This preview shows page 1. Sign up to view the full content.
Unformatted text preview: CARRI Research Report 10 BEHAVIORAL SCIENCE PERSPECTIVES ON RESILIENCE Fran H. Norris
Dartmouth Medical School
National Center for PTSD
National Center for Disaster Mental Health Research
National Consortium for the Study of Terrorism and Responses to Terrorism Report Prepared for
The Community and Regional Resilience Institute (CARRI)
Oak Ridge National Laboratory June 2010 Behavioral Science Perspectives on R esilience RESEARCH FIN D IN GS ABOUT COMMUN ITY
A N D REGION AL RESILIEN CE
O ne of the comm itm ents of the Com m unity and Regional Resilience In stitute (CARRI) is to
u nd erstand w hat resilience is and how to get there, based on research evidence.
A s one resource for this effort, CARRI has com m issioned a num ber of summ arie s of existing
k now led ge about resilience, arising from a num ber of d ifferent research trad itions. This paper is
o ne in a series of such sum m aries, w hich w ill be integrated w ith new resilience explorations in
s everal CARRI partner cities and w ith further d i scussions w ith the research com m unity and
o ther stakehold ers to serve as the know led ge base for the in stitute .
For further inform ation about CARRI’s research com ponent, contact Thomas J. W ilbanks,
w [email protected] v , or Sherry B. Wright, w [email protected] . CA RRI R esearch R eport 10 iii Behavioral Science Perspectives on Resilience COMMUN ITY AN D REGION AL RESILIEN CE IN STITUTE
O ak Rid ge N ational Laboratory’s (ORN L) Com munity and Regional Resilience In stitute
(CARRI) is a program of the Congressionally fund ed Southeast Region Research Initiative.
C ARRI is a r egional program w ith national im plications for how com m unities and regions
p repare for, respond to, and recover from catastrophic events. CARRI w ill d evelop the
p rocesses and tools w ith w hich com m unities and regions can better prepare to w ithstand the
e ffects of natural and h um an -mad e d isasters by collaboratively d eveloping an und erstanding of
com m unity resilience that is accurate, d efensible, w elcom ed , and applicable to com m unities
a cross the region and the nation.
C ARRI is presently w orking w ith thre e partner com m unities in the Southeast: Gulfport,
M ississippi; Charleston/ Low Country, South Carolina; and the Mem phis, Tennessee, urban
a rea. These partner comm unities w ill help CARRI d efine com m unity resilience and test it at the
com m unity level. Using input from the partner com m unities, lessons learned from around the
n ation, and the guid ance of ORN L-convened researchers w ho are experts in the d iverse
d isciplines that com prise resilience, CARRI w ill develop a com m unity resilience fram ew ork that
o utlines p rocesses and tools that com m unities can use to becom e m ore resilient. Of critical
im portance, CARRI w ill d em onstrate that resilient com m unities gain econom ically from
r esilience investm ents.
From its beginning, CARRI w as d esigned to com bine com m unity eng agement activities
w ith research activities. Resilient com m unities are the objective, but research is critical to ensure
t hat CARRI’s und erstand ing is based on know led ge -based evid ence and not just ad hoc id eas —
w e w ant to get it right. To help w ith this, C ARRI has com missioned a series of sum m aries on
t he current state of resilience know led ge by lead ing experts in the field . This kind of interactive
linkage betw een research and practice is very rare.
In ad d ition to its partner com m unities and national and local research team s, CARRI has
e stablished a robust social netw ork of private businesses, government agencies, and non governm ental associations. This netw ork is critical to the CARRI research and engagem ent
p rocess and provides CARRI the valuable inform a tion necessary to ensure that w e rem ain on
t he right path. Frequent conversation w ith business lead ers, governm ent officials, and volunteer
o rganizations provid e a bottom -up know led ge from practitioners and stakehold ers w ith real w orld , on -the -ground e xpe rience. We accept that this program cannot truly und erstand
com m unity resilience based only on stud ies in a laboratory or university. CARRI seeks to
e xpand this social netw ork at every opportunity and gains from each new contact. w w w .resilientUS.org CA RRI R esearch R eport 10 v Behavioral Science Perspectives on Resilience LIST OF RESEARCH REPORTS BY N UMBER C ARRI Report 1: Susan L. Cutter, Lind sey Barnes, Melissa Berry, Christopher Burton, Elijah
Evans, Eric Tate, and Jennifer Webb, Community and Regional Resilience:
Perspectives from Hazards, Disasters, and Emergency M anagement ,
Septem ber 2008. C ARRI Report 2: Susanne C. Mo ser, R esilience in the Face of Global Environmental Change,
Septem ber 2008. C ARRI Report 3: C raig Colten, Robert Kates, and Shirley Laska, Community Resilience: Lessons
from N ew Orleans and Hurricane Katrina, Septem ber 2008. C ARRI Report 4: Betty H earn Mo rrow , Community Resilience: A Social Justice Perspective,
Septem ber 2008. C ARRI Report 5: Lance Gund erson, Comparing Ecological and Human Community Resilience,
January 2009. C ARRI Report 6: Kathleen Tierney, D isaster Response: Research Findings and Their Implications
for Resilience M easures, M arch 2009. C ARRI Report 7: Thom as J. Wilbanks , H ow Geographic Scale M atters in Seeking Community
R esilience, A ugust 2009. C ARRI Report 8: A d am Rose , Economic Resilience to Disasters, N ovem ber 2009. C ARRI Report 9: C raig Colten, Building Community Resilience: A Summary of Case Studies from
Charleston, Gulfport, and M emphis, A pril 2010. CA RRI R esearch R eport 10 v ii Behavioral Science Perspectives on Resilience CON TEN TS
RESEARCH FIN DIN GS ABOUT COMMUN ITY AN D REGION AL RESILIEN CE ....................... iii
C OMMUN ITY AN D REGION AL RESILIEN CE IN STITUTE ............................................................. v
LIST OF RESEARCH REPORTS BY N UMBER....................................................................................v ii
LIST OF FIGURES ..................................................................................................................................... xi
BEH AVIORAL SCIEN CE PERSPECTIVES ON RESILIEN CE: IN TRODUCTION
TO TH E REVIEW ....................................................................................................................................... 1
1. O VERVIEW: RESILIEN CE FRAMEWORKS, C ORE C ON CEPTS, A N D P RIN CIPLES
FROM BEH AVIORAL SCIEN CE ..................................................................................................... 3
1.1 M eanings of Resilience ............................................................................................................ 3
1.2 P s ychological Manifestations of Resilience .......................................................................... 4
1.3 P revalence of Resilience .......................................................................................................... 5
1.4 Sources of Resilience ............................................................................................................... 7
1.4.1 A ttachm ents and Social Relationships ..................................................................... 7
1.4.2 A gency .......................................................................................................................... 8
1.4.3 Intelligence, Problem -Solving, Inform ation -Processing, and SelfRegulatory System s .................................................................................................... 9
1.4.4 M icrosyste m s, Comm unities, and Macrosystem s .................................................. 9
1.5 C onservation of Resources for Resilience .......................................................................... 10 2. A N TICIPATION A N D REDUCTION O F VULN ERABILITIES: H OW IN DIVIDUALS
P ERCEIVE A N D REDUCE RISK .................................................................................................... 11
2.1 Roots of Anticipation: Cognitive Schem as, Assum ptions, and Illusions ....................... 11
2.2 O ptim istic Bias ....................................................................................................................... 12
2.3 Risk Perception ....................................................................................................................... 12
2.4 From Risk Perception to Action: the Precaution Ad option Process ............................... 13
2.5 O ther Psychological Influences on Precautionary Behavior ........................................... 14
2.6 Red ucing Vulnerability through Col lective Action .......................................................... 16
2.7 C losing N ote for Section 2 .................................................................................................... 17 3. RESPON SE A N D RECOVERY: H OW IN DIVIDUALS REACT TO, C OPE W ITH ,
A N D RECOVER FROM STRESSFUL EVEN TS ............................................................................ 18
3.1 O verview of the Stress Process ............................................................................................ 18
3.2 C ognitive Appraisals of Stress ............................................................................................. 19
3.3 W ays of Coping ...................................................................................................................... 21
3.3.1 C oping Self-Efficacy .................................................................................................. 23
3.4 Social Support Dynamics ...................................................................................................... 24
3.5 C losing N ote for Section 3 .................................................................................................... 26 4. IMPLICATION S O F BEH AVIORAL SCIEN CE: RESILIEN CE A N D
IN TERVEN TION .............................................................................................................................. 27 5. REFEREN CES ................................................................................................................................... 30 CA RRI R esearch R eport 10 ix Behavioral Science Perspectives on Resilience LIST OF FIGURES
Relationship betw een perceived risk and behavior o ver tim e ....................................... 14
The d irect effects of injustice, id entity, and efficacy (paths a, b, and c) on
collective action; id entity also has ind irect effects through injustice ( d*a) and
e fficacy (e*c) ........................................................................................................................... 17
M od el of s tress resistance and resilience .......................................................................... 19
Social support d eterioration d eterrence m od el ................................................................ 26
A m od el of the process w hereby psychosocial stress ind uces psychopathology
a nd som e conceptions of h ow to counteract this process ............................................... 28 CA RRI R esearch R eport 10 xi Behavioral Science Perspectives on Resilience BEHAVIORAL SCIEN CE PERSPECTIVES ON RESILIEN CE:
IN TROD UCTION TO THE REVIEW
D ecad es of research on the hum an stress response, both experim ental and observational in
m ethod , have yield ed a w ealth of know led ge that is d irectly relevant to und erstand ing the
h um an capacity to ad apt to a w id e variety of stres sful events and situations. The behavioral
s ciences, especially the various d isciplines of psychology, have been concerned prim arily w ith
t he resilience of ind ivid uals, in particular w ith und erstand ing the biological, cognitive,
e m otional, and behavioral p rocesses that protect persons from harm and prom ote their w ell being and grow th.
G iven CARRI’s m ission to prom ote community r esilience, it is im portant that lead ing
s cholars in behavioral science have recognized that the resilience of ind ividuals is d epen d ent
u pon the resilience of the com m unities in w hich they are em bed d ed . Such an ecological
p erspective is especially appropriate in the context of collective stressors, such as d isaster. After
a ll, com m unity m embers’ postd isaster w ell -being d epend s in part o n the effectiveness o f
o rganizational responses, and u ltim ately the purpose of d isaster m anagement is to ensure the
h ealth and s afe ty of the public. For this reason, the prevalence of psychological w ellness in the
com m unity is one im portant criterion for e valuating com m unity resilience in the afterm ath of
e xtrem e events. Within lim its, postd isaster d istress is n orm al, b ut w hen highly preval ent,
s evere, or persistent, it p oints to system ic needs, societal p roblem s , or response failures . In
s hort, com m unity m ental health is the sine qua non o f com m unity resilience .
It is equally im portant to recognize that the resilience of com m unities is d epend ent upon the
r esilience of the ind ivid uals w ho com pose them . Influences betw een individ ual and comm unity
a re m utual, transactional, and bid irectional. Effective prom otion of c om m unity resilience
t herefore requires know led ge of the various psychological factors that influence ind ivid ual and
collective behavior before and after ad verse events. Accord ingly, the purpose of this review is
t o sum m arize perspectives from behavioral sc ience that are useful for understand ing the
r esilience of ind ivid uals. The review gives particular attention to those concepts that are likely
t o be m ost specific to a behavioral science perspective, relative to sociological or natural science
p erspectives . Thus, the review focuses heavily on how hum an cognitions (e.g., core belief
s ystem s, appraisals, attributions) influence action (i.e., goal -directed behavior).
The review is organized into four prim ary parts. Section 1 p resents an overview of resil ience
p erspectives in behavioral science. H ere the em phasis is on the big picture: What is resilience,
h ow is it m anifested , how com m on is it, and w here d oes it com e from ? The next tw o s ections o f
t he review are organized accord ing to the four ―chara cteristics‖ of a resilient com m unity
o utlined by CARRI. It a nt icipat es p roblems, opportunities , a nd the potential for surprise.
It reduces v ulnerabilities to d evelopm ent paths, socio econom ic cond itions , a nd
id entified threats.
It responds e ffectively, fairly and legitim ately.
It recovers r apid ly, safely , a nd fairly. A s w ill be seen, t he behavioral science literature m ak es clear linkages betw een the
characteristics o utlined by CARRI; for exam ple, a nticipat ion of harm lead s to a ctions that
r ed uce v ulnerabilities. Section 2 o f this review (Anticipation and Red uction of Vulnerabilities)
d escribes core belief system s (schem as) that und erlie the hum an ability to anticipate problem s
a nd opportunities a nd focuses on precaution ad option and collecti ve action as illustrative CA RRI R esearch R eport 10 1 Behavioral Science Perspectives on Resilience b ehaviors that aim to red uce vulnerabilities of individ uals and groups. Section 3 (Response and
Recovery) d escribes core concepts from stress research that ind icate how ind ivid uals respond to
a d versity, w ith a focus on cognit ive appraisals, w ays of coping, and social support d ynam ics .
The final section, Section 4 (Im plications), d escribes how the concept of resilience creates a
coherent fram ew ork for integrating interventions that vary in their tim ing and target.
Before p roceed ing, I should ad d a note about how this review w as cond ucted . The m aterial
cam e from three prim ary sources: a system atic search for recently published articles on
p sychological resilience; the author’s pre existing know led ge of key w ritings in stres s research
a nd theory over the past 25 years; and a specific search for relevant review s and m eta -analyses
in journals such as Psychological Bulletin a nd Clinical Psychology Review . The challenge w as to be
n either too narrow nor too broad in scope. A revie w limited to the results of a search on
― resilience‖ w ould be severely flaw ed . Relatively little of the relevant w ork in behavioral
s cience explicitly m entions resilience, lead ing m any previous w riters to conclud e that resilience
h as been und er -recognized a nd und er -stud ied . H ow ever, a careful read ing of the stress and
coping literature challenges this view and instead points to a long -stand ing concern w ith factors
t hat prom ote hum an ad aptation, d evelopm ent, and w ell -being. Thus this review sum m arizes
m ajor w orks that id entify the cognitive, em otional, and behavioral und erpinnings of stress
a nticipation, mitigation, response, and recovery regard less of w hether or not a resilience fram e
w as used by the authors. On the other hand , it w as not possible to review t he thousand s of
s tud ies that have ad d ressed this r eport’s m ajor them es. For this reason, I relied heavily on key
conceptual papers, prior review s, and notew orthy m eta -analyses to id entify the m ost im portant
p erspectives to sum m arize. While I aim ed t o be specific enough to show the attention to
m echanism that is characteristic of behavioral science, I glossed over som e nuances that seem ed
less essential. I tried to strike a balance betw een old and new , by includ ing references to classic
p apers, as w ell as recent ad vancem ents. 2 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience 1. O VERVIEW: RESILIEN CE FRAMEWORKS, CORE CON CEPTS, AN D
PRIN CIPLES FROM BEHA VIORAL SCIEN CE 1.1 M eaning s o f Resilience
In the behavioral sciences, resilience has been d efined variously as the process of, capacity
for, or o utcom e of successful ad aptation after traum a , ad versity, o r severe stress. In his
influential A merican Psychologist a rticle, G eorge Bonanno (2004) a rgued that it is im portant to
d istinguish resilience from recovery. R ecovery , h e argued , connotes a t raject ory (pattern of
change) in w hich t he individ ual’s n orm al fu nctioning d eteriorates , usually for a p eriod of
s everal m onths, a nd t hen grad ually returns to pre -event levels. By contrast, resilience r eflects a n
a bility to m aintain a s table equilibrium . Resilie nt ind ivid uals m ay experience transient
p erturbations in norm al functioning (e.g., restless sleep) but generally exhibit a s table trajectory
o f healthy functioning a fter stress (p. 20).
This notion that resilience m ust be und erstood as a longitud inal t rajectory o f good
fun ctioning h as since b ecom e a m ajor them e in behavioral science resilience theory and
r esearch. The absence of d ysfunction a t one point in tim e d oes not e n sure t hat it w as absent
p reviously or w ill not occur at som e later point i n tim e. In fact, N orris and colleagues (2008,
p . 130) e mbed d ed this observation into their central d efinition of resilience as ― a process linking
a s et of ad aptive capacities to a positive trajectory of functioning and ad aptation after a
d isturbance .” In t his d efinition, N orris and colleagues carefully did not equate resilience w ith
t he outcom e, but rather w ith the process (trajectory) linking resources (adaptive capacities) to
o utcom es (ad aptation).
There are s om e sem antic nuances that m ust be reckoned w i th . W hereas Bonanno
e m phasized psychological stability a s the found ation of resilience, others have argued t hat
r esilience is better characterized as adaptability t han as s tability (Ad ger, 2000; Klein, N icholls, &
Thom alia, 2003). In other w ord s, resilience is a process of ―bouncing back‖ from harm rather
t han im m unity from harm (Garm ezy, 1991; Layne et al., 2007). This im age can be traced to its
o rigins in mathematics and physics, w here resistance w as d efined as t he force (stress) required
t o d isplace a system from equilibrium , w hereas resilience w as d efined as the tim e required for
t he system to return to equilibrium once d isplaced (Bod in & Wim an, 2004). In physics, resilience
h as little to d o w ith how large the i nitial d isplacem ent is or even how severe the oscillations are
b ut is m ore precisely the speed w ith w hich hom eostasis is restor ed . Applying this analogy to
t he hum an stress response, w e should use the concept of resistance (not resilience) to d escribe
s itu ations w here d ysfunction is m inim al because coping resources have ef fectively blocked
t he s tressor.
N orris, Tracy, and Galea (2009) elaborated on this them e further by suggest ing, as have
s om e others (Layne et al., 2007), that resilience m ay be best und er stood and m easured as one
m em ber of a set o f possible trajectories that m ay follow exposure to traum a o r severe stress . In
a d d ition to resistance, resilience, and recovery t rajectories, there are at least three other trajectories
o f potential interest in r esearch about the consequences of stressful events: relapsing/remitting, in
w hich severe d istress d isplay s a cyclical course; delayed dysfunction , in w hich post -traum atic
s tress d isord er (PTSD) o r som e other traum a -related d isord er em erges after consid erab le tim e
h as passed; and chronic dysfunction , w here an initial stress reaction persists. D espite recent
a d vances, cross -sectional stud ies still m ake up the vast m ajority of research o n the consequences
o f m ajor e vents and generally cannot d istinguish these t rajectories from one another. A t
m inim um , t hree w aves o f d ata are required t o d raw inferences about t he full range of potential
p ostevent t rajectories. CA RRI R esearch R eport 10 3 Behavioral Science Perspectives on Resilience Lepore and Revenson (2006) presented a som ew hat d ifferent p erspective o n the t hem e of
p ost -event t rajectories. Resilience , in their view , can take any of t hree forms: r esistance,
r ecovery, o r reconfiguration. They characterized resistance as being like the tree that stand s still,
u nd isturbed , in the face of the w ind . Resistant people exhibit norm al fu nctioning before, d uring,
a nd after the stressor occurs. They characterized recovery a s b eing like the tree t hat b end s to
a ccom m od ate t he w ind . The stressor d isrupts the person’s norm al state of functioning, but
w hen t he s tressor passes, t he p erson resum es h is or her n orm al or pre -stressor level of
functioning. They view ed this p rocess a s a continuum , and w hile the sooner rec overy occurs the
b etter, persons w ho are slow to recover are still resilient relative to those w ho d o not recover at
a ll. Lepore and Revenson t hen a d ded reconfiguration t o the m ix: in this case, the t ree d oes not
s im ply m ake tem porary accom m od ation s a nd then resum e its o riginal shape; instead , it
changes its shape. Likew ise, ind ivid uals m ay reconfigure their beliefs and behaviors in a
m a nner tha t allow s them to ad apt . ― Post -traum atic grow th‖ is one exam ple of such a
t ransform ation.
D istinctions betw een various post -event sym ptom tra jectories are im portant t o consider
b ecause it is possible that r esistance, resilience, and recovery h ave d ifferent d eterm inants and
a re best prom oted by d ifferent intervention strategies introd uced at d ifferent tim es (Layne et al.,
2007; N orris et al., 2009). I w ill revisit this point in Section 4. N onetheless, for m ost o f this
r eview , resistan ce and r esilience w ill b e consid ered together as optim al patterns o f change in
t he context of stressful events. 1.2 Psychological M anifestations of Resilience
W hat is t he end result of r esilience ? Ad aptation is the theoretical result, but how is it
m anifest ed ? Th ere is not yet a consensus on this point in behavioral science. In d evelopm en tal
p sychology (Masten & O brad ovic, 2008), resilience encom passe s three d istinct phenom ena:
(1) a chieving better than expected outcom es given risk; (2) sustaining com petence/ funct ioning
u nd er ad verse cond itions; and (3) regaining norm al functioning follow ing a p eriod of exposure
t o traum a or ad versity. In the m ental health arena, r esearchers have v ariously p roposed that the
a ppropriate ind ivid ual outcom e is m inim al im pairm ent of fu nctioning d espite d istress, rapid
r ecovery from d istress, no d istr ess at all, and ad versia l grow th (Bonanno, 2005; Linley & Joseph,
2005; Litz, 2005). In practice, resilience often has been operationalized as the absence of
p sychopathology, w hich is highly inad equate (Layne et al., 2007; Rutter, 1993). A strict focus on
p revalence of d isord ers m ay cause us to erroneously or prem aturely conclud e that a com m unity
h as recovered from an event w hen there is still consid erable d istress, d ysfunction, or
d issatisfa ction present , a nd it dism isses the ad verse consequences of suffering that d oes not
q ualify for s trict d efinitions of pathology (N orris et al., 2008).
N orris et al. (2008) s uggest ed t hat ―w ellness‖ is a viable indicator that adaptation has
o ccurred on the ind ivid ual level. They b ase d t his conclusion largely on the w ritings of Emory
C ow en (1994; 2000), w ho contend ed that psychological w ellness provid es a broad and
integrative frame fo r psychological questions and activities, includ ing those related to prim ary
p revention, em pow erm ent, and resilience. By arguing that ―w ellness m ust be a m atter of prim e
concern at all tim es, not just w hen it fails,‖ Cow en (2000, p. 80) aim ed to m ove psych ology
b eyond a focus on repairing d y sfunction t o a prim e focus on prom oting w ellness. N orris and
colleagues d efine d p sychological w ellness accord ing to four criteria: (1) absence of
p sychopathology; (2) healthy patterns of behavior; (3) ad equate role funct ioning at hom e,
s chool, and/ or w ork; and (4) high quality of life. Quality of life captures how people generally
feel about their lives as a w hole and in d om ains of w ork, family life, health, leisure, and 4 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience n eighborhood (e.g ., Zautra & Bachrach, 2000). This d efinition of w ellness is not id entical to
C ow en’s (2000, p. 83) but consistent w ith it, as w ell as w ith areas of concern for public health
a nd behavioral health policies and program s. A criterion of w ellness serves to rem ind us that
w e m ust attend to d isa ster victim s’ abund ant problem s in living that m ay interfere w ith their
q uality of life (N orris et al., 2002a; 2002b). In m aking their recom m end ations for future research,
Layne a nd colleagues (2007) argued that w e m ust a pproach resilience and other post -event
t rajectories as d om ain -specific phenom ena. As this recom m end ation suggests, one m ight
o bserve resilience in one d om ain o f w ellness (e.g., psychopathology) but not necessarily in
a nother (e.g., q uality of life).
The criterion of w ellness can be extend e d from the ind ivid ual t o the com munity level
w ithout d ifficulty. Although a com m unity is not m erely the sum total (or average) of its
m em bers, com m unity -level ad aptation can b e und erstood as ―population w ellness,‖ a high
p revalence of w ellness in the com m unity, d efined as high and non -d isparate levels of m ental
a nd behavioral health, role functioning, and quality of life in constituent populations. If
e m ergency m anagem ent system s function effectively to protect lives, red uce injuries, m inimize
d am age to p ublic utilities, and connect com m unity m em bers to necessary services, it is
r easonable to expect the population to rem ain w ell. W ellness levels in the com m unity can a nd
s hould b e m onitored in post -d isaster need s assessm ent and surveillance initiatives to g uid e
r esource a llocation . 1.3 Prevalence of Resilience
A s Bonanno (2004, p. 23) noted in the s econd key point o f his article , a va ilable research
s uggests that r esilience is not rare but is actually quite com m on . One of the precursors o f the
r ecent explosion of interest in a d ult resilience w as the finding across several large -sca le
e pid em iologic investigations (Breslau et al., 1998; Kessler et al., 1995; N orris, 1992) that rates of
p osttraum atic stress disord er ( PTSD ) in the general population w ere d ramatically low er than
r ates of exposure t o potentially traum atic events . D epend ing on the range of events assessed ,
t hese and sim ilar studies found that from 50% t o 90% of the U.S. population had e xperienced an
e vent that might be consid ered traum a tic, such as a physical or sexual assault, a n injury -causing
a ccid ent, or t he s ud d en unexpected d eath of a loved one. In the Detroit Area Survey (Breslau et
a l., 1998), one of the strongest stud ies m ethod ologically, the conditional r isk of P TSD (the
p revalence of PTSD given exposure) w as 18% in w om en and 10% in m en w hen estim ated on the
b asis of the respond ent’s w orst event. In other w ord s, m ore than four o ut o f five traum a exposed p eople n ever d evelop PTSD , although they usually d o experience som e sym ptom s o f
d istress. P TSD t ake s a chronic course in about one -third o f t he p ersons w ho d evelop it (Kessler
e t al., 1995).
Longitud inal stud ies exam ining p articular stressors and populations h ave p r ovid ed further
e vid ence that r esilience a nd resistance are com m on rather than rare. For exam ple, O rcutt,
Erikson, and Wolfe (2004) stud ied a pred om inantly m ale sam ple of Gulf War veterans w ithin
5 d ays of their return to the United States and approxim ately 2 a nd 6 y ears later. Tw o d istinct
g row th curv es characterized the d ata. The largest group of veterans show ed low levels of PTSD
s ym ptom s initially and little change. The other group show ed slightly higher sym ptom s
initially, follow ed by significant increases over tim e. These trend s w ould be m ost cons istent
w ith the r esistance and delayed d ysfunction t rajectories d efined by N orris et al. (2009), but the
long intervals m ay have m ad e it d ifficult to capture othe r potential patterns . O ’Donnell, Elliott,
Lau, and Cream er (2007) stud ied a pred om inantly m al e sam ple of injury survivors assessed
p rior to hospital d ischarge and 3 and 12 m onths post -event. Patterns in their d ata also pointed CA RRI R esearch R eport 10 5 Behavioral Science Perspectives on Resilience t o a larger resistant subgroup (these persons w ere low in PTSD sym ptom s at all tim e points)
a nd a sm aller subgroup w ith ch ronic d ysfunction (these persons had higher levels of sym ptom s
initially and grew m ore sym ptom atic over tim e). O tt, Lueger, Kelber, and Prigerson (2007)
cluster analyzed longitud inal d ata (collected on average a t 4, 9 and 18 m onths post -event) from
a p red o m inantly fem ale sam ple of old er bereaved spouses. The largest group of par ticipants
follow ed a r ecovery t rajectory b y being high in sym ptom s at Tim e 1 and then show ing slo w ,
s tead y im provem ents . The second largest group follow ed a resilient trajectory, and t he t hird
a nd sm allest group show ed chronic d ysfunction .
Resistance or resilience w as clearly exhibited by the old er ad ults stud ied by Bonanno,
W ortm an, and N eese (2004). The investigators t racked d epres sive sym ptom s in 185 o ld er ad ults
b eginning on average 3 years prior to d eath of spouse (bereavem ent) a nd again 6 and 18 m onths
a fter t heir s pouse’s d eath. N early half of the sam ple (46%) show ed low levels o f d epression a t
a ll three tim e -points (be fore ber eavem ent and 6 and 18 m onths after) . O nly 11% show ed a
m arked increase in d epress ion 6 m onths a fter the d eath follow ed by a s ubsequent return to
n orm al a t 18 m onths (the expected ―typical‖ grief pattern) , a nd o nly 16% show ed a s ubstantial
increase in d ep re ssion a t 6 m onths t hat rem ained elevated a t 18 m onths (chronic, unrelenting
g rief). The m ajority of stud y participants d id experience som e yearning and em otional pangs ,
b ut t hese feelings w ere transient and interfe re d relatively little w ith d ay -to -d ay functioning.
In interpreting these finding s, it should be noted that Bonanno et al. d efined change in
d epression conservatively as m ore than one standard d eviati on (SD ) o n t he psychological
m easure. In m ost stud ies, one S D w ould b e characterized a s a large change . W ith this point in
m ind , t heir find ings are interest ing t o contrast to th ose of H obfoll et al. (2009). H obfoll and
colleagues e xamined trajector ies of distress d uring ongoing terrorism am ong 709 Jew s and
A rabs in Israel, longitud inally, 1 y ear a part. These investigators d efined ―d istress‖ s om ew hat
liberally as the presence of m ore than one s ym ptom of traum atic stress or d epression a t the tim e
o f assessm ent . Longitud inally, 22% of participants w ere classified as show ing resistance
(d istressed at n either time -point), 14% as show ing resilience (d i stressed at Tim e 1 but not
Tim e 2), 54% as show ing chronic d istress (distressed a t both tim e -points), and 10% as show ing
d elayed d istress (d istressed at Tim e 2 but not Time 1). A re the d ram atic d ifferences b etw een
H obfoll’s and Bonanno’s find ings the result of d ifferences in events , populations, or m ethod ?
These d ifferences cannot be d isentangled w hen results d epend upon investigator -d erived a nd
inconsistent d efinitions. M oreover, the second stud y also raise s the issue of w hether the
r esilience concept even applie s w hen t he threat of exposure to violence persists .
N orris et al.’s (2009) trajectory analysis is especially relevant to this review , and thus w ill be
d escribed in som e detail. G roup -based m od eling proced ures w ere used t o id entify v arious
t rajectories o f d istress in the afterm ath of t w o very d ifferent d isasters : t he 1999 flood a nd
m ud slid es in Villaherm osa and Teziutlán (Mexico) a nd the 2001 terrorist attacks in N ew York
C ity . Th is stud y afford ed several a d vantages relative to prior research. First, each stud y had
four -w ave longitud inal d esigns t hat allow ed the full range of potential trajectories to be
id entified . Second , both the Mexico and N ew York stud ies w ere population b ase d . M ost
p reviou s stud ies on this topic had focused exclusively on selected groups, such as m ale combat
v eterans, hospitalized injury survivors, or old er bere aved spouses. Third , although the stud ies
lack ed pred isaster m easures, the m easures of post -traum at ic stress w ere event s pecific and thus
o vercam e the problem to the extent possible. A lthough p osttraum atic stress (m ore precisely, the
lack thereof) is incom plete as a m anifestation of resilience , it has useful properties for research
b ecause of its hypot hetical zero -point before the event. Finally, r eplication of the m odels across
d isaster types and settings increases confid ence in the generalizability of r esults. 6 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience Even though b oth d isasters w ere quite s erious, from one -third (Mexico) to one -half (N ew
York) of participants exhibited resistance b y never show ing m ore than m ild d istress,
o perationally d efined as ≤ 3 e vent -related s ym ptom s (17 specific sym ptoms w ere m easured in
e ach sam ple at each tim e point). Although t he authors characterized th is trajectory as resistance,
it is sim ilar to the d efinition of resilience provid ed by Bonanno (2004 ). The p revalence of
resilience w as 32% in Mexico and 10% in N ew York . The resilient g roup s n ever show ed severe
d istress but rather began w ith m od erate d ist ress, rapid ly im proved to mild d istress, and
s ubsequently show ed no d istress . The prevalence of recovery in Mexico w as 11%; this group
im proved from severe to m od erate d istress d uring the course of the stud y, w hich conclud ed at
2 y ears post -d isaster; extra polation of the recovery trajectory ind icates that it m ight take another
y ear or tw o (i.e., until 3–4 years post -d isaster) for this group to reach the mild levels of distress
e xperienced by the resistant and resilient groups at 2 years post. The p revalence of recovery in
N Y (9%) w as about the sam e as in Mexico, and most of this group’s im provem ent took place
b etw een 2 and 3.5 years post. Chronic dysfunction t rajectories w ere evid ent in both M exico (22%)
a nd N ew York (13%), but delayed dysfunction w as observed only in N ew York (14%).
If one d oes som e ad d itional ―lum ping‖ to red uce d etail, the find ings from Mexico and N ew
York are im pressively sim ilar: In Mexico and N ew York, respectively, 67% and 64% show ed
r esistance or resilience (the tw o m ost d e sirable patterns), 11% and 9% sh ow ed r ecovery, and
22% and 27% s how ed chronic or d elayed d ysfunction. O n the b asis of these results, it might b e
r easonable to propose a ― 7:1:2‖ expectation for g eneral populations post -disaster : just und er
70% o f ad ults w ill show resistance or resilience ; 10% w ill recover, but m ore slow ly; and 20%
w ill show m ore end uring problem s. 1.4 S ources of Resilience
In his 2004 article, Bonanno’s t hird a nd final k ey point w as that there are m ultiple and
s om etim es unexpected pathw ays t o resilience. This statem ent s uccinctly sum m arized a w ealth
o f research exam ining the influence of r isk and protective factors in the context of stress. A s he
s tated , there is truly n o single w ay o f m aintaining equilibrium .
D evelopm ental psychologists have been particularly influential in calling attention to the
d iverse w ays that child ren m ay overcom e ad versity to d eve lop and function appropriately,
b eginning w ith the sem inal research of W erner and Sm ith (1982). D evelopm enta l researchers
w ho have follow ed t he lives of at -risk child ren have o bserved everything from serious m ental
illness t o rem arkable success in ad ulthood . Recently, Masten and Obrad ovic (2008) review ed
t he r esearch on resilience in hum an developm ent w ith t he a im of inform ing plans for d isaster
r esponse and recovery. They n oted that r esilience arises from interaction s a cross m ultiple
d om ains of life , and o utlined the a d aptive systems t hat are believed to be m ost im portant for
h um an resilience. Their list provid e d an excellent outline , s o I have follow ed it here, ad d ing
a d d itional relevant concepts and com mentary along the w ay. This list of base resources sets the
s tage for the t hem es t hat a re d eveloped m ore fully in Sections 2 a nd 3 o f this review . 1.4.1 A ttachments and Social Relationships
Relationships provid e the found ation for resilience. For young child ren, t he p otential for
r esilience d epend s prim arily on t he q uality of early relationships w ith parent figures. Later,
friend s and rom antic partners provid e this b ase o f security . A ll planning for d isaster m ust
a ccount for the attachm ent system and how such relationships are likely to m otivate behavior
a nd p rovid e a s ense of security (Masten & Obrad ovic, 2008). CA RRI R esearch R eport 10 7 Behavioral Science Perspectives on Resilience A s ubstantial bod y of research has d eveloped regard ing the influence of social support on
ind ivid uals’ ability to cope w ith and recover from significant sources of stress (Brew in ,
A nd rew s, & Valentine, 2000; Taylor & Stanton, 2007) including disaste rs and terrorist attacks
(N orris et al., 2002b ). In ad d ition, it has b een argued that Social C apital, d efined as the
a ggregate of the actual or potential resources linked to the possession of a d urable netw ork of
r elationships (Bourd ieu, 1985), is one of t he four essential ad aptive capacities that create
com m unity resilience (Norris et al., 2008). The m echanism s by w hich social support plays its
p rotective role a re m ore com plex than often recognized because social relationships can
t hem selves be dam aged by d isasters. This w ork is r eview ed in Section 3 (Response a nd
Recovery). 1.4.2 A gen cy
― Agency‖ is an um brella term for the various w ays in w hich people exert personal control
o ver their lives, futures, events, and environm ents. A b ehavioral science per spective on
r esilience m ust begin and end w ith the position that ind ivid uals are active agents in their ow n
lives. They m ake plans and behave intentionally to bring about d esired outcom es. H um an
b eings d o not sim ply respond passively to rew ard s and punishm ents or to larger sociological
forces. H um ans a re m otivated to ad apt, although m otivation can be extinguished by prolonged
e xposure to unresponsive environm ents or uncontrollable events. P eople w ith positive view s o f
t heir ow n efficacy exert m ore effort to succeed and are m ore likely to persist in face of ad versity.
A gency can be achieved either behaviorally or cognitively (Band ura, 1982). W hereas
― b ehavioral control‖ occurs w hen people take actions that forest all o r m od ify aversive events,
― cognitive control‖ occurs w hen people believe they can manage threats, should they arrive.
The fundam ental im portance of cognitive control in contem porary psychology has given rise to
s everal specific theoretical fram ew orks that d iffer in their nuances and sem antics but share
m uch in com m on at their core. For an und erstand ing of resilience, three of the m ost im portant
concepts are self-efficacy , optimism , and hope.
Self-efficacy beliefs are a person’s perceived ability to organi ze and execute courses of action
t o achieve desired outcomes. The psychologist w ho has w ritten most extensively about self efficacy is Albert Bandura. In a particularly influential paper (1982, p. 122), he w rote, ―Efficacy in
d ealing w ith the environment i s not a fixed act or simply a matter of know ing w hat to do. Rather
it involves a generative capability in w hich component cognitive, social, and behavioral skills
m ust be organized into integrated courses of action to serve innumerable purposes.‖ Self-efficacy
h as been shown to influence precautionary behavior, as w ell as how people respond emotionally
a nd behaviorally to stress. These specific functions w ill be revisited in Section s 2 a nd 3.
The related concept of optim ism has been d efined in tw o w ay s. First, it is an a ttributional
s tyle o f m aking external, v ariable, and specific a ttributions for negative outcom es rather than
internal, stable, and global o nes (Seligm an, 1991). In attr ibution theory, m ore broad ly, people
a ssess the causes of events on three d im ensions. The internality dim ension captures w hether
o utcom es are ascribed to internal factors (personal characteristics) or to external factors; the
s tability d im ension captures w hether outcom es a re ascribed to end uring factors (im m utable
t raits/ characteristics) or transient factors; and the generality factor captures w hether the causes
a re believed to operate in m any situations or only a few (Band ura, 1982). Optim ists tend to see
t heir failures as d ue to aspects of the specific situation that are unlikely to be repeated .
Second , optimism refers to g eneralized expectancies for s uccess (Scheier & Carver, 1985).
O ptim ism has been show n to be a key protective factor for disaster recov ery (e.g., Carr et al.,
1997; Dougall et al., 2001). H ow ever, the role of optim ism in influencing risk perception and 8 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience p recaution ad option is more com plex, and it creates challenges for p ractitioners w ho seek to
increase the public’s p repared ness for d isast er. This them e is explored m ore fully in Section 2.
H ope subsum es som e elem ents of both self -efficacy and optim ism . A s a mechanism relevant
t o und erstanding resilience, t he goal-direction m od el o f Charles Snyd er (Feld m an & Snyd er,
2005; Snyd er et al., 2000) is especially relevant. H ope, accord ing to Snyd er, h as tw o elem ents:
(1) pathways – t he p erceived capacity to prod uce routes t o goals a nd (2) agency – t he t houghts
t hat people have regard ing their ability to begin and continue m ovem ent on t he p athw ays to
t heir g oals . Snyd er et al. (2000) argued that p revention is a n act of hope, a ― positive, em pow ered
v iew of one’s ability to act so as to attain better tom orrow s.‖ They noted that interventions need
t o d o m ore than enhance agentic thinking. They n eed to provid e r outes for m oving forw ard
(the p ath w ays). 1.4.3 Intelligence, Problem -S olvi ng, Information -Processing , and Self -Regulatory
M asten and Obrad ovic (2008) observed that b etter cognitive and problem -solving skills
h ave been im plicated in nearly every stud y com paring ad aptive and m alad aptive groups o f at risk youth, includ ing M asten’s (2001) s tud y of a school sam ple follow ed over 20 years.
Intelligent behavior rests on health y b rain d evelopm ent and functioning as w ell as learning
p rocesse s and experiences. O vercom ing ad versity often requires ind ivid uals to regulate their
t houghts, feelings , a nd behaviors in ord er to continue functioning. Fear and anxiety can
interfere w ith higher cognitive abilities.
Inform ation processing m od els have beco me hugely influential in the field of traum a
r esearch and PTSD treatm ent. These m od els m ake clear pred ictions about how the nature of an
ind ivid ual’s initial response t o a crisis can influence long -term outcom es. In contem porary
b ehavioral science, the stu d y of regulatory systems for controlling arousal, affect, and attention
is usually inte rtw ined w ith the stud y of inform ation processing. This theme w ill be explored in
Section 3. 1.4.4 M icrosystems, Communities, and Macrosystems
― Microsystem s‖ refer t o the fam ilies , p eer n etw orks, and settings (classroom s, w ork) in
w hich ind ivid uals are em bed d ed . A s M asten and Obrad ovic (2008, p. 9) sum m arized , ― H um ans
a re social, and their ad aptive functioning is em bed d ed in a com plex array of interd epend ent
r elation ships and social system s that also serve many regula tory and protective roles.‖ Fam ily
r outines, rituals, beliefs, narratives, a nd v alues w ork to re gulate and protect individ uals. The
r esilience of ind ivid uals and fam ilies is also linked t o t he r esilience o f larger sy stem s . This
p erspective und erlies the grow ing interest of behavioral scientists in the concept of com m unity
r esilience (e.g., N orris et al., 2008).
M asten a nd Obrad ovic (2008) n oted that ―macro system s ‖ (e.g., culture, med ia) have r arely
b een incorp orated into behavioral stud ies, but can be assum ed to influence t he r esilience of
ind ivid uals . H ow ever, the influence of med ia has becom e a topic of some controversy w ith
r egard s to und erstanding the im pact of terrorism a nd d isasters o n persons w ho a re not directly
h arm ed . There is am ple evid ence (e.g., Ahern et al., 2002; Bernstein et al., 2007; Lau, Lau, Kim ,
& Tsui, 2006; Silver et al., 2002) t hat post -d isaster s ym ptom s are positively correlated w ith
e xposure to extrem e m ed ia im ages (e.g., people jum ping from the tow ers on 9/ 11/ 2001; d ead
b od ies in the afterm ath of the 2004 tsunam i; the iconic d ead child in footage of the Oklahom a
C ity bom bing ). W hat these relations m ean is not yet clear, and investigators generally caution CA RRI R esearch R eport 10 9 Behavioral Science Perspectives on Resilience a gainst attributing caus ality. Marshall and colleagues (2007) proposed that m ed ia effects are
e xplained by ―relative risk appraisals‖ that m ed iate the relation betw een an event and its
m eaning . This point w ill be revisited in Section 2 o f this review . 1.5 Conservation of Re sources for Resilience
M asten’s list could be characterized as a m enu of r esources t hat combine a nd interact t o
create m ultiple pathw ays to r esilience. Borrow ing H obfoll’s (1989) d efinition o f resources ,
N orris et al. (2008, p. 134) likew ise conclud ed t hat r esilience ― depend s on a host of objects,
cond itions, characteristics, and energies that people value – t hat is, r esources .‖ They also
a rgued , how ever, that pre -event resources m ust possess one or m ore d ynam ic attributes in
o rd er to engend er resilience. D raw ing upon the earlier w ork of Bruneau et al. (2003), they
characterized these d ynam ic attributes a s robustness, the ability of the resource to w ithstand
s tress w ithout suffering d egrad ation; redundancy , the extent to w hich elements are substitutable
in t he event of d isruption or d egrad ation; and rapidity , the speed w ith w hich the resource can be
a ccessed and used . By high lighting the necessity of these a ttributes , N orris and colleagues
a im ed to integrate resilience perspectives w ith evid ence show ing tha t resources are not static –
t hey evolve, strengthen, w eaken, and rebound – a nd these trajectories are of interest i n their
o w n right (H obfoll, 1989).
This is im portant because r esilience can fail w hen the necessary r esources are them selves
d am aged or d is rupted by the stressor. The concept of resource loss has becom e central in stress
t heory, prim arily because of the influenc e of H obfoll’s (1989, 1998) theory of ―conservation of
r esources‖ (COR). The bas ic tenet of COR theory is that ind ivid uals strive to o btain, retain,
p rotect, and foste r those things that they value, w hich are term ed r esources . In H obfoll’s theory,
s tress occurs w hen resources are threatened , w hen resources are lost, or w hen individ uals fail to
g ain resources follow ing a significant inv estm ent of other resources. COR theory has becom e
h ighly influential in disaster research because d isasters and terrorism threaten a host of object
r esources (hou sing), personal resources (optim ism , safety), social resources (com panionship),
a nd energies (m oney, free tim e). This phenom enon w herein resources are them selves harm ed
b y the stressors they are presum ed to buffer severely limits the protection resources can afford .
In recent w ork, H obfoll and colleagues (i n press) have aim ed to m ore explicitly a p ply COR
t heory to explain h um an r esilience . In particular, they highlighted the principle of resource
investment – p eople m ust invest resources in ord er to protect against resource loss, recover from
lo sses, and gain resources. P ersons w ith greater resources a re therefore less vulnerable to
r esource loss and m ore capable of generating gain. They characterize d r esilience as a n activ e
p rocess t hat dem ands t he investm ent of p ersonal and social resources . H obfoll and colleagues
a lso posited that, a lthough resource loss is m ore potent than resource gain, t he s alience of gain
increases und er cond itions of resource loss. Und er cond itions of high loss (such as in the
a fterm ath of disasters), efforts that result in e ven s m all gains can e licit hope and le ad to further
g oal-d irected efforts. A dd itionally, H obfoll and colleagues d escribed the potential for r esource
loss and gain spirals: Initial loss begets future loss. A s they noted , s econd ary stressors after
t raum atic events continue to tax personal resour ces , a nd thus t he d em and on resources
continues until resources can be gained and loss halted . Efforts to enhance r es ource gain cycles
a nd to m inimize resource loss cycles should facilitate hum an resilience. Resource loss / gain is a
crucial them e for u nd ers tand ing stress r ecovery, and o ne m ore sp ecific application (the
m obilization and d eterioration of social support) w ill be review ed in Section 3. 10 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience 2. A N TICIPATION AN D RED UCTION OF VULN ERABILITIES:
H OW IN D IVID UALS PERCEIVE A N D RED UCE RISK
A ccord ing t o CARRI, the first characteristic o f a resilient com m unity is that it anti cipates
p roblem s and opportunities , and the second characteristic is that it takes action to red uce
v ulnerabilities . In Section 2 o f this review , I s umm arize research and theor y from behavioral
s cience that is relevant to und erstand ing how ind ivid uals anticipate or perceive r isk and engage
in b ehaviors d esign to red uce perceived vulnerabilities . 2.1 Roots of Anticipation: Cognitive Schemas, A ssumptions , and Illusions
A s it is for com m unity resilience, t he a bility to anticipate risk is central to ind ivid ual
r esilience , although the process is com pli cated by a host of b iases that regulate hum an thought
a nd em otion. Research on cognitive ―schem as‖ has been am ong the richest areas o f p rogress in
t he past few t hree d ecades in psychology. By necessity, the w orks cited in this section represent
o nly a small fraction of the total b od y of w ork on hum an thought .
A s chem a refers to an a bstract know led ge structure that is stored in m em ory; it is the rich
n etw ork of inform ation a bout a given stim ulus d om ain that provid es the basis for anticipating
t he future (Janoff-Bulm an, 1989; Taylor & Brow n, 1988). It guid es w hat w e notice and rem em ber
a nd how w e interpret new inform ation. Schem as are str ongly held expectations , and b iases
o perate to m aintain stability. Becaus e w e a re conserv ative about changing schem as, w e
g enerally prefer to assimilate rather than accom mod ate d iscrepant inform ation. Inform ation
p roces sing is full of shortcuts, e rrors , and s elf-serving interpretations o f d ata . Largely, these
p rocesses a re ad aptive, but not entirely.
Janoff-Bulm an a nd Frieze (1983) ad vanced schema theory in stress research by arguing that
v ictim s’ d istress w as d ue to t he shattering of t he b asic assum pt ions t hey held about them sel ves
a nd their w orld . There are t hree m ajor s ets of assum ptions (or schem a ) t hat m ost of us hold . The
first is the assum ption of invulnerability, the belief that it ―can’t happen to m e.‖ As w ill be
d iscussed in m ore d etail shortl y , p eople und erestim ate t he likelihood of personally
e xperiencing negative events . The second assum ption is of the w orld as m eaningful – e vents are
com p rehensible a nd controllable. This assum ption builds on the ―just w orld h ypothesis,‖ the
n otion that ―peo ple get w hat they d eserve ‖ (Lerner, 1980). The third assum ption is of the self as
p ositive – w e tend to v iew o urselves as w orthy, decent people, a nd have h igh s elf-esteem .
C oping w ith traum a is in a large part a proc ess of rebuild ing the assum ptive w orld .
In som e w ays, Janoff-Bulm an a nd Frieze ’s (1983) contribution set the stage for Taylor and
Brow n ’s (1988) a rticle t hat had w id espread influence on the w ay m ental health w as u nd erstoo d
in the behavioral sciences . Taylor and Brow n s umm arized a w ealth of experim ental evid ence to
s upport their conclusion that m ental health m ay rest m ore on illu sion than reality . They argued
t hat o verly positive self-evaluations, exaggerated p erceptions of control , a nd u nrealistic
o ptim ism a re charact e ristic and ad aptive properties of norm al hum an thought. These themes
correspond ed r oughly to the three assum ptions (self a s positive, w orld as m eaningful,
invulnerability) outlined b y Janoff-Bulm an a nd Frieze , but Taylor and Brow n em phasized that
t hese r osy assum ptions w ere not necessarily rooted in objective fact. For the topic of
― Anticipation,‖ an illusion of particular relevance is unrealistic optim ism . P eople typically
b elieve that the future w ill be better than the present. This perspective is good for m ental health,
b ut unrealistic optim ism m ay lead people to ignore legitimate risks in their environm ent. JanoffBulm an a nd Frieze (1983) likew ise noted t hat the assum ption of invulnerability can be
m alad aptive if it keeps people from e ngaging in precau tionary b ehaviors. CA RRI R esearch R eport 10 11 Behavioral Science Perspectives on Resilience 2.2 O ptimistic Bias
The concept of unrealistic optim is m can be m ore specifically und erstood and m easured as
― o ptim istic bias.‖ Optimistic bias (also know n as com parative optimism ) is the tend ency to see
o thers as m ore vulnerable to risk t han oneself. O ptim istic bias h as been a rich area for research
s ince Weinstein (1980) first introd uced the concept. A recent paper by H arris, G riffin , and
M urray (2008) provid ed an excellent re view of this research, and accord ingly t he rest o f this
s ection relies h eavily on their e xceptionally clear t ests o f the original tenets of the theory. They
cond ucted their research u sing s ophisticated m ulti-level m od eling over seven s am ples .
First, they tested the fund am ental question of w hether people d o, in fact, believe t hat
n egative events are less likely to happen to them than to other people. This hypothesis w as
s upported . A veraged across all sam ples, for an event that is average in frequency, average in
controllability, a nd a verage in severity, t he expected self-risk rating w as 4.3 o n a 10 point scale,
t he exp ected other -risk rating w as 5.2, and the d ifference w as 0.9 – a s ubstantial and significant
o verall op timistic bias .
Second , they tested w hether optim istic bias (m easured by the d ifference betw een self -risk
r atings and other -risk ratings) w as confined to only particular types of events. While bias w as
less pronounced w ith higher frequency events tha n w ith low er frequency events, it w as
e lim inated only for extrem el y com m on event s . It w as influenced little by the relative
u nd esirab ility of events; t hat is, the bias held across m ost events.
Third , they tested w hether optim istic bias w as confined to o nly particular types of people.
W einstein had proposed that ind ivid ua ls w ith h igh self-esteem w ould s how lower s elf-risk
r atings and t hus m ore o ptim istic bias . This w as s upported ; h ow ever, the bias w as not unique to
h igh -esteem ind ivid uals. C onversely, individ uals w ith high t rait a nxiety w ere hypothesized to
s how higher s elf-risk ratings and thus less o ptimistic bias. In accord , anxiety w as positively and
s ignificant ly r elated to self-risk, but t he findings also suggested that o nly very extrem e ly
a nxious ind ivid uals w ould s how a n absence of com parative optimism .
Fourth, Weinstein p roposed that the p erceived controllability o f events w ould influence the
e xtent of o ptim istic b ias. This w as a lso s trongly supported by the find ings: the greater the
p erceived controllability of a negative event, the greater the tend ency for people t o believe that
t heir ow n chances of experiencing the event w ere less t han t hose o f the a verage p erson .
In sum m ary, H arris et al. (2008, p. 1235) found evid ence for a ― robust optim istic bias.‖ The
b ias w as s t rongest for controllable events and r are events, a nd for ind ividuals w ith particularly
h igh self-esteem or low trait anxiety, but it w as lim ited neither to particular types of events nor
t o particular types of persons . 2.3 Risk Perception
O ptim istic bias is one face t of the larger construct of risk perception. Summ arizing m any
y ears of t hought and research on r isk perception, P aul Slovic and colleagues (2004) d iscussed
t w o fund am ental w ays in w hich hum an beings com prehend risk: t he ―analytic system ‖ is r ule
b ased , s low , a nd e ffor tful; t he ― experient ial system ‖ is intuitive, fast, r elatively a utom a tic, and
t he m ost natural and com m on w ay to respond to risk. The latter r elies on im ages and
a ssociations, lin ked by experience to positive or negative e ffect. Slovic et al. strongly d ispute d
t he notion t hat affective responses are irrational . Rather, they argue d , analytic a nd experiential
s ystem s w o rk in parallel and d epend on each o ther for guid ance.
Feelings (em otions, affect) t hat beco me salient in the process of jud ging risk d epend on b oth
t he ind ivid ual and situation . Affective im pression s o f the context can be easier and m ore 12 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience e fficient than w eighing a ll of the pros and cons and r etrieving relevant exam ples from m em ory.
A ffect, in a sense, becomes a m en tal shortcut, som etim es referred t o as a ―heuristic‖ in this field .
Support for a n ― affect heuristic‖ com es from a d iverse set of stud ies, a few of w hich w ere
r ev iew ed by Slovic et al. (2004). In particular, feelings of d read o r outrage, w hich are a ssociated
w ith t he v oluntariness, lethali ty, controllability, a nd fairness o f events and threats, strongly
influence p ublic perception and acceptance of risk for a w id e range of hazard s. The risk
a ssociated w ith d read ed e vents (those a ssociated w ith strong feelings , s uch as nuclear
a ccid ents) is u nacceptable even if t he likelihood o f their occurrence is r emote.
M arshall and colleagues (2007) speculated on the role of risk perceptions a nd affect
h euristics in explaining the far -reaching consequences of terrorist attacks, such as the
Septem ber 11, 2001, a ttacks on the World Trad e Center. Their goal w as to explain w hy P TSD
s ym ptom s related to 9-11 occu rred in large numbers of ind ivid uals w ho d id not fit trad itional
d efinition s o f exposure, and they argued that r ela tive risk appraisals provid ed t he m issing
p sychological link . Basing their thinking in part on t he w ork of Slovic (e.g., Slovic et al., 2004),
M arshall and colleagues coined t he term ― relative risk appraisal ‖ b ecause it em phasizes the
com parative process through w hich an environm ental event is appraised in relation to prior
e xpe riences and risk expectations, includ ing s tand ard s for acceptable risk (th us the qualifying
t erm , relative). M oreover, t errorist attacks w ould activate a ffect h euristics associated w ith
d read ed events b ecause t hey are u ncontrollable and e ngend er risks that are not fully visible or
k now n. The authors a rgued that 9-11 carried high ― signal potential,‖ w hich is a w arning that a
n ew ongoing threat has entered t he e nvironment. H igh im pact e vents w ith high negative signal
p ot ential m ay be the exceptional events that can prod uce PTSD a nd other a d verse m ental health
consequences e ven in persons w ho w ere not directly exposed . In a sense, returning to the
o riginal them e of this section, these events are e motionally p ow erful enough t o shatter the
p erceptions of invulnerab ility and unrealistic optim ism t hat o therw ise p ervad e hum an thought
a bout risk . 2.4 From Risk Perception to Action: the Precaution Adoption Process
O f course, an im portant question – a nd one that has been of interest in both psychology and
p ublic health – is the extent to w hich risk perceptions influence p reventive behavior. A good
e xam ple of s urvey r esearch on this topic w as Brew er et al.’s (2004) t ests of t he relations betw een
p erceived risk and behavior ov e r tim e in the context of Lym e d isease. Because an
u nd erstanding of t he attitud e –behavior relation is central to CARRI’s themes of anticipation
a nd red uct ion of vulnerabilities, I h ave r eprod uced the figure u sed by the a uthors to outline t he
h ypothesized effects over tim e (Figure 1).
Brew er et al. argued that it w as im portant to clarify the d istinctions betw een three ―easily
confused hypotheses.‖ The behavior motivation hypothesis is that p erceptions of risk cause people
t o take protective action ; t he risk reappraisal hypothesis s tates that w hen people take actions
t hought to be effective, the y low er their risk perceptions; and the accuracy hypothesis is that r isk
p erceptions accurately reflect risk behavior. These relationships cannot be d isentangled in cross sectional d ata and m ust be tested longitudinally. All thr ee hypotheses w ere supported :
P articipants w ith higher initial risk perceptions w ere m uch m ore likely than others to get
v accinated a gainst Lym e d isease o ver the next few m onths ; b eing vaccinated led to r ed uction in
r isk perception; and p eople w ho got v accinated correctly believed that their risk of future
infection w as low er than that of people not vaccinated . CA RRI R esearch R eport 10 13 Behavioral Science Perspectives on Resilience + Perceived risk,
Time 1 Behavior
motivation Perceived risk,
Time 2 Risk
reappraisal + - - Accuracy Preventive
Time 2 Figure 1. Relationship betw een perceived risk and behavior over time . Variables are in
b old font ; p rocesses a re in italics; the sign (+/ -) is the d irection (positive, negative) of the
o bserved correlation. W ith kind perm ission from Springer Science+Business Med ia: A nnals of
Behavioral M edicine, ― Risk perceptions and their relation to risk behavior,‖ vol. 27, 2004, p . 126,
N . Brew er, N . Weinstein, C. Cuite, and J. H errington. 2.5 O ther Psychological Influences on Precautionary Behavior
The r elatively recent exam ple o f Brew er et al.’s research illustrates themes that have been
p rom inent in the prevention literature for som e tim e, but there are num erous other factors that
h ave been propose d or found to be related to precaution ad option. That is, perceived risk is one
factor, but only one factor that influences behavior (perhaps it is necessary, but not sufficient).
O ne of the earliest and best know n researchers in t h is area w as Ronald Roger s (e.g.,
M ad d ux & Rogers, 1983), a s ocial psychologist w hose w ork focused on fear appeals, persuasion,
a nd a ttitud e change. Fear appeals w ork to increase threat , thereby increasing protection
m otivation. The 1983 article by Mad d ux and Rogers, a classic w ork in this field , revised Rogers’
o riginal theory of protection m otivation by incorporating self -efficacy into t he p red ictive
m od els. Influenced by Band ura’s v iew s that w ere introd uced in Section 1 o f t his report,
M ad d ux and Rogers argued that a ll processes of psychological change s hould o perate through
a lteration of a n ind ivid ual’s expectancies of personal m astery or efficacy. U sing a factorial
laboratory experim ent, Mad d ux and Rogers investigated t he effects of t hreat (a m anipulation o f
t he probab ilit y of the threat’s occurrence) together w ith the effects of coping response efficacy (a
m anipulation of the effectiveness of the recom mend ed coping response) , and self-efficacy (a
m anipulation regarding th e relative ease or d ifficulty involved in the coping response) on
intentions to ad opt a recom m end ed preventive b ehavior. O ne exam ple m ight help to clarify the
cond itions : cigarette s m oking is likely or unlikely to lead to hea rt d isease (threat); cessation o f
s m oking now is likely or unlikely to un -d o the d am age and red uce risk (coping response
e fficacy); and quitting sm oking is easy or d ifficult to d o (self-efficacy). 14 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience The find ings w ere instructive . First, the p robability of t he t hreat’s occurrence and the
e ffectiveness of a coping response both had m ain effects on intentions to ad opt a recom m end ed
p reventive behavior. Se lf-efficacy h ad b oth m ain a nd interactive effect s on intentions . In fact,
s elf-efficacy w as the s ingle m ost pow erful pred ictor of behavioral intentions. The im portant
p oint here is that messages d esigned to encourage precautions cannot focus solely on threats.
Fear a lone may have m inim al effect on behavior unless individ uals are also inform ed about
s pecific a ctions that can be taken to red uce the threat and a re given reason t o believe they h ave
t he ability t o perform the action. This fund am ental principle from behavioral science is violated
w hen w arning s h ave no counterpart in how ind ivid uals are expected to b ehave.
Roger’s prote ction m otivation theory is only o ne of several prom inent theories of p recaution
a d option. N eil Weinstein (1993) com pared and contrasted four com peting theori es : the health
b elief m od el, the theory of reasoned action ( revised as the t heory of planned behavior),
p rotection m otivation theory, and subjective expected utility the ory. A lthough the focus w as on
h ealth behavior, conceptually there is m uch overlap betw een health behavior, hazard
p repared ness, and other self -protective behavior s (N orris, 1997). The nuances that d ifferentiate
t he theories, for exam ple , their com binatorial rules, are im portant for investigators und ertaking
r esearch on precautions, but for the purp oses of this report, a sim pler outline of pred ictors
s hould suffice . A s d iscussed b y Weinstein (1993, p p . 325-6) in the first section of his p aper, the
p rim ary v ariables, found in m ost theories, are t hese : Perceived severity – t he ind ivid ual’s evaluation of the und esirability of a particular
o utcom e assum ing no change in behavior; in d ifferent theories, this concept has been
labeled perceived severity, negative utility, and negative evaluation, but the und erlying
m eanings o f these term s a re the sam e.
Perceived probability – t he perceived likelihood that a particular ne gative outcom e w ill
o ccur ; this factor is variously referred to as perceived vulnerability, perceived
s usceptibility, subjective probability, and expectancy in d ifferent theories.
Perceived effect – t he perceived effecti veness o f the precaution; e ssentially , t he expected
b enefit is the d ifference betw een (a) beliefs about the severity and probability of an
o utcom e assum ing no change in behavior and (b) beliefs about the severity and
p robability of an outcome assum ing the ad option of a protective m easure.
Expected cost – t he expecte d costs of ad opting the precaution, includ ing tim e, effort,
m oney, inconvenience, and loss of satisfactions obtained from the current behavior.
S elf-efficacy – t he person’s perceived ability to perform the behavior, t hat is, the ease or
d ifficulty of the action; this factor is em phasized m ost strongly in protection m otivatio n
t heory , as w as d iscussed above, but t he concept is im plicit in m ost theories w ithin the
category of costs/ barriers to behavioral change.
S ocial influence – t he extent to w hich other people in the ind ivid ual’s social circle w ant
h im or her to perform a given behavior (norm ative beliefs) in com binat ion w ith the
ind ivid ual’s m otivation t o com ply w ith each of their preferences; this factor is unique to
t he theory of reasoned acti on, w hich m ay be one of the s trengths of that theory. N one of the m od els pred icts am ount of behavior b ut rather its likelihood . A lso, in research,
t he focus is often on pred icting behavioral intentions r ather than behavior per s e. CA RRI R esearch R eport 10 15 Behavioral Science Perspectives on Resilience 2.6 Reducing Vulnerability through Collective Action
C om m unity resilience d epend s not only on individ ual s’ m otivations t o protect them selves
b ut a lso on their m otivations to protect their comm unities. Thus, I searched for behavioral
s cience research t hat might elucid ate the factors that le a d ind ivid uals to participate in collective
a ction t o red uce vulnerabilities or d isadvantages . I w as fortunate to find a recent m eta -analysis
p ublished in Psychological Bulletin (the highest ranking journal for such a nalyses) b y v an
Z om eren, Postm es, and Spears (2008). This w as an am azingly rich arti cle for any read er w ith
interes t in this topic; all of the research sum m arized in this section o n collective action w as
d raw n from this paper, to w hich the read er is refer red for original citations and a
com prehensive reference list .
Van Zom eren a nd colleagues d escribed the k ey challenge for their m eta -analysis a s one of
b rid ging s ubjective (psychological) and social (structural) perspectives on w hen, w hy, and how
p eople en gage in social protest. They aim ed t o integrate three perspectives that focus on
subjective injustice, efficacy , and identity a s t he prim ary p r ed ictors of collective action.
P erceived injustice r efers to ind ivid uals’ s ubjective experience of unjus t d isad v antage.
A ccord ing to relative d eprivation theory , it is only w hen social com parisons result in a subjective
s ense of injustice that collective action to red ress inju stice is likely to occur. G roup -based
e m otions , s uch a s anger, often intervene. The m eta -an alysis exam ined (a) w hether people’s
s ubjective experience of injustice in term s of group -based inequality or d eprivation pred icts
collective action and (b) w hether the affective experience of injustice produces stronger effect s
t han non -affective p erception of injustice. Both pred ictions held a cross stud ies in the m eta analysis .
Efficacy r eflects t he id ea being that people engage in collective action if they believe this w ill
m ake it m ore likely that rel evant goals are achieved . Efficacy em erged a s a significant pred ictor
a cross stud ies in the m eta -analysis, but it w as m ore im portant for pred icting behavior related to
― incid ental d isad vantages‖ (e.g., protesting the build ing of a plant in one’s neighborhood) than
for ―structural d isadvantages‖ (e. g., d iscrimination). Incid ental d isad vantages revolve around
issue -based or situation -based d isadvantages, and m ay be the m ost appropriate research to
consid er for und erstand ing h ow ind ivid uals m ay becom e involved in actions to boost their
com m unity’s resilience to hazard s.
The third factor explored by the authors w as identity . A ccord ing to s ocial id entity theory,
p eople s trive for and benefit from positive social id entities ass ociated w ith group m em berships,
a nd s ocial identity serves to m obilize people fo r social change. W hen m embers of a low er s tatus
g roup perceive their s tatus d ifferential to b e illegitim ate and unstable, t hey are m ore likely to
id entify w ith the group and t o engage in collective action to change their intergroup status
d ifferential. In t he m eta -analysis perform ed by the authors, politicized id entities , such as those
a ssociated w ith social movem ent organizations, w ere m ore strongly link ed to collective action
t han w ere n on -politicized s ocial id entities . N evertheless, non -politicized id en tities still
p red icted s ubstantial v ariance in outcom e m easures , s uggesting that social id entity in general
p red icts collective action .
Van Zom eren et al. (2008) proposed a new m od el, the ―integrative social identity m od el of
collective action‖ (SIMCA), t hat accounts for relations betw een the three pred ictors as w ell as
for their effects on collective action. In SIMCA (p. 511), social id entity is t he m ost fund amental
p recursor o f collective a ction because it has b oth d irect and ind irect effects o n action . This is
illustrated in Figure 2, reprod uced from their paper (Figure 2, p. 521) b ut m od ified to em phasize
t he d istinction betw een d irect and ind irect effects . Id entity influences perceived injustice
b ecause it p rovid es t he b as is for a g roup -based exp erience of injustice. It influences e fficacy 16 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience b ecause a s tronger sense o f id entity em pow ers ind ivid uals to act. In the subset of studies that
h ad all relevant m easures, the SIMCA m od el provid ed t he best fit to the data . Van Zom eren a nd
colleagues conclud ed t hat social id entity is ―at the very heart of explanations of collective
a ction.‖ d Perceived
Injustice b Social Identity e a Perceived
Efficacy Collective Action c Figure 2. The direct effects of injustice, identity, and efficacy (paths a , b , and c) on
collective action ; i dentity also has indirect effects through injustice (d*a ) and efficacy (e*c).
C op yright 2008 by the Am erican Psychological Association. Reprod uced w ith perm ission.
M . Van Zom eren , T. Postm es, and R. Spears (2008, p. 521). ― Tow ard an integrative social
id entity m od el of collective action: A quantitative research synthesis of three socio psychological perspectives,‖ Psychological Bulletin , vol. 134, July 2008. DOI:10.1037/ 00332909.134.4.504. The use of APA inform ation d oes n ot im ply end orsem ent by APA.
The centrality of identity in SIMCA s uggests that sense of community a nd place attachment ,
t w o prim ary foci of research in com m unity and environm ental psychology, are potentially
r elevant for und erstand ing collective action t o red uce vulnerabil ities. N orris et al. (2008)
d iscussed these tw o concepts a s central elem ents of Social Capital , o ne o f the four ad aptive
capacities r equired for com m unity resilience . Sense of com m unity is an attitud e of bonding
(trust and belonging) w i th other m em bers of one’s group or locale (Perkins et al., 2002, p. 37).
Ind ivid uals w ho possess a s trong s ense of com m unity tend to show concern for com m unity
issues, respect for and service to others, a nd a d eep s ense of connection. P lace attachm ent is
closely related to one’s sense of com m unity . It implies an em otional connection to one’s
n eighborhood or city, som ew hat apart from connections to the s pecific people w ho live there .
Brow n and Perkins (1992) argued that place attachm ents are integral to self -d efinitions; these
a ttachm ents are holistic and m ulti -faceted and provid e stability. P lace attachm ent often
u nd erlies citizens’ efforts to revitalize a com m unity (Perkins et al., 2002) a nd thus m ay be
e specially relevant w hen com m unity m em bers perceive in cide ntal d isadvantages that
r equire a ction. 2.7 Closing N ote for S ection 2
In closing Section 2 o f this behavioral science review w ith a m od el of collective b ehavior, I
s hould be clear that colle ctive action ultim ately rests on t he p erceptions, m otivation , a nd
b ehavior of ind ivid uals w ho see them selves as agents for change in their com m unity. This is
o ne concept that illustrates the transactional nature of ind ivid ual and comm unity resilience. In
t he N orris et al. (2008) mod el of com m unity r esilience (p. 136), collective action w as a key
e lem ent o f Com m unity Com petence, one of the four ad aptive capacities required for
com m unity resilience. W e now see t hat collective action r ests heavily on ind ivid ual p erceptions CA RRI R esearch R eport 10 17 Behavioral Science Perspectives on Resilience o f injustice, efficacy, and i d entit y . A s N orris et al. n oted , r esid ents of end angered com m unities
m ust be able to learn about their risks and options and w ork together flexibly and creatively to
s olve problem s. The inclusion of C om m un ity Com petence in the com m unity resilience m od el
r e flects b ehavioral science’s fund am ental concern w ith a gency, m astery, and control (see also
Brow n & Kulig, 1996). A s evid enced in concepts such as self-efficacy and o ptim istic bias , h um an
a gency h as p ervad e d b ehavioral science r esearch and theory related t o CARRI’s them es of
a nticipatio n and red uction of vulnerabilities . If I w ere to s um m arize Section 2 in a single
s entence, it w ould be that a n ind ivid ual’s capacity for m eanin gful, intentional action is w hat
u nd erlies his or her p otential to a nticipate p roblem s and opportunities and to transform his or
h er environm ent to red uce vulnerabilities . 3. RESPON SE AN D RECOVERY: H OW IN D IVID UALS REACT TO, COPE WITH,
A N D RECOVER FROM STRESSFUL EVEN TS
In Section 2, I review ed research and theory from be havioral science relevant to
u nd erstanding CARRI’s first tw o characteristics of a resil ient com m unity: (1) It a nticipates
p roblem s and o pportunities a nd (2) takes action to red uce vulnerabilities . In this section, I
r eview research and theory from behavior al science relevant to understand ing the rem aining
t w o characteristics: (3) It r espond s e ffectively and (4) r ecovers r apid ly . Stress response and
r ecovery has been a huge topic of research in the behavioral sciences for about 40 years, a nd
m uch has been learned about how individ uals respond t o a nd recover from s tressors of various
k ind s, inclu d ing d isasters and terrorist attacks . 3.1 O verview of the Stress Process
M od els of the stress process typically consid er (a) chara cteristics of th e stressor,
(b) a ppraisals of the stressor, (c) the response to or effects of the stressor, and (d ) various
cond itions that influence the relations betw een the stressor, stress appraisal, and stress
r esponse. In stress -d iathesis theory, level of exposure i s proposed to interact w ith pre existing
v ulnerabilities to influence the stress response. The influence of this theory is evid ent in som e of
t he earliest published fram ew orks for d isaster research (see Benight, McFarlane, & N orris, 2006,
for further d iscus sion of these points). A s an exam ple, N orris et al.’s (2008) m od el of stress
r esistance and resilience is p rovid ed in Figure 3. This m od el o w es it s fund am ental structure t o
Barbara Dohrenw end ’s (1978) e arly m od el o f psychosocial stress . I w ill revisit her m od el in
Section 4 b ecause it w as a forw ard -thinking attem pt to relate the stress process to d ifferent
p oints an d types of intervention . Like m ost m od els o f the stress process, Norris et al.
e m phasizes interactions b etw een the str essor and r esources . Resistance occurs w hen resources
a re sufficiently s trong (robust, red und ant, or rapid ) t o buffer or counteract the im med iate
e ffects of the stressor such that no d ysfunction occurs. Resilience occurs w hen resources are
s ufficie ntly str ong t o buffer or counteract the effects of the stressor such that a r eturn to
functioning o ccurs. Vulnerability occurs w hen resources a re n ot sufficie ntly strong t o create
r esistance or resilience, resulting in persistent d ysfunction. The m ore severe, e nd uring, and
s urprising the stressor, the stronger the resources m ust be to create resistance or resilience. 18 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience Pre-Event
Environment crisis? resistance Post-Event
deterioration vulnerability Persistent
Rapidity Figure 3. M odel of stress resistance and resilience . S ource: N orris et al. (2008, p. 130).
The process b egins w hen a stressor, such as a d isaster, occurs. Stressors are aversive
circum stances that threaten w ell-being or functioni ng . Stressors vary on a num ber o f
d im ensions, includ ing s everity, d uration, and surprise. Severity of exposure is a consistent risk
factor for adverse psychosocial consequences in t he afterm ath of d isasters (N orris et al., 2002 a;
2002b ). Specific stressors that have been found to a ffect post -disaster w ell-bein g includ e
b ereavem ent, injury to self or family m ember, life threat, property d am age, financial loss,
com m unity d estruction, and d isplacem ent. O bjectively d efined stressors are m od ified by stress
a ppraisals and related cognitions, coping responses, and so cial support. It is som ew hat artificial
t o separate these topics into ―response‖ and ―recovery‖ because appraisals, coping, and social
s upport all occur continuously, over tim e, from the first m om ents of an event until its effects
h ave d issipated . 3.2 Cog nitive Appraisals of Stress
C ontem porary und erstand ings of stress ow e m uch to the groundbreaking w ork of Richard
Lazarus and Susan Folkm an. Folkman (1984) provid ed an excellent overview and sum m ary of
t he key features of Lazarus’ cognitive theory of stress (Lazarus & Folkm an, 1984). This theory
p osits tw o central processes , cognitive appraisal and coping . In this theory, stress w as defined
a s a relationship b etw een person and environm ent that is appraised b y the person as taxing or
e xceed ing his or her resources and as end angering his or her w ell -being. N ote that stress is
n ot a p roperty of the person or environm ent, not a stim ulus or response, but a relationship
b etw een person and environm ent. The m eaning o f an event is d etermined by an individ ual’s
There are tw o major form s of appraisals (Folkm an, 1984). In a primary appraisal, the p erson
e valuates the significance of t he t ransaction w ith respect to h is or her w ell-being: Is it irrelevant,
b enign –positive, or stressful? Stressful appraisals includ e (a) h arm / loss – t he injury or d am age
is a lread y d one ; (b) t hreat – t he p otential for harm/ loss e xists; and (c) challenge – t here is an CA RRI R esearch R eport 10 19 Behavioral Science Perspectives on Resilience o pportunity for grow th, m astery, or gain . Which appr aisal occurs is d eterm ined by an a rray of
p erson a nd situation factors, includ ing b eliefs (pre existing notions about reality that serve as t he
p erceptual lens), g eneralized control e xpectancies, and com mitm ents (w hat is im portant, w hat
h as m eaning). In a secondary appraisal, the p erson evaluates h is or her coping resources and
o ptions: W hat can I d o? This appraisal is i nfluenced by situational jud gm ents about t he
p ossibility of control. (I w ill return to the topic of ―coping‖ in a subsequent section of Section 3.)
P rim ary a ppraisals of an event’s m eaning are especially relevant for und erstand ing
r esponses to d ifferent types of d isasters. Som e behavioral researchers have argued that in the
case of terrorism , the objective, observable experience of an e vent should be d e -em phasized ,
r elative to the subjective, psychological experience of w hat that event im plies for the future, as
e vid enced in perceptions of threat (Marshall et al., 2007). From this perspective, w hat the event
m eans to the individ ual is m ore im portant t han his or her tangible losses . The recent attention to
u nd erstanding t errorism is actually not the first tim e that the disaster field has struggled w ith
e xposure d efinitions and d eterm ined that psychological m eanings cannot be ignored . For
e xam ple, the nuclear accid ent at Three Mile Island (TMI) in 1979 posed similar d ilemm as.
A lthough there w as an extensive and frightening em ergency evacuation, resid ents experienced
n o property d am ages, no injuries, and m inim al exposure to rad iation. H ow c ould this event
t hen be a disaster? What accounte d for the chronic stress and p sychological d istress evid ent in
TMI (e.g., Baum , Gatchel, & Schae ffer, 1983)? Discussions of these consequences often
e m phasized the long -term uncertainties regard ing the natur e of the im pact, fears that there
m ight be unseen or d elayed consequences, and anxieties regard ing threat of recurrence (Bolin,
1985). The ad verse m eanings of d isasters are not lim ited to threat and uncertainty regard ing
future harm . The long -stand ing concern about the heightened potential of hum an -caused
d isasters (includ ing various failures of technology) to affect mental health arises essentially
b ecause of w hat these events m ean to people about the larger society. As Bolin (1985 , p. 24)
n oted , hum an -cau sed d isasters ―represent in the eyes of victim s a callousness, carelessness,
intentionality, or insensitivity o n the part of others‖ a nd m ake the sense of victimization
intense. Resid ents of areas afflicted by technological accid ents often bitterly d ebate t he severity
o f the threat, and antagonism s may yield high levels of anger, alienation, and m istrust (Kaniasty
& N orris, 1999).
The process of appraising events begins imm ed iately but continues throughout t he stress
r esponse and recovery pha ses. D algleish (2004) review ed contem porary cognitive theories of
P TSD (the prim ary a d verse o utcom e of interest follow ing traum a exposure) t hat have evolved
from t he recognition of cognitive a ppraisals a s central t o t he stress response . H e id entified four
t heories that h ave influence tod ay. The first is schema theory , w hich w as d iscussed here in
Section 2. To recap, schem a theory (e.g., Janoff-Bulm an & Frieze, 1983) em phasizes the basic
a ssum ptions people hold. Schem as a re strongly held expectations that guid e w hat w e notice
a nd rem em ber and how w e interpret new information. A ccord ing to Dalgleish, schem a theory
h as tw o prim ary e xplanatory principles: (1) past experience is t he filter t hrough w hich new
e xperiences a re p rocessed ; and (2) n ew m aterial that is sufficien t ly inconsistent w ith past
e xperience is d isruptive and leads schem as to either assimilate t he information o r becom e
changed (accom m od ation). The second t heory is actually a set of associative network theories t hat
e m phasize the connectivity b etw een variou s m ental representations o f the w orld . Disparate
p ieces of inform ation can activate each other and lead to generation of affect a nd em otion . For
e xam ple, Foa , Steketee, and Rothbaum (1989) d escribed a fear netw ork com pos ed of
(a) info rm ation a bout the feared object ; (b) info rmation a bout t he ind ivid ual’s cognitive,
b ehavioral, and physiological r eactions to the feared object ; and (c) info rmation t hat links
(connects) t hese stim ulus and response elem ents together . The third theory, Brew in, Dal gleish, 20 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience a nd Joseph’s (1996) dual representation theory , e m phasizes the role of mem ory in p rocessing
s tressors over tim e . This theory em phasizes the distinctions betw een (a) v erbally accessible
m em ories (VAM) t hat can be d eliberately retrieved and a re fully contextualized in
a utobio graphical m em ory a nd (b) situationally accessible m em ories (SAM) t hat cannot be
d elib erately accessed and are not contextualized but rather are t riggered by stim uli reminiscent
o f t he t raum a. If SAM can be m ade m ore accessible an d integrated into VAM, the person can
r ecover from the event more effectively. The fourth and final theory is Ehlers and C lark ’s (2000)
cognitive model of PTSD , in w hich the n otion of current threat is central. P TSD becom es
p ersistent w hen individ uals p rocess the traum a in a w ay t hat lead s to a sense of current t hreat
r ather than a m em ory of a t hreat experienced in the past. The initial t hreat tied to the original
t raum a is transform ed through o ngoing a ppraisal p rocesses a nd m em ory representations .
A ppr aisals of ongoing threat lead t o fear, t he d ominant em otion in PTSD .
Ehlers and Clark’s (2000) cogn itive m od el of PTSD h ighlights the continuing r ole of
cognitive appraisals in stress response and recovery. In short, p eople w ho d evelop PTSD after
t raum a e xposure can be d istinguished from those w ho d o not by their excessively negative
appraisals of the event, its sequellae, or both. Ehlers and Clark (2000) nicely illustrated their points
in a table (Table 1, p . 322) that show ed particular negative appraisals related to t he event or its
s equellae. A few of these exam ples are p rovid ed as follow s: Fact that traum a happened : N ow here is safe .
Traum a happened to me: I a ttract d isaster .
Behaviors, em otion s d uring event: I d eserved it, I cannot cope w ith stress .
Emotional num bing: I’m d ead insid e, I’ll never be able to relate to people again .
D ifficulty concentrating: M y brain is d amaged , I’ll lose m y job.
O ther people’s positive response: They think I’m to o w eak to cope on m y ow n.
O ther people’s negative response: N obody is there for me. Thus one’s initial appraisal o f an event may foretell long -term outcomes (and thus t he
likelihood of resilience) . For exam ple, Ehlers et al. (1998) found that the nature o f the em otional
r esponse during a m otor vehicle crash p r ed icted t he likelihood that the survivor w ould later
d evelop PTSD . Stud ies of d isasters and t errorist attacks (N orris et al., 2002b ) have a lso found
t hat self-reported panic d uring the period of im pact is a particularly pow erful pred ictor of
s ubsequent ad verse m ental health consequences, includ ing d epression and anxiety, as w ell as
P TSD. Severely negative appraisals (panic, ―catastrophizing‖) prom pt d ysfunctional cognitions
a nd behavior s that have t he short -term aim of reducing d istress but have t he long -term
consequence of creating and m aintaining disord er (Foa et al., 1999). A ccord ing to Ehlers and
C lark, the factors that influence cognitive processing d uring t he t raum a includ e characteristics
o f the traum a , p revious experience , intellectual ability (if low , processing is less conceptual),
p rior beliefs , and s t ate factors, such as alcohol, exertion, a nd a rousal. The authors recom m end ed
t hinking through how s urvivors’ interact ions w ith police , hospital procedures, m ed ication, and
o ther m atters could be mod ified to lessen the likelihood of p roblem atic appraisals. ― C ognitive
r estructuring ‖ m ay be a particularly h elpful form of PTSD treatm ent b ecause it challenges
inaccurate, unhe althy beliefs and threats. 3.3 Ways of Coping
A t the outset, it m ay be useful to d istinguish coping resourc es from coping processes
(Taylor & St anton, 2007). Coping resources a re those r elatively stable p ersonal charact eristics that CA RRI R esearch R eport 10 21 Behavioral Science Perspectives on Resilience p rotect individ uals from stress , for exam ple , optim ism , m astery, s elf-efficacy, s elf-esteem , and
p erceived s ocial support. N ote the overlap betw een m em bers of this list and the s ources of
r esilience d escribed in Section 1. One illustrative program of w ork in this vei n – a nd one that
a ugm ents previous t hemes in Section 2 – is a s eries of stud ies cond ucted by Bonanno and
colleagues to explore the id ea that ―self -enhancing biases‖ m ay be particularly efficacious und er
cond itions of ad versity, a tim e w hen threats to t he self are prom inent. Their interest in t his topic
s tem m ed from Taylor and Brow n’s (1988) thesis regard ing illusions and mental health, in w hich
s uch self-enhancing biases featured prom inently. In a stud y cond ucted after the civil w ar in
Bosnia (Bonanno, Field , Kovacevic, & Kaltm an, 2002), participants w ho rated them selves m ore
favorably than their peers rated them (self -enhancers) also tend ed to rat e them selves as better
a d justed , and these d iscrepancies also correlated positively w ith the ratings provid ed by m ental
h ealth experts. These find ings supported speculations that self -enhancement is ad aptive. In a
longitud inal stud y of persons w ho w ere in or near the World Trad e Center on Septem ber 11,
2001 (Bonnano, Rennicke, & Dekel, 2005), average self -enhan cem ent scores w ere higher in the
g roup of participants w ho show ed little d istress over tim e (d efined as resilience in this stud y)
t han in groups w ho w ere d istressed at one or the other or both of the tim e -points in the stud y.
Coping processes s ubsum e a variety of thoughts, feelings, and behaviors r eflecting the
ind ivid ual’s efforts to ―m anage, master, tolerate, red uce, or m inimize the d em and s of a stressful
e nvironm ent‖ (Taylor & Stanton, 2007, p. 378). O ften, reference is m ad e to ―w ays of cop ing,‖
b asic categories used t o classify how people m anage stress . W ays of coping can be (1) prim ary –
coping efforts t o influence objective events or cond itions ; (2) s econd ary – coping e fforts a im ed
a t m axim izing one’s fit to current conditions ; or (3) relinquished – a bsence of any coping
a ttem pt, includ ing acceptance and adjustm ent (Skinner, Ed ge, Alt m an, & Sherw ood , 2003).
M ost com m only, researchers d istinguish betw een problem -focused coping (efforts d irected tow ard
t he stressful situation) and emotion -focused coping (efforts d irected tow ard red ucing d istress).
The form er have been more consistently associated w ith positive outcom es in research (e.g.,
low er psychological distress, better health) than have the latter ( Taylor & Stanton, 2007). JanoffBulm an a nd Frieze (1983) view ed coping as the p rocess of rebuild ing o ne’s assum ptive w orld ,
o f com ing to term s w ith shattered assum ptions. Survivors seek to red efine t he event to
m inimize threat and t he p erception o f oneself as a victim . This w ay of coping can involve
com paring o neself t o less fortunate others , creating hypothetical w orse w orld s , construing
b enefit , and m anufacturing norm ative stand ard s of ad justm ent .
Skinner et al.’s (2003) e xtraord inary Psychological Bulletin a rticle s u m m arized a variety of
conceptual issue s that have plagued coping t heory and r esea rch alm ost from the beginning.
They n oted the lack of consensus about core coping strategies, and they recom m end ed
a band oning the three m ost com m on classification schem es: pro blem -focused vs. em otion focused coping; approach vs. avoid ance; and cognitive vs. behavioral coping efforts. Instead ,
t hey ad vocated for d eveloping a hierarchical understand ing of coping actions. At the low est
level o f the hierarchy are instances of copin g, t he ―countless changing real -tim e responses that
ind ivid uals use in d ealing w ith specific stressful transaction s‖ (p. 217). Instances are concrete
e xam ples of w ays of coping, w hich in turn, are nested w ithin h igher -ord er families of coping.
Finally, fam ilies of coping can be organized into even higher -ord er ad aptive processes, called
concerns in this scheme. The three prim ary concerns are organized around challenges and
t hreats to achieving/ protecting (1) com petence o r control, (2) a ttachm ent or r elated n ess, a nd
(3) a utonom y . O n the basis of an exhaustive search and analysis of past research on coping,
Skinner et al. (2003) id entified 12 families of coping that appear to be m ost consistently
im portant. These are listed a s follow s , in a m od ified version of Figure 4 (p. 245) from their
p aper. For exam ple, planning is a w ay of coping w ithin the family of problem -solving, w hich 22 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience e ncom passes one class of approaches to managing challenges or threats to o ne’s competence
(Adaptive Processes) Inform ation seeking
(find ad d itional contingencies) Read ing, observing, asking
o thers H elp lessness
(find lim its of actions) C onfu sion, cognitive
interference or exhau stion
C ognitive avoid ance,
b ehavioral a void ance, d enial Self-reliance
(p rotect available social resou rces) Em otional regu lation,
b ehavioral regu lation Su p p ort Seeking
(u se available social resou rces) P roxim ity seeking, yearning,
o ther alliance D elegation
(find lim its of resou rces) C om p laining, self-p ity Isolation
(w ithd raw from u nsu p p ortive
context) Social w ithd raw al,
concealm ent of p roblem A ccom m od ation
(flexibly ad ju st p references and
o p tions) C ognitive restru ctu ring,
m inim ization, accep tance N egotiation
(find new op tions) Bargaining, p ersu asion,
p riority -setting Su bm ission
(give u p p references) Ru m ination, intru sive
t hou ghts O p p osition
(rem ove constraints) A u tonom y
(efforts to coord inate
p references and available
o p tions) Strategizing, instru m ental
a ction, p lanning Escap e
(flee d anger o r e nvironm ent) A ttachm ent/ Related ness
(efforts to coord inate
r eliance and social
r esou rces available) Illustrative Ways of Coping P roblem -solving
(ad ju st action to be effective )
C om p etence/ Control
(efforts to coord inate
a ctions and contingencies in
t he environm ent) Families of Coping
(Functions) O ther -blam e, p rojection,
a ggression 3.3.1 Coping Self -Efficacy
A v ariation on the them e of w ays of coping is coping self-efficacy . A s noted earlier in this
r eport, self-efficacy is a fund am ental expression of hum an agenc y. ― Whatever other factors
s erve as guid es and m otivators, they are rooted in the core belief that one has the pow er to
p rod uce d esired effects by one’s actions, otherw ise one has little incentive to act or to persev ere
in the face of d ifficulties ‖ (Benigh t & Band ura, 2004, p . 1131). W ithin the context of stress and
copin g, self-efficacy perceptions are often labeled coping self-efficacy. G eneral self-efficacy is a
t rait -like factor, w hich relates to capacities to m anage d em and s across d ifferent situations.
C oping s elf-efficacy is usually d efined and m easured w ithin particular d om ains; situation specific beliefs of coping efficacy are m ore proximally related to behaviors, although these
b eliefs m ay be influenced by the general trait (Benight & Band ura, 2004) .
O ne w ay of thinking about coping self-efficacy is that it has less to d o w ith how o ne copes –
t hat is, it m atters little w hether the w ay of coping is problem solving, avoid ance, or cognitive
r estructuring – a nd m ore to d o w ith w hether or not o ne believe s he or she can m anage the CA RRI R esearch R eport 10 23 Behavioral Science Perspectives on Resilience p articular problem or threat . This concept has m uch face valid ity. Chip Ben ight h as been
e specially influential in ad vancing the im portance of coping self -efficacy in the context of
d isasters. Through a series of stud ies of event s ran ging from hurricanes in Florid a to flood s in
C olorad o to t he bom bing in Oklahom a, Benight h as show n that coping self-efficacy is p roxim al
t o w ell-being and often m ed iate s t he effects of other coping resources, such as d ispositional
o ptim ism and social s upport (Benight & Band ura, 2004). For exam ple, in this view , social
s upport influences mental health prim arily because it raises the recipient’s self -efficacy to
m anage environm ental dem and s.
In a recent m eta -analysis , Lusczcynska, Benight, and Cieslak (2009) s ystematically r eview ed
r esearch for relations betw een self-efficacy and psychological outcome s w ithin the context of
m ass traum a. They found m ed ium to large effects of self -efficacy on d istress and PTSD
s ym ptom s in cross -sectional stud ies and larg e effects in longitud inal stud ies. 3.4 S ocial Support D ynamics
D ecad es of research have consistently d em onstrated that social support red uces d istress and
p rom otes ad justm ent und er cond itions of stress (Taylor & Stanton, 2007). Social support refers
t o so cial interactions that provid e individ uals w ith actual assistance and em bed them into a w eb
o f social relationships perceived to be loving, caring, and read ily available in tim es of need
(Barrera, 1986). This general d efinition points to three m ajor facets of social support: social
embeddedness (quantity and types of relationships w ith others), received support (actual receipt of
h elp), and perceived support (the belief that help w ould be available if ne eded , together w ith a
s ense of belonging and being cared for ). In a nutshell, perceived support r efers to helping
b ehavior that m ight h appen, received support r efers to helping behavior that does h appen, and
social embeddedness r epresents the m ost basic st ructural com ponent from w hich these functional
com ponents em erge. In this s ection , I w ill first d iscuss s ocial support conceptually , a nd then I
w ill s um m arize results from e m pirical tests of its functions in the afterm ath of d isaster . This
conceptual sum m a ry b orrow s heavily from previous review s of the d isaster social support
literature (Kaniasty & N orris, 1999; 2004).
Social support is a pow erful protective factor for all types of s tress and t raum a (Brew in,
A nd rew s, & Valentine, 2000; Cobb, 1976; Cohen & W ills, 1985; Cutrona & Russell, 1990; Sarason
e t al., 1991) but is com plicated after d is asters . Initially, there is a strong m obilization o f social
s upport, but later, paradoxically, there is a deterioration o f social support. In the im m ed iate
a fterm ath of d isasters, high levels of m utua l helping m aterialize . Tem porarily, at l east, people
s pontaneously share e xperiences a nd a re u plifted by the recognition that others care. This phase
d uring w hich the m obilization of support pred om inates a nd help is abund ant h as earned a
v ariety of heartw arm ing labels, s uch as ―altruistic com m unity,‖ ― heroic phase ,‖ ―honeym oon
p hase,‖ and ―post disaster utopia.‖ In the longer period that follow s, the realities of loss and the
form id able challenges of recovery m ust be faced . Th e heightened level of helping and cohesion
s eld om last. The attentive m ed ia and generous outsid ers leave to another crisis. With the
p assage of tim e, camarad erie is replaced by grief, anger, and d isillusionment.
M any things can lead to postd isaster d ecline s in p erceived s ocia l support . Because d isasters
a ffect entire com m unities , the need for support m ay sim ply exceed its availability, causing
e xpectations of support to be v iolated . Relocation and job loss – a nd even death follow ing the
m ost severe events – r emove im portant others from victims' supportive environm ents. Disaster
v ictim s often aband on routine social activities, leaving few er opportunities for com panionship
a nd leisu re . Social netw orks become saturated w ith stories of and feelings about the event and
m ay escape interacting. Whereas victims w ant and need to be listened to, they and others in their 24 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience social environments may not necessarily w ish to be the listeners. P hysical fatigue, em otional
irritability, and scarcity of resources increase the potential for interpersonal conflicts and social
w ithd raw al. D ifferent groups of victims may som etimes find themselves at odds.
In t ests of the ir social support deterioration m odel, Ka niasty and Norris (1993) show ed that
d eclines in perceived support occurred after disasters and that these declines affected m ental
h ealth a dversely . These tests w ere based on data collected from victims o f floods in Kentucky for
w hom b oth predisaster and p ostdisaster data w ere available. Flood v ictims reported low er
p erceived social support after the flood t han t hey h ad r eported before the flood. The erosion of
social support w as one path through w hich the disaster led to distress . That is, flood victims
experienced the impact of the disaster both directly and indirectly, through deterioration of their
social support. Since this time, evidence has continued to grow regarding the a dverse impact of
d isasters on perceived s upport a nd social embedd edness (e.g., N orris et al., 2005) a s w ell as
r egarding t he p rotections afforded by p ostdisaster social support over time (e.g., Kaniasty &
N orris, 2008).
The deterioration of support is, fortunately, not inevitable. Extending their previous model of
social support det erioration, N orris and Kaniasty (1996) proposed a nd tested a m od el of social
support deterioration deterrence (SSDD). The hypothesized relations betw een exposure, social
s upport, and d istress are show n in the various paths of the SSDD m od el, a s show n in Figure 4.
In the SSDD m od el, the m obilization of support is evid enced in the hypothesized positive effect
o f severity of exposure on received s ocial support; the m ore severe the exposure, the m ore help is
r eceived . The d eterior ation of support is evid enced in the hypothesized n egative effect of
s everity of exposure on perceived s ocial support; the m ore severe the exposure, the greater the
likelihood that perceptions of social support availability and belonging w ill suffer.
― Dete rioration d eterrence‖ occurs because r eceived support protects a gainst d eclines in
p erceived support (prot ective assistance); the m obilization of received support t o a greater or
lesser extent offsets o r counteracts t he forces that lead to the d eterioratio n of perceived support.
A ll effects of received support (actual help received from family/ friend s) operated through the
m aintenance of positive perceptions/ expectations of social support and belonging. The
m obilization of support to an extent, alt hough not com pletely, counteracted the d eterioration of
U sing d ata collected 12 and 24 m onths after H urricane H ugo and 6 and 28 m onth s after
H urricane And rew , and the m ethod s of structural equation m od eling, N orris and Kaniasty
(1996) found stro ng evidence for the hypothesized SSDD m od el. Although m ore severe d isaster
e xposure w as associated w ith low er perceptions of s upport, the t otal effects of d isaster on
p erceived support w ere less severe than t hey might have been b ecause exposure w as positively
a ssociated w ith received support, and received support w as positive ly a ssociated w ith
s ubsequent perceived support. In other w ord s, survivors w ho receive d v ery hig h levels of help
follow ing the hurricanes w ere p rotected against salien t eros ion in expectations of s upport. This
finding indicates that the m ore w e can d o to help d isaster survivors m obilize – a nd sustain –
s ocial s upport, the m ore resilient they w ill be . CA RRI R esearch R eport 10 25 Behavioral Science Perspectives on Resilience Received
assistance (+) Perceived
mobilization (+) protective
deterioration (-) Severity of
(+) Figure 4. S ocial s upport d eterioration d eterrence m odel . Solid lines ind icate paths
h ypothesized (and show n) to be significant; the dashed line ind icates the p ath hypothesized
(and show n) to be non -significant. S ource: N orris and Kaniasty (1996, p. 502).
Tangible, inform ational, and e m otional form s of help a re all need ed after d isasters (Kaniasty
& N orris, 2004). Tangible support m ay be the easiest form to provid e. Ind eed , both
g overnm ental and non -governm ental agencies provid e victim s of d isaster w ith essential shelter,
food , m oney, and loans to hasten physical a nd fiscal recovery. Fam ily and friend s’ o rientation
t o do s om ething further augm ents the abund ance of tangible support. H ow ever, inform ational
s upport m ay be even m ore im portant than tangible support after hum an -caused d isasters
characterized by invisibility, confusion, and uncertainty. Emotional support is also i m portant
a fter d isasters, as it is in m ost tim es. N otw ithstand ing the essential role governm ent agencies
a nd other form al sources of support play in the afterm ath of d isasters, the greatest challenge lies
in fostering naturally occurring social resources, w hich are m ost vital for d isaster survivors ,
e specially w ith regard to the exchange of em otional support.
The SSDD m od el d ocuments processes w herein helping activities counteract the forces of
s upport d eterioration. U nfortunately, our research has also sho w n that various factors interfere
w ith the ad equacy of support receipt . D isaster -stricken com m unities are not alw ays ruled in the
m ost egalitarian w ay (Kaniasty & N orris, 1995 ). Ideally, the m obilization of support in a
com m unity follow s the rule of relative needs, as represented in the figure by the path from
s everity of exposure to received support. Sim ply put, the m ost support goes to those w ho need
it the m ost. Som etim es , how ever, the d istribution of support follow s the rule of relative advantage.
Facto rs such as ethnicity and econom ic status are key variables affecting r eceipt of support a fter
d isasters. Socially and econom ically d isad vantaged groups are frequently too overburd ened to
p rovid e am ple help to other m em bers in tim e of ad d itional need . 3.5 Closing N ote for S ection 3
It is often observed that ―w hat is stressful for one ind ivid ual is not necessarily so for
a nother.‖ This truism is strained in the d isaster context, w here t he stress is profound , far reaching, and objectively inarguable. Yet , ind ividual d ifferences in resilience and recovery
cannot be d enied . Individ uals w ho possess stronger psychosocial resources before the event are
likely to show greater resilience after. Their post -event appraisals, coping efforts, and social
interactions further shape and d efine the rapid ity of their ad aptation. Im portantly, the stress 26 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience p rocess is am enable to change, and this is w here our a ttention t urns for the final part of
t his r eport. 4. IMPLICATION S OF BEHAVIORAL SCIEN CE:
RESILIEN CE AN D IN TERVEN TION W hile behavioral scientists often conclud e papers by d raw ing an im plication or tw o for
intervention, certain w riters have a im ed m ore intentionally to d erive lessons from research that
a re useful for p rogram s or policy. M asten and Obrad ovic (2008), for e xam ple, r eview ed the
literature on resilience to d raw principles for d isaster planning. C ogently, they observed that
― the best surveillance, equipm ent, com m unication system s, antiviral supplies, m ilitary, and
e m ergency services in the w orld w ill not be ef fective w ithout equal attention to the issues posed
b y hum an behavior under cond itions of life -threatening d anger to child ren and families (p. 9).‖
These authors discussed the im portance of considering the nature of the threat and the
d evelopm ental tim ing o f the experience. They noted that t he experiences and responses of
ind ivid uals w ill be influenced by the functioning of system s in w hich they are em bed d ed and
p articularly by t he b ehavior of t he p eople they trust or w ho provid e a s ecure base in an
a ttachm ent relationship. They recom m end ed that planners id entify the m ost likely ―first
r espond ers‖ for vulnerable popula tions, for exam ple , parents and teachers in the case of
child ren, and help them prepare to respond effectively.
For conclud ing this r eport, I w ill return to a ―big picture‖ per spective on resilience. Over
30 y ears ago, Barbara Dohrenw end (1978, p. 2) presented ―a m od el of the process w hereby
p sychosocial stress ind uces psychopathology and som e conceptions of how to counteract this
p rocess.‖ This m od el is reprod uced in Figure 5.
The heart of Dohrenw end ’s m od el is the id ea of the t ransient stress reaction , a reaction that
r esem bles d isord er but is inherently transient or self -lim iting. Three possible outcom es follow
t he n atural stress reaction: (1) growth, in w hich the ind ivid ual uses the event to d evelop further
p sychologically ; (2) status quo, in w hich the ind ivid ual returns to pre -event psychological state;
a nd (3) disorder, in w hich the stress reaction persists and ap pears to become self-sustaining. The
n ext layer of the m od el consisted of the various personal and situational assets that com e into
p lay at various stages of the stress sequence, potentially influencing the very occurrence of
s tress as w ell as the final o utcom e (i.e., grow th, status quo, or d isord er). The final layer of
D ohrenw end’s m od el illustrated potential points of intervention, ranging from political action
t o corrective therapy. One of the appealing features of this m od el w as how seem ingly disparate
a ctivities took on, in Dohrenw end ’s w ord s, ―a satisfying coherence and d irected ness.‖ The
p revention/ intervention activities w ere all d irected at und erm ining the process w hereby stress
g enerates psychopathology, but t hey tackled it at different points. Th is m essage rem ains ever so
critical for a field (resilience) guid ed by m ultiple d isciplinary perspectives.
N orris and Thom pson (1995) elaborated on Dohrenw end’s fram ew ork w ith a focus on
t raum atic events, especially d isasters. The final layer of their m od el organized potential points
o f intervention along tw o d im ensions. The first dim ension w as the t iming of the intervention .
H ere the authors distinguished betw een interventions taking place before the crisis (prim ary
p revention), d uring or shortly follow in g the crisis (second ary prevention), or after the crisis
(tertiary prevention). The second dim ension w as the level of the intervention . H ere the authors
d istinguished betw een interventions targeting ind ivid ual -level attitud es and practices or
s ocietal-leve l attitud es and practices. CA RRI R esearch R eport 10 27 Behavioral Science Perspectives on Resilience Political
action SITUATION IN
development SITUATIONAL MEDIATORS
MATERIAL SUPPORTS, SOCIAL
SUPPORTS, ETC. Stressful life
CHARACTERISTICS OF Psychological
growth Transient stress
reaction No substantial
change PSYCHOLOGICAL MEDIATORS Psychopathology ASPIRATIONS, VALUES, COPING
ABILITIES, ETC. PERSON Education &
socialization Individual skill
therapies Figure 5. A m odel of the process w hereby psychosocial stress induces psychopathology
and some conceptions of how to counteract this process. W ith kind permission from Springer
Science+Business Med ia: A merican Journal of Community Psychology , ― Social stress and
com m unity psychology,‖ v ol. 6, 1978, p. 2, B. S. D ohrenw end .
It is not d ifficult to translate these earlier m od els into contem porary ones that use a
r esilience fram e. Prim ary prevention activities t aking place before the s tressful event or d isaster
a im to boost resistance; second ary prevention activities t aking p lace d uring or s hortly after the
e vent a im to boost resilience; and tertiary prevention activities t aking place later in tim e aim to
foster recovery . W hereas resistance is the hypothetical id eal, the best p ossible outcom e is not
a lw ays resistance, nor is it al w ays resilience. A ppropriate interventions should in crease the
likelihood of r esilience am ong people w ho w ere n ot resistant and the likelihoo d of recovery
a m ong people w ho w e re not r esilient . All prevention strategies are necessary to provid e a
continuum of postd isaster care (N orris & Rosen, 2009; N orris et a l., 2009).
Behavioral scientists a re increasingly in terested in intervention r esearch , includ ing tests of
p articular psychotherapies for survivors of disasters and terrorism . For exam ple, H am blen and
colleagues (2009) tested the effectiveness of C ognitiv e Behavior Therapy for P ostd isaster
D istress (CBT-PD) in a sam ple of s urvivors o f H urricane Katrina . This m anualized intervention
e m phasizes cognitive restructuring, a technique that challenges inaccurate, unhealthy beliefs.
P articipants w ere assessed at r eferral, pretreatm ent, interm edia te treatment, and posttreatment
a nd show ed sig nificant and large im provem ents . Brew in and colleagues described a ―screen
a nd treat‖ approach that w as im plem ented after the 2005 Lond on bom bings. Outcom e d ata
r evealed large effect size s for cognitive behavioral treatm ents of persons w ho screened positive
for PTSD. 28 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience H ow ever, d espite some progress, it rem ains largely true as of this w riting that the evid ence
b ase for psychological interventions in the aftermath of d isasters is w eak. Because of a variety of
r eal-w orld challenges, clinical interventions are rarely tested , and programs are rarely e valuated
(but s ee N orris & Rosen, 2009). A few years ago, n oting that it w as unlikely that evid ence from
clinical trials w ould b e forth com ing s oon , the N ational Center for PTSD organized a w orldw id e
p anel of experts and charged them to reach consensus on intervention principles. This led to the
p ublication o f a seminal Psychiatry p aper b y Stevan H obfoll and a long list of collaborators
(2007), in w hich t he authors d re w recom m endations from the em pirical literature about
im m ed iate and mid -term interventions in the a fterm ath of m ass traum a . They d efined
intervention in the broadest sense, ranging from provision of com m unity support and public
h ealth m essaging to clinical assessm ent and intensive therapy. They looked for intervention related foci that w ere best supported by the literature as prom oting stress -resistant a nd resilient
o utcom es follow ing exposure to extrem e stress.
O rganizing their conclusions as ―five essential elem ents of m ass traum a intervention,‖ these
e xperts believed that post -d isaster interventions should aim to prom ote s afety, calm ing,
e fficacy, connectedness, and hope. The overlap of this list o f essential elements w ith t he
constructs review ed in this report is notable. The panel’s concern for promoting safety a rose from
investigation s r elating to both o bjective and subjective risk . For exam ple, t he belief that t he
w orld is d angerous is a d ysfunctional cognition that m ed iate s d evelopm ent of PTSD (see
Section 3). Safety includ es safety from bad new s, r um ors , a nd other factors that m ay increase
t hreat perception. A s for promoting calming, the pa nel observed that exposure to m ass traum a
o ften results in m arked increases in em otionality. Som e anxiety is norm al. The question is
w hether such arousal increases and rem ains at such a level as to interfere w ith s leep, eating, and
p erform ance . ― N orm alizat ion‖ is a key intervention principle to enhance calm ing. The elem ent
o f promoting efficacy should com e as no surprise; t he im portance of having a s ense of control
o ver outcom es h as alread y been discussed several tim es in this report. H obfoll et al . n oted t hat
t he principle can be extend ed to collective efficacy, w hich is the sense that one belongs to a
g roup that is likely to experience positive outcomes. A s support for the goal of promoting
connectedness, the authors referenced the trem endous bod y of resea rch on t he central im portance
o f social support, but noted the com plexit ies of sustaining support h ighlighted by t he Social
Support Deterioration D eterrence m od el (N orris & Kaniasty, 1996). Finally, the authors
r eview ed e vid ence for t he elem ent o f promoting hope follow ing m ass trauma. Disagreeing to an
e xtent w ith Snyd er’s construal of hope as com posed of agency and pathw ays (see Section 1),
H obfoll et al. (2007) argued that hope has a religious connotation for m any and is not
n ecessarily actio n o riented . Such ind ivid uals find hope not through agency but through belief
in G od .
In greater d etail than I can present here, H obfoll and colleagues outlined a series of public
h ealth m easures and individ ual/ group m easures that could be taken to prom o te s afety,
calm ing, efficacy, connected ness, and hope. They id entified approaches that should not be used
a s w ell as those that w ere recom m end ed . This paper should be required read ing for anyone
charged w ith d eveloping interventions to prom ote the resilien ce of ind ividuals after d isasters or
t errorist attacks.
H obfoll et al.’s Psychiatry a rticle w as follow ed by a series of invited commentaries, prepared
b y experts in the traum a field . These papers d rew ad d itional conclusions for program s and
p olicies. In t heir contribution to this set, N orris and Stevens (2007) d iscussed areas of overlap
b etw een the H obfoll et al. fram ew ork and the N orris et al. (2008) com m unity resilience m od el.
In particular, they highlighted how efforts to prom ote community r esilience w o uld a chieve the
five essential elem ents of m ass traum a intervention. This observati on circles back to the point CA RRI R esearch R eport 10 29 Behavioral Science Perspectives on Resilience m ad e in introd ucing this review regard ing the mutual, transactional relationship betw een
ind ivid ual and com m unity. The resilience of one cann ot b e d etermined in isolation from
t he o ther. 5. REFEREN CES
A d ger, W. (2000). Social and ecological resilience: Are they related? Progress in Human
Geography, 24, 347–364.
A hern, J., Galea, S., Resnick, H .S., Kilpatrick, D.G., Bucuvalas, M., Gold , J., et al. (2002).
Television images and psychological sym ptom s after the Septem ber 11 terrorist attacks.
Psychiatry, 65(4), 289–300.
Bandura, A. (1982). Self-efficacy mechanism in human agency. A merican Psychologist, 37, 122–147.
Barrera, M. (1986). Distinctions be tw een social support concepts, m easures, and m od els.
A merican Journal of Community Psychology, 14, 413–445.
Baum , A., Gatchel, R., & Schaeffer, M. (1983). Emotional, behavioral and physiological effects at
Three Mile Island . Journal of Consulting and Clinical Psychology, 51, 565–572.
Benight, C. C., & Band ura, A. (2004). Social cognitive theory of posttraum atic recovery: The role
o f perceived self-efficacy. Behaviour Research and Therapy , 42, 1129–1148.
Benight, C., McFarlane, A., & N orris, F. (2006). Form u lating questions about postd isaster m ental
h ealth. In F. N orris, S. Galea, M. Fried man, & P. Watson (Ed s.), M ethods for disaster mental
health research (pp. 62–77). N ew York: Guilford Press.
Bernstein, K. T., Ahern, J., Tracy, M., Boscarino, J. A ., Vlahov, D., & Galea, S. (2007). Television
w atching and the risk of incid ent probably posttraum atic stress disord er: a prospective
e valuation. Journal of N ervous and M ental Disease, 195(1), 41–47.
Bod in, P., & Wiman, B. (2004). Resilience and other s tability concepts in ecology: N otes on their
o rigin, validity, and usefulness. ESS Bulletin, 2, 33–43.
Bolin, R. (1985). Disaster characteristics and psychosocial im pacts. In B. Sow d er (Ed .), D isasters
and mental health: Selected contemporary perspectives (pp. 3–28). Rockville, MD: N IMH .
Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the human
capacity to thrive after extremely aversive events? A merican Psychologist, 59(1), 20–28.
Bonanno, G. (2005). Clarifying and extend ing the construct of ad ult resilience. A merican
Psychologist, 60, 265–267.
Bonanno, G. A., Field , N . P., Kovacevic, A., & Kaltm an, S. (2002). Self-enhancem ent as a buffer
a gainst extrem e ad versity: Civil w ar in Bosnia and traum atic loss in the United State s.
Personality and Social Psychology Bulletin, 28(2),184–196.
Bonanno, G. A., Rennicke, C., & Dekel, S. (2005). Self -enhancem ent am ong high -exposure
s urvivors of the Septem ber 11th terrorist attack: Resilience or social m aladjustm ent?
Journal of Personality and Social Psychology, 88(6), 984–998.
Bourd ieu, P. (1985). The form s of capital . In J. Richard son (Ed .), H andbook of theory and research
for the sociology of education (p. 248). N ew York: Greenw ood .
Breslau, N ., Kessler, R., Chilcoat, H ., Schultz, L., Davis, G., & And reski, P. (1998). Traum a and
p osttraum atic stress disord er in the com m unity. The 1996 Detroit Area Survey of
Traum a. A rchives of General Psychiatry, 55, 627–632.
Brew er, N ., Weinstein, N., Cuite, C., & H errington, J. (2004). Risk perceptions and their relation
t o risk behavior. A nnals of Behavioral M edicine, 27, 125–130. 30 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience Brew in, C., And rew s, B., & Valentine, J. D. (2000). Meta -analysis of risk factors for Posttraum atic
Stress Disord er in traum a -exposed ad ults. Journal of Consulting and Clinical Psychology,
Brew in, C., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of post traum atic
s tress d isord er. Psychological Review , 103, 670–686.
Brow n, B., & Perkins, D. (1992). Disruptions i n place attachm ents. In I. Altm an, & S. Low (Ed s.),
Place attachment (pp. 279–304). N ew York: Plenum Press.
Brow n, D., & Kulig, J. (1996/ 97). The concept of resiliency: Theoretical lessons from comm unity
r esearch. H ealth and Canadian Society, 4, 29–52.
Bru neau, M., Chang, S., Eguchi, R., Lee, G., O’Rourke, T., Reinhorn, A., et al. (2003). A
fram ew ork to quantitatively assess and enhance the seism ic resilience of com m unities.
Earthquake Spectra, 19, 733-752.
C arr, V., Lew in, T., Kenard y, J., Webster, R., H az ell, P., Carter, G., et al. (1997). Psychosocial
s equelae of the 1989 N ew castle earthquake: III. Role of vulnerability factors in post d isaster m orbidity. Psychological M edicine, 27, 179–190.
C obb, S. (1976). Social support as a m od erator of life stress. Psychosomatic M edicine, 38, 300–314.
C ohen, S., & Wills, T. A . (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin, 98, 310–357.
C ow en, E. (1994). The enhancem ent of psychological w ellness: Challenges and opportunities.
A m erican Journal of Community Psychology, 22, 149–179.
C ow en, E. (2000). Com munity psychology and routes to psychological w ellness. In J. Rappaport
& E. Seid m an (Ed s .), H andbook of community psychology (pp. 79–99). Dord recht,
N etherland s: Kluw er Acad em ic Publishers.
C utrona, C., & Russell, D. (1990). Type of social support and specific stress: Tow ard a theory of
o ptim al m atching. In B. R. Sarason, I. G. Sarason, & G. R. Pierce (Ed s.), S ocial support: A n
interactional view . N ew York: Wiley.
D algleish, T. (2004). Cognitive approaches to posttraum atic stress disord er: The evolution of
m ultirepresentational theorizing. Psychological Bulletin, 130, 228–260.
D ohrenw end, B. S. (1978). Social stress and com munity psychology. A merican Journal of
Community Psychology, 6, 1–14.
D ougall, A., H yman, K., H ayw ard , M., Mcfeeley, S., & Baum , A. (2001). Optim ism and
t raum atic stress: The importance of social support and coping. Journal of A pplied Social
Psychology, 31, 223–245.
Ehlers, A., & Clark, D. M. (2000). A cognitive m odel of posttraum atic stress d isord er, Behaviour
R esearch and Therapy, 38, 319–345.
Ehlers, A., Mayou, R., & Bryant, B. (1998). Psychological predictors of chronic po sttraum atic
s tress d isord er after m otor vehicle accid ents. Journal of A bnormal Psychology, 107, 508-519.
Feld m an, D., & Snyd er, C. R. (2005). H ope and the m eaningful life: Theoretical and em pirical
a ssociations betw een goal -d irected thinking and life m eani ng. Journal of Social and
Clinical Psychology, 24, 401–421.
Foa, E., Ehlers, A., Clark, D. M., Tolin, D. F., & Orsillo, S. M. (1999). The Posttraum atic
C ognitions Inventory (PTCI): d evelopm ent and valid ation. Psychological Assessment , 11,
Foa, E., Steketee, G., & Rothbaum , B. (1989). Behavioral/ cognitive conceptualization of post traum atic stress disord er. Behavior Therapy, 20, 155–176.
Folkm an, S. (1984). Personal control and stress and coping processes: A theoretical analysis.
Journal of Personality and Social Psychology , 46(4), 839–852. CA RRI R esearch R eport 10 31 Behavioral Science Perspectives on Resilience G arm ezy N . (1991). Resilience and vulnerability to ad verse d evelopm ental outcom es associated
w ith poverty. A merican Behavioral Scientist, 34(4), 416–430.
H am blen, J. L, N orris, F., P ietruszkiew icz, S., Gibson, L. E, N aturale, A., & Louis, C. (2009).
C ognitive Behavioral Therapy for Postd isaster Distress: A com m unity based treatm ent
p rogram for survivors of H urricane Katrina. A dministration and Policy in M ental Health
and M ental Health Services Research, 36, 206–214.
H arris, P., Griffin, D., & Murray, S. (2008). Testing the lim its of optimistic bias: Event and
p erson m od erators in a m ultilevel fram ew ork. Journal of Personal and Social Psychology,
H obfoll, S. E. (1989). Conservation of resources: A new a ttem pt at conceptualizing stress.
A merican Psychologist, 44, 513–524.
H obfoll, S. E. (1998). S tress, culture and community: The psychology and philosophy of stress . N ew
York: Plenum Press.
H obfoll, S. E., H all, B., H orsey, K., & Lam oureux, B. (in press). Resilience in the face of
t errorism : Linking resource investm ent w ith engagem ent. In S. Southw ick, D. Charney,
M . Fried m an, & B. Litz (Ed s.), R esilience: Responding to Challenges across the Lifespan.
H obfoll, S., Palm ieri, P., Jo hnson, R., Canetti-N isim , B., H all, B., & Galea, S. (2009). Trajectories
o f resilience, resistance, and d istress d uring ongoing terrorism : The case of Jew s and
A rabs in Israel. Journal of Consulting and Clinical Psychology, 77, 138–148.
H obfoll, S., Watso n, P., Bell, C., Bryant, R., Brym er, M., Fried m an, M. J., et al. (2007). Five
e ssential elem ents of imm ed iate and m id -term mass traum a intervention: Empirical
e vid ence. Psychiatry 70, 283–315.
Janoff-Bulm an, R. (1989). Assum ptive w orld s and the stress of t raum atic events: Applications of
t he schem a construct. S ocial Cognition, 7, 113–136.
Janoff-Bulm an, R., & Frieze, I. (1983). A theoretical perspective for und erstand ing reactions to
v ictimization. Journal of Social Issues, 39, 1–17.
Kaniasty, K., & N orris, F. (1993). A test of the support d eterioration m od el in the context of
n atural d isaster. Journal of Personality and Social Psychology, 64, 395–408.
Kaniasty, K., & N orris, F. (1995). In search of altruistic com m unity: Patterns of social support
m obilization follow ing H urricane H ugo. A merican Journal of Community Psychology, 23,
Kaniasty, K., & N orris, F. H . (1999). The experience of d isaster: Individ uals and com m unities
s haring traum a. In, R. Gist & B. Lubin (Ed s.), R esponse to disaster: Psychosocial, community,
and ecological approaches. P hilad elphia, PA: Brunner/ Maze.
Kaniasty, K., & N orris, F. H . (2004). Social support in the afterm ath of disasters, catastrophes,
a nd acts of terrorism : Altruistic, overw helm ed , uncertain, antagonistic, a nd patriotic
com m unities. In R. Ursano, A. N orw ood , & C. Fullerton (Ed s.), Bioterrorism: Psychological
and public health interventions. Cam brid ge, UK.: Cam brid ge University Press.
Kaniasty, K., & N orris, F. (2008). Longitud inal linkages betw een perceived s ocial support and
p sychological distress: A test of sequential m od el of social causation and social selection.
Journal of Traumatic Stress, 21, 274–281.
Kessler, R. C., Sonnega, A., Brom et, E., H ughes, M., & N elson, C. B. (1995). Posttraum atic stress
d iso rd er in the N ational Com orbid ity Survey. A rchives of General Psychiatry, 52(12),
Klein, R., N icholls, R., & Thom alla, F. (2003). Resilience to natural hazard s: H ow useful is this
concept? Environmental Hazards, 5, 35–45. 32 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience Lau, J.T.F., Lau, M., Kim , J.H ., & Tsui, H .Y. (2006). Im pacts of med ia coverage on the
com m unity stress level in H ong Kong after the tsunam i on 26 Decem ber 2004. Journal of
Epidemiology and Community Health, 60(8), 675–682.
Layne, C. M., Warren, J. S., Watson, P. J., & Shalev, A . Y. (2007). Risk, vulnerability, resistance,
a nd resilience: Tow ard an integrative conceptualization of posttraum atic ad aptation. In
M . Fried m an, T. Keane, & P. Resick (Ed s). H andbook of PTSD: Science and practice
(pp. 497–520) N ew York: Guilford Press.
Lazarus, R., & Folkman, S. (1984). S tress, appraisal, and coping. N ew York: Springer.
Lepore, S., & Revenson, T. (2006). Resilience and posttraum atic grow th: Recovery, resistance, &
r econfiguration. In L. Calhoun & R. Ted eschi (Ed s.), T he handbook of posttraumatic growth:
R esearch and practice (pp. 24–46). M ahw ah, N J: Erlbaum .
Lerner, M. (1980). The belief in a just w orld . N ew York: Plenum .
Linley, P., & Joseph, S. (2005). The hum an capacity for grow th through adversity. A merican
Psychologist, 60, 262–264.
Litz, B. (2005). Has resilience to severe trauma been underestimated? A merican Psychologist, 60, 262.
Lucszcynska, A., Benight, C., & Cieslak, R. (2009). Self -efficacy and health -related outcom es of
collective traum a: A system atic review . European Psychologist, 14, 49–60.
M ad d ux, J., & Rogers, R. (1983). Protection m otivation and self -efficacy: A revised theory of fear
a ppeals and attitud e change. Journal of Experimental Social Psychology, 19, 469–479.
M arshall, R.D., Bryant, R. A ., Am sel, L. V., Suh, E. J., Cook, J. M ., & N eria, Y. (2007). The
p sychology of ongoing threat: r elative risk appraisal, the Septem ber 11 attacks, and
t errorism -related fears. A merican Psychologist, 62(4), 304–316.
M asten, A. S. (2001). Ordinary m agic: Resilience processes in d evelopm ent. A merican
Psychologist, 56(3), 227–238.
M asten, A., & Obrad ovic, J. (2008). D isaster preparation and recovery: Lessons from research on
r esilience in hum an d evelopm ent. Ecology and Society, 13, 9 [on -line].
N orris, F. (1992). The epid em iology of traum a: Frequency and im pact of different potentially
t raum atic events on d ifferent d em ographic groups. Journal of Consulting and Clinical
Psychology, 60, 409–418.
N orris, F. (1997).The frequency and structure of prec autionary behavior in the d om ains of
h azard prepared ness, crim e prevention, vehicular safety, and health m aintenance. H ealth
Psychology, 16, 566–575.
N orris, F., Baker, C., Murphy, A., & Kaniasty, K. (2005). Social support m obilization and
d eterioration after Mexico’s 1999 flood : Effects of context, gend er, and time . A merican
Journal of Community Psychology , 36, 15–28.
N orris, F. H ., Fried man, M. J., & Watson, P. J. (2002a ). 60,000 d isaster victim s speak: Part II.
Sum m ary and im plications of the disaster m e ntal health research. Psychiatry , 65, 240-260.
N orris, F. H ., Fried man, M. J., Watson, P. J., Byrne, C. M., Diaz, E., & Kaniasty, K. (2002 b ). 60,000
d isaster victim s speak: Part I. An em pirical review of the em pirical literature, 1981 –2001.
Psychiatry , 65, 207–239.
N orris, F. H ., & Kaniasty, K. (1996). Received and perceived social support in tim es of stress: A
t est of the social support d eterioration d eterrence m od el. Journal of Personality and Social
Psychology , 71, 498–511.
N orris, F., & Rosen, C. R. (2009). Innovations in d isaster m ental health services and evaluation:
N ational, state, and local responses to H urricane Katrina (Introd uction to the special
issue). A dministration and Policy in M ental Health and M ental Health Services Research,
36,159–164. CA RRI R esearch R eport 10 33 Behavioral Science Perspectives on Resilience N orris, F., & Stevens, S. (2007). Com m unity resilience and the principles of m ass traum a
intervention. Psychiatry, 70, 320–328.
N orris, F., Stevens, S., Pfefferbaum , B., Wyche, K., & P fefferbaum , R. (2008). Com m unity
r esilience as a m etaphor, theory, set of capacities, and strategy for d isaster read iness.
A merican Journal of Community Psychology, 41, 127–150.
N orris, F., & Thom pson, M. (1995). Applying comm unity psychology to the prev ention of
t raum a and traum atic life events. In J. Freed y & S. H obfoll (Ed s.) T raumatic stress: From
t heory to practice (pp. 49–71). N ew York: Plenum .
N orris, F., Tracy, M., & Galea, S. (2009). Looking for resilience: Und erstand ing the longitud inal
t rajecto ries of responses to stress. S ocial Science & M edicine, 68, 2190–2198.
O 'Donnell, M. L., Elliott, P., Lau, W., & Cream er, M. C. (2007). PTSD sym ptom trajectories: From
e arly to chronic response. Behaviour Research and Therapy, 45(3), 601–606.
O rcutt, H . K., Erickson, D. J., & Wolfe, J. (2004). The course of PTSD sym ptom s am ong gulf w ar
v eterans: A grow th mixture m od eling approach. Journal of Traumatic Stress, 17, 195–202.
O tt, C. H ., Lueger, R. J., Kelber, S. T., & Prigerson, H . G. (2007). Spousal bereave m ent in old er
a d ults: Com m on, resilient, and chronic grief w ith d efining characteristics. Journal of
N ervous and M ental Disease, 195(4), 332–341.
P erkins, D., H ughey, J., & Speer, P. (2002). Comm unity psychology perspectives on social
capital theory and com munity d evelopm ent practice. Journal of the Community
D evelopment Society, 33, 33–52.
Rutter, M. (1993). Resilience: som e conceptual consid erations. Journal of Adolescent Health, 14,
Sarason, B. R., Pierce, G. R., Shearin, E. N ., Sarason, I. G ., Waltz, J.A., & Poppe, L. (1991).
P erceived social support and w orking m od els of self and actual others. Journal of
Personality and Social Psychology, 60, 273–287.
Scheier, M. F., & Carver, C. S. (1985). Optimism , coping, and health: Assessm ent and
im p lications of generalized outcom e expectancies. H ealth Psychology, 4, 219–247.
Seligm an, M. (1991). Learned optimism . N ew York: Knopf.
Silver, R.C., H olm an, E. A ., McIntosh, D. N ., Poulin, M., & Gil-Rivas, V. (2002). N ationw id e
longitud inal stud y of psychological responses to Septem ber 11. Journal of the A merican
M edical A ssociation, 288(10), 1235–1244.
Skinner, E., Ed ge, K., Altm an, J., & Sherw ood , H . (2003). Searching for the structure of coping: A
r eview and critique of category system s for classifying w ays of coping. Psychological
Bulletin, 129, 216–269.
Slovic, P., Finucane, M., Peters, E., & MacGregor, D. (2004). Risk as analysis and risk as feelings:
Som e thoughts about affect, reason, risk, and rationality. R isk A nalysis, 24, 311–322.
Snyd er, C. R., Feld m an, D., Taylor, J., Schro ed er, L., & Ad am s, V. (2000). The roles of hopeful
t hinking in preventing problem s and enhancing strengths. A pplied & Preventive
Psychology, 9, 249–270.
Taylor, S. E., & Brow n, J. D . (1988). Illusion and w ell -being: A s ocial psychological perspective
o n m e ntal health. Psychological Bulletin, 103, 193–210.
Taylor, S. E., & Stanton, A. L. (2007). Coping resources, coping processes, and m ental health.
A nnual Review of Clinical Psychology, 3, 377–401.
Van Zom eren, M., Postmes, T., & Spears, R. (2008). Tow ard an integrative social identity m od el
o f collective action: A quantitative research synthesis of three socio -psychological
p erspectives. Psychological Bulletin, 134, 504–535.
W einslein, N . (1980). Unrealistic optim ism about future life events. Journal of Personality and
S ocial Psychology, 39, 806–820. 34 CA RRI R esearch R eport 10 Behavioral Science Perspectives on Resilience W einstein, N . (1993). Four com peting theories of health -protective behavior. H ealth Psychology,
W erner, E. S., & Sm ith, R. S. (1982). V ulnerable but invincible: A longitu dinal study of resilient
children and youth. N ew York: Ad am s, Bannister, & Cox.
Z autra, A., & Bachrach, K. (2000). Psychological d ysfunction and w ell -being: Public health and
s ocial indicator approaches. In J. Rappaport & E. Seid m an (Ed s.), H andbook of com munity
psychology (pp. 165–186). Dord recht, N etherland s: Kluw er Acad emic Publishers. CA RRI R esearch R eport 10 35 Community and Regional Resilience Institute
National Security Directorate
P.O. Box 2008
Oak Ridge National Laboratory
Oak Ridge, TN 37831-6252
www.ResilientUS.org ORNL 2011-G00800/jcn ...
View Full Document