 
1989) as losses incurred in natural disasters including the loss of physical and mental 
health would naturally create stress.  
 
32 
 
Disasters and Family Relationships 
Some of this research supporting the COR model takes place within the disaster 
context (Catapano et al. 2001; Suar et al., 2002), while some of it does not (Dekel & 
Solomon, 2006; Renshaw et al., 2008, Trief et al., 2006). Some of these studies are 
inclusive of family relationship variables (Catapano et al., 2001). However, no studies 
exist which primarily investigate the influence of disaster on family closeness. Studies 
connecting natural disasters with family functioning have included family closeness as a 
protective factor against mental distress post disaster (Wickrama & Wickrama, 2007), 
and positive parent child relationships as predictive of lower mental health distress after 
disaster (Wickrama & Kaspar, 2007). Family variables have also been included in studies 
to understand what coping styles families use predict the best mental health outcomes for 
family members (Vigil & Geary, 2008), and also what parent-child interaction styles 
before disaster predict mental health symptom distress (Proctor et al., 2007). These 
disaster studies are often conducted rather soon after the disaster occurrence to 
understand the breadth of immediate effects that ensue (Catapano et al., 2001; Wickrama 
& Kaspar, 2007; Wickrama & Wickrama, 2007).  
Wickrama and Wickrama (2007) studied the family context of mental health risk 
in tsunami affected mothers in Sri Lanka. The sample was comprised of 325 mothers 
from Sri Lanka whose children were mostly over age 10. Respondent families were 
identified and contacted through voter registration records. They were recruited from 2 
villages in Sri Lanka, Kudawella-W and Kudawella-O. Site selection from over 100 
villages in the Hambantota district was based on poverty levels, rurality and accessibility. 
 
33 
 
All families contacted for the study were living in temporary housing in their own 
village. The data was collected 3-4 months after the tsunami struck.   
 The researchers included a measure of tsunami exposure to assess the degree of 
destruction survivors? experiences to their physical possessions and lives of family and 
close individuals. First, a damages index was created (? = .76) to measure the severity of 
tsunami damages to property. Six questions, all rated on a Lickert type scale ranging from 
1 (not at all) to 4 (complete) asked about the extent of damage to the home, livestock, 
vehicles and other possessions. Another 2 item scale was created to indicate the amount 
of life destruction sustained by survivors. It included questions regarding the number of 
deaths and injuries to family members, close relatives and neighbors. Pre-tsunami 
poverty was assessed using 7 items from a scale adapted from Conger et al., (1994). The 
first six items identified whether the family had money for the things they needed 
including leisure activities, food, clothing and equipment. The last question asked about 
whether the pre-tsunami income matched family expenses or not. 
The questions were answered on a five-point scale ranging from ?strongly disagree? to 
?strongly agree.?  
 The researchers measured secondary tsunami risks to understand what conditions 
might explain mental health outcomes. Displacement duration was indicated by asking 
how long families had been out of their homes in days. Next, family conflict, was 
measured by self report of increased family conflict and violence, and decreased support 
from family members following the tsunami. Child problems, another risk factor, were 
measured by asking about the presence of child health problems, and child behavior 
 
34 
 
problems that currently influenced activities following the tsunami. Family conflict and 
child problems were measured on the same 4 point Lickert scale used for property 
destruction. The two scales were highly correlated (r = .26, p < 0.01) and were combined 
to capture a latent construct of family problems. Resilience factors were also taken into 
account and included ?familism,? marital status, number of children, community support, 
religious participation, and hardiness. Fourteen items on a 5 point scale ranging from 
?strongly disagree? to ?strongly agree? were adapted from a scale by Portes and Rumbaut 
(2001) to measure family closeness and cohesion, or ?familism.? Respondents also 
indicated whether they participated in community activities pre-tsunami. They were also 
asked regarding neighbors before the tsunami, specifically about whether neighbors 
looked after each other in times of trouble and looked out for one another?s children. The 
community and neighbor items were combined as a latent construct of community 
support. Religious activity was captured with only one item asking about frequency of 
attendance at religious services. Mental hardiness was assessed by questions pertaining to 
respondent hope for the future and confidence in their ability to persevere. 
 Mental health outcomes (e.g. Depression, PTSD) were assessed using pre-
established measures. Depression was indicated by a translated version of the Center for 
epidemiological Studies Depression Scale (CES-D; Radloff, 1977). PTSD symptom 
levels were assessed using diagnostic interview questions from the DSM-IV (American 
Psychiatric Association, 1994).  
Wickrama and Wickrama (2007) found that family problems mediated symptoms 
of depression and PTSD and property destruction. The researchers used t coefficients to 
 
35 
 
test for significant effects in the sample. All t statistics greater than 1.96 for the degrees 
of freedom in the models were significant. The direct effect of property destruction was 
significantly predictive of depressive symptoms (?=.69, t = 6.51), after adding family 
problems to the model, the influence of property destruction was significantly reduced (? 
= .22). The direct effect of property destruction was also significantly predictive of PTSD 
symptoms (? = .26, t = 4.52). After adding family problems to the model, the influence of 
property destruction was again significantly reduced (? = .11) (Wickrama & Wickrama, 
2007). Family problems in previous research have been defined as a ?secondary risk? 
post disaster (Kaniasty & Norris, 1993), and perhaps wrongly so in light of the findings 
of Wickrama and Wickrama (2007).  
However, family problems (Wickrama & Wickrama, 2007) did not mediate 
between life destruction (captured by items indicated mothers? report of the number of 
deaths/injuries sustained among neighbors and family) and the mental health outcomes of 
PTSD and depression. Perhaps this is an indication that the brief conceptualization of 
family problems failed to capture all the ways the family dynamic could be influenced.  
 Resilience factors shown to be protective against depression included familism (? 
= -0.016), community support (? = -0.08), and hardiness (? = -0.22). Those factors found 
to be protective of mothers against PTSD included a smaller number of children (? = -
0.17), religious participation (? = -0.16), and hardiness (? = -0.26).  
 Religion as a resilience factor in the study by Wickrama and Wickrama (2007) 
was measured by one items only indicating frequency of participation. This item did not 
address what attributions, or other religious involvement may have served as a buffer. It 
 
36 
 
appears however, that religion may serve as moderator between tsunami exposure and 
mothers? mental health. To investigate this question, the researchers split the sample at 
the median into high and low religion participation groups. PTSD symptoms were more 
frequent in the low participation group (M = 17.03, SD = 4.36) than in the high 
participation group (M = 14.32, SD = 4.25; p < 0.05). The relationship between property 
destruction and PTSD symptoms was also higher among the low religious group (? = 
0.26) than in the high religious participation group (? = 0.03). This difference was 
significant at a level of p < 0.05 as was the difference in the relationship between life 
destruction and PTSD symptoms. In the group that was considered high on religious 
participation, the relationship between life destruction and PTSD symptoms was ? = 0.06 
whereas for those who were low on religious participation, the same relationship was 
higher (? = 0.20). When a median split analysis was performed findings were similar in 
that the relationship between number of children and PTSD symptoms was ? = -0.28 for 
those in the high familism group, and ? = -0.01in the low familism group. Mothers in the 
low familism group also had significantly higher PTSD scores than those in the high 
familism group (p < 0.05) (Wickrama & Wickrama, 2007).  
 These findings suggest important potential conclusions regarding families and 
disaster. First, mediation analyses suggested that disasters take effect on mental health at 
least in part through relationship disturbances.  If this is true, it?s unfortunate that we 
don?t know more about how relationships are affected since these problems could be a 
major contributor to mental health issues. Additionally, religious participation appears to 
be an important protective factor against PTSD for mothers who have experienced 
 
37 
 
disaster. More specifically, these findings indicate that mothers who attend religious 
services frequently, tend to experience a lesser degree of posttraumatic stress regardless 
of the intensity of disaster exposure than those who do not attend religious services as 
frequently. However, the moderation of family cohesion on the relationship between the 
number of children and PTSD indicates that mothers with high family cohesion 
experience greater PTSD symptoms when they have a higher number of children. This 
could be explained in that those with high levels of cohesion are likely highly invested in 
their families and may be more stressed by the tsunami experience when caring for more 
children. Even so, the average levels of PTSD symptoms were lower on average among 
mothers in the high family cohesion group as opposed to the low family cohesion group. 
This suggests that family cohesion is protective against PTSD in the disaster context. 
Similar findings were uncovered regarding marital status. Those mothers that were 
married experienced less depressive and PTSD symptoms than unmarried mothers on 
average. Further study of disaster is warranted within the family context given that close 
family ties tend to protect disaster exposed mothers from PTSD. Another impetus for 
further study is that family problems appear to serve as a vehicle for disaster exposure to 
threaten mental health. It would be interesting to rearrange research questions to 
understand how family closeness and other family variables as outcomes are influenced 
and protected when controlling for mental health problems. 
 Using the same sample, Wickrama and Kaspar (2007) made several hypotheses 
regarding tsunami exposure and the family. First, they hypothesized that high levels of 
tsunami exposure would relate to high levels of PTSD symptoms in affected adolescents. 
 
38 
 
Secondly, the researchers hypothesized that high levels of tsunami exposure would also 
relate to high levels of depressive symptoms. A third hypothesis suggested that 
psychosocial losses (days of displacement, social losses, family losses and mother 
depression) would exacerbate depressive and posttraumatic stress symptoms. Fourth, it 
was hypothesized that parent child relationships primarily characterized as positive would 
associate with lower levels of tsunami related PTSD and depressive symptoms in 
adolescents. The fifth and final hypothesis stated that the depressive symptoms of 
mothers would moderate, in an amplifying fashion, the relationship between psychosocial 
losses and PTSD and depressive symptoms.  
Measures used in Wickrama and Kaspar?s 2007 study to indicate tsunami 
exposure (i.e. property destruction and life destruction) were the same as in Wickrama 
and Wickrama?s 2007 study. Psychosocial losses were measured by several indicators 
including the number of days that the adolescent was displaced from home. Social losses 
were indicated by an item asking about the disruption of playgroups, and other 
community and peer groups. Family loss was a variable constructed (? = .62) to indicate 
the level of damages incurred on families both emotionally and fiscally. Specifically, 
respondents were asked to rate the level of or existence of problems influencing daily 
activities such as increased family violence/conflicts, increased parental alcohol 
consumption, damage to parental occupation, and parental mental health problems. 
Responses were given on a 1-4 scale ranging from 1 (not at all) and 4 (complete).  
Parental child relationship quality (? = .75) was assessed by mothers? responses to 
five items asking about getting along with the child, understanding the child, child 
 
39 
 
obedience, child level of demand, and child insistence on having their way.  This scale 
yielded a high internal consistency coefficient (? = .85). Depression of mothers was 
measured with the CES-D and demonstrated adequate internal consistency (? = .71) 
(Radloff, 1977). Ten items from the Youth Self Report (YSR) precipitating from the 
Child Behavior Checklist (CBCL; Achenbach) were used to assess depressive symptoms 
in adolescents (? = .80). PTSD symptoms in both mothers (? = .61) and adolescents (? = 
.65) were assessed using17 items from the DSM-IV diagnostic interview (American 
Psychiatric Association, 1994). 
Results supported hypotheses 1 and 2. Property destruction predicted variance in 
depressive symptoms (? = .28, p < .01) and PTSD symptoms (? = .24, p < .01). Loss of 
lives also contributed to adolescent PTSD symptoms (? = .19, p < .01) and depressive 
symptoms (? = .13, p < .01). Taken together, property destruction and loss of lives 
(tsunami exposure) accounted for 10% of the variance in both PTSD and depressive 
symptoms in affected adolescents. Hypothesis 3 was also supported. Psychosocial losses 
predicted variance in adolescent depressive symptoms even when controlling for tsunami 
exposure. Specifically, number of days displaced (? = .20, p < .01), social losses (? = .10, 
p < .01), family loss (? = .30, p < .01), and depressive symptoms of the adolescent?s 
mother (? = .23, p < .01) accounted for 38% of the variance in depressive symptoms. 
These same predictors accounted for 26% of the variance in PTSD symptoms although 
only family loss (? = .29, p < .01) and mother?s depression (? = .28, p < .01) significantly 
predicted adolescent PTSD.  
 
40 
 
Hypothesis 4 was also supported by the results of Wickrama and Kaspar?s study 
(2007). A positive relationship between mother and adolescent was related to 
significantly lower PTSD (? = -.10, p < .01) and depressive (? = -.12, p < .01) symptoms. 
Finally, hypothesis 5 was supported in part. To test for an amplifying moderation effect 
of mother?s depression on the relationship between psychosocial losses and adolescent 
depressive and PTSD symptoms, the sample of mothers was split at the mean of 
depression scores. In adolescents with mothers who demonstrated higher depression 
scores, social loss (? = .16, p < .01), family loss (? = .36, p < .01), and days of being 
displaced from home (? = .20, p < .01) related to depressive symptoms. Family loss was 
significantly related to PTSD symptom levels (? = .35, p < .01) for adolescents whose 
mothers had high depression scores. For adolescents with mothers who had lower 
depression scores, psychosocial losses didn?t relate significantly to symptom levels of 
PTSD or depression (Wickrama & Kaspar, 2007). 
Most scales in this study were adequate with regard to internal consistency and so 
likely yield results that indicate important patterns between tsunami exposure and mental 
health. Finding that adolescent symptom levels are related to the degree of disaster 
exposure is in keeping with other research findings regarding adult survivors of disaster 
(Suar et al., 2002).  Family loss was found to contribute to PTSD and depression in 
tsunami exposed adolescents. The way this variable was conceptualized was diverse as it 
included consideration of family violence, and occupational disruption of parents. Indeed, 
the low internal consistency coefficient revealed that the items comprising this measure 
may be more distinct than congruent in measuring the effects of disaster on families and 
 
41 
 
should perhaps be separate constructs in research. The finding that family loss 
contributed to adolescent symptoms has excellent face validity. However, research 
literature would benefit from a deeper exploration of how these distinct troubling family 
losses influences positive family functioning (e.g. closeness and cohesion). 
Further research studying the family context after disaster and adolescent mental 
health was completed two months after Hurricane Katrina ravaged the southeast United 
States. Researchers Vigil and Geary (2008) studied family coping styles and the influence 
of those styles on adolescent mental health. Specifically, the researchers wanted to know 
if family coping styles affect the relationship between hurricane exposure and 
psychological well being. They also wanted to know how family coping strategies and 
psychological well-being of affected and non-affected adolescent compared.  
The sample consisted of 81 (59 female) adolescents aged 12-17 years (M = 14.4 
years). Families living in relocation camps for about 3 weeks were recruited for study. 
Data from a control group (31 of the original sample) was collected almost a year after 
the hurricane. Initial analyses failed to detect significant demographic differences in the 
two samples. 
In order to assess family coping style, 10 items from the Family Crisis Oriented 
Personal Evaluation Scale were used (McCubbin, Larsen & Olson, 1987). These items 
captured pieces of all five original subscales (seeking social support, reframing, seeking 
spiritual support, mobilization, and passive appraisal). However, Cronbach?s alpha 
analyses revealed sub-adequate levels of reliability in this sample (between <.40 and .60) 
raising questions as to the utility of this scale to accurately assess family functioning in 
 
42 
 
response to disaster. Global self-esteem of adolescents was measured using the 
Rosenberg Self-Esteem Scale (RSES; Rosenberg, 1965). Its 10 items are scored on a 
lickert type scale ranging from 1 (rarely or never) to 4 (almost always) and demonstrated 
marginally adequate internal consistency (? = .69). Subjective psychological distress was 
assessed using the Impact of Events Scale-Revised (IES-R; Weiss & Marmar, 1997). The 
22 items were revised to refer to Hurricane Katrina specifically and assess PTSD 
symptoms in the past week. Responses ranged from 1 (rarely or no days) to 5 (almost 
always or 7 days) and exhibited good internal consistency (? = .93). Anxiety symptoms 
were measured using Reynolds and Richmond?s (1978) Revised Children?s Manifest 
Anxiety Scale (RCMAS) with 37 yes or no questions (? = .87). Finally, depressive 
symptoms (in the last week) were assessed using the Center for epidemiologic Studies 
Depression Scale (CES-D; Radloff, 1977). The 20 questions were answered with 
responses ranging from 1 (rarely or less than 1 day) to 4 (almost always or 5-7 days). 
The scale demonstrated good internal consistency (? = .83).  
To address their first question, Vigil and Geary (2008) compared the Katrina-
affected and non-affected samples in terms of problem solving or coping behavior. The 
group affected by Katrina reported higher ?family mobilizing? coping scores than the 
non-affected group. Family mobilizing as a coping style involves seeking community and 
institutional support at higher levels than other styles. The Katrina-affected group also 
demonstrated significantly lower self-esteem (t = -3.05, p < .01), higher distress 
(avoidance) (t = 2.14, p < .05) and depressive symptoms (t = 2.02, p < .05). What this 
indicates is that disaster affected families in this sample coped with stress by using more 
 
43 
 
community and institutional resources to solve problems than non-affected families. 
Also, the teens in the affected group exhibited poorer mental health. Vigil and Geary 
(2008) then investigated whether the relationship between Hurricane exposure (i.e. being 
in the affected group) and mental health outcomes was moderated by family coping style. 
In order to answer this second question, Vigil and Geary used simultaneous 
regression analyses and found that previously significant relationships between 
psychological distress scores (b = 8.65, p < .05), self-esteem scores (b = -2.97, p < .01), 
and depression (b = 4.50, p < .05) and group category (i.e. whether or not the adolescent 
was in the affected group) became insignificant when mobilizing as a family coping style 
was controlled for. Further analyses revealed ?mobilized? coping as a mediator of the 
relationship between lower self-esteem, psychological distress and having experienced 
Hurricane Katrina. Mediation between depression and Hurricane experience approached 
significance as well (Z = 1.91, p = .06) (Vigil & Geary, 2008).  
These findings indicate that those families whose primary coping style involved 
seeking community and institutional support had adolescents with worse mental health 
problems. However, this could be a spurious finding as it would be safe to assume that 
families whose adolescents have problems are more likely to seek community and 
institutional support. This was manifest in a mediation effect observed of adolescent 
hurricane experience between lower self-esteem and higher symptoms of distress (Vigil 
& Geary, 2008).  
Proctor et al. (2007) also studied family context in relation to disaster response in 
children. In an investigation of family conflict before the 1994 Northridge earthquake hit, 
 
44 
 
affective quality in family interaction was being studied. Roughly 9 months after the 
earthquake, participating families were mailed surveys to complete regarding disaster 
exposure. Most of the two parent families (117 of 144) completed the follow up 
assessment post-earthquake. Parents in the sample had been together for an average of 
9.6 years (SD = 3.5) and lived together with the target child at the time of the pre-
earthquake assessment. Children were roughly 6.5 years old (SD = .68) and ranged in age 
from 4-8. Pre earthquake behavioral observations afforded researchers with information 
regarding the affective quality of parent-child relationships. Families were filmed during 
a 10 minute structured play task and behaviors were coded as either negative (e.g. 
pressure, impatience, negative mood) or positive (responsiveness, positive reinforcement, 
and positive mood). Observations had high inter-rater reliability with internal consistency 
coefficients ranging from .84 to .90. 
After the earthquake, mothers reported on earthquake impact by giving 
information relating to family physical injury, home damages, displacement from the 
home, material losses, financial expenses, changes in living conditions, lost employment, 
increased workload, altered commute, and children being unable to attend daycare or 
school. Scores on this impact index ranged from 0-17 (M = 4.8, SD = 4.4), the highest 
possible score being 32 (? = .80). Child distress was measured using mother report of 23 
symptoms of child distress after the quake including assessment of PTSD symptoms (? = 
.95). Mothers reported 2 times post earthquake, once within the first month afterward, 
and once 8 months post quake to assess for persistence in child distress. Items were 
scored on a scale from 1 (none) to 4 (a lot). Scores were summed and the 8 month scores 
 
45 
 
were used in mediation and moderation analyses to represent persistent symptoms. 
Finally, post-earthquake parental stress was also assessed by asking mothers to respond 
to 72 items measuring symptoms of PTSD, depression and somatic concerns on the same 
scale as child distress. Mothers reported for themselves and their husbands. Indicators of 
relationship difficulties captured by marital conflict behavior items were also included 
although marital difficulties were not mentioned in the analyses. Mothers also reported 
on these symptoms twice, once a month after the earthquake, and again 8 months 
afterward (? = .94). The researchers used mothers? one month scores in meditational 
analyses as they were interested in determining what effect immediate parent stress had 
on persistent child distress.  
The researchers found that 59% of the sample reported home damage, 50% lost 
material possessions, 39% spent a mean of $10,298 on post disaster expenses, 33% had 
an altered commute after the disaster, 20% had children who were kept out of school for 
over a week, 12% experienced injury of a family member, 5% had to move, and 5% 
reported that one parent lost employment. Additionally, 26% received financial assistance 
from the Federal Emergency Management Agency.  
Children displayed significantly higher distress scores one month after the quake 
(M = 15.40) than 8 months afterward (M = 7.66) as was expected (t (116) = 10.0, p < 
.001). The researchers tested whether affective quality of interaction before the 
earthquake moderated the relationship between earthquake impact and persistent child 
distress. In all regression equations for girls and boys, earthquake impact as a main effect 
was a significant predictor of persistent distress. Additionally, mothers? negative behavior 
 
46 
 
significantly predicted girls? persistent distress, and taken together with earthquake 
impact predicted 29% of the variance in girls? persistent distress. A significant interaction 
was found between father negative behavior and earthquake impact in predicting girls 
persisting distress (? = -.24, p <.05). This suggests that fathers? negative interaction 
behavior pre-earthquake moderated the relationship between girls? persistent distress and 
earthquake impact. More specifically, girls? experiencing negative father behavior before 
the earthquake displayed higher persisting distress scores across a range of earthquake 
impact. Whereas those experiencing low negative behavior from fathers varied in distress 
more as a function of earthquake impact.  This suggests that family relationships existing 
before disaster for girls are important in predicting mental health outcomes.  
To test whether earthquake impact influenced child psychological well-being 
through parental stress, the researchers performed meditational analyses. Full mediation 
was supported for boys in that the beta coefficient was significantly reduced from a 
significant .42 to a non-significant .06 when controlling for parent post-earthquake stress. 
In fact, 86% of the effect of earthquake impact influenced boys? distress through parental 
stress. Partial mediation was supported for girls as well, while earthquake impact was still 
a significant predictor of persisting distress after controlling for parental stress, the beta 
coefficient was still reduced from .48 to .34. Twenty-nine percent of the effect of 
earthquake impact influenced girls through parental stress.  
The researchers performed post hoc analysis to assess whether mothers? positive 
behavior with boys? and fathers? negative behavior with girls acted as a moderator for the 
mediator pathway between earthquake impact and post-earthquake parental stress and 
 
47 
 
child distress. There was a significant interaction between mothers? positive behavior and 
earthquake impact (? = .28, p < .05) in prediction of parenting stress which mediated the 
relationship between earthquake impact and child distress. This finding indicated that 
parents who displayed positive parenting behavior were highly stressed as a function of 
earthquake impact (? = .72, p < .001) whereas parents showing low amounts of positive 
parenting behavior showed no significant relationship between earthquake impact and 
parental stress (Proctor et al., 2007). This indicates that parents exhibiting low amounts of 
positive parenting did not experience stress as a function of disaster impact. What this 
may mean is that these parents stress levels were not proportional to the impact they 
experienced, but fluctuated as a function of something else. 
Family responses to disaster influence persistent distress, and also, family 
interaction before disaster predicts adverse mental health outcomes for children 
afterward. Additionally, from this study it appears that the intensity of earthquake impact, 
including assessments of injury, influences child mental health through parental stress 
response to these conditions. This finding supports the need to control for mental health 
status in studies of disaster. Controlling for mental health will help to increase 
understanding of how resultant disaster conditions such as injury and illness influence not 
only mental health of survivors, but family dynamics after disaster. This study indicated 
that family interaction and parental distress affect how survivors are influenced by natural 
disasters. However we still need to know more about how these behavior interactions 
may have changed as a result of disaster. The way that families cope with stress and 
interact appears to be influential in predicting mental health outcomes for children. It is 
 
48 
 
also evident that parental stress and adverse mental health influences the coping ability of 
children and adolescents (Proctor et al., 2007; Vigil & Geary, 2008). 
Disaster and Long Term Mental Health Outcomes 
Researchers have documented mental health status and other variables resultant of 
disasters in the short term context. The data analyzed in Wickrama and Wickrama (2007) 
and in Wickrama and Kaspar (2007) was collected 3-4 months after the tsunami hit. The 
study by Suar et al. (2002) retrieved data from respondents 3 months after the Orissa, 
India supercyclone. Information from survivors of the Hat Yai flood (Assanangkornchai 
et al., 2004) was obtained even more immediately to the disaster, just 10 weeks. Catapano 
et al. (2001) studied psychological outcomes of landslide survivors a year after it 
happened and found that the degree of loss was significantly associated with negative 
mental health outcomes. Also, researchers studying the effects of war trauma on marriage 
found similar detrimental results both 3 months (Renshaw et al., 2008) and 3 decades 
(Dekel & Solomon, 2006) after soldiers returned home. This suggests that traumatic 
experiences influence long term mental health outcomes. Research has been conducted 
which addresses the long term mental health implications of experiencing a disaster as 
well (Crabbs & Heffron, 1981; Kaniasty & Norris, 1993; Lindy & Titchener, 1983; North 
et al., 2008; Wu et al., 2006). 
 Crabbs and Heffron (1981) indicated that what amount of time needs to pass to 
assess the long term versus the short term effects of disaster is ambiguous. However they 
also state that short term effects are often considered in days or weeks while long term 
effects are investigated after months or years. Other differences help to clarify what may 
 
49 
 
constitute short term versus long term effects of natural disaster such as what survivors 
need. For example, immediate needs such as food and clothing tend to represent short 
term needs and results of disaster while emotional responses and mental distress 
surrounding reminders of the event may suggest a long term effect.  
 North, Hong, Suris, and Spitznagel (2008) studied survivors of the 
Oakland/Berkeley Firestorm that took place in 1991. Their purpose was to understand the 
sequelae of mental health disorders sustained by survivors. The selected participants for 
their study by identifying residential areas in the direct path of the firestorm leaving 
homes damaged or destroyed. Letters of invitation to participate were sent to 191 of these 
homes, but as many of them were no longer habited because of evacuation, the 
researchers were unsure as to who of those invited actually received letters. Of the 86 
households responding to their invitation, 69 individuals (one from each home) were then 
interviewed. These interviews were conducted 4, 16, and 39 months after the firestorm 
hit. For the final sample, Hong et al. (2008) used data from the 62 respondents who 
completed all follow up interviews.  
 In order to assess how the experience of this disaster influenced mental health 
over time of survivors, the researchers used the DSM-III-R Diagnostic Interview 
Schedule (Robins, Helzer, Cottler, & Goldring, 1989). They also used the disaster 
supplement to this schedule (DIS; Robins & Smith, 1983) this measure and its 
supplement was used at all interview points.  Only sections of the DIS were used, the 
researchers based their decisions on what items to administer based on previous research 
to gather the most pertinent information from respondents. Sections of the DIS used 
 
50 
 
included demographic information, and several assessments of psychiatric disorders 
including PTSD and depression. The disaster supplement to the DIS assessed survivor 
exposure to the fire, danger, and family or friend exposure. It also included indicators of 
negative life events before and after the disaster. The Temperament and Character 
Inventory was used (TCI; Cloninger, 1992) at 16 months post disaster. It is a 240 item 
self report instrument to measure different domains of personality. These included harm 
avoidance, novelty seeking, reward dependence, and persistence. There were also three 
aspects of character measured with this instrument that included being self-directed, 
cooperative, and self-transcendent.  
 To analyze the data they collected, the researchers compared mental health 
outcomes in respondents based on their level of exposure to the disaster and other 
demographic indicators. The final respondent sample was mostly married, well-educated, 
older, non-ethnic females.  Four months after the firestorm, the researchers noted that 
most respondents (with three exceptions) with psychiatric disorder diagnoses after the 
disaster had them before the firestorm occurred. However, there was evidence of post-
traumatic symptoms and distress was wide-spread. The number of PTSD symptoms 
experienced was positively significantly associated with being female (M=4.8, SD=5.8 
vs. M=2.1, SD=3.4; t = 3.07, df = 56, p=.003).  At the three year follow up, one of those 
diagnosed with PTSD after disaster had recovered, but three more cases met diagnostic 
criteria, and one case met diagnostic criteria at the one year follow up meaning that three 
years after the disaster, the rate of PTSD among the sample was increased to 11%. These 
cases all had PTSD symptoms at the first time of measurement, but did not exceed the 
 
51 
 
diagnostic threshold, and 5 of the 7 had a psychiatric disorder pre-disaster. This suggests 
that perhaps disaster exacerbated mental health vulnerabilities rather than serving to 
produce mental health problems at the diagnosable level. A similar finding held true with 
major depression, with 4 additional cases being identified at the first follow-up interview. 
However, three of these four cases moved into remission at the three year interview, as 
well as the original 8 cases identified at the first interview. This suggests that depressive 
symptoms subsided over time in this population perhaps indicating that depressive 
symptoms are strongest right after a disaster. The researchers found that reports of self 
harm decreased a significant 32% (p=.003) over the follow up interviews, as did the 
report of feeling ?very upset? (15% decrease; p=.007).  
 Survivors of the firestorm exhibited personality differences based on community 
comparisons. The researchers also found that certain personality characteristics were 
related to adverse mental health diagnoses. Those who had experienced the firestorm 
tended to score significantly higher on personality sub-scores of cooperation, and self-
directedness. Survivors also tended to score lower on self-transcendence. The harm 
avoidance sub-scale was significantly related to a PTSD diagnosis. 
 What these findings suggest is that the survivors in this sample tended to be 
resilient over time with regard to mental health diagnoses, but that the experience of the 
disaster may have exacerbated pre-existing mental health vulnerabilities for some. Crabbs 
and Heffron (1981) noted that after a natural disaster, marriage partners with 
underdeveloped coping skills may be more likely to attend to their own needs and 
problems while neglecting their partners? which could reasonably be expected to create 
 
52 
 
problems in marital satisfaction. This suggests that what North et al. (2008) found 
suggesting that disasters may exacerbate mental health vulnerability may be true for 
vulnerable marriages as well. 
 Also, it appears that over time, depressive symptoms tended to subside in the 
sample suggesting that perhaps depression fluctuates as a function of chronological 
proximity to the disaster. The researchers made a specific point to distinguish between 
psychiatric diagnoses and symptoms noting that each are eased with different 
interventions, but both need consideration and attention in response to disaster. 
 With regard to personality findings, survivors of the disaster tended to be more 
cooperative and self-directed than their peers who had not experienced a disaster. The fire 
storm survivors and community sample did not differ significantly in their harm 
avoidance tendencies suggesting that experiencing a disaster, may not contribute to this 
personality characteristic.  There is need to continue this line of research to learn more 
regarding disaster related family and marital problems long after a disaster has occurred.  
 While disaster may exacerbate pre-existing vulnerability in mental health status 
(North et al., 2008) and marital functioning (Crabbs & Heffron, 1983), other evidence 
suggests that experiencing a disaster may create vulnerability. A study involving 222 
older adults who had survived a severe flood in Kentucky served as a test of the social 
support deterioration model. This model posits that psychosocial loss operates as a 
vehicle for disaster to influence mental health outcomes. It is in fitting with the COR 
(Hobfoll, 1989) model which suggests loss to be the culprit creating stress when changes 
occur.  
 
53 
 
Originally the sample of adults was involved in another study, and the occurrence of 
the flood prompted further inquisition afterwards to assess social support deterioration after the 
flood. The final sample was split into three sub-groups based on the level of exposure 
respondents had to the disaster. Thirty-seven respondents were considered primary 
victims because they experienced direct losses of a physical or material nature. 
Secondary victims of which there were 106, did not sustain direct loss, but resided in an 
affected area. Finally, 79 nonvictims were those respondents who lived in nearby 
counties and didn?t experience flooding. Data was collected three months before the 
flood, and two times afterward. Time three was 9-12 months post flood.  
Three types of social support were considered in this study, social embeddedness, 
perceived support from nonkin, and perceived support from kin. Social embeddedness 
was defined in this study as the ?number of connections people have with significant 
others in their social environments? (Kaniasty, Norris, 1993, p. 400). Perceived social 
support is defined as whether or not a person thinks they are reliably connected to others, 
so that in the event that support is needed, it would be provided. To assess these 
variables, the researchers used nine items from the Louisville Social Support Scale 
(LSSS; Norris, Murrell, 1987).  Depression was measured using the CES-D (Radloff, 
1977) demonstrating an internal consistency equal to .89. Personal loss was measured on 
a 5 point likert scale ranging from 1 (minimal damage) to 5 (total destruction) and was 
subjective in nature. Community destruction was assessed by examining data from the 
Kentucky Division of Disaster and Emergency Services to indicate the number of homes 
 
54 
 
destroyed in relation to the population of each county. In the analyses, the researchers 
controlled for sex, marital status, and education. 
Kaniasty and Norris hypothesized that disaster exposure would influence depression 
both indirectly and directly through the loss of social support (social embeddedness, perceived 
social support from nonkin, and perceived social support from kin). To test this, they 
employed structural equation modeling to test the relationships between these constructs. 
The researchers fit the first model to the data which exhibited good fit (GFI = .97). 
Personal loss at time 2 influenced depression at time 2 only (t = 2.61, p < .01) and 
community destruction at time 2 only influenced depression at time 3 (t = 3.11, p < .005).  
A second model was fit three times for each type of social support. For social 
embeddedness (GFI = .94), mediation was not supported because personal loss did not 
have a significant effect on embeddedness. For nonkin support (GFI = .95), mediation 
was supported between personal loss and depression at time2. Personal loss was 
positively related to nonkin support at time 2 (? = .33, p < .001), and nonkin support at 
time 2 was also significantly related to depression at time 2 (? = .18, p < .01).  The path 
between personal loss to time 2 depression came close to 0 (? = .03, p < .39) when 
nonkin support (time 2) was included in the model. When it was not in the model, the 
relationship between personal loss and time 2 depression was ? = .11 , p < .05. For kin 
support (GFI = .96) mediation was not supported either. The models were refit with 
depression at time 3 as the outcome variable. Community destruction was used as disaster 
stress to predict depression. Results indicated that embeddedness mediated the 
relationship between community destruction and depression at time 3. Nonkin support 
 
55 
 
also mediated between community destruction and depression, kin support did not. These 
results suggest that community destruction led to a deterioration of social support which 
led to greater distress. This study evidences the need for greater understanding of how 
disasters influence support systems long after the disaster has taken place. This study did 
not indicate that a loss of kin support mediated between destruction and mental distress, 
but it did show that community destruction was negatively related to kin support. This 
suggests that the destruction caused by disaster influences family support negatively. It 
may be that other complications from disaster including injury and illness whose effects 
persist over time effect family support or other indications of family relationship 
functioning. 
Physical complications spawned from an earthquake in Taiwan predicted a lower 
quality of life for survivors three years after the disaster (Wu et al., 2006). The instrument 
used to assess quality of life pertained to mental health and physical functioning and not 
relational outcomes. However, if physical problems influence mental health, it is 
reasonable to expect that family and marital relationships may be affected as well.  To 
confirm or refute this expectation, studies including variables of marital and family 
functioning long term after disasters need to be performed. 
It is clear that natural disasters often invoke major changes on the lives of those 
who experience them, both immediately (Suar et al., 2002; Wickrama & Wickrama, 
2007) and long term (North et al., 2006; Wu et al., 2006). Findings evidencing both the 
rise and decline of mental health distress after disaster (North et al., 2006) and make it 
difficult to conclude how disasters affect mental health long after the disaster, to say 
 
56 
 
nothing of physical health, or family relationships. However, findings indicate that 
circumstances inflicted by disaster (e.g. physical health problems) do influence life 
quality long after the disaster (Wu et al., 2006). Also, even mental health distress 
symptoms appear to both rise and fall over time after disaster, it seems that generally, 
symptom distress eases over time (North et al., 2006). Studies investigating cross-
sectional samples of survivors of disasters some time after their occurrence, and 
continued longitudinal study will help to fill in the gaps regarding how disaster exposure 
relates to families long after disasters happen. 
Religious coping 
Disasters leave destruction as they occur, such that family relationships, physical 
health, and mental health of those affected are disturbed (Catapano et al., 2001; Suar et 
al., 2002; Wickrama & Wickrama, 2007). Problems resulting from these losses would be 
expected when considering natural disasters from the lens of the COR model (Hobfoll, 
1989). The losses that accompany disaster tend to provoke stress, and it is safe to assume 
that they do so through the changes that precipitated them. 
The way that those affected by disasters make sense of losing resources influences 
their ability to cope with the stress those losses invoke (Assanangkornchai et al., 2004). 
The efficacy of religion as a facilitator of coping appears to depend on the way it is used 
to explain negative events. Some religious attitudes may aid survivors through a reframe 
of disaster events and by providing hope for the future, although others may invoke doubt 
or anger (Smith et al., 2000).   
 
57 
 
 Reframing loss, and changing the way one thinks about valued resources are some 
of the ways that people cope with stress. These cognitive shifts are used to minimize loss 
in addition to attempts to replace lost resources or expecting a net gain of resources 
through continued endurance. One may choose to lessen the value of the resource(s) they 
have lost (Hobfoll, 1989) 
.  The population considered in the present study is unique with regard to value 
system. The Sri Lankan mothers studied are largely Buddhist. Devaluing resources may 
not require a change in fundamental values as a major tenet of Buddhism is the inherent 
devaluation of all things. In fact, in order to create a sense of wellbeing, one has to 
abandon wants and lusts, and overcome them (Thera, 1993). This orientation may place 
the current study population at an interesting advantage in terms of their ability to use this 
method of managing resource loss. So although it has been suggested that changing 
values to deal with loss can become problematic (Hobfoll, 1989), this may be a moot 
point in a Buddhist population as putting little value on things is an inherent value. 
Unfortunately, we do not know much about how Buddhist philosophy is applied 
or how those applications influence coping with the tsunami. In fact ??there have been 
no systematic evaluations of the relevance of Buddhist ideas and strategies to the 
understanding and management of trauma reactions following the tsunami?? (De Silva, 
2006 p. 284). However, more recent research has begun to include basic measurements of 
frequency of religious activity in tsunami studies (Wickrama, & Wickrama, 2007). It is 
important to give attention to the ways in which non-theistic philosophies operate in 
times of stress, especially with regard to the tsunami given the massive amount of people 
 
58 
 
it affected (Anderson, 2007; Wickrama & Wickrama, 2007). This is especially true since 
Buddhism is a largely influential religious tradition in the Eastern world (Hsu, O?Connor, 
& Lee, 2009). Much of what we know regarding religion and stress focuses around a 
Judeo-Christian perspective (Ai, Peterson & Huang, 2003; Grame et al., 1999; Shaw, 
Joseph & Linley, 2005). There is research providing evidence of the workings of theistic 
faiths in the natural disaster context (Menzel Baker, Hunt & Rittenburg, 2007; Smith et 
al., 2000). Findings indicate that religious activities may show an investment that many 
may make for an expected return of a deeper closeness with God, an increased closeness 
with one?s congregation or church, or an increase in spirituality. For example, prayer and 
church attendance were related to positive religious outcomes after a natural disaster in 
the American Midwest (Smith et al., 2000). Buddhist philosophy is largely ritualistic (De 
Silva, 2006), and there is a need for studies to investigate how Buddhist religious 
activities relate to disaster outcomes in addition to Judeo-Christian religious activities.   
It is paramount to understand differences in outcomes regarding the application of 
religious beliefs (in addition to religious activities) to disaster losses, as different 
philosophies may relate to different outcomes. Also, it is clear that intra-faith differences 
exist depending on how one chooses to fit their beliefs to make sense of a given situation. 
For example, an attitude of blaming God for losses has been shown to relate negatively to 
mental health outcomes following disaster (Smith et al., 2000). This indicates a need to 
understand how religion is used to cognitively construct meaning of events in addition to 
knowing about respondent religious affiliation.   
 
59 
 
A study by Davidson, Connor, and Lee (2005) attempted to understand how the 
use of karma and reincarnation related to resilience. The sample was taken from the 
United States and included 1,969 respondents who offered information regarding 
traumatic experiences, religious beliefs and resilience via internet. The sample was 
contacted through a random digit dialing procedure. The sample was adjusted using post 
stratification weights so as to render the sample nationally representative. 
General health was assessed using respondent rating of their physical and mental 
health as excellent, good, fair, or poor. Most respondents rated themselves as excellent or 
good, so two level variables were created by combining excellent and good to be ?good? 
and fair and poor to be ?poor.? Resilience was measured using 11 of 25 original items 
from the Connor-Davidson Resilience Scale (CD-RISC; Connor & Davidson, 2003). 
Control, commitment, challenge, goal-orientation, self-esteem, adaptability, social skills, 
humor, strengthening through stress, and endurance of pain are the concepts captured 
within the selected items. Using a 5 point lickert scale, respondents indicated how true 
each statement was in the last month ranging from 1 (not at all) to 4 (true nearly all of 
the time). The maximum score was 44, and the mean score in the sample was 33.7.  
Respondents? history of traumatic events and trauma related distress was assessed 
by asking respondents to identify whether they had experienced a range of traumatic 
events including rape, physical harm caused by another person, incest etc. Respondents 
were then asked about whether the event was still bothersome, if the perpetrator had been 
forgiven, and if feelings of anger or hatred were present. Participants rated these options 
from 0 (not at all) to 4 (very much). PTSD symptom severity was measured for the 
 
60 
 
previous week using items 1-8 of the Davidson Trauma Scale (DTS; Davidson, 1996). 
Spirituality was assessed using a 13 item questionnaire asking about general spiritual 
beliefs of a Western orientation, and beliefs regarding reincarnation and karma. 
Participants were asked to agree or disagree with each statement about these beliefs on a 
6 point scale.  
The researchers performed bivariate analyses with t-tests as well as Pearson Chi 
Square tests to illuminate differences between those who agreed and disagreed with 
karmic belief. Significant predictors of karmic belief were discovered using multivariate 
logistic regression. In all only 5% of the total sample endorsed a belief in karma. This is 
not surprising given that the study was conducted in the United States. Those who 
reported a karmic belief were more likely to be non-white than those who did not 
(Pearson Chi square = 17.63, df = 1, p = 0.000).  They were also more likely to be 
unmarried (Pearson Chi square = 4.10, df = 1, p = 0.043) and to indicate that they had 
poor health (Pearson Chi square = 4.91, df = 1, p = 0.027) physically, and mentally 
(Pearson Chi square = 5.39, df = 1, p = 0.020). Those who held karmic beliefs reported 
more traumatic experiences (t = -2.85, p = 0.004). Physical and emotional abuse was 
more likely among the group endorsing karma (Pearson Chi square = 5.72, p = 0.017), 
and that group was also more likely to have experienced the violent death of a family 
member (Pearson Chi square = 4.41, df = 1, p = 0.036). Rape was also more prevalent 
among those who reported a belief in karma (Pearson Chi square = 13.62, df = 1, p = 
0.000).  
 
61 
 
The researchers fit a logistic regression model to detect if the number of traumas 
experienced was predictive of karmic beliefs. Davidson et al. (2005) controlled for race, 
marital status, physical and mental health and found that the only significant predictor of 
karmic belief that remained was race (Odds Ratio = 2.14, 95% Confidence Interval 1.48, 
3.92). The researchers found that non-white respondents were over twice as likely to 
believe in karma than were white respondents in this sample. A final logistic regression 
model was fit to determine the relationship between PTSD severity and karmic belief. 
The researchers found that race was still the only significant predictor of karmic belief 
after controlling for marital status, physical health, and mental health (Odds Ratio = 2.05, 
95% Confidence Interval 1.09, 3.86) (Davidson et al., 2005).  
The researchers found that karmic belief related significantly to poor physical and 
mental health. Further research is needed to understand why this relationship exists. This 
study was done in the United States where Buddhism is minority religion, and researchers 
have documented findings which connect minority religious membership with negative 
outcomes (Assanangkornchai et al., 2004). Davidson et al. (2005) found that race was the 
most significant predictor of karmic belief in their study, this makes sense since 
Buddhism is a prominent religion in the Eastern world and is not dominant in Western 
culture. Still, the research literature needs to move beyond connecting what factors are 
associated with karmic belief to understand how other Buddhist beliefs and practices 
interact with trauma outcomes. Research regarding Buddhist belief and resilience would 
be incomplete without similar studies conducted in populations where Buddhism is more 
 
62 
 
prevalent. Studies with predominantly Buddhist populations to detect intra-faith variance 
in practice, belief and resilience through trauma are needed. 
De Silva (2006) wrote concerning cultural observations after the 2004 tsunami hit 
with regard to Buddhist belief and practice. Specifically, relief efforts were noted to have 
stemmed from several Buddhist monks who offered humanitarian aid by providing 
shelter and supplies to the needy. These demonstrations embody karuna or compassion 
from loving kindness and equality. Buddhists affected by the tsunami also held varying 
explanations and sense-making strategies regarding the tsunami. Some held to the 
Buddhist concept of anicca or the notion of the impermanence of all things and allowed 
this concept to guide their construction meaning around the tsunami. Others felt that the 
tsunami and the losses which trailed behind it were due to karma, a reaping of 
consequences for one?s actions. De Silva calls for further investigation of disaster impact, 
religious belief and subsequent mental health outcomes of survivors (De Silva, 2006). 
Research does exist which has considered disaster impact and the psychological and 
religious outcomes of positive and negative religious coping in Western populations 
(Smith et al., 2000).  
In 1993, a large flood hit the Midwestern United States. Smith et al. (2000) 
distributed questionnaires among affected communities in Illinois and Missouri 6 weeks 
and 6 months after the flood. Participants included 209 adults, the majority of which were 
white (95.7%), married (76.6%), and female (62.2%). Religious groups represented 
included Protestant (75% of the sample) and Catholic (25% of the sample). Of the 
original 209 respondents, 131 responded to the follow-up questionnaire.  
 
63 
 
Religious dispositions of survivors were measured by assessing average church 
attendance, frequency of prayer and the salience religion held in respondent lives. Each 
of these was measured with a single item question utilized in previous research 
(Pargament et al., 1990; Pargament et al., 1994). Religious attributions were assessed 
using three items to understand to what cause respondents attributed the flood. They were 
asked if survivors felt the flood was due to ?God?s anger or punishment,? ?God?s love or 
reward,? or ?God?s Will or Purpose? responses ranged from 1 (not at all) to 5 
(completely) (Pargament & Hahn, 1986). Religious coping activities were measured using 
a 35 item scale developed by Pargament et al. (1990) to assess 6 types of religious 
coping. The discontent subscale, (? = .76) involved being angry with and distant from 
God and the church. The good deeds subscale (? = .81) measured attempts to live a more 
religious life, while the plead subscale (? = .76) entailed praying for miracles and 
bargaining with God. The religious focus subscale indicated whether respondents used 
religion as a diversion from problems (? = .69), and the subscale capturing religious 
support (? = .77) measured support seeking from clergy or other church members in 
one?s congregation. Finally, the spiritually based subscale (? = .89) was comprised of 
items focused on the existence of an intimate partnering with God. Respondents 
answered these subscale questions on a 4-point scale ranging from 1 (not at all) to 4 (a 
great deal). Internal consistency for the scale at time 1 was ? = .81 and ? = .83 for time 2. 
Psychological outcome of survivors was measured using the 12-item General 
Health Questionnaire (GHQ-12; Goldberg, & Williams, 1988). This measure was used to 
detect psychological distress symptoms in the last month using a 4 point lickert scale 
 
64 
 
from 1 (less so than usual) to 4 (much more than usual). Religious outcomes were 
measured using a 3 item scale developed by Pargament et al., (1990) to ask about 
relationship changes with God, and one?s church post flooding, and also about spiritual 
growth post disaster. Responses were scored on a 5-point scale from 1 (not at all) to 5 
(greatly). Alpha levels at times 1 and 2 were ? = .78, .76 respectively. 
Finally, flood exposure was measured using a scale developed primarily for Smith 
et al.?s (2000) study assess the degree of home displacement, property damage and 
personal injury sustained by survivors. The scale also assessed secondary flood exposure 
by asking about respondents? inconvenience experienced, how much family members had 
been affected and how much the community had been affected. These indicators were 
assessed by asking respondents to detail on a scale from 1 (none at all) to 5 (a great deal) 
the extent to which they were exposed in each of these six ways. Internal consistency was 
? = .82. Demographic information was gathered including age, gender, education, 
income, marital status and race.  
Multiple regression models were fit by the researchers to detect the influence of 
religious variables in predicting religious and psychological outcomes while controlling 
for demographic characteristics and flood exposure. Religious and Psychological 
outcomes at times one and two were dependent variables in the regression models. 
Results indicated that flood exposure explained 9% of the variance in psychological 
outcome at time 1, and 13% at time 2. Demographic variables were not significant 
predictors of any outcomes, and flood exposure was not predictive of any significant 
 
65 
 
variance in Religious outcome at times 1 or 2. The religious predictor variables accounted 
for 17% of the variance at time 1 and 15% of variance at time 2 of psychological 
outcomes. A general trend revealed that positive attributions regarding the flood (God?s 
love or reward) combined with positive religious coping (indicated by the spiritually 
based, religious focus, good deeds and religious support subscales) was associated with 
better psychological outcomes. Alternatively, attribution of the flood to God?s anger or 
punishment and negative religious coping (indicated by the plead and discontent 
subscales) were associated with poorer psychological outcomes.  
Religious outcome variables were predicted in large part (43%) by the religious 
predictor variables at time one, and 26% of the variance in religious outcomes was 
explained by the religious predictor variables at time two. Positive religious coping was 
associated with better religious outcomes than negative religious coping. Attribution of 
the flood to God?s love or reward and positive religious coping related to better religious 
outcomes.  
Mediation was supported in that positive religious coping when controlled yielded 
the relationship between religion prior to the flood and religious outcomes null at times 1 
and 2. Religion before the flood and religious outcomes were significantly correlated (r = 
.32, .26 p < .01) but the beta relationship in the multiple regression model shifted to a non 
significant ? = 0, .04 when controlling for positive religious coping at times 1 and 2 
respectively. A similar result was shown for the relationship between pre-flood religion 
and psychological outcomes at time 1.  
 
66 
 
There was partial support for mediation between pre-flood religion, and 
psychological outcomes at time 1 by attribution of the flood to God?s love or reward. 
This was shown in that pre-flood religion and psychological outcomes at time 1 were 
correlated (although not significantly; r = .12) but the beta coefficient after controlling for 
positive attribution decreased to  ? = .09 (Smith et al., 2000). 
The results of  Smith et al.?s work (2000) indicates that using positive attributions 
to explain trauma and religious coping methods involving social activity predict better 
outcomes religiously and psychologically. Also, the results of the study by Smith et al. 
(2000) indicate that one?s religiosity prior to a disaster may influence religious and 
psychological outcomes through the coping mechanisms used after the disaster. If this is 
actually the case, further investigation of religious behavior is warranted in the aftermath 
of disaster as it may be that those actions predict religious outcomes and psychological 
welfare. Attributing the flood positively to God?s love or reward partially mediated 
between pre-flood religion and psychological outcomes. This suggests that the research 
literature would benefit from understanding more about religiously based cognitive 
constructions employed to explain disaster. It would be beneficial also to explore these 
same questions and paths of influence with other religious faiths and populations beyond 
majority culture in the Western world.  
Some research has been conducted to understand the effects of different religious 
attributions in non-Western culture. Research findings of a study in El Salvador gives 
some information about how survivors outside the Western world used religious 
 
67 
 
attribution to make sense of a disaster. Perez-Sales, Cervellon, Vazquez, Vidales and 
Gaborit (2005) conducted a study with a disaster displaced population after  the 
earthquakes in El Salvador in 2001 to investigate what conditions post-disaster related to 
resilience.  
Of the 115 survivors interviewed, roughly 75% had some explanation for what 
had occurred. Of those, 57% attributed the earthquakes as a punishment from God. This 
punishment was thought to be catalyzed by poor behavior, violence, or a lack of prayer 
and respect. Not many survivors felt that the quakes were the result of natural causes 
(31.4%) (Perez-Sales et al., 2005). The main focus of this study was not to investigate the 
effects of these attributions. As such, the results did not focus on how these attributions 
related to mental health or other outcomes post disaster. Knowing that survivors had 
these assumptions and had made meaning this way is important. However, the research 
literature needs to expand these findings to understand what outcomes these attributions 
may influence. 
Understanding the tenets of multiple religious philosophies and their cognitive 
application to disaster deserves further attention. Buddhism is comprised of concepts 
which likely shape the way Buddhists view the world and life events. Buddhist 
philosophy is one of the major traditions of world religion. It was founded in the 6
th
 
century B.C. by Siddhartha Gautama (Hsu, O?Connor & Lee, 2009). Known for its 
promotion for an abdication of desire, a sense of well being is established by 
abandonment of burning desires, and an overcoming of wants within Buddhist 
philosophy. Thusly, misery generates from a proliferation of lusts and unbridled desires. 
 
68 
 
Refusing to subscribe to the notion of victory or defeat is also prized (Thera, 1993).   The 
concepts of samsara, nirvana, and karma are the most basic beliefs within Buddhism. 
Samsara means ?wheel of life? and references varying levels of realms of existence 
passed through in a sequence of reincarnation. Nirvana refers to a final releasing from 
these phases of reincarnation (i.e. a series of birth-death-birth). Karma indicates the 
consequences of the chosen behavior that one ascribes to while in one or another phase of 
life or samsara. Good deeds and actions tend to produce a higher level of existence in the 
life to come and vice versa (See Hsu, O?Connor, & Lee, 2009 for summary). 
Doing good includes the directive to those of Buddhist faith to be service oriented 
and willing to help and support one another (Rahula, 1996). This tenet was found enacted 
in reaction to the tsunami in that Buddhist monks took in several who were made 
homeless by the disaster and were provided with food and shelter regardless of religious 
denomination (De Silva, 2006). This display of service was likely a great help as previous 
research has indicated the important role religious leader have in helping to encourage the 
natural tendency toward resilience through support (Carballo et al., 2005). 
As part of the religious assessment included indications of various religious 
activities, a brief explanation of them is pertinent. Jataka stories, similar to parables in 
that they teach difficult principles with stories, are literary depictions of the former lives 
of the Buddha being reincarnated as human and animal, originally passed down orally for 
generations (Wickremeratne & Bond, 2006). Seth-Kavi, or a ?blessing verse? is 
essentially a religious poem for recitation in a benedictory fashion (Guruge, 2003). Bodhi 
Puja consists of offerings made to the Bo tree and aid in mind development (Bond, 1992). 
 
69 
 
Dana, is a ceremonial offering of food to Buddhist monks, and Bana refers to the 
deliverance of sermons (Seneviratne, 1999). The context in which Buddhists participate 
in these rituals may influence the way they respond to disaster. Mental health responses 
to trauma have been explained in part by a variety of predictors, including religion (Smith 
et al., 2000). However, variance in disaster outcomes, including family problems is not 
fully understood. Continued examination of disaster outcomes including physical health 
problems may help to explain the emergence of family problems after a disaster. 
Furthermore, continued study of Buddhist practice and may further explain family 
relationship outcomes after a disaster.   
 
Based on this literature review, I present the following 3 hypotheses for this study: 
Hypothesis 1: Physical health problems from the tsunami that continue to affect 
 daily activities will be negatively related to marital satisfaction and family 
 closeness.  
 Hypothesis 2: Physical health problems from the tsunami that continue to affect 
daily activities will be negatively related to marital satisfaction and family 
closeness controlling for the effects of PTSD and depression. 
 Hypothesis 3: Religious attributions and religious family activities will moderate 
the relationship between physical health problems from the tsunami that continue 
to affect daily activities and marital satisfaction and family closeness controlling 
for PTSD and depression.  
  
70 
 
METHOD 
Sample 
 Participants were mothers living in Southern Sri Lanka who were affected by the 
2004 Tsunami. The data used in the present study were obtained from another study 
entitled Health Risk and Resiliency in Tsunami Affected Mothers and Adolescents 
conducted in 2008. Secondary analyses were performed to answer the questions 
addressed in the present study. Women were identified using voter registration records 
which are updated annually by an officer in the village-who is supervised by the district 
officer. From these records, which are publicly available, eligible participants were 
identified by their having an adolescent child in the home. Trained, recruited, native, 
female interviewers conducted in-home interviews and administered translated 8
th
 grade 
level (English to Sinhalese) questionnaires. This convenience sample consists of 170 Sri 
Lankan mothers living in Polhena in Southern Sri Lanka. This village that was exposed to 
the devastating effects of the 2004 tsunami is located in the Matara district.  
Measures 
Marital Satisfaction was conceptualized by two items on the Tsunami Mental 
Health Study?ADULT Questionnaire (Wickrama & Kaspar, 2007). The first item asked, 
?Overall, how happy are you with your relationship?? This question was answered on a 
6-point Likert scale with responses ranging from ?Extremely unhappy? to ?Extremely 
happy.? The other item was, ?Overall, how satisfied are you with your relationship?
71 
 
This question was also answered on a 6-point Likert scale ranging from ?Not at 
all satisfied? to ?Completely satisfied.? The mean composite of these items demonstrated 
an adequate Cronbach?s alpha (?=.75) in this sample.   
Family Closeness was measured by a scale used in previous research to assess 
support and cohesion in family members (Portes & Rumbaut, 2001). These items created 
a latent construct, ?familism,? which was deemed a resilience factor for mothers? mental 
health after the tsunami (Wickrama & Wickrama, 2007). The participants were asked to 
rank the validity of a series of 14 statements about their family with a range of 5 
responses from ?strongly disagree? to ?strongly agree.? Examples of items include ?We 
help each other in times of trouble,? ?There are lots of bad feelings in our family,? ?We 
get along well together.? Cronbach?s alpha was ?=.82 in the sample. 
 Persisting physical health problems resulting from the Tsunami were assessed by 
asking respondents to rank the applicability of three statements with responses ranging on 
a 5-point Likert scale from ?strongly disagree? to ?strongly agree.? The statements were 
?Tsunami related physical health problems interfere with my daily activities, even 
today,? ?My spouse?s Tsunami related physical health problems interfere with my daily 
activities, even today,? and ?My children?s Tsunami related physical health problems 
interfere with my daily activities, even today.? Cronbach?s alpha for these items was 
marginally adequate ?=.58.  
 To measure current symptoms of PTSD, 16 items derived from the DSM-IV 
(American Psychiatric Association, 1994) diagnostic interview were administered. This 
scale has been used in previous research relating to trauma and marital variables 
 
72 
 
(Renshaw, Rodrigues & Jones, 2008). Use of these items in previous research with a 
similar sample (Wickrama & Wickrama, 2007) omitted two of the original PTSD items. 
They were as follows: ?Since the event, have you ever felt you have no future? and 
?Have you been unusually irritable or did you have outbursts of anger.?  This was done to 
improve internal consistency of the measure in a study using a Sri Lankan study of 
tsunami survivors (Wickrama & Wickrama, 2007). Given the sample similarity in this 
study, the same steps were taken in the current study. Remaining items asked about 
symptoms occurring in the last 4 months of respondents and included such questions as 
?Did you feel distant or cut off from others?? ?Did you feel a lot worse when you were 
reminded of the event?? Respondents were asked to give a yes or no answer to each 
question. Internal consistency for the sample in the present study was ?=.81.  
 To assess current depressive symptoms within the sample, 20 items from the 
Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) were used. 
This measure has been used previously in research relating to traumatic experiences and 
marital functioning (Renshaw, Rodrigues & Jones, 2008). Previous use of this measure, 
translated in cross-cultural examinations of mental health, indicates this as a viable 
measure of depression in terms of its psychometric properties (Wickrama & Kaspar, 
2007). The responses to the 20 items were averaged to indicate levels of depression in 
respondents. Scores over 16 points indicate clinical cut-off (Radloff, 1977). Questions 
asked about respondent experiences in the past week such as ?I had trouble keeping my 
mind on what I was doing? and ?I felt depressed.? Answers ranged on a 4-point Likert 
 
73 
 
scale from ?rarely, or none of the time? to ?most or all of the time.? Cronbach?s alpha for 
the current sample was ?=.84.  
 Religiosity was examined in terms of two aspects of religiosity. First the use of 
religious attribution to make sense of the tsunami was be assessed by the following yes or 
no questions ?Do you believe the tsunami occurred due to Karma?? ?Do you believe the 
tsunami consequences were unavoidable due to your Karma?? and ?Were you not 
effected or surprised by the tsunami because it is one instance of Samsara?? (?=68). Also, 
items to assess the frequency of religious activity in a relational context (i.e. with family, 
spouse) were included. These items included, ?How often do you go to the temple with 
your husband?? and ?How often do you go to the temple with other members of your 
family?? Responses to these items ranged from ?almost never? to ?almost daily? on a 5-
point Likert scale (?=.67).  
Plan of Analysis  
 Three hypotheses were tested. The hypotheses of this study were:  
 1. Physical health problems from the tsunami that continue to affect daily 
activities will be negatively related to marital satisfaction and family closeness.  
 2. Physical health problems from the tsunami that continue to affect daily 
activities will be negatively related to marital satisfaction and family closeness 
controlling for the effects of PTSD and depression. 
 3. Religious attributions and religious family activities will moderate the 
relationship between physical health problems from the tsunami that continue to 
 
74 
 
affect daily activities and marital satisfaction and family closeness controlling for 
PTSD and depression (see Figure 1).  
To test the first hypothesis, marital satisfaction and family closeness were regressed on 
physical health problems (see Figure 2). To test the second hypothesis, marital 
satisfaction and family closeness were regressed on physical health problems controlling 
for PTSD and depression (see Figure 3). To test the third hypothesis, marital satisfaction, 
family closeness, and religious attributions were regressed on physical health problems 
controlling PTSD and depression (see Figure 4). Then the same regression model was fit 
including religious family activities (see Figure 5). Religious attributions were not kept in 
that model because it did not relate significantly to any of the outcomes (see figure 4). 
Finally, a multiplicative interaction term between physical health problems and religious 
family activities was added to test whether family religious activities moderated the 
relationships between physical health problems and family outcomes (see Figure 6).  
 
 
 
75 
 
Figure 1: Hypothesized path model of physical health problems, PTSD, depression, marital 
satisfaction, family closeness, religious attributions and family religious activities.  
 
 
 
 
 
 
 
 
 
 
 
 
 
Physical Health 
Problems from the 
Tsunami 
Posttraumatic Stress 
Disorder (PTSD) 
Depression 
Family Closeness 
Religious Attributions Family Religious 
Activities 
Marital Satisfaction 
 
 In order to assess the relationships among the predictors, outcomes, and 
moderators, the statistical software, SPSS, was used. The data were then fit to five 
different path models using the statistical software, Mplus. Mplus was selected to fit 
these models because using path analysis allows for simultaneous estimations of the 
relationships between multiple variables.   
 
 
 
 
 
76 
 
RESULTS 
Univariate Analysis 
 After reverse coding selected items within the measures of marital satisfaction, 
family closeness, PTSD, and depression, mean composites of the variables were created. 
This was done by summing the items in each scale and then dividing by the total number 
of items. Cronbach?s alpha for the scales were then estimated (see Table 1).  
Table 1 
Estimated Cronbach Alphas for Sample (N=170) 
Measure         Cronbach 
Alpha 
Marital Satisfaction         .75 
Family Closeness         .82 
Physical Health Problems from the Tsunami      .58 
PTSD           .81 
Depression          .84 
Religious Attributions         .68 
Religious Family Activities        .67 
 
Inspection of histograms of each variable, revealed that all of the variables appear 
symmetric and to be roughly normal in their distribution, although the PTSD scale 
revealed one outlying respondent. The average scale scores, and other univariate statistics 
are presented in Table 2. 
  
 
 
 
77 
 
Table 2 
Univariate Statistics for Marital Satisfaction (MS), Family Closeness (FAMILY), 
Physical Health Problems from the Tsunami (PHYSHTH), PTSD, Depression (CESD), 
Religious Attributions (ATTR), and Religious Family Activities (FAMACT) 
Scales  N M SD Median Skewness Range       Kurtosis  
MS           160      4.10 .87    4.00      -.81    1-6              2.11 
FAMILY       170      4.21 .53    4.29      -.71  2.8-5              -.13 
PHYSHTH    170      2.65 .87    2.67       .15              1-4.7              -.53 
PTSD            163      1.45 .25    1.44       .39  1-2.5            .99 
CESD            170        .98 .49      .85       .78  .35-2.4            -.22 
ATTR            170      1.42 .39    1.33       .33    1-2              -1.36 
FAMACT      168      3.72 .93    4.00      -.93                1-6               1.20 
 
Bivariate Analysis 
 To determine what relationships the variables had with one another, a bivariate 
analysis was performed. Pearson correlations were computed between study variables 
and are presented in Table 3. These calculations indicated 11 significant relationships 
(see Table 3). The strongest of these included physical health problems with PTSD 
(r=.27, p=.001), physical health problems with depression (r=.41, p<.001), PTSD and 
depression (r=.47, p<.001), religious attributions and PTSD (r= -.35, p<.001), and 
religious attributions and depression (r= -.21, p=.010). It is expected that PTSD and 
depression would be related as these disorders tend to have high comorbidity in previous 
research (Suar et al., 2002). Also, physical health problems correlating with PTSD and 
depression is not surprising given past studies (Catapano et al., 2001). Finally, these 
preliminary correlations imply that religious attributions will be significantly negatively 
related to PTSD and depression. This hypothesis will be tested with the fitting of path 
analyses to the proposed models.  
 
 
78 
 
Table 3 
Pearson Correlation Coefficients for Marital Satisfaction (MS), Family Closeness 
(FAMILY), Physical Health Problems from the Tsunami (PHYSHTH), PTSD, Depression 
(CESD), Religious Attributions (ATTR), and Religious Family Activities (FAMACT) 
(N=170) 
            MS     FAMILY   PHYSHTH      PTSD     CESD    ATTR      FAMACT 
  
MS           1.00 
FAMILY         .16* 1.00  
PHYSHTH    -.18*     - .20*             1.00 
PTSD             .07        -.06             .27**        1.00    
CESD             .07        -.18*   .41**        .47**       1.00 
ATTR            -.15~       .10  .00        -.35**       -.21**       1.00 
FAMACT       .10          .16*             .05               .13            -.07         .03 1.00 
~p<.10 *p<.05 ** p<.01 
 
Multivariate Analysis 
 To address the hypotheses, a series of path models were fit using Mplus software. 
In the first model, marital satisfaction, and family closeness were regressed on physical 
health problems from the tsunami. In the second model, marital satisfaction and family 
closeness were regressed on physical health problems controlling for the presence of 
posttraumatic stress disorder (PTSD) and depression simultaneously. In the third model, 
marital satisfaction and family closeness were regressed on physical health problems, 
PTSD, depression, and religious activities participated in with family members. In the 
next model, religious family activities were included among the predictors. Religious 
attributions was dropped from further analysis because it did not reveal any significant 
relationships in model 3 (see Figure 4). In the fifth and final model, a multiplicative 
interaction term was added to the regression equation to detect possible moderation of 
religious family activities on the relationship between physical health problems, 
 
79 
 
controlling for PTSD, and depression. In other words, in the fifth model (see Figure 6), 
marital satisfaction and family closeness were regressed on physical health problems, 
PTSD, depression, religious activities participated in with family members and the 
interaction between physical health and family religious activities. 
 Before examining model results, test of model fit were considered to assess 
whether the models fit the data. The fit indices consulted included the Tucker-Lewis 
Index (TLI), the Comparative Fit Index (CFI), Chi-Square (?
2
) with its accompanying 
degrees of freedom (df) and p-value, and finally the Root Mean Square Error of 
Approximation (RMSEA). Because all of the models had 0 degrees of freedom, they all 
had a chi-square value of 0, TLI and CFI values of 1, and RMSEA  values of 0 (p =0.00) 
indicating perfect fit of the models to the data.  
  
Hypothesis 1 
 The first hypothesis was formulated to suggest that physical health problems 
would predict significant variance in marital satisfaction and family closeness even four 
years after the tsunami. Specifically, given the literature review, I hypothesized that 
physical health problems would relate negatively to marital satisfaction and family 
closeness. The variable, physical health problems from the tsunami, was simultaneously 
regressed on marital satisfaction and family closeness (see Figure 2). Results showed that 
physical health problems from the tsunami had a significant negative relationship with 
marital satisfaction (? = -0.17, p<.05). This indicates that high levels of physical health 
problems 4 years after the tsunami associate with low levels of marital satisfaction 
 
80 
 
controlling for all else in the model. Also, physical health problems from the tsunami 
related negatively to family closeness (? = -0.11, p<.05). In other words, high levels of 
physical health problems from the tsunami associate with low levels of family closeness 
controlling for all else in the model. Hypothesis 1 was supported by the model. Physical 
health problems from the tsunami accounted for 2.8% of the variance in marital 
satisfaction, and 3.2% of the variance in family closeness. 
 
81 
 
Figure 2: Path model of non-standardized parameter estimates of physical health problems from 
the Tsunami and marital satisfaction and family closeness (Model 1) (N=170) 
 
             R
2
=2.8% 
 
  
? = -.169* .759(-.17*)  
 
           .050(.16*) 
             
           R
2
=3.2%
  
? = -.110* -.082(-.20*) 
*p < .05 
Note: Estimates in parentheses are correlations, estimates outside parentheses are covariances 
 
Physical Health 
Problems from the 
Tsunami 
Family Closeness 
Marital Satisfaction 
 
Hypothesis 2 
 The second hypothesis suggested that physical health problems persisting four 
years after the tsunami would be negatively related to marital satisfaction and family 
closeness even when controlling for PTSD and depressive symptoms. To test this 
hypothesis, physical health problems from the tsunami, PTSD, and depression were 
 
82 
 
regressed on marital satisfaction and family closeness simultaneously (Figure 3). The 
negative relationship between physical health problems and marital satisfaction was 
stronger when PTSD and depression were controlled (? = -.21 p<.05). This suggests that 
even when accounting for PTSD and depression, physical health problems still relates 
negatively to marital satisfaction 4 years after the occurrence of the tsunami. When 
controlling for PTSD and depression, physical health problems still related marginally 
significantly to family closeness (? = -.09 p<.10) although the relationship was not as 
significant or as high in magnitude in model one. Even so, model two supported the 
second hypothesis. In other words, high levels of physical health problems are related to 
low levels of marital satisfaction and family closeness controlling for all else in the 
model. Depression, PTSD, physical health problems from the tsunami account for 3.9% 
of the variance in marital satisfaction, and 4.0% of the variance in family closeness. 
 
83 
 
Figure 3: Path model of non-standardized parameter estimates of physical health problems from 
the Tsunami, PTSD, Depression, and marital satisfaction and family closeness (Model 2) 
(N=170) 
 
 R
2
=3.9% 
  
    .059 
      .049(.16*) 
                                                                                                                                         R
2
=4.0% 
       .057 
 
 
 
~p < .10 *p < .05 
 
Note: Estimates in parentheses are correlations, estimates outside parentheses are covariances 
 
? = -.21* -.124(-.18*) 
? = -.112 
.150 
? = 0.105 
? = -0.276 
? = -.090~ .282(-.20*) 
? = .036 
Physical Health 
Problems from the 
Tsunami 
Posttraumatic Stress 
Disorder (PTSD) 
Depression 
Family Closeness 
Marital Satisfaction 
 
 
 
Hypothesis 3 
 
 Hypothesis three posited that religious attributions or religious family activities 
would moderate the relationships between physical health problems from the tsunami and 
marital satisfaction and family closeness. To test this hypothesis, incremental models 
needed to be fit. First, physical health problems from the tsunami, PTSD, depression, and 
religious attributions were regressed on marital satisfaction and family closeness 
simultaneously (Figure 4). Results of this model indicated that physical health problems 
were negatively related to marital satisfaction (? = -.20 p<.05) and marginally related to 
family closeness (? = -.10 p<.10) controlling for all else in the model. No other paths in 
 
84 
 
the model had slopes reaching significance, including religious attributions. In order to 
assess if family religious activities was significantly related to the outcomes, it was added 
in the next model (see Figure 5), but religious attributions was not included. Taken 
together, physical health problems from the tsunami, PTSD, depression, and religious 
attributions explained 4.1% of the variance in marital satisfaction, and 4.3% of the 
variance in family closeness (see Figure 4).  
 
85 
 
Figure 4: Path model of non-standardized parameter estimates of physical health problems from 
the Tsunami, PTSD, depression, religious attributions, and marital satisfaction and family 
closeness (Model 3) (N=170)        R
2
=4.1% 
                                                ? = -.20* -.125(-.17*) 
                      ? = .13 
 ? = .11 
  ? = -.14 
           .050(.16*) 
 ? = -.10~ -.082 (-.20*) 
                                                                                                                                      R
2
=4.3%  
  ? = .10 
                      ? = -.11 
 ? = .076 
 
~p < .10 
*p < .05 
Note: Estimates in parentheses are?correlations,?estimates?outside parentheses are covariances 
 
Physical Health 
Problems from the 
Tsunami 
Posttraumatic Stress 
Disorder (PTSD) 
Depression 
?
?
Family Closeness 
Religious Attributions 
?
?
Marital Satisfaction 
 
Next, a model was fit to determine if religious family activities related significantly to the 
outcome variables (see Figure 5). Results of the model indicated that physical health 
problems were still negatively related to marital satisfaction (? = -.21 p<.05) and 
marginally related to family closeness as well (? = -.10 p<.10) controlling for all else in 
 
86 
 
the model. Additionally, model results showed that family religious activities was 
significantly related to family closeness (? = .09 p<.05), although not related to marital 
satisfaction. What this indicates is that high levels of family religious activities relate to 
high levels of family closeness controlling for all else in the model. Also, high levels of 
physical health problems relate to lower levels of marital satisfaction and family 
closeness controlling for all else in the model (see Figure 5). Taken together, physical 
health problems from the tsunami, PTSD, depression, and family religious activities 
account for 5.5% of the variance in marital satisfaction, and 6.4% of the variance in 
family closeness. 
 
87 
 
Figure 5: Path model of non-standardized parameter estimates of physical health problems from 
the Tsunami, PTSD, depression, religious family activities, and marital satisfaction and family 
closeness (Model 4) (N=170)                                                                                          R
2
=5.5% 
                                                 ? = -.21* -.122(-.17*) 
                         ? = .20 
                                 ? = .14 
 
                                                                                                  ? = .12                              .050(.16*) 
                                                                                              ? = -.10~ -.082(-.20*) 
R
2
=6.4%  
                                                                                        ? = .03 
                      ? = 0.08 
                                                ? = .09* .075 (.16*) 
 
~p < .10 
*p < .05 
Note: Estimates in parentheses are?correlations,?estimates?outside parentheses are covariances 
 
?
Physical Health 
Problems from the 
Tsunami 
Posttraumatic Stress 
Disorder (PTSD) 
Depression 
?
?
Family Closeness 
Religious Family 
Activities 
?
?
Marital Satisfaction 
 
 A final model was fit to see if family religious activities moderated the 
relationship between physical health problems, PTSD, depression and the outcome 
variables. In order to test this, a multiplicative interaction term between physical health 
 
88 
 
problems and religious family activities was created and added to the model as a 
predictor. In this final model, it was found that physical health problems related 
negatively to marital satisfaction (? = -.70 p < .05), and related marginally to family 
closeness as well (? = -.37 p < .10). This indicates that high levels of physical health 
problems relate to low levels of family closeness and marital satisfaction controlling for 
all else in the model. Because the interaction term did not significantly relate to marital 
satisfaction or family closeness, family religious activities did not moderate the 
relationship between physical health problems and marital satisfaction and family 
closeness controlling for all else in the model. Taken together, the model results indicated 
that physical health problems from the tsunami, PTSD, depression, family religious 
activities and the interaction between family religious activities and physical health 
problems explained 7.1% of the variance in marital satisfaction, and 5.3% of the variance 
in family closeness four years after the tsunami (see Figure 6).   
 
89 
 
Figure 6: Path model of non-standardized parameter estimates of physical health problems from 
the Tsunami, PTSD, depression, religious family activities, physical health problems*religious 
family activities, and marital satisfaction and family closeness (Model 5) 
 
   R
2
=7.1% 
 ? = -.70* -.123(-.17*) 
                                               ? = .14 
 
                                                                                           ? =.14 
                                                                                              ? = -.20 
 
 ? = .13   .049(.16*) 
                                                                                                              ? = -.37~ -.082(-.20*)  R
2
=5.3%    
   
                                                                                                  ? = -.09 
                                                                                             ? = -.08 
 
                                                                                   ? = -.10 
 
 
 
  ? = .08 
 
~p < .10 
*p < .05 
 
Note: Estimates in parentheses are correlations, estimates outside parentheses are covariances 
Physical Health 
Problems from the 
Tsunami 
Posttraumatic Stress 
Disorder (PTSD) 
Depression 
 
 
 
 
Family Closeness 
Family Religious 
Activities 
 
 
 
 
Marital Satisfaction 
Physical Health 
Problems * Family 
Religious Activities 
 
  
 
 
 
 
90 
 
DISCUSSION 
 
 The most interesting findings yielded by the analysis include the persistence of 
physical health problems being significantly related to the family outcomes as the 
incremental models progressed and included more control variables. Secondly, although 
family religious activities did not moderate the relationship between physical health 
problems and marital satisfaction and family closeness, it was still a significant predictor. 
Specifically, family religious activities were positively related to family closeness 
controlling for PTSD and depression, while physical health problems were negatively 
related to both family outcomes in the same model. What this indicates is that even four 
years after the tsunami, physical health problems relate negatively to family relationship 
outcomes in disaster survivors. Also, even when accounting for the negative influence of 
physical health problems, family religious activities relate positively to family closeness. 
What this may implicate is that even though 4 years have passed since the devastating 
tsunami, there are still existing problems with family relationships. Also, there are factors 
that appear protective of family relationships. This speaks to the need for continued 
intervention efforts for families of survivors.  
 Research ought to continue to discover further what disaster circumstances relate 
to negative relational outcomes, and what relates positively to those outcomes. Although 
it does not appear that religious family activities buffer the relationship between physical 
91 
 
health problems, it may be that family religion buffer against other disaster outcomes. 
While these are the most interesting findings of this study, other findings are summarized 
below. 
Summary of Results 
 Hypothesis 1: Physical health problems from the tsunami that continue to affect 
daily activities will be negatively related to marital satisfaction and family closeness. 
Results from model 1 indicate that this hypothesis was supported. It was found that 
physical health problems related negatively to marital satisfaction and to family 
closeness. In other words, high levels of physical health problems were related to low 
levels of marital satisfaction and family closeness. Physical health problems were also 
correlated negatively with both family closeness and marital satisfaction in the sample. 
However, the levels of family closeness and marital satisfaction are unknown previous to 
the disaster, so it would be spurious to suggest any causal link from physical health 
problems to the family outcomes.   
 The COR model (Hobfoll, 1989) is also supported by this finding which dictates 
that stress is evoked from loss of resources. The loss of physical health has been shown in 
previous studies to relate negatively to relationship functioning (Booth & Johnson, 1994). 
Physical health problems lingering from the 2004 tsunami may influence marital 
satisfaction because of the limitations that such health problems may introduce. Booth 
and Johnson (1994) noted that physical health problems took their toll on marital 
relationships in part through an increase of problem behaviors within the marriage, and a 
 
92 
 
decrease in positive interaction. It could also be that the presence of health problems and 
marital problems together affect the severity of health problems.   
 The relationship between physical health problems persisting from the tsunami 
and marital satisfaction is also in keeping with Trief et al. (2006) who found that 
heightened marital stress was related to higher symptom distress. Trief et al. (2006) also 
found that higher rates of marital problems were related to lower ability to cope with 
physical illness. This relationship suggests that perhaps not only do physical health 
problems contribute to marital problems, but marital problems may contribute a 
weakened ability to manage the stress and losses accompanying physical health 
problems.  
 High family differentiation was associated with higher levels of coping ability, 
and lower symptom severity as found by Murray et al. (2007). This suggests that those 
who cope positively with physical health problems have appropriate family boundaries. 
The finding that health problems were related negatively to family closeness makes since 
considering Murray et al.?s (2007) findings. This suggests that problems with family 
closeness, or low cohesion is associated with physical health problems or health problem 
influences on daily activities. Murray et al. (2007) also found that higher levels of 
differentiation were associated with lower amounts of stressful life events outside 
physical illness. Survivors of the tsunami may be at greater risk for family closeness 
problems then because they are more likely to have experienced multiple stressful life 
events as a result of the tsunami. This may further explain why the negative relationship 
exists between physical health problems from the tsunami and lower family cohesion. 
 
93 
 
 Catapano et al. (2001) reported that that majority of respondents who were highly 
exposed to disaster destruction (including personal and family member injury) 
experienced family tensions (52%). Assangkornchai et al. (2004) reported a smaller 
incidence of family problems post disaster, but did find that some survivors reported 
increased family quarrels nonetheless (8%). These tensions could perhaps be due to a 
lessened tolerance for normative inconvenience and stress. This lessened tolerance could 
come from the fact that those exposed severely to disaster are more likely to have lost 
resources that would normally be utilized to buffer against stress.  
 A basic assumption of the COR model is that stress stems from resource loss, so 
people do what they can to avoid resource loss, in other words, to conserve resources. 
This happens by compensating for lost resources, or building up reserves in times when 
resources are not taxed. When resources are taxed, either the actual lack or resources, or 
the perception that one cannot meet challenges because of a lack of resources creates 
stress (Hobfoll, 1989). It is this taxing that perhaps explains the rise of tensions and 
quarrels within the family. One possible source of contention is rearranging and 
reorganizing family functioning with fewer resources. Physical health problems 
stemming from the tsunami represent such a tax of resources. Those with persisting 
physical limitations may become unable to contribute to the family in the way they used 
to. The responsibility for household tasks or earning money may have to be redistributed 
to adjust. The fact that physical health problems from the tsunami creating problems in 
daily activities persisted over four years afterward may signify that this adjustment may 
take longer than would be assumed.   
 
94 
 
 
  Hypothesis 2: Physical health problems from the tsunami that continue to affect 
daily activities will be negatively related to marital satisfaction and family closeness 
controlling for the effects of PTSD and depression. Model 2 was fit to address this 
hypothesis and was found to support it. When PTSD and depression were added to the 
path model as predictors of marital satisfaction and family closeness, the influence of 
physical health problems was still significant. Depression and PTSD were not found to 
significantly relate to marital satisfaction controlling for all else in the model. This 
finding implies that when controlling for physical health problems, PTSD and depression 
do not significantly predict marital satisfaction or depression. I did not test for a possible 
mediation by physical health problems in the relationship between mental health 
problems and marital satisfaction and family closeness. This was because PTSD and 
depression were not significantly correlated with marital satisfaction, and PTSD was not 
significantly correlated with family cohesion. However, depression was correlated with 
family closeness (r = -.18, p < .05). 
 Wickrama and Wickrama (2007) found that family problems mediated the 
relationship between property destruction and mental health problems. What Wickrama 
and Wickrama found is that family problems tended to be a vehicle for property 
destruction to influence mental health. They found that family problems did not mediate 
between life destruction (injury and death) and mental health problems. Perhaps this is 
because mental health status relates to family relationships through physical health 
problems. However, this mediation was not investigated in the present study.  
 
95 
 
 Trief et al. (2006) found that marital problems were related to increased 
depression in a sample with diabetes. This suggests that even when controlling for 
physical health problems, depression and marital problems were negatively related. This 
opposes the findings of the present study, but perhaps the relationships between physical 
health problems, mental health status and family relationships are uniquely influenced by 
the disaster or other highly stressful context. This idea finds possible support in that 
respondents in Dekel and Solomon?s (2006) study who were former POW?s, regardless 
of mental health status, on average demonstrated a greater likelihood to have low martial 
adjustment than those who were not POW?s. This suggests that controlling for mental 
health, the traumatic experiences of being a POW (which may have included physical 
health complications) relate to lower marital adjustment.  
 Respondents in Wickrama and Wickrama?s (2007) study had lower mental health 
problems on average when they had high levels of family cohesion. From the current 
study findings, we do not know how PTSD and depression relate to marital satisfaction 
and family closeness without controlling for physical health problems, but Wickrama and 
Wickrama (2007) suggest that family cohesion is protective of tsunami survivors against 
mental health problems. Wickrama and Kaspar (2007) found that positive mother-child 
relationships were related to better child mental health outcomes after a disaster, as was 
mothers? mental health. Proctor et al. (2007) noted that poor parenting behaviors pre-
disaster related to increased distress in children after the disaster, and that disaster impact 
related to child persisting distress through parental stress. This suggests that family 
member reactions to disaster and family relationships before the disaster that are positive 
 
96 
 
tend to protect survivors from mental health struggles. This means that research findings 
suggest that perhaps family closeness is a resource that buffers against mental health 
problems in disaster, but results of the present study suggest that family closeness may be 
depleted by resource losses at the same time.   
 No previous studies have been conducted which specifically analyze the 
relationship between physical injury and family outcomes after a disaster. While this 
relationship needs further investigation with regard to mental health status, it seems that 
physical health problems are an important predictor of family relationship problems after 
a disaster. 
 It is also clear that physical health problems predict family relationship problems 
long after disaster given that this sample was taken almost 4 years after the tsunami. This 
is in support of other pre-existing research suggesting that physical health problems do 
indeed persist after disasters and relate to lower quality of life assessed physically and 
mentally (Wu et al., 2006). North et al. (2008) also found that while the intensity of 
symptoms may decrease over time after disaster, those symptoms may still exist long 
afterward.  
  
 Hypothesis 3: Religious attributions and religious family activities will moderate 
the relationship between physical health problems from the tsunami that continue to 
affect daily activities and marital satisfaction and family closeness controlling for PTSD 
and depression. 
 
97 
 
             The COR model (Hobfoll, 1989) suggests that there are ways to minimize 
resource loss that involve cognitive reframing strategies as well. Religious attributions 
used to explain losses tend to have mixed relationships with mental health and religious 
outcomes. Participation in religious activities with family, friends and religious leaders 
may also be a way that disaster survivors mobilize social resources to help offset resource 
loss (Smith et al., 2000). 
  Before testing for moderation, Model 3 was fit to determine if religious 
attributions related to family closeness and marital satisfaction. The results of Model 3 
indicate that religious attributions did not relate significantly to marital satisfaction or 
family closeness. As such, moderation by religious attribution was not tested. Model 4 
was fit to determine if family religious activities related to the relational outcomes. 
Family religious activity significantly related to family closeness such that higher levels 
of family religious activity related positively to family closeness. Controlling for all else 
in the model, physical health problems from the tsunami had a significant negative 
relationship with marital satisfaction and marginally significant negative relationship 
with family closeness.  
  The finding that religious family activity was positively related to marital 
satisfaction and family closeness makes sense in light of Wickrama and Wickrama?s 
(2007) results that those participating in religious activity had lower levels on average of 
depression and PTSD post disaster. It could be that participating with one?s family in 
religious activity provides an element of social support that could be protective. In 
reference to the COR model, this family support may represent a compensatory resource. 
 
98 
 
The importance of social support is manifest in that Wickrama and Kaspar (2007) found 
that disruption of social support was related to higher levels of depression in adolescents 
after the tsunami.  
  Religious attributions were not found to be significantly related to marital 
satisfaction or family closeness controlling for all else in model 3. As such, religious 
attributions were excluded from further analysis. Previous research has connected the 
quality of religious attribution to religious and mental health outcomes (Smith et al., 
2000), but no studies exists to my knowledge that examine the connection between 
religious attributions and relational outcomes in the disaster context. However, model 3 
yielded results indicating physical health as relating significantly to marital satisfaction 
and marginally to family closeness even when controlling for mental health status, and 
religious attributions.  
  Models 5 and 6 were fit to test the possible moderation of religious family 
activity. Neither of the paths from the interaction term were significant, leaving the third 
hypothesis unsupported. Religious participation has been found in previous research 
(Wickrama & Wickrama, 2007) to moderate the relationship between PTSD and tsunami 
exposure (which includes family and personal injury). However, Wickrama and 
Wickrama (2007) did not find that religious participation moderated between depression 
and tsunami exposure. Even so, mental health outcomes appeared to be buffered by 
religious participation (Wickrama & Wickrama, 2007). Also, adverse mental health 
outcomes coexist with family problems post disaster (Catapano et al., 2001) so perhaps 
what is protective of mental health may also be protective of family outcomes. It was also 
 
99 
 
found that Buddhists in one study were less likely than non-Buddhists to exhibit PTSD 
symptoms after disaster (Assanangkornchai et al., 2004) although that may have had to 
do with being part of a majority religion and obtaining social support as a result. Given 
these relationships, it was hypothesized that religious attributions or family religious 
activity would buffer family outcomes against the incidence of lingering physical health 
problems post-tsunami. The fact that model 3 found religious attributions to be an 
insignificant predictor, and that moderation was not supported by Model 6 suggests 
perhaps that the constructs of religiosity were not measured with adequate detail. For 
instance, religious attribution may have been captured with a greater complexity by 
including statements that, while in keeping with Buddhism, exhibited a greater sense of 
inner locus of control. The statements used to measure religious attribution (see 
Appendix F) tended toward a conceptualization of external locus of control. Respondents 
affirming these statements not only endorsed religious attributions concerning the 
tsunami, but carried the sentiment of externalization. This externalization was exhibited 
by attributing the tsunami as ?unavoidable? or ?due to samsara.? A greater external locus 
of control when constructing attributions around disaster has been shown to relate to 
higher mental distress (Suar et al., 2002). The response option asking respondents to 
indicate whether they attributed the incidence of the tsunami to karma may suggest more 
control of the respondent in their life consequences. However, a sentiment of 
consignment is still implied that is in keeping with an externalized locus of control.  
  Another possible reason for moderation not being supported in Model 6 could 
also be that religiosity moderates between other aspects of disaster exposure and not 
 
100 
 
physical health problems. Disaster exposure has been measured in more comprehensive 
ways in the past to include such experiences as the death of a family member or loved 
one (Assanangkornchai et al., 2004), or home damage and loss (Suar et al., 2002). This 
may mean that even though physical health problems appear to be important in their 
relation to marital satisfaction and family closeness problems, religiosity may not 
moderate or buffer that particular relationship in this sample. Even so, it appear that 
family religious activities are beneficial to family relationships in tsunami survivors.   
Implications of Research Findings  
 Physical health problems were found to be significantly related to marital 
satisfaction and family functioning years after the tsunami. This suggests that disaster 
survivors experience the results of disaster not only emotionally, but physically and for 
years afterward. Special attention to the long term daily interferences of physical health 
problems in survivor families is needed for those seeking to assist in disaster recovery. 
While efforts in disaster recovery are often focused on meeting the immediate physical 
needs of survivors (De Silva, 2006), and debriefing survivors emotionally and mentally 
(Bronish et al., 2006) it appears that attention to family relationships, and chronic health 
problems is also needed. The emergence of tensions and quarrels among family members 
has been documented (Assanangkornchai et al., 2004; Catapano et al., 2001). Even so, it 
appears that research efforts post disaster, are often mainly focused at the individual 
(Assanangkornchai et al., 2004) and community levels (Catapano et al., 2001) rather than 
at the family level. Emerging research focusing more intently on family relationship 
 
101 
 
outcomes and risks (Wickrama & Wickrama et al. 2007) may begin to shift the focus of 
intervention development post disaster toward families which appears to be warranted.  
 Perhaps the most significant implication of this study is the large amount of 
variance still left unexplained within marital satisfaction and family closeness after the 
tsunami. Well over 90% of the variance in these outcomes after disaster still needs 
explaining. It is likely that many other factors contribute to family outcomes after a 
disaster. Even though attention ought to be given to physical health problems, when 
attempting to assist families in their adjustment after a disaster, a thorough assessment of 
all the challenges faced by families will need to be considered.  This is especially the case 
since physical health problems relate to adverse family outcomes 4 years after the 
disaster. The fact that this substantial amount of time has passed since the disaster may in 
part explain the small amount of variance in family outcomes by physical health 
problems. As previous research indicates, chronological proximity to disaster tends to 
relate to more intense adverse mental health symptoms (North et al., 2006). It could be 
then that physical health problems may explain a greater amount of variance in family 
closeness and marital satisfaction if assessed closer to the time of the disaster. 
 The COR model was generally supported in that disruption of the resource of 
physical health tends to relate negatively with family outcomes. However, other 
resources that are lost and potentially used to buffer losses by disaster survivors need to 
be investigated. 
 
 
 
102 
 
Future Research 
 In addition to considering what other resources besides physical health are 
disturbed by disaster, perhaps the first step in expanding the research on family outcomes 
of natural disaster outcomes would be to include other family members as respondents. 
The current study included only women (mothers). This sample was constructed in this 
way due to the ?familistic? or cohesive nature of Sinhalese families. Health problems of 
respondents, their husbands, and their children which influence daily activities years after 
the tsunami appear relate negatively to their marital satisfaction and family closeness. In 
order to better understand how having a physical health problem differs from having a 
spouse or child with a physical health problem, future research would benefit from 
distinguishing who has the physical health problem in the family. Asking all family 
members to report on family outcomes would also give a more complete perspective of 
how family closeness and marital satisfaction relate to the outcomes of disaster. For now, 
the results of the present study simply indicate that health problems within the family 
tend to have an adverse relationship with marital satisfaction and family closeness after 
disaster. 
 Further research on the construct of physical health problems from natural 
disasters as well as other resource losses such as home damage, occupation disruption 
need to be considered in more detail with regard to family outcomes. Previous disaster 
studies often are concerned with injury from disasters (Wickrama & Wickrama, 2007), 
and occasionally address illness (Assanangkornchai et al., 2004). However, there is 
 
103 
 
limited work on the physical limitations which may accompany injuries and illnesses of 
disaster and complicate the daily lives of survivors for years afterward (Wu et al. 2006). 
It is possible that physical health limitations may interact with other disaster outcomes to 
synergize losses. For example, physical health problems may contribute to occupational 
disruptions and compound the difficulties presented by challenges such as home damage 
and destruction. However, this synergy may be more easily detected in the short term 
after disaster since the passage of 4 years may have allowed some survivors to adjust or 
compensate for some of their resource losses. Research focusing on how these losses 
relate to family outcomes will be an important contribution, especially as studies also 
include considerations of how physical health problems relate to family outcomes post 
disaster. 
 Previous research has shown that often the most reliable predictors of adverse 
mental health outcomes after a disaster, are mental health outcomes before a disaster 
(North et al., 2006). Future research should continue to assess pre-disaster mental health 
status as accurately as possible. Furthermore, if pre-existing mental morbidity predicts 
adverse mental health outcomes post-disaster, an inclusion of pre-disaster family 
closeness and marital satisfaction may reveal whether disasters create family problems or 
perhaps exacerbate pre-existing vulnerability. While asking respondents to self report on 
retrospection of family functioning and marital satisfaction may be spurious, it may 
provide a preliminary baseline level to be compared. Religious activity before a disaster 
has been shown to be connected with religious and mental health outcomes afterward 
 
104 
 
(Smith et al.). Although religious attributions and family religious activity didn?t 
moderate between physical health problems from the tsunami and family outcomes, 
moderation ought to be investigated with other losses that may predict variance in family 
outcomes after disasters. 
 Existing studies of the effects of religious attributions in Western culture included 
attributions of a positive and negative nature (Smith et al., 2000). It was found that the 
quality of attribution (which often signified either an internal or external locus of control) 
influenced the relationship these attributions had with mental health outcomes. Future 
research with Buddhist populations should include more varied attributions than the 
present study offered for consideration. Believing the tsunami to have been unavoidable 
due to karma may be an indication of an external locus of control, which has been shown 
to relate to higher PTSD symptom levels (Suar et al., 2002). An inclusion of a more 
neutral or positive attribution, such as that regarding the tsunami being due to anicca, or 
the impermanence of all things (De Silva, 2006) may yield important results for relational 
outcomes. This is especially important because previous research has connected the 
quality of religious attribution to religious and mental health outcomes (Smith et al., 
2000), but no studies exists to my knowledge that examine the connection between 
religious attributions and relational outcomes in the disaster context. 
Strengths 
 One strength of this study is that physical health problems from the tsunami was 
used as the main predictor. This is a strong point because physical injuries and financial 
losses are regularly composited to construct measures of disaster exposure (Proctor et al., 
 
105 
 
2007; Suar et al., 2002; Wickrama & Kaspar, 2007; Wickrama & Wickrama, 2007). It 
may be that physical health problems and financial losses could have similar relationships 
with family outcomes. However, separating physical health challenges and financial 
challenges from disaster may evidence distinct relationships with family outcomes which 
still need investigation. This investigation is just beginning as family relationships have 
not been studied recently as main outcomes in disaster research although family 
outcomes are peripherally included at times (Catapano et al., 2001; Wickrama & 
Wickrama et al., 2007). Also, conducting this study 4 years after the tsunami, revealed 
that physical health problems exist which continue disturb daily activities, and explain 
variance in family relationships. This finding is a strength of the current study because it 
prompts further investigation of what other circumstances persist years after disaster. The 
fact that physical health problems continue to influence daily activities and related 
negatively to relational outcomes highlights the fact that intervention is still likely 
warranted at this time for survivors. 
 This study contributed to a greater understanding of the effects of disaster in that 
it considered marital satisfaction as an outcome which has not been done before although 
some studies have included information on marital status in their consideration of disaster 
outcomes and a status of being married has been shown to be protective against adverse 
mental health outcomes (Wickrama & Wickrama, 2007). It appears that physical health 
problems from the tsunami related negatively to marital satisfaction, and further research 
would benefit from further consideration of what other disaster circumstances relate to 
marital satisfaction. 
 
106 
 
 Another strength of this study is that it used a more detailed conceptualization of 
religiosity than some other disaster studies. While Western religious attributions and 
activities have been studied in the disaster context (Smith et al., 2000), studies with 
Eastern populations where Buddhism is the majority religion had not been previously 
conducted in response to the tsunami. Religiosity of tsunami survivors has been assessed, 
and found to be a protective factor against mental health problems in tsunami survivors 
(Wickrama & Wickrama et al., 2007) but religious attributions were not addressed, and 
neither was family religious activity.  
Limitations 
 One of the main limitations of this study is that it relied on the self report of 
respondents. While the measures used to assess depression (CESD; Radloff, 1977), and 
PTSD (DSM-IV diagnostic interview) (American Psychiatric Association, 1994) were 
reliable in this sample, the measures of some other study constructs were not as reliable. 
Particularly the religious scales which were not pre-established measures were lower in 
internal consistency.   
 Another limitation of this study is found in some of the sample characteristics. 
This sample was comprised of only women, and was relatively small. Given that all 
respondents lived in the same village, the results give a picture of the tsunami experience 
for these respondents, but may not be generalizable to other populations. There was also 
no information available regarding basic demographic characteristics such as education 
levels, age etc. so it is not possible to know what populations these results may be 
compared to.  
 
107 
 
 The data was cross-sectional and didn?t include measures of retrospective family 
functioning or marital satisfaction. This rendered me unable to differentiate between 
family functioning and marital satisfaction, mental health or religiosity prior to the 
tsunami so changes were not detectable. 
Conclusion 
 Even given the limitations of this study, some conclusions are able to be drawn 
with regard to the tsunami. It appears that physical health problems from the tsunami 
relate negatively to marital satisfaction and family functioning, even when controlling for 
mental health status, family religious activity and the interaction between family religious 
activity and physical health problems in this sample. It is likely physical health problems 
play an important role in understanding how families and marriages function and vice 
versa after a disaster.  Religious attributions were not a significant predictor, and 
religious family activities didn?t moderate these relationships in this sample which could 
be due to measurement error, or underdeveloped instruments. However, family religious 
activities related positively to family closeness. As such, it may be concluded that 
religious activities participated in with family members contribute positively to family 
outcomes, or the reverse may be true.  
 Fitting path models to the data allowed for an exploration of the relationships 
between these constructs controlling for all else in the models, and the findings may have 
important implications for those assisting disaster survivors. Attention to family 
relationships may allow for helpers to identify new and useful points of intervention at 
the family level. Considering physical health problems when working with families after 
 
108 
 
a disaster may also open other avenues by which to understand the effects of disaster 
exposure on marriages and families. Even when considering these findings, there is still 
an enormous amount of variance in marital satisfaction and family closeness left to be 
explained after disaster strikes, especially years afterward. Future studies should continue 
to investigate marital satisfaction, family closeness and other indications of family 
functioning as they seek to understand how physical injury and other disaster outcomes 
relate to families long term after disaster. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
109 
 
REFERENCES 
Achenbach, T. (1991). Manual for the youth self-report and 1991 profile. Burlington, 
VT: University of Vermont, Department of Psychiatry. 
Ai, A.L., Peterson, C., & Huang, B. (2003). The Effect of Religious?Spiritual Coping 
on Positive Attitudes of Adult Muslim Refugees From Kosovo and Bosnia. The 
International Journal for the Psychology of Religion, 13, 29-47.  
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental 
disorders (4
th
 ed.). Washington, DC: Author.  
Anderson, K.M. (2004). Charting a course into the unknown: Banda Aceh, Indonesia, 
tsunami, 2004. Perspectives in Psychiatric Care, 43, 47-51. 
Assanangkornchai, S. Tangboonngam, S., & Edwards, J.G. (2004). The flooding of Hat 
Yai: predictors of adverse emotional responses to natural disaster. Stress and 
Health, 20, 81-89.  
Azarian, A., & Skriptchenko-G.V. (1998). Children in natural disasters: An experience of 
the 1998 earthquake in Armenia. The Academy of Experts in Traumatic Stress. 
http://www.aaets.org/arts/art38.htm. 
Beck, A.T. (1988). Depression. Philadelphia, PA: University of Pennsylvania Press. 
Blumstein, P., & Schwartz, P. (1983). American Couples. New York: William Morrow. 
Bond, G.D. (1992). The Buddhist revival in Sri Lanka: religious tradition, 
reinterpretation and response. New Delhi, India: Motilal Banarsidass Publishers. 
110 
 
Booth, A., & Johnson, D.R. (1994). Declining health and marital quality. Journal of 
Marriage and the Family, 56, 218-223.  
Boss, P. (1992). Primacy of Perception in family stress theory and measurement. Journal 
of Family Psychology, 2, 113-119.  
Briere, J., & Elliott, D. (2000). Prevalence, characteristics, and long-term sequelae of 
natural disaster exposure in the general population. Journal of Traumatic Stress, 
13,661-679. 
Bronisch, T., Maragkos, M., Freyer, C., Muller-Cyran, A., Butollo, W., Weimbs, R., & 
Platiel, P. (2006). Clinical insights: Crisis intervention after the tsunami in Phuket 
and Khao Lak. Crisis, 27, 42-47. 
Burkhardt, C.S., Clark, S.R., & Bennett, R. M. (2001). Pain coping strategies and quality 
of life in women with fibromyalgia: Does age make a difference? Journal of 
Musculoskeletal Pain, 5, 5-21.  
Caldera, T., Palma, L., Penayo, U., & Kullgren,  G. (2001). Psychological impact of the 
hurricane Mitch in Nicaragua in a one-year perspective. Soc Psychiatry Psychiatr 
Epidemiol, 36, 108-114.  
Calhoun, L.G., Cann, A., Tedeschi, R.G., & McMillan, J. (2000). A correlational test of 
the relationship between posttraumatic growth, religion, and cognitive processing. 
Journal of Traumatic Stress, 13, 521-527.  
Calier, I.V.E., Lamberts, R.D., Van Uchelen, A.J., Gersons, B.P.R. (1998). Clinical 
utility of a brief diagnostic test for posttraumatic stress disorder. Psychosom Med, 
60, 42-47.  
 
111 
 
Carballo, M., Heal, B., & Hernandez, M. (2005). Psychosocial aspects of the Tsunami. 
Journal of the Royal Society of Medicine, 98, 396-399.  
Carballo, M., Heal, B., & Horbaty, G. (2006).  Impact of the tsunami on psychological 
health and well-being. International Review of Psychiatry, 18, 217-223. 
Catapano, F., Malafronte, R., Lepre, F., Cozzolino, P., Arnone, R., Lorenzo, E., Tartaglia, 
G., Starace, F., Magliano, L., & Maj, M. (2001). Psychological consequences of 
the 1998 landslide in Sarno, Italy: a community study. Acta Psychiatr Scand, 104, 
438-442.  
Chandra, V., Pandav, R., & Bhugra, D. (2006). Mental health and psychosocial support 
after the tsunami: observations across affected nations. International Review of 
Psychiatry, 18, 205-211. 
Charlson, M.E., Pompei, P., Ales, K.L., & MacKenzie, C.R. (1987). A new method of 
classifying prognostic comorbidity in longitudinal studies: Development and 
validation. Journal of Chronic Diseases, 40, 373-383.  
Cloninger, C.R. (1992). The temperament and characteriInventory, version 9. St. Louis: 
Center for psychobiology of personality. 
Cohan, C.L., & Cole, S.W. (2002). Life course transitions and natural disaster: Marriage, 
birth, and divorce following hurricane Hugo. Journal of Family Psychology, 16, 
14-25. 
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. 
Journal of Health and Social Behavior, 24, 385-396.  
 
112 
 
Conger, R.D., Ge, X., Elder Jr., G.H., Lorenz, F.O., Simons, R.L., & Whitbeck, L.B. 
(1994). Economic stress, coercive family process, and developmental problems of 
adolescents. Child Development, 65, 541-561.  
Conger, R.D., Rueter, M.A., & Elder, Jr. G.H. (1999).Couple Resilience to Economic 
Pressure. Journal of Personality and Social Psychology, 76, 54-71. 
Connor, K.M., & Davidson, J.R.T. (2003). Development of a new resilience scale: The 
Connor-Davidson Resilience Scale (CD-RISC). Depress Anxiety, 18, 76-82. 
Centre for Research on the Epidemiology of Disasters (CRED), (2005). Trends in natural 
disasters. (2005). In UNEP/GRID-Arendal Maps and Graphics Library. Retrieved 
22:27, May 2, 2009 from http://maps.grida.no/go/graphic/trends-in-natural-
disasters. 
Chen, L. (1997). Grief as a transcendent function and teacher of spiritual growth. 
Pastoral Psychology, 46, 79-84. 
Crabbs, M.A., Heffron, E. (1981). Loss Associated with a Natural Disaster. The 
Personnel and Guidance Journal, 378-382. 
Dakin, J., & Wampler, R. (2008). Money doesn't buy happiness, but it helps: Marital 
satisfaction, psychological distress, and demographic differences between low- 
and middle-income clinic couples. American Journal of Family Therapy, 36, 300-
311.  
Davidson, J.R.T. (1996). The Davidson trauma scale. MultiHealth system, Inc., Toronto, 
Ontario. 
 
113 
 
Davidson, J.R.T., Connor, K.M., Lee, L. (2005). Beliefs in karma and reincarnation 
among survivors of violent trauma: A community survey. Soc Psychiatry 
Psychiatr Epidemiol, 40, 120-125.  
Dekel, R., & Solomon, Z. (2006). Marital relations among former prisoners of war: 
contribution of posttraumatic stress disorder, aggression, and sexual satisfaction. 
Journal of Family Psychology, 4, 709-712.  
De Silva, P. (2006). The tsunami and its aftermath in Sri Lanka: explorations of a 
Buddhist perspective. International Review of Psychiatry, 18, 281-287.  
Dorahy, M. J., Lewis, C. A., Schumaker, J. F., Akuamoah-Boateng, R., Duze, M. C., & 
Sibiya, T. E. (1998). A cross-cultural analysis of religion and life satisfaction. Mental 
Health, Religion and Culture, 1, 37-43. 
Eysenck, H.J. (1959). Manual for Maudsley Personality Inventory. Reprinted in India, 1993. 
Author.  
Exline, J.J., Smyth, J.M., Gregory, J., Hockemeyer, J., & Tulloch, H. (2005). Religious 
Framing by Individuals With PTSD When Writing About Traumatic Experiences. 
International Journal for the Psychology of Religion, 15, 17-33. 
Fontanesi, F., Gobeti, C., Zimmermann-Tansella, C., Tansella, M. (1985). Validation of 
the Italian version of the GHQ in a general practice setting. Psychol Med, 15, 411-
415.  
Freeman, C., Carlson, J., & Sperry, L. (1993). Adlerian marital therapy strategies with 
middle income couples facing financial stress. The American Journal of Family 
Therapy, 21(4), 324-332. 
 
114 
 
Goldberg, D.P. (1972). The detection of psychiatric illness by questionnaire. Maudsley 
Monograph no. 21. London: Oxford Press. 
Goldberg, D., & Williams, P. (1988). A user?s guide to the General Health 
Questionnaire. Windor: Berkshie, England: Nfer-Nelson. 
Grame, C.J., Tortorici, J.S., Healey, B.J., Dillingham, J.H., & Winklebaur, P. (1999). 
Addressing spiritual and religious issues of clients with a history of psychological 
trauma. Bulletin of the Menninger Clinic, 63, 223-239.  
Gurland, B., Golden, R.R., Teresi, J.A., & Challop, J. (1984). The SHORT-CARE: An 
efficient instrument for the assessment of depression, dementia and disability. 
Journal of Gerontology, 39, 166-169.  
Guruge, A.W.P. (2003). Serendipity of Andrew George: Sequel to free at last in paradise. 
Bloomington: AuthorHouse.  
Hendrick, S.S. (1988). A generic measure of relationship satisfaction. Journal of 
Marriage and the Family, 50, 93-98.  
Herman, J.L. (1992). Trauma and recovery. New York: Basic Books.  
Hill R. (1958). Genetic features of families under stress. Social Casework,49, 139-150. 
Hobfoll, S. (1989). Conservation of resources a new attempt at conceptualizing stress. 
American Psychologist, 44, 513-523. 
Hobfoll, S., & Spielberger, C.D. (1992a). Process of family stress: a response to Boss 
(1992) an Kazak (1992). Journal of Family Psychology, 6, 125-127.  
Hobfoll, S., & Spielberger, C.D. (1992b). Family stress: Integrating theory and 
measurement. Journal of Family Psychology, 6, 99-112. 
 
115 
 
Horowitz, M., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of 
subjective stress. Psychosocial Medicine, 41, 209-218.  
Hsu, C., O?Connor, M., & Lee, S. (2009). Understandings of death and dying for people 
of Chinese origin. Death Studies, 33, 153-174.  
International Centre for Migration and Health (IMCH). 2005. Public health consequences 
of the tsunami: Impact on displaced people. Volume I, November. 
Kaniasty, K, & Norris, F.H. (1993). A test of the social support deterioration model in the 
context of natural disaster. Journal of Personality and Social Psychology, 64, 
395-408. 
Kazak, A.E. (1992). Stress, change, and families: theoretical and methodological 
considerations. Journal of Family Psychology, 6, 120-124.  
Kroger, J., & Adair, V. (2008). Symbolic meanings of valued personal objects in identity 
transitions of late adulthood. Identity: An international journal of theory, 8, 5-24. 
Lavee, Y., McCubbin, H.I., & Patterson, J.M. (1985). The double ABCX model of family 
stress and adaptation: an empirical test by analysis of structural equations with 
latent variables. Journal of Marriage and the Family, 47, 811-825.  
McCubbin, H., Larsen, A., & Olson, D. (1987). F-COPES: Family Crisis Oriented 
Personal Evaluation Scales. In H.I. McCubbin & A. I. Thompson (Eds.), Family 
assessment inventories for research and practice (pp. 193-207). Madison: 
University of Wisconsin-Madison.  
 
116 
 
McMillen, J.C., Smith, E.M., & Fisher, R.H. (1997). Perceived benefit and mental health 
after three types of disaster. Journal of Consulting and Clinical Psychology, 65, 
733-739.  
Menzel Baker, S.M., Hunt, D.M., & Rittenburg, T.L. (2007). Consumer vulnerability as a 
shared experience: tornado recovery process in Wright, Wyoming. Journal of 
Public Policy & Marketing, 26, 6-19. 
Murray, Jr.T.L., Murray, C.E., & Daniels, M.H. (2007). Stress and family relationship 
functioning as indicators of the severity of fibromyalgia symptoms: a regression 
analysis. Stress and Health, 23, 3-8.  
Najarian, L.M., Goenjian, A.K., Pelcovitz, D., Mandel, F., & Najarian, B. (1996). 
Reolcation after a disaster: Posttraumait stress disorder in Armenia after the 
earthquake. Journal of the American Academy of Child & Adolescent Psychiatry, 
35, 374-383. 
Nilchaikovit, T., Sukying, C., & Silpakit, C. (1996). Reliability and validity of the Thai 
version of the General Health Questionnaire. Journal of the Psychiatric 
Association of Thailand, 41, 2-17.  
Norris, F., Murrell, S. (1987). Transitory impact of life-event stress on psychological 
symptoms in older adults. Journal of Health and Social Behavior, 28, 197-211.  
North, R.J., Holahan, C.J., Moos, R.H., & Cronkite, R.C. (2008). Family support, family 
income, and happiness: A 10-year perspective. Journal of Familiy Psychology, 
22, 475-483.  
 
117 
 
North, C.S., Hong, B.A., Suris, A. & Spitznagel, E.L. (2008). Distinguishing distress and 
psychopathology among survivors of the Oakland/Berkeley firestorm. Psychiatry, 
71, 35-45. 
O?Reilly, B.K. (1996) Case study: treatment of post traumatic stress disorder from a 
Christian perspective. Psyche and Faith Beyond Professionalism, 9, 53-57. 
Pargament, K.I. (1996). Religious methods of coping: Resources for the conservation and 
transformation of significance. In Religion and the clinical practice of 
psychology, E.P. Shafranske (Ed.), pp. 215-239. Washington, D.C.: American 
Psychological Association. 
Pargament, K.I., Ensing, D.S., Falgout, K., Olsen, H., Reilly, B., Van Haitsma, K., & 
Warren, R. (1990). God help me (I): Religious coping effects as predictors of 
outcomes to significant negative life events. American Journal of Community 
Psychology, 18, 793-824.  
Pargament, K.I., & Hahn, J. (1986). God and the just world: Causal and coping 
attributions to God in health situations. Journal for the Scientific Study of 
Religion, 25,193-207.  
Pargament, K.I., Ishler, K., Dubow, E., Stanik, P., Cullman, R.R., Crowe, P., Albert, M., 
& Royster, B.J. (1994). Methods of religious coping with the Gulf War: Cross-
sectional and longitudinal analyses. Journal for the Scientific Study of Religion, 
33, 347-361.  
Park, C.L., Cohen, L.H., & Murch, R.L. (1996). Assessment and prediction of stress-
related growth. Journal of Personality, 64, 71-105.  
 
118 
 
Pearlin, L.I., & Schooler, C. (1978). The structure of coping. Journal of Health and 
Social Behavior, 19, 2-21.  
Perez-Sales, P., Cervellon, P., Vazquez, C., Vidales, D., Gaborit, M. (2005). Post-
traumatic factors and resilience: The role of shelter management and survivors? 
attitudes after the earthquakes in El Salvador (2001). Journal of Community & 
Applied Social Psychology, 15, 368-382.  
Polonsky, W.H., Fisher, L., Earles, J., Dudl, R.J., Lees, J., Mullan, J., et al. (2005). 
Assessing psychosocial distress in diabetes: Development of the diabetes distress 
scale. Diabetes Care, 28, 626-631.  
Portes, A., & Rumbaut, R.G. (2001). Legacies. The story of the immigrant second 
generation. Berkeley, CA: University of California Press.  
Proctor, L.J., Fauchier, A., Oliver, P.H., Ramos, M.C., Rios, M.A., & Margolin, G. 
(2007). Family context and young children?s responses to earthquake. Journal of 
Child Psychology and Psychiatry, 48, 941-949.  
Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the 
general population. Applied Psychological Measurement, 1, 385-401.  
Rahula, W. (1996). What Buddha Taught. Nedimala, Dehiwala, Sri Lanka: Buddhist 
Cultural Centre.  
Renshaw, K.D., Rodrigues, C.S., & Jones, D.H. (2008). Psychological symptoms and 
marital satisfaction in spouses of operation Iraqi freedom veterans: Relationships 
with spouses? perceptions of veterans? experiences and symptoms. Journal of 
Family Psychology, 22, 586-594.  
 
119 
 
Reynolds, C.R., & Richmond, B.O. (1978). What I think and feel: A revised measure of 
children?s manifest anxiety. Journal of Abnormal Child Psychology, 6, 271-280. 
Robins, L.N., Helzer, J.E., Cottler, L., & Goldring, E. (1989). NIMH diagnostic interview 
schedule, version III-revised. St. Louis: Washington University. 
Robins, L.N., Smith, E.M. (1983). The diagnostic interview schedule/disaster 
supplement. St Louis: Washington University. 
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton 
University Press.  
Rotter, J.B. (1966). Generalized expectancies for internal versus external control of 
reinforcement. Psychological Monographs, 80 (Whole No. 609). 
Rudnick, A. (1997). Psychotherapeutic effects of religion. Psychosomatics, 38, 576-578. 
Sarason, J.G.,Johnson, J.H., & Siegal, J.M. (1978). Assessing the impact of life changes: 
Development of the Life Experiences Survey. Journal of Consulting and Clinical 
Psychology, 46, 932-946.   
Sattler, D.N., DeAlvarado, A.M.G, & DeCastro, N.B. (2006).  El Salvador earthquakes: 
Relationships among acute stress disorder symptoms, depression, traumatic even 
exposure, and resorce loss. Journal of Traumatic Stress, 19, 879-893. 
Seneviratne, H.L. (1999). The work of kings: the new Buddhism in Sri Lanka. Chicago: 
University of Chicago Press.  
Shaw, A., Joseph, S., & Linley, P.A. (2005).  Religion, Spirituality, and posttraumatic 
growth: a systematic review. Mental Health, Religion & Culture, 8, 1-11. 
 
120 
 
Skowron, E.A., & Schmitt, T.A. (2003). Assessing interpersonal fusion: Reliability and 
validity of a new DSI fusion with others subscale. Journal of Marital and Family 
Therapy, 29, 209-222.  
Smith, B.W., Pargament, K.I., Brant, C., & Oliver, J.M. (2000). Noah revisited: religious 
coping by church members and the impact of the 1993 midwest flood. Journal of 
Community Psychology, 28,169-186. 
Soonthornchaiya, R., & Dancy, B.L. (2006). Perceptions of depression among elderly 
Thai immigrants. Issues in Mental Health Nursing, 27, 681-698.  
Spanier, G.B. (1976). Measuring dyadic adjustment: New scales for assessing the quality 
of marriage and similar dyads. Journal of Marriage and the Family, 38,15-28.   
Speilberger, C.D., Gorsuch, R.L., & Lushene, R.E. (1970). STAI manual for the state-
trait anxiety inventory. California: Consulting Psychologists Press.  
Suar, D., Mandal, M.K., & Khuntia, R. (2002). Supercyclone in Orissa: An assessment of 
psychological status of survivors. Journal of Traumatic Stress, 15, 313-319.  
Thera, N. (1993). The Dhammapada: PALI text & translation with stories in brief & 
notes. Taipei, Taiwan: The Corporate Body of the Buddha Educational 
Foundation.  
Trief, P.M., Morin, P.C., Izquierdo, R., Teresi, J., Starren, J., Shea, S., & Weinstock, R.S. 
(2006). Marital quality and diabetes outcomes: The IDEATel project. Families, 
Systems, & Health, 24, 318-331.  
Tsunami Evaluation Coalition (TEC). (2005). Initial Findings, December 2005. 
 
121 
 
Vigil, J.M., & Geary, D.C. (2008). A Preliminary Investigation of Family Coping Styles 
and Psychological Well-Being Among Adolescent Survivors of Hurricane 
Katrina. Journal of Family Psychology, 22, 176-180.  
Wade, S.L., Carey, J., & Wolfe, C.R. (2006). An online family intervention to reduce 
parental distress following pediatric brain injury. Journal of Consulting and 
Clinical Psychology, 74, 445-454. 
Weidmann, A., Fehm, L., & Fydrich, T. (2008). Covering the tsunami disaster: 
subsequent post-traumatic and depressive symptoms and associated social factors. 
Stress and Health, 24, 129-135. 
Weiss, D. & Marmar, C. (1997). The Impact of Event Scale-Revised. In J. Wilson & T. 
Keane (Eds.), Assessing psychological trauma and PTSD (pp. 399-411). New 
York: Guilford Press. 
Wickrama, K., & Kaspar, V. (2007). Family context of mental health risk in Tsunami-
exposed adolescents: Findings from a pilot study in Sri Lanka. Social Science & 
Medicine, 64, 713-723. 
Wickrama, K.A.S., & Wickrama, K.A. (2008). Family context of mental health risk in 
Tsunami affected mothers: Finding from a pilot study in Sri Lanka. Social Science 
& Medicine, 66, 994-1007. 
Wickremeratne, S., & Bond, G.D. (2006). Buddha in Sri Lanka: Remembered Yesterdays. 
New York: SUNY Press.  
Wu, H.C., Pesus, C., Chou, F.H.C., Su, C.Y., Tsai, K.Y., Ou-Yang, W.C., Su, T.T.P., 
Chao, S.S., Sun, W.J., & Chen, M.C. (2006). Survey of quality of life and related 
 
122 
 
risk factors for a Taiwanese village population 3 years post-earthquake. The 
Authors Journal compliation: The Royal Australian and New Zealand College of 
Psychiatrists, 355-361.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
123 
 
APPENDIX A 
Marital Satisfaction 
We would like to know overall how happy or satisfied you are with your marital 
relationship 
 
p19a.    Overall, how happy are you with your relationship? (circle one) 
1. Extremely unhappy 
2. Very unhappy 
3. Unhappy 
4. Happy 
5. Very happy 
6. Extremely happy 
 
p19b.    Overall, how satisfied are you with your relationship? (circle one) 
1. Not at all satisfied 
2. Not very satisfied 
3. Somewhat dissatisfied 
4. Somewhat satisfied 
5. Very satisfied   
6. Completely satisfied 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
124 
 
APPENDIX B 
Family Functioning 
Please indicate by circling the number that best reflects how strongly you agree or 
disagree with these statements about your family during the past 4 years. 
 
                                                                                 Strongly                                            Stron ly g
                                                                                                                         Disagree   Disagree Neutral Agree Agree 
 
 
p18a.    We help each other in times of trouble    1         2        3       4        5 
p18b.     We can talk to each other about sad  
 feelings    1         2        3       4        5 
p18c.    We talk to each other about how we feel    1         2        3       4        5 
p18d.    We tell each other how we feel about one  
 another     1         2        3       4        5 
p18e.     Each person in the family is accepted for  
  who they are     1         2        3       4        5 
p18f.     There are lots of bad feelings in our family    1         2        3       4        5 
p18g.    We are able to solve problems in our family  1         2        3       4        5 
p18h.     We get along well together     1         2        3       4        5 
 
p18i.      We confide in each other      1         2        3       4        5   
p18j.      In times of crisis, we can turn to each other  
  for support      1         2        3       4        5 
p18k.     We avoid discussing our fears and concerns  1         2        3       4        5 
p18l.      We feel accepted for what we are       1         2        3       4        5 
p18m.    Making decisions is a problem for our family1         2        3       4        5 
p18n.    We laugh, joke, and have good times  
together        1         2        3       4        5 
 
 
 
125 
 
APPENDIX C 
Physical Health Problems from the tsunami 
Scale=1strongly disagree 2 disagree 3 neutral 4 agree 5 strongly agree  
 
p9f. Tsunami related physical health problems interfere  
with my daily activities, even today     1 2 3 4 5  
p9i. My spouse?s Tsunami related physical health problems interfere  
 with my daily activities, even today     1 2 3 4 5 
p9j. My children?s Tsunami related physical health problems interfere  
 with my daily activities, even today     1 2 3 4 5 
        
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
126 
 
APPENDIX D 
DSM-IV Diagnostic Interview for PTSD 
 Instructions: Sometimes things happen to people that would upset or frighten 
almost  anyone. Many such things like being injured, seeing other people hurt or killed, 
becoming lost or separated from people you know, or seeing houses destroyed happened 
to people during the Tsunami. We are NOT going to ask you to describe these events, but 
we are interested in the effect such events might have had on you, recently. 
 
Did you experience any of the above or similar events that upset or frightened you? If 
yes, please continue; if not, thank you for participating in this study. 
 
Recently (during the past four months): circle one 
 
p12a.    Did you find that traumatic events kept coming back to you in some way?  
NO(2) YES(1) 
 
p12b.    Did you keep thinking about the event even when you did not want to?      
NO(2) YES(1) 
 
p12c.    Did you have bad dreams about the event?                                                   
NO(2) YES(1) 
 
p12d.    Did you find yourself acting or feeling as though you were back at that time?                                
NO(2) YES(1) 
 
p12e.    Did you feel a lot worse when you were reminded of the event?                 
NO(2) YES(1) 
 
Recently, (during the past four months): circle one 
 
p12g.    Did you make a special effort to avoid thinking (or talking) about what happened, 
or getting upset about it?  
NO(2) YES(1) 
p12h.    Did you stay away from things that would remind you of the event?  
NO(2) YES(1) 
 
127 
 
p12i.     Did you have trouble remembering some important part of what happened?  
NO(2) YES(1) 
p12j.     Were you much less interested in things that used to be important to you, like 
sports, hobbies, social activities?  
NO(2) YES(1) 
p12k.    Did you feel distant or cut off from others?  
NO(2) YES(1) 
p12l.     Were you unable to feel strong emotions or were you emotionally numb (e.g., 
couldn't feel happy, sad, or excited about things)?  
NO(2) YES(1) 
p12m.   Have you ever felt that you have no future?  
NO(2) YES(1) 
 
p12n.    Did you have trouble sleeping? 
NO(2) YES(1)  
 
p12o.    Were you unusually irritable or did you have outbursts of anger?  
NO(2) YES(1) 
 
p12p.    Did you have trouble concentrating?  
NO(2) YES(1) 
 
p12q.    Were you watchful or on guard even when there was no reason to be?  
NO(2) YES(1) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
128 
 
APPENDIX E 
Center for epidemiological Studies Depression Scale (CESD) 
 
 Instructions: In this section, we are going to present to you some sentences that 
say something about how people sometimes feel. Please read each sentence and indicate 
the number that best reflects how often you have felt this way IN THE PAST 7 DAYS.  
 
0                     1                  2    3 
 Rarely or none      Some or a little        Occasionally or a Most or all  
    of the time             of the time    moderate amount of the time of the time 
(less than 1 day)      (1 to 2 days)               (3 to 4 days) (5 to 7 days) 
 
p11a.    I was bothered by things that usually don?t bother me.  
p11b.    I did not feel like eating; my appetite was poor.  
p11c.    I felt that I could not shake off the blues even with help  
from my family and friends.   
p11d.    I felt that I was just as good as other people.  
p11e.    I had trouble keeping my mind on what I was doing.  
p11f.     I felt depressed.  
p11g.    I felt that everything I did was an effort.  
p11h.    I felt hopeful about the future.  
p11i.     I thought my life had been a failure.  
p11j.     I felt fearful.  
p11k.    My sleep was restless.  
p11l.     I was happy.  
p11m.   I talked less than usual.  
p11n.    I felt lonely.  
p11o.    People were unfriendly.  
p11p.    I enjoyed life.  
p11q.    I had crying spells.  
129 
 
p11r.     I felt sad.  
p11s.    I felt that people disliked me.  
p11t.     I could not ?get going.?  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
130 
 
APPENDIX F 
Religious Attributions 
p14s. Do you believe the tsunami occurred due to Karma? 
1 NO 2 YES 
 
p14t. Do you believe the tsunami consequences were unavoidable due to your Karma? 
1 NO 2 YES 
 
p14u. Were you not effected or surprised by the tsunami because it is one instance of 
Samsara? 
1 NO 2 YES 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
131 
 
APPENDIX G 
Family Religious Activities 
 
 
14f. How often do you go to the temple with your husband? 
 
1 almost never 2 monthly 3 weekly 4 more than once a week 5 almost daily 
 
14g. How often do you go to the temple with other members of your family? 
 
1 almost never 2 monthly 3 weekly 4 more than once a week 5 almost daily