Development and Psychopathology, 2 (1990), 425-444

Copyright © 1991 Cambridge University Press

Printed in the United States of America

Resilience and development:

Contributions from the study of children
who overcome adversity

ANN S. MASTEN, KARIN M. BEST, AND NORMAN GARMEZY

University of Minnesota

Abstract
This article reviews the research on resilience in order to delineate its significance and potential for
understanding normal development. Resilience refers to the process of, capacity for, or outcome of successful
adaptation despite challenging or threatening circumstances. Three resilience phenomena are reviewed: (a)
good outcomes in high-risk children, (b) sustained competence in children under stress, and (c) recovery from
trauma. It is concluded that human psychological development is highly buffered and that long-lasting
consequences of adversity usually are associated with either organic damage or severe interference in the
normative protective processes embedded in the caregiving system. Children who experience chronic adversity
fare better or recover more successfully when they have a positive relationship with a competent adult, they are
good learners and problem-solvers, they are engaging to other people, and they have areas of competence and
perceived efficacy valued by self or society. Future studies of resilience will need to focus on processes that
facilitate adaptation. Such studies have the potential to illuminate the range and self-righting properties of,
constraints on, and linkages among different aspects of cognitive, emotional, and social development.

One of the fundamental tenets of develop- & Braswell, in press). The goal of this arti-
mental psychopathology is that atypical cle is to examine the significance of resil-
and normative development are best stud- ience research for the study of normal de-
ied in concert, each in the context of the velopmental processes and for promoting
other, the study of one informing and en- better developmental outcomes in children
riching the study of the other (Cicchetti, at risk.
1984, 1989, 1990; Masten & Braswell, in Recent empirical study of resilience is
press; Sroufe & Rutter, 1984). Concomi- closely tied to the emergence of develop-
tantly, it has been asserted that the study of mental psychopathology. They share many
children who overcome risk and adversity common roots (Masten, 1989). Studies of
enhances the understanding of both normal children at risk for psychopathology played
development and maladjustment (Masten a key role in bringing both the phenome-
non of resilience and the integrative frame-
work of developmental psychopathology to
Preparation of this article was facilitated by grant the forefront of theoretical and research at-
#MH33222 from the National Institute of Mental tention (Anthony, 1974; Garmezy, 1971,
Health and a McKnight-Land Grant Professorship 1974a, 1974b, 1985; Masten, 1989; Masten
award to Dr. Masten from the University of Minne- & Garmezy, 1985; Rutter, 1979b, 1990).
sota. The authors express their deep appreciation to Investigators trained in psychopathology
Auke Tellegen and their other colleagues on Project
Competence for many thought-provoking conversa-
joined forces with developmentalists to
tions over the years and to Michael Rutter for his tackle difficult conceptual and method-
contributions to their thinking about resilience. ological obstacles to the study of individual
426 A. S. Masten, K. M. Best, andN. Garmezy

differences over time in children at risk Resilience as Overcoming the Odds

(Garmezy, 1974a; Garmezy & Phipps-
Yonas, 1984; Garmezy & Streitman, 1974; The concept of resilience that grew directly
Masten & Bras well, in press). Attempts to out of risk research refers to successful ad-
monitor the quality of adaptation across aptation despite the odds against good de-
the lifespan focused attention on the con- velopment. Risk factors are statistical cor-
struct of competence and its relation to relates of poor or negative outcomes. Thus,
psychopathology. Investigators began to for example, poverty, low maternal educa-
note children who were flourishing despite tion, low socioeconomic status, low birth-
their risk status; these children, labeled "re- weight, family instability, and schizophre-
silient" or "stress-resistant," became the fo- nia in a biological mother have been
cus of researchers seeking to understand in- associated with lower academic achieve-
dividual variations in human behavioral ment, more emotional or behavioral prob-
responses to adversity or challenge. lems, lower work achievement, and/or
Resilience refers to the process of, ca- trouble with the law (Farran & McKinney,
pacity for, or outcome of successful adap- 1986; Kopp & Krakow, 1983; Masten &
tation despite challenging or threatening Garmezy, 1985; Watt, Anthony, Wynne, &
circumstances. Psychological resilience is Rolf, 1984). In epidemiological and clinical
concerned with behavioral adaptation, usu- research, the consequences of "cumulative
ally defined in terms of internal states of risk" have been studied; risk factors of-
well-being or effective functioning in the ten co-occur, and when they do, they
environment or both. "Protective factors" appear to carry additive and possibly ex-
moderate the effects of individual vulnera- ponential risks (Masten, Morison, Tel-
bilities or environmental hazards so that legen, & Pellegrini, 1990; O'Dougherty,
the adaptational trajectory is more positive Wright, Garmezy, Loewenson, & Torres,
than would be the case if the protective fac- 1983; Rutter, 1979b; Sameroff & Seifer,
tor were not operational. Protective factors 1990).
do not necessarily yield resilience: it is con- Risk factors are often identified retro-
ceivable that protective processes, such as spectively, as when victims of toxic shock
efforts by a parent to foster adaptation, syndrome were found to have elevated
may not be adequate if the vulnerability of rates of using certain types of tampons, or
the individual or the severity of the adver- low-achieving children were found to have
sity is too great to overcome. higher rates of perinatal brain damage and
In the psychological literature, resilience disadvantaged parents. Prospective studies
has been used to describe three distinct often corroborate an overall risk effect,
kinds of phenomena: (1) good outcomes that is, more problems are observed among
despite high-risk status, (2) sustained com- the risk group; at the same time, such in-
petence under threat, and (3) recovery from vestigations usually reveal diversity in out-
trauma. Studies across these phenomena comes, including a group of children who
have yielded similar as well as unique find- develop well.
ings. Both are important: similarities point The Study of Adult Development at
toward universal human capacities and Harvard Medical School, for example, has
protective factors for development; unique indicated diversity and considerable resil-
findings may guide the design of specific ience in socioeconomically disadvantaged
intervention and prevention programs for males (Felsman & Vaillant, 1987; Long &
children in particular risky situations. Vaillant, 1984). This longitudinal study has
Studies of these phenomena will be re- followed 456 inner-city children since 1940,
viewed to examine the contribution resil- when they were first recruited in Boston at
ience research has made and can make the ages of 12-16. Boyhood Competence
to the understanding of normal develop- was the best overall predictor of adult ad-
ment. justment in middle age. For men from the
Resilience and development 427

highest risk backgrounds in terms of psy- comes than the original risk factor of "re-
chosocial disadvantage and multiproblem productive casualty" (phase 1) or the risk
families, upward mobility was associated of obvious brain damage (phase 2) (Rutter,
with childhood IQ and a composite vari- 1990). Better measurement has led to more
able called Childhood Coping, which in- precise identification of risk variables that
dexed adequacy in social, family, and constrain adaptation.
school functioning. Perhaps the most im- When risk is defined by an unrefined
portant finding of this study for the present status variable, such as premature birth or
review was the overall pattern of resilience low income or marital status, it is difficult
in what has been labeled the "underclass," to study resilience because the exact nature
leading Long and Vaillant (1984) to con- of the risk is unclear. Studies of risk for
clude, "The transmission of disorganiza- schizophrenia provide another example of
tion and alienation that seems inevitable this problem.
when a disadvantaged cohort is studied ret-
rospectively appears to be the exception
rather than the norm in a prospective study Children at risk for schizophrenia
that locates the successes as well as the fail- Studies of children of mothers with schizo-
ures" (p. 344). phrenia played a crucial role in the emer-
gence of developmental psychopathology
and resilience as prominent theoretical and
Three phases of risk research
research topics (Garmezy, 1974a; Garmezy
The perinatal risk literature exemplifies the & Streitman, 1974; Masten, 1989; Masten
phases of research that often characterize & Braswell, in press). A substantial geneti-
the study of risk factors. The first phase cally oriented literature indicates that these
includes the initial identification of a risk children have a relatively high probabil-
factor, often based on retrospective data. ity of eventually developing this disorder
Thirty years ago, Pasamanick and Kno- (around 10 times higher than the general
bloch (1960, 1961) noted that cerebral population risk of about 1%), yet 90% or
palsy, epilepsy, mental retardation, reading more of these children at risk will not mani-
disabilities, and behavior disorders were fest schizophrenia (Gottesman & Shields,
linked to perinatal complications, which in 1982). Since there are not yet specific mark-
turn were associated with low socioeco- ers established for identifying which chil-
nomic status: they popularized the concept dren may be vulnerable to this disorder,
of "reproductive casualty." The second the 90% who do not develop the disorder
phase of risk research is marked by pro- usually are not viewed as "resilient." Some
spective studies of the potential risk group may be, but without more exact specifica-
and often yields rather different conclu- tion of risk, resilience cannot be ascer-
sions. When Sameroff and Chandler (1975) tained. When genetic or other specific
reviewed the data on perinatal risk 15 years markers of vulnerability to schizophrenia
after Pasamanick and Knobloch, they con- are identified, it will be possible to study
cluded that, in the absence of clear organic resilience more precisely with respect to this
brain damage, perinatal complications per disorder (Masten, 1989).
se carried little long-term risk for psycho- There are other, more general, ways in
pathology. Moreover, socioeconomic sta- which children living with a parent who has
tus appeared to outweigh reproductive schizophrenia are at risk, and in this case
risks as a long-term predictor of outcome. the risks can be delineated. Parents with
Refinement characterizes the third phase of severe mental disorders may have difficult-
risk research. Brain imaging and other im- ies providing physical and emotional care
proved technologies have provided more or may fail to show consistency of care be-
specific evidence of organic damage, which cause of the symptoms of their illness. Hos-
now appears to be a better predictor of out- pitalization of the parent may necessitate
428 A. S. Masten, K. M. Best, and N. Garmezy

separation from children. Investigations of In a search for differentiating character-

children of mentally ill parents indicate istics, these resilient children were com-
that the severity of the illness and its effects pared with a comparable high-risk sample
on caregiving may have more of an impact who had developed problems. The results
on a child's development than the particu- raise important questions about the con-
lar type of illness (Rodnick & Goldstein, struct of risk itself. For example, resilient
1974; Sameroff & Seifer, 1990). The most children received and appeared to elicit
widely noted correlate of favorable child greater attention as babies. They were more
outcomes for children living with a parent active, and socially responsive, viewed as
who has a serious mental illness is warm good-natured (boys) and cuddly (girls) by
relationships with adults in and/or outside their mothers. They also had fewer congen-
the family; other correlates are child attri- ital defects. These results could be interpre-
butes of good intellectual skills and appeal- ted to suggest that a robust constitution, an
ingness to adults (Bleuler, 1984; Fisher, easy temperament, and good parenting
Kokes, Cole, Perkins, & Wynne, 1987; are protective factors, since these qualities
Heston, 1960; Radke-Yarrow & Sherman, were associated with better outcomes
1990). among the high-risk sample. On the other
hand, the same data could be interpreted
to signify that these children had better
Cumulative risk outcomes because they initially had more
Many risk studies have focused on complex assets and fewer liabilities.
risk-laden environments or multiple/cumu-
lative risks that co-occur rather than Proximal and distal risk. Baldwin, Bald-
single-risk factors or status variables. In win, and Cole (1990) proposed that proxi-
these studies, it is sometimes difficult to mal risk be distinguished from distal risk in
separate resilience from lower overall risk. studies of multiple-risk factors. Distal-risk
The difficulty is well-illustrated in a classic factors, such as social class, are not experi-
study of resilience in children with mul- enced directly by a child: they are mediated
tiple-risk factors described by Werner and by proximal variables. Proximal-risk fac-
Smith (1982) in their book, Vulnerable but tors, such as inadequate nutrition, discord
Invincible: A Longitudinal Study of Resil- between parents, or antisocial behaviors
ient Children and Youth. This study was in a parent, impinge directly on a child.
part of an ongoing longitudinal project de- For example, the distal-risk factor of
signed to follow a cohort of infants born in poverty or economic hardship may well be
1955 on the Hawaiian island of Kauai. A mediated by the proximal-risk factor of in-
set of key risk factors in this study effective parenting or, if one takes the posi-
consistently predicted serious problems at tive view, the protective factor of compe-
later points in development: risks included tent parenting (Elder, Nguyen, & Caspi,
chronic poverty, low maternal education, 1985; Lempers, Clark-Lempers, & Simons,
family conflict or instability, and moder- 1989).
ate-to-severe perinatal stress. About one- When a child at risk because of a distal
third of the subjects had four or more of factor such as poverty is protected by good
these risk factors. Of these, about one- parenting, it may be more accurate to de-
third (one-tenth of the total sample) were scribe the family rather than the child as
identified as resilient because they devel- resilient (Baldwin et al., 1990). In their
oped none of the problems associated with study of resilient families from the Roches-
risk in the overall sample. Instead, they ap- ter Longitudinal Study (Sameroff & Seifer,
peared to be developing well, as indicated 1990), Baldwin et al. (1990) compared the
by assessments at age 10 and age 18 (Wer- behaviors of parents who had competent
ner & Smith, 1982) and, subsequently, at children but who lived under different
age 30 (Werner, 1989). conditions. They compared families from
Resilience and development 429

low-risk, middle-class neighborhoods with IQ, as an index of academic aptitude,

families from high-risk, urban neighbor- may function as a protective factor for
hoods. Risk factors included low SES, mi- disadvantaged children in modern society
nority status, and father absence. The two because of the benefits associated with aca-
groups of effective parents shared certain demic achievement in societies that empha-
qualities, including warmth and placing a size schooling. For example, in technologi-
high value on responsible behavior in their cal societies, school success and all the
children. They differed in other ways, how- positive concomitants of this success may
ever. Parents of children at high risk were be a key mechanism by which children at
less democratic, and placed a higher value risk for antisocial behavior (because of
on self-control than did effective parents of their own antisocial behavior in early child-
low-risk children. The differences in par- hood or the criminality of their parents)
enting styles appeared to reflect the rearing avoid intraindividual or transgenerational
context outside the family: a higher level of continuity (Garmezy & Masten, in press;
monitoring and control was probably re- Hirschi & Hindelang, 1977). In a cross-
quired to protect children and foster devel- sectional study of a 30-year-old Danish
opment in the dangerous environment. birth cohort, Kandel et al. (1988) found
that high-risk offspring of severely criminal
fathers who avoided criminal records had
Protective factors for children at risk higher IQ scores. They found that this
When distal risk is high, as it usually is in high-risk group with no criminal record
poor urban neighborhoods, the importance had a higher mean IQ than three other
of parenting, schooling, and intellectual comparison groups, including a low-risk
skills appears to increase, as if in a compen- group, suggesting that higher IQ may have
satory fashion. Thus, higher IQ scores, compensated for greater risk potential. A
competent adult parenting figures, and ef- prospective study in a New Zealand birth
fective schools are frequently identified as cohort found higher childhood IQ scores
correlates of good outcomes in high-risk predictive of low delinquency rates in ado-
samples, whether in Boston (Long & Vail- lescence for both boys and girls among all
lant, 1984), Rochester (Baldwin et al., risk groups (White, Moffitt, & Silva, 1989).
1990), or London (Rutter, Maughan, Mor- In this case, risk was defined by antisocial
timore, Ouston, & Smith, 1979). behavior at age 5. High-risk children who
Positive school experiences also appear avoided delinquency had IQ scores compa-
to be protective for proximal risk. Rutter rable to (but not higher than) low-risk non-
and Quinton (1984) found that positive delinquent adolescents. Consistent with
school experience was associated with bet- many other studies of delinquent behavior
(West & Farrington, 1973), rates of delin-
ter marital and work outcomes in a group
quency in both studies were higher in the
of institutionally reared women, but not in
high-risk groups and, in the New Zealand
a comparison group. Rutter (1990) sug-
study, higher in the males than the females.
gested that positive school experiences
compensated for the less positive rearing On the other hand, one of the risk mech-
experiences of such women and that such anisms for aggressive children may be the
experiences were not differentially related deleterious effects of their behavior on
to the outcome in the comparison group their own intellectual development. Based
subjects because they had experiences of on data from their 22-year prospective
success and accomplishment at home. study of aggression, Huesmann, Eron, and
Other work by Rutter (1979a), Werner, Yarmel (1987) suggested a transactional
(1990), and Werner and Smith (1982) indi- model in which low intelligence interferes
cated that favorable school experiences with early learning of nonaggressive
lessen the effects of stressful home environ- problem-solving skills and coping strate-
ments. gies; aggression, in turn, interferes with
430 A. S. Masten, K. M. Best, andN. Garmezy

learning opportunities at school and with leagues in Topeka exerted a strong influ-
peers, thereby impeding intellectual devel- ence on the formulation of this concept of
opment. resilience. They distinguished two types of
Another protective factor associated coping: Coping I referred to the "capacity
with qualities and systems that extend be- to cope with opportunities, challenges,
yond the child and the family is the church. frustrations, threats in the environment";
Although the role of the church has not and Coping II referred to the "maintenance
been examined as systematically as have of internal integration" (Murphy & Mori-
other protective factors, church member- arty, 1976). In Murphy's model, resilience
ship and faith in a higher power are re- implied both forms of coping and resulted
ported to function as protective factors in from multiple mechanisms, innate and
diverse high-risk situations (Comer, 1988; learned. Among these protective processes
Ianni, 1989; Pinkney, 1987; Rosenblatt, were biological homeostasis, mastery moti-
1983; Werner, 1990). Baldwin et al. (1990) vation, and successful experiences with
noted that among the inner-city minority challenging tasks. This model drew heavily
families with competent children, member- on psychoanalytic theory, Robert White's
ship in a church community appeared to be concept of competence, and the rich case
an important aspect of family life. Simi- observations of a cohort of children grow-
larly, Werner (1989) observed that religious ing up in Topeka (Masten & Garmezy,
faith characterized many of the resilient 1985).
Kauai risk group in adulthood. Carefully controlled studies of resilience
Pargament (1987) described the func- in children under stress are difficult to
tions of religion in terms of coping pro- achieve because of the nature of the phe-
cesses that depend on the individual's be- nomena. For obvious reasons, prospective
liefs, the nature of the event, and the studies that begin prior to the stress expo-
context in which it occurs. Religion may sure are rare and serendipitous. Thus, most
influence appraisals of stressful situations studies are retrospective, and prospective
or fears of death, availability of social sup- studies usually begin after the stressor has
port resources, or choices of coping behav- occurred or it is already in progress.
ior (e.g., prayer vs. alcohol consumption).
Pargament also notes that deeper spiritual
commitments or closeness to a church com- Divorce
munity or seeking of religious support may A research area with great relevance in
develop from adverse life experiences. today's society is the study of children's
responses to divorce. Studies of divorce
suggest that there is a crisis period of read-
Resilience as Stress-Resistance
justment immediately following marital
In the literature on stress and coping in disruption, followed by gradual improve-
children, a somewhat different meaning of ment in child functioning (Emery, 1988;
resilience is evident, pertaining to sustained Hetherington, Stanley-Hagan, & Ander-
competent functioning despite severely chal- son, 1989; Peterson & Zill, 1986; Waller-
lenging circumstances. In this case, acute stein & Kelly, 1980). Boys appear to react
or chronic major life stressors are the risk with more conduct problems than do girls,
factors, and the research has focused on although boys may be responding to ongo-
current or short-term functioning. Resil- ing or long-term family conflict and custo-
ience in this situation implies effective cop- dial arrangements rather than marital dis-
ing, meaning efforts to restore or maintain solution per se (Block, Block, & Gjerde,
internal or external equilibrium under sig- 1986, 1988; Emery, 1988; Peterson & Zill,
nificant threat by means of human activi- 1986). Children also appear to fare better
ties, including thought and action. when they live with the same-sex parent.
The studies of Lois Murphy and her col- Because mothers still have custody in 90%
Resilience and development 431

or more of divorces (Emery, 1988), girls Garmezy and Rutter (Garmezy, 1985; Gar-
may be protected by the traditional custody mezy & Rutter, 1985). Although the studies
arrangements. Girls also may be less ex- are often individually flawed or method-
posed to parental conflict in families and ologically weak, the data across dramati-
receive more comforting from mothers cally different situations and cultures con-
than boys (Hetherington et al., 1989). Both verge to suggest several protective factors
boys and girls fare better in divorced fami- that appear to buffer psychological devel-
lies when interparental conflict is low, the opment.
relationship with each parent is good, and Across accounts, the most important
parenting quality is high (Emery, 1988; Pe- and consistent protective factor is that of
terson & Zill, 1986). Divorce also may lead adults caring for children during or after
to more responsible behavior in children major stressors. Brown and Harris (1978)
whose help may be more necessary in found greater vulnerability to depression in
single- than dual-parent households (Demo adult women who lost a parent in child-
& Acock, 1988). hood; however, it later became clear that
It is difficult to separate developmental the quality of caregiving the child received
effects of divorce from confounding vari- after the loss, rather than the loss itself,
ables related to age of child at marital sepa- was the key predictor of vulnerability to
ration, severity of marital conflict, changes depression (Brown, Harris, & Bifulco,
in financial situations, and age of parents. 1986; Rutter, 1990). Children in disasters
Perhaps for this reason, results for age- and war adapt better when their parents
related effects and long-lasting effects of function adequately in the presence of
divorce do not present a clear picture. stress and when they are not separated
Overall, preschoolers appear to be more from their parents (Garmezy & Rutter,
negatively affected in the short term (Em- 1985). Children living with parents in
ery, 1988), but the effects of adolescents homes with severe marital conflict appear
may be more enduring (Kalter, Riemer, to be protected if there is a good relation-
Brickman, & Chen, 1985; Wallerstein, ship with one of the parents (Rutter, 1990).
1984, 1985; Wallerstein, Corbin, & Lewis, Another salient protective factor impli-
1988). cated in these reports is belief in one's ef-
Given the prevalence of divorce, it is im- fectiveness, particularly in terms of self-
portant to examine more thoroughly the efficacy and self-confidence (Garmezy,
possibility of age and sex differences across 1985; Rutter, 1979b; Werner, 1990). Self-
children and across time. From the prem- efficacy may function protectively by moti-
ises of developmental psychopathology vating attempts at adaptation, in contrast
(Masten & Bras well, in press), one would to the passivity that often accompanies de-
predict that vulnerability and resilience will spair (Bandura, 1982, 1986). Self-efficacy
shift over time. Careful longitudinal stud- increases as a result of mastery experiences;
ies of change in family structure, parent in turn, feelings of self-efficacy increase the
behavior, and child behavior have the po- likelihood of instrumental behavior. Resil-
tential to illuminate processes related to ient children may enter a situation more
developmental tasks, cognitive reinterpre- prepared for effective action by virtue of
tation of past events, and changing devel- their self-confidence; subsequently, suc-
opmental needs for parenting by each sex cessful mastery of a difficult situation
parent for boys and girls. would be expected to increase self-efficacy
and reinforce efforts to take action.
Self-efficacy processes may underlie the
Protective factors in children under stress
phenomenon of "required helpfulness"
Many other diverse and scattered reports (Rachman, 1979). It has been observed that
of children's responses to severe or chronic resilient high-risk children sometimes have
danger or loss have been summarized by been called upon to help the family or com-
432 A. S. Masten, K. M. Best, and N. Garmezy

munity during adversity (Bleuler, 1978; El- and measures of intelligent behaviors, in-
der, 1974; Garmezy, 1985; Werner, 1990). cluding multifaceted IQ tests, also serve as
Elder (1974) reported that children of the markers of the capacity for adaptation un-
Great Depression may have benefited from der adversity.
the need for their assistance; children ral- Internal locus of control also has been
lied to contribute to the family welfare. associated with competence under stress
Effective parents may generate or in- (Nowicki & Strickland, 1973; O'Grady &
crease self-efficacy in multiple ways by Metz, 1987; Werner & Smith, 1982). Al-
mechanisms suggested by Bandura (1982) though internal locus of control may be the
and others (Amato & Ochiltree, 1986): they most advantageous stance under most cir-
may model effective action, provide oppor- cumstances, a flexible perspective may be
tunities for their children to experience the most adaptive; rigid internal locus of
mastery, and verbally persuade their chil- control in the face of uncontrollable devas-
dren of the children's own effectiveness. tation may make individuals vulnerable to
Attachment theory offers another per- emotional and somatic distress. In norma-
spective on the processes by which respon- tive situations, the importance of flexibility
sive parenting may lead to self-efficacy may not be evident.
(Bowlby, 1969; Sroufe, 1979): good care The functional significance of individual
leads the child to view the self as a lovable, differences in temperament also have been
worthwhile person; it provides a "secure considered in studies of stress resistance or
base" for exploration, thereby increasing stress vulnerability. Suomi (1987) argued
the likelihood for mastery experiences and that both human infants and rhesus mon-
fostering the development of autonomy. keys vary substantially and with intraindi-
Of course, such parenting is likely to be vidual consistency during ontogeny in their
provided by parents who also foster self- reactions to environmental novelty or chal-
efficacy in other ways, described earlier. lenges. The "high-reactive" monkeys Suomi
Faith and religious beliefs may also studied show characteristic physiological
function in analogous ways. Religion may responses to challenge and distinctive be-
enhance positive beliefs about the self. haviors that change with development.
Concomitantly, religion provides social These reactions are elicited by threatening
connectedness derived from abstract rela- or challenging situations; no differences
tionships with religious protective figures are evident in familiar, nonthreatening sit-
and concrete relationships provided by sup- uations. Individual differences in stress re-
portive members within a religious commu- activity could constitute a vulnerability or
nity. protective factor, depending on the situa-
Know-how, intellectual skills, or social tion.
cognitive abilities are often correlated with The role of context has not always been
resilience (Garmezy, 1985; Masten, 1989; emphasized in studies of human tempera-
Werner, 1990). Resilient children appear to ment. It has been suggested that "easy"
be resourceful and effective problem- temperament is a protective factor and
solvers. More intellectually able children "difficult" temperament is a vulnerability
often have socioeconomic and educational factor for adaptation to adversity. In their
advantages, as well as more able parents. Kauai study, Werner and Smith (1982)
They may also be effective niche-seekers found that resilient children differed in
(Scarr & McCartney, 1983). They may be temperamental qualities from other high-
able to quickly and accurately discern dan- risk infants who later developed problems:
ger, find safe escape routes or havens, and as indicated earlier, they were described as
locate sources of help. Sternberg (1985) de- more active and good-natured or cuddly.
fined intelligence in terms of the mental ac- Other investigators have found that diffi-
tivity directed at adaptation; studies of re- cult temperament is associated with greater
silient children support the idea that tests abuse in an abusive family and more acci-
Resilience and development 433

dents (Berger, 1985; Carey, 1982; Huttu- become possible: older children can under-
nen & Nyman, 1982). stand that separations are temporary and
Rutter (1989) cautioned that the advan- caregivers return, they usually have more
tages and disadvantages of easy or difficult experience with temporary separations,
temperament may depend on the circum- and they can be prepared in various ways
stances, citing a provocative study of Masai for the strange situation of hospital care.
infants by de Vries (1984), who reported Under high-stress conditions, boys are
that, in the life-threatening environment of reported to show more disruptive or ag-
a severe drought, difficult babies were gressive behavior than girls, and girls more
more likely to survive than easy babies. The anxiety and depression than boys. In gen-
"squeaky-wheel" hypothesis may offer an eral, girls have been described as more re-
explanation: demanding people may re- silient than boys in childhood but as more
ceive the care and resources essential to sur-vulnerable in adolescence (Rutter, 1989;
vival under harsh conditions, de Vries Werner & Smith, 1982). It is not yet clear
(1984) also suggested that "difficult" in- whether these apparent differences reflect
fants may be less of a problem in the Masai constitutional changes in vulnerability,
culture, where assertiveness is a highly val- age-related changes in societal expecta-
ued trait and where large extended families tions, or inaccurate overgeneralizations.
provide many caregivers. Girls do appear to have a biological edge
in early development, as indexed by lower
Age and sex differences. Age and sex dif- mortality rates and lower rates of various
ferences in adaptation have often been birth defects and neurological problems
noted in studies of adaptation to major (Rutter, 1989). They may also be favored
stressors. Older children or adolescents in the social environment of some societies
have been reported to have stronger and or some situations. In the United States,
longer lasting reactions to major disasters young girls may have more same-sex com-
than very young children, at least in the petent role models at hand, as provided by
absence of harm to the primary caregiver the preponderance of female teachers in el-
(Block, Silber, & Perry, 1956; Gleser, ementary school or by the usual custody
Green, & Winget, 1981; Silber, Perry, & arrangements following divorce. In adoles-
Bloch, 1958). Older children may be more cence, societal expectations, sexual pres-
aware of the magnitude and implications sures, and biological changes may turn the
of the situation and be more connected tables so that girls have greater challenges,
to a wide network of people who have fewer resources, or greater vulnerability
been affected by the disaster than would (Masten, 1988; Werner & Smith, 1982). In
younger children. the literature on divorce, it has been sug-
gested that delayed effects of early stress
Vulnerability and resistance to stressful
may be manifested in girls during adoles-
experiences may shift as a function of on-
cence (Kalteret al., 1985).
going developmental changes. Very young
babies manifest less reaction to strange sit- The patterns of age-related sex differ-
uations and to temporary separations from ences in studies of stress are provocatively
caregivers than do older babies. Yet older similar to patterns observed across the liter-
children also show less distress in response ature on psychopathology (Masten & Bras-
to separations. Concomitantly, children well, in press; Rutter, 1985): (a) disorders
6 months to 4 years old show greater dis- that begin in early or middle childhood of-
tress during hospital admissions than do ten show heightened prevalence in males
younger or older children (Rutter, 1985). (e.g., autism, conduct disorder); (b) some
Vulnerability appears to begin with the at- disorders commonly arising in adolescence
tachment to a specific caregiver, typically show a preponderance of females (e.g., an-
arising around 6 months of age. Vulnera- orexia nervosa, depression); and (c) boys
bility wanes as cognitive coping strategies more often show symptoms of aggressive/
434 A. S. Masten, K. M. Best, andN. Garmezy

disruptive behavior across all ages. These dictor of such short-term symptoms: chil-
relatively consistent findings raise impor- dren pinned down several hours during the
tant questions about the linkages among siege near a child who was killed on the
processes of development, adaptation, and playground showed more symptoms than
psychopathology that may be related to sex children trapped inside the school building,
and gender and that need to be examined who showed more symptoms than children
thoroughly in future research. already at home at the time. Children who
knew the victim showed more symptoms;
children in grave danger who did not real-
Resilience as Recovery
ize what was going on had fewer trauma
When adversity is prolonged or severe, re- symptoms than did children aware of the
silience refers to the phenomenon of recov- danger.
ery when the immediate danger or stress Terr (1979) also found few sex and age
recedes, not to invulnerability. Children of differences in her clinical study of children
the Holocaust, children subjected to abuse aged 5-14 who experienced the Chowchilla
and neglect by caregivers, and children who bus kidnapping. Studies with a greater age
have lost a parent attest to the human ca- range, however, have revealed age differ-
pacity for recovery (Anthony & Cohler, ences in the expression of anxiety (Garmezy
1987; Dugan & Coles, 1989; Garmezy, & Rutter, 1985). For example, following
1985; Garmezy & Rutter, 1983; Zimrin, disasters, preschoolers are more likely to
1986). cling and wet the bed than are older chil-
dren.
Acute trauma and recovery Studies of disasters in which some chil-
dren have been separated from parents and
Acute trauma dramatically illustrates the others have not made it clear that the pres-
recovery concept of resilience. By defini- ence of a parent plays a critical role in chil-
tion, a traumatic event is expected to over- dren's responses (Garmezy & Rutter, 1985).
whelm coping resources: no one is expected MacFarlane (1987) reported that children's
to maintain high-level psychological well- separation from their parents immediately
being or competence under conditions of after a devastating Australian bushfire was
extreme threat or in the immediate after- a stronger predictor of trauma symptoms
math of disaster. Behaviors among chil- than was the degree of exposure to danger.
dren during and after life-threatening It is widely reported that the response
events show many similarities to behavior and functioning of parents during and after
reported among adults in analogous situa- a stressor can have a profound effect on
tions, with some age-specific variability child behavior. Several mechanisms may
and characteristics. Responses often in- account for this phenomenon, each of
clude extreme anxiety and hyperarousal, which reflects normative developmental
sleep problems, nightmares, exaggerated processes. One possible explanation is the
startle responses, intrusive images or process of social referencing by which hu-
thoughts, and avoidance of reminders of mans, beginning in infancy, gather impor-
the experience (Eth & Pynoos, 1985; Gar- tant information about the environment.
mezy & Rutter, 1985; Yule & Williams, Infants uncertain of the threat posed by an
1990). object or a situation will look to the mother
Pynoos et al. (1987) found no age or sex for information (how dangerous is this?);
differences in trauma symptoms 1 month if mothers provide a clear signal of danger,
after a terrifying sniper incident experi- such as by their facial expressions, the be-
enced by 5-13-year-old children attending havior of the infant is affected (Feinman,
a school in California. Psychological prox- in press; Hornik & Gunnar, 1988). Analo-
imity to the threat rather than individual gously, it seems reasonable to posit that
differences appeared to be the best pre- parental behavior during disasters will in-
Resilience and development 435

fluence children's interpretations of the sit- (Cicchetti & Carlson, 1989; Curtiss, 1977;
uation and their subsequent behavior: an Lane, 1976; Skuse, 1984a, 1984b) and child
obviously terrified parent is terrifying to a survivors of massive trauma in Nazi con-
child. Children may assimilate observed centration camps (Moskovitz, 1985) and in
coping behaviors, as well. Hesse and Cic- Cambodia during the Pol Pot regime of
chetti (1982) argued that children are likely 1975-1979 (Kinzie, Sack, Angell, Manson,
to assimilate the emotional repertoire, lan- & Rath, 1986). Studies of these children
guage, and coping style of their parents. not only document the possibility of recov-
Perhaps most importantly, trauma may af- ery but also reveal patterns of vulnerability
fect the parent's basic parenting behaviors, and recovery that illuminate developmental
possibly eroding the quality of care and processes.
support available to the child. Because of their rarity, studies of the
Over time, in most cases of acute most extreme and deviant situations of
trauma, symptoms abate (Garmezy & Rut- maltreatment are limited to individual case
ter, 1985). Unless a child's circumstances studies. Studies of these children have pro-
are permanently altered (e.g., the family vided provocative data for scholars of lan-
has been killed) or the child ic permanently guage development, socialization, and de-
altered (e.g., brain injury), recovery is velopmental psychopathology. Although it
likely to occur. Long-term, follow-up stud- is impossible to know the developmental
ies of acute trauma are rare, but the few potential or level of these children prior to
systematic studies, such as the follow-up deprivation, careful study of their recovery
of families who survived the Buffalo Creek has provided clues about human potential
dam disaster (Gleser et al., 1981), suggest for resilience and the limits of resilience in
that recovery is shaped by individual differ- terms of different aspects of development.
ences in the child and variations in the re- Skuse (1984b) summarized the observa-
covery environment (Garmezy & Rutter, tions from nine cases of severe deprivation
1985). Age is one factor: older children reported in the literature. Although the de-
showed more long-term effects after Buf- gree and combination of malnutrition,
falo Creek. The supportive qualities of the abuse, sensory deprivation, and physical
family are another: supportive families restraint varied, as did the duration and age
were associated with better recovery. Four of onset of deprivation, Skuse was able to
years after the Chowchilla incident, Terr draw several compelling conclusions from
(1983) reported that children with troubled these cases. First, it was evident that mo-
families showed more symptoms than chil- tor, cognitive, and social development are
dren with more competent families. severely affected by extremely adverse envi-
Recovery appears to be most difficult ronments. Second, recovery of function in
when acute psychosocial trauma includes these areas, often exponential in rate, pro-
the loss of adequate care or when children vided evidence that many aspects of cogni-
experience horrific acts perpetrated by or tive and social development are robust and
against their attachment figures. Pervasive, self-righting or "canalized" (Waddington,
long-lasting symptoms of Post-Traumatic 1966) and not constrained by critical peri-
Stress Disorder (PTSD) have been reported ods for development. Third, language was
in children who have witnessed the murder, the most vulnerable cognitive function, at
rape, or suicide of a parent (Eth & Pynoos, least of those faculties studied. At discov-
1985; Malmquist, 1986). ery, language (particularly expressive lan-
guage) was severely impaired in all cases.
Total absence of speech or language com-
Maltreatment and recovery prehension at time of discovery was associ-
ated with poor prognosis, perhaps because
Vivid testimony to the phenomena of resil-
organic damage was likely in such cases.
ience and vulnerability in development is Fourth, severe malnutrition was associated
found in studies of maltreated children
436 A. S. Masten, K. M. Best, andN. Garmezy

with limited recovery. Finally, Skuse con- the organization and regulation of the self,
cluded that the crucial ingredient in the re- behavior, and affect, as well as with symp-
covery environment was the loving care of toms that resemble chronic PTSD, espe-
the foster mother. cially in cases of borderline personality dis-
Children subjected to severe but less ex- order (Putman, 1989; van der Kolk, 1987).
treme maltreatment have been studied in Such associations have led investigators to
greater numbers; this allows for more sys- hypothesize that severe, repeated abuse in
tematic and generalizable research. Recent the preschool years is a causal factor in the
investigations of maltreatment conceptual- development of these disorders. This hy-
ized within the framework of develop- pothesis is congruent with recent theoreti-
mental psychopathology illustrate the cal perspectives and data linking maltreat-
power of such research to elucidate both ment in early childhood with problems in
normative and pathological development the attachment relationship and the organi-
and to inform intervention efforts (Cic- zation and regulation of the self (Cicchetti,
chetti, 1990; Cicchetti & Carlson, 1989). 1990).
Based on developmental theory, Cic- Researchers working with maltreated in-
chetti and other investigators have hypoth- fants and toddlers have noted the striking
esized that maltreated children will show attachment behaviors of abused toddlers
difficulties in linguistic and self-develop- (Crittenden & Ainsworth, 1989). In the
ment related to problems in the social/ Strange Situation procedure, the behavior
emotional functioning of the mother-child of a high percentage of abused toddlers
relationship (Cicchetti, 1990). Initial find- does not fit the usual attachment catego-
ings have supported hypothesized linkages ries. Their behavior is often bizarre and
between socioemotional and cognitive/lin- marked by disorganization or inconsis-
guistic development, revealing possible tency. These children meet the criteria for
processes by which maltreatment may in- a new attachment class, "disorganized dis-
fluence development. These findings also oriented" (type "D"), described by Main
indirectly test hypotheses concerning the and Solomon in 1986 (Carlson, Cicchetti,
role of the social environment in normative Barnett, & Braunwald, 1989; Crittenden &
development of language and the self (Cic- Ainsworth, 1989). Although available data
chetti, 1990). are limited, some toddlers with "disorga-
One of the most exciting areas of inquiry nized" attachment appear to become more
emerging from developmental thinking and organized with time (Crittenden & Ains-
research about the effects of trauma in early worth, 1989). Given prolonged, traumatic
childhood concerns the connection among abuse, however, the organization of the
early trauma, the development of the self- self-system may be permanently affected.
system, and personality or PTSD. Develop- Research summarized by Cicchetti (1990)
mentalists studying maltreated toddlers also suggests that maltreated toddlers have
have observed signs of disorganization and difficulties with the integration of the self-
delay in self-development (Cicchetti, 1990; system. Prospective studies of physically
Cicchetti & Carlson, 1989). Concurrently, and sexually abused children, which track
there has been a resurgence of interest in the patterns of development in the self-system
study of adult disorders strongly associated and the social/emotional underpinnings of
both with trauma in early childhood and personality, could reveal specific vulnera-
with disordered functioning of the self- bilities and protective factors for recovery
system. in these children. In turn, the study of re-
"Borderline personality" and "multiple covery patterns has the potential for fur-
personality" long have been associated with ther fine-tuning of hypotheses about the in-
a history of early trauma (Putnam, 1989; tegration and organization of behavior.
van der Kolk, 1987). These disorders are For example, it will be important to study
also associated with marked difficulties in closely the coordination of recovery in so-
Resilience and development 437

cial, emotional, cognitive, and relationship Kaufman and Zigler (1989), who reviewed
areas of functioning. the evidence, estimated the risk of intergen-
Maltreatment by the primary caregiver erational continuation of abuse at around
in early childhood appears to jeopardize 30% —a substantial enough risk that arti-
the organization and development of the cles about good parents who were abused
attachment relationship, the self, and the as children refer to "breaking the cycle of
regulation and integration of emotional, abuse" (e.g., Egeland, Jacobvitz, & Sroufe,
cognitive, motivational, and social behav- 1988). The correlates of good parenting
ior. These are the major tasks of adapta- found in these adults have been summa-
tion in this period of development; signifi- rized by Kaufman and Zigler (1989). These
cant interference in the achievement of correlates, which offer clues to possible
these developmental milestones may com- protective processes for recovery, include
promise future development (Cicchetti, the following: a good relationship with a
1990; Masten & Braswell, in press; Sroufe caregiving adult in childhood, high IQ, spe-
& Rutter, 1984). It is not surprising that cial talents, physical attractiveness, social
early maltreatment is a risk factor for a skills, a supportive spouse, current finan-
wide variety of later adaptational problems cial security, social supports, strong reli-
(Cicchetti & Carlson, 1989; Wolfe, 1987). gious affiliations, positive school experi-
Yet maltreated children do recover. ences, and therapy (p. 139).
Studies of abused infants and preschoolers Studies of institutional rearing also pro-
(Crittenden, 1985; Farber & Egeland, 1987) vide evidence of the differential effects var-
and studies of maternal deprivation or fail- ious aspects of adversity have on domains
ure to thrive (Rutter, 1979; Masten & of development. Careful studies of children
O'Connor, 1989) document that maltreated reared in institutions for part of their lives
infants and children improve when the indicate that cognitive development is ad-
caregiving environment improves. Simi- versely affected by barren (however clean)
larly, Hodges and Tizard (1989a, 1989b) surroundings and facilitated by stimulating
found that the emotional and social adjust- perceptual and linguistic learning environs.
ment of adolescents initially raised in insti- Discontinuous care provided by a series of
tutions (until age 2) reflected their place- caregivers, on the other hand, does not ap-
ment milieu. Children reared in adoptive pear to damage intellectual development,
homes were relatively better adjusted, as at least not in obvious ways, but it may
indicated by a variety of measures, than have important consequences for social
those returned to their biological parents. and emotional development (Hodges & Ti-
Both groups are significantly less well ad- zard, 1989a, 1989b; Rutter, 1979a; Tizard
justed than comparison children who expe- & Hodges, 1978; Tizard & Rees, 1974).
rienced stable home care. These findings These findings suggest that social/emo-
highlight the role of sensitive caregiving in tional and cognitive domains of develop-
development and also the capacity for re- ment and functioning may be vulnerable
covery. to different environmental vicissitudes. In
Many maltreated children become effec- addition to documenting necessary and suf-
tive parents, despite the considerable risk ficient qualities of the caregiving environ-
for intergenerational transmission of abuse. ment, these results have theoretical impli-
The development of good parenting skills cations for understanding the role of
in such children when they have children environmental influences on development
of their own is one of the most important and practical implications for designing
indications of long-term resilience. Parent- programs and policies related to institu-
ing is one of the major tasks of adult devel- tional care.
opment and, theoretically, an area of po- Like maltreated children, women reared
tential vulnerability in children who have in institutions also can become good par-
not internalized good models of parenting. ents, despite their risk for parenting diffi-
438 A. S. Masten, K. M. Best, andN. Garmezy

culties (Rutter, 1990). Quinton and Rutter sity seriously impedes the quality of care-
(1984) found that a supportive spouse (im- giving needed by the child. Severe interfer-
plicitly, a good marriage) was associated ence in parental care can result from lack
with good parenting by women reared in of opportunity to form specific attach-
institutions. ments, as in constantly changing caregiv-
ers, from impaired functioning in caregiv-
Conclusions ers, as when the only available parenting
figure is incapacitated by stress, psychosis
Studies of psychosocial resilience sup- or substance abuse, or from abusive treat-
port the view that human psychological de- ment or neglect by caregivers. For older
velopment is highly buffered and self- children, social connections independent of
righting. Long-lasting effects of childhood caregiving, such as those found in schools
adversity are almost always related to ei- and communities, may compensate for
ther organic damage or severe interference family inadequacies.
in the protective processes embedded in the Children who experience chronic adver-
caregiving system. sity fare better or recover more successfully
Organic deficits, whether caused by in- when (a) they receive good and stable care
jury, toxic substances, or malnutrition, from someone or, in the case of older chil-
may reduce adaptability through multiple dren, when they have a positive relation-
mechanisms. A basic risk mechanism may ship with a competent adult; (b) they are
be learning impairments. Other examples good learners and problem-solvers; (c) they
are reduced problem-solving aptitude, in- are engaging to other people, adult care-
creased emotional lability, or diminished givers for infants, other adults and peers
appeal to adults. for older children; and (d) they have an
Parents and other caregivers function as area of competence and perceived efficacy,
the first environmental protective agents of valued by themselves or society, whether it
development. Under normative conditions, is academic, athletic, artistic, or mechan-
these relationships appear to foster several ical.
protective psychological processes. Parents Compensatory or protective systems or
nurture mastery motivation and self-esteem institutions in a child's life appear to oper-
as well as physical growth. Parents provide ate through mechanisms analogous to pro-
information, learning opportunities, be- tective functions in the family. Schools and
havioral models, and connections to other churches, for example, provide competent
resources. When these transactional pro- role models and nurturing adults who sup-
tective processes are absent or are severely port and reinforce the development of
limited for prolonged periods, a child may skills that improve problem-solving, moti-
be significantly handicapped in subsequent vation, and later socioeconomic opportuni-
adaptation by low self-esteem, inadequate ties. They also impart values congruent
information or social know-how, a disincli- with living in society. Schools provide a va-
nation to learn or interact with the world, riety of arenas for building competence
and a distrust of people as resources. Di- and, concomitantly, self-efficacy and self-
minished parenting effectiveness and lack esteem. Education itself is protective when
of parental monitoring are associated with knowledge and problem-solving skills are
noncompliant behavior in children. A fostered.
child's noncompliant behavior can have se- Resilience is not the province of one sex
rious, long-lasting consequences, contrib- nor does it develop or increase in a linear
uting to the development of unsocialized fashion with age (Horowitz, 1989; Mac-
aggression thats turns other people away, coby, 1983). However, resilience is related
including teachers and peers, or increasing to sex and to development. The vulnerabili-
the probability of accidental injuries. ties and advantages of one sex over the
Resilience is the exception when adver- other may shift with developmentally re-
Resilience and development 439

lated changes in cognition, emotion, and knowledge, however, also bring new fears,
the social environment. They also may vary expectations, and dangers.
with the cultural context. Studies of resilience highlight the impor-
Human cognitive abilities influence tance of context in protective processes. Ef-
problem-solving, the appraisal of experi- fective parents are characterized by the de-
ences, and other facets of adaptation. In- gree to which their behaviors facilitate
fants cannot solve problems by planning, development for a particular child in a
reading, or asking for help, although the given cultural and historical context, not
attachment system usually functions to en- by some fixed set of "good" parenting be-
sure that young children elicit assistance haviors. Similarly, individual behaviors (or
from adults. Infants also may be protected coping strategies) that prove to be protec-
from the stress or consequences of certain tive in some environments may not work in
adversities by limited awareness of the sig- a different context. Variations in tempera-
nificance of events or limited memories of ment and talents may protect the adaptabil-
trauma or limited integration of adverse ex- ity of the species as a whole; the same dis-
periences into cognitive schema (Kagan, position in individuals may be an asset in
1983; Rutter, 1985). Similarly, preschool one culture or situation and a liability in
children may initially show strong reactions others (Rutter, 1989). There are many dif-
to divorce caused by its effects on their par- ferent reasons a child may be engaging to
ents and their care, as well as separation adults: a child may be physically attractive,
anxiety; however, in the long run, pre- talented, socially skilled, cuddly, or simply
schoolers may show fewer sequelae because "my child." There are also many ways to
they do not fully comprehend or integrate experience mastery in interacting with the
the situation into their systems of beliefs environment.
and expectations. Resilience, like adaptation in general, is
Young children cannot suffer the angst always contextual. Therefore, coping ef-
of existential reflections nor can they be forts will be tailored to the situation and
stressed directly by a trusted peer's be- will vary across resilient people. The most
trayal. Young children may be affected by successful "stress inoculation" efforts have
such stressors indirectly, however, if their been tailored to quite specific situations,
caregivers are affected. Betrayal by a fam- such as terrorist attacks (Ayalon, 1983)
ily friend, for example, could affect the or medical procedures (Ferguson, 1979;
mood and caregiving behavior of family Wolfer & Visintainer, 1979). Fostering
members interacting with the child. Young general resilience to adversity is synony-
children are highly dependent on care for mous with facilitating psychological devel-
protection and, therefore, they are vulnera- opment.
ble to declines in the quality of care.
With age, generally, the psychological
Future directions for research
buffering systems for human development
increasingly come under the control of the The first wave of resilience research has
individual (Cicchetti & Schneider-Rosen, provided an important but limited under-
1986; Scarr & McCartney, 1983). This standing of variation in children's adapta-
allows older children and adolescents in ad- tion to vicissitudes. A fairly congruent pic-
verse family environments to seek protec- ture of resilience has emerged, but it is
tive relationships outside the immediate en- painted in broad strokes. It is the task of
vironment and to seek niches better suited future investigators to portray resilience in
to their development. Over time, as a func- research questions that shift from the
tion of maturation and experience, children "what" questions of description to the
develop new physical and thinking skills, "how" questions of underlying processes
and new social resources to draw upon in that influence adaptation. The need to shift
meeting challenges. Increasing control and to focusing on risk and protective processes
440 A. S. Masten, K. M. Best, andN. Garmezy

rather than factors has been convincingly and personality formation — has and will
articulated by Rutter (1985, 1987, 1990). continue to illuminate understanding of re-
Close examination of resilience patterns silience. Studies of development in nonnor-
is needed to understand the effects of ad- mative situations, particularly those heav-
versity on different aspects of development ily laden with challenge, have and will
and to understand the relations of develop- provide data on the necessary and suffi-
ment to adaptation. Developmental level, cient qualities of caregiving environments,
individual differences, relationships, the the significant functions of parenting and
nature of challenges, and the qualities of other normative protective processes, the
the recovery environment will influence the range and variation of human responses to
adaptation of children exposed to adverse challenge, the self-righting properties of
experiences. Specific links between charac- development and constraints on those
teristics of adversity and developmental properties, and the linkages or indepen-
sequelae rarely have been made; the poten- dence among different aspects of cognitive,
tial of studies that explore such links is emotional, and social development.
illustrated by the results from studies of The study of resilience has restored im-
maltreated children and institutional rear- portant and once neglected pieces to the
ing. puzzle of human adaptation. The greatest
Basic psychological research — designed gains, however, will come from integrating
to delineate the normative development of these pieces into the full context of human
specific aspects of interpersonal relation- development and the diversity of environ-
ships, mastery motivation, problem-solving ments in which such development inevita-
skills, cognitive/emotional development, bly must proceed.

References
Amato, P. R., & Ochiltree, G. (1986). Family re- Bleuler, M. (1984). Different forms of childhood
sources and the development of child competence. stress and patterns of adult psychiatric outcome.
Journal of Marriage and the Family, 48, 47-56. In N. F. Watt, E. J. Anthony, & J. E. Rolf (Eds.),
Anthony, E. J. (1974). The syndrome of the psycho- Children at risk for schizophrenia (pp. 537-542).
logically invulnerable child. In E. J. Anthony & Cambridge: Cambridge University Press.
C. Koupernik (Eds.), The child in his family: Chil- Bloch, D. A., Silber, A., & Perry, S. E. (1956). Some
dren at psychiatric risk (pp. 529-545). New York: factors in the emotional reaction of children to
Wiley. disaster. American Journal of Psychiatry, 113,
Anthony, E. J., & Cohler, B. J. (Eds.). (1987). The 416-422.
invulnerable child. New York: Guilford. Block, J., Block, J. H., & Gjerde, P. F. (1988). Paren-
Ayalon, O. (1983). Coping with terrorism: The Israeli tal functioning and the home environment in fami-
case. In D. Mechenbaum & M. E. Jaremko (Eds.), lies of divorce: Prospective and concurrent analy-
Stress reduction and prevention (pp. 293-339). ses. Journal of the American Academy of Child
New York: Plenum. and Adolescent Psychiatry, 27, 207-213.
Baldwin, A. L., Baldwin, C , & Cole, R. E. (1990). Block, J. H., Block, J., & Gjerde, P. F. (1986). The
Stress-resistant families and stress-resistant chil- personality of children prior to divorce: A pro-
dren. In J. Rolf, A. S. Masten, D. Cicchetti, spective study. Child Development, 57, 827-840.
K. H. Neuchterlein, & S. Weintraub (Eds.), Risk Bowlby, J. (1969). Attachment and loss. Vol. I. At-
and protective factors in the development ofpsy- tachment. New York: Basic Books.
chopathology (pp. 257-280). New York: Cam- Brown, G. W., & Harris, T. O. (1978). Social origins
bridge University Press. of depression. New York: The Free Press.
Bandura, A. (1982). Self-efficacy mechanism in hu- Brown, G. W., Harris, T. O., & Bifulco, A. (1986).
man agency. American Psychologist, 37, 122-147. Long-term effects of early loss of parent. In M.
Bandura, A. (1986). Social foundations of thought Rutter, C. E. Izard, & P. B. Read (Eds.), Depres-
and action. Englewood Cliffs, NJ: Prentice-Hall. sion in young people: Developmental and clinical
Berger, M. (1985). Temperament and individual dif- perspectives (pp. 251-296). New York: Guilford.
ferences. In M. Rutter & L. Hersov (Eds.), Child Carey, W. B. (1982). Clinical use of temperament data
and adolescent psychiatry: Modern approaches in pediatrics. In R. Porter & G. M. Collins (Eds.),
(pp. 3-16). Oxford: Blackwell Scientific Publica- Temperamental differences in infants and young
tions. children (pp. 191-201). London: Pitman.
Bleuler, M. (1978). The schizophrenic disorders: Carlson, J., Cicchetti, D., Barnett, D., & Braunwald,
Long-term patient and family studies. New Haven: K. G. (1989). Finding order in disorganization:
Yale University Press. Lessons from research on maltreated infants' at-
Resilience and development 441

tachments to their caregivers. In D. Cicchetti & among abused and neglected children. In E. J. An-
V. Carlson (Eds.), Child maltreatment: Theory thony & B. Cohler (Eds.), The invulnerable child
and research on the causes and consequences of (pp. 253-288). New York: Guilford.
child abuse and neglect (pp. 494-528). Cambridge: Farran, D. C , & McKinney, J. D. (Eds.). (1986). Risk
Cambridge University Press. in intellectual and psychosocial development. Or-
Cicchetti, D. (1984). The emergence of developmental lando, FL: Academic.
psychopathology. Child Development, 55, 1-7. Feinman, S. (in press). Social referencing and the so-
Cicchetti, D. (1989). Developmental psychopathol- cial construction of reality in infancy. New York:
ogy: Past, present, and future. In D. Cicchetti Plenum.
(Ed.). The emergence of a discipline: Rochester Felsman, J. K., & Vaillant, G. E. (1987). Resilient
Symposium on Developmental Psychopathology children as adults: A 40-year study. In E. J. An-
(Vol. 1, pp. 1-12). Hillsdale, NJ: Erlbaum. thony & B. J. Cohler (Eds.), The invulnerable
Cicchetti, D. (1990). The organization and coherence child (pp. 289-314). New York: Guilford.
of socioemotional, cognitive, and representational Ferguson, B. F. (1979). Preparing young children for
development: Illustrations through a develop- hospitalization: A comparison of two methods.
mental psychopathology perspective on Down syn- Pediatrics, 64, 656-664.
drome and child maltreatment. In R. Thompson Fisher, L., Kokes, R. F., Cole, R. E., Perkins, P. M.,
(Ed.), Socioemotional development. Nebraska & Wynne, L. C. (1987). Competent children at
Symposium on Motivation (pp. 259-366). Lincoln: risk: A study of well-functioning offspring of dis-
University of Nebraska Press. turbed parents. In E. J. Anthony & B. J. Cohler
Cicchetti, D..& Carlson, V. (Eds.). (1989). Child mal- (Eds.), The invulnerable child (pp. 211-228). New
treatment: Theory and research on the causes and York: Guilford.
consequences of child abuse and neglect. New Garmezy, N. (1971). Vulnerability research and the
York: Cambridge University Press. issue of primary prevention. American Journal of
Cicchetti, D., & Schneider-Rosen, K. (1986). An orga- Orthopsychiatry, 41, 101-116.
nizational approach to childhood depression. In Garmezy, N. (1974a). Children at risk: The search for
M. Rutter, C. Izard, & P. B. Read (Eds.), Depres- the antecedents of schizophrenia. Part II: Ongoing
sion in young people: Developmental and clinical research programs, issues, and interventions.
perspectives (pp. 71-134). New York: Guilford Schizophrenia Bulletin, 9, 55-125.
Press. Garmezy, N. (1974b). The study of competence in
Comer, J. P. (1988). Maggie's American Dream. New children at risk for severe psychopathology. In
York: New American Library. E. J. Anthony & C. Koupernik (Eds.), The child
Crittenden, P. M. (1985). Maltreated infants: Vulner- in his family. Vol. 3: Children at psychiatric risk
ability and resilience. Journal of Child Psychology (pp. 77-97). New York: Wiley.
and Psychiatry, 26, 85-96. Garmezy, N. (1985). Stress-resistant children: The
Crittenden, P. M., & Ainsworth, M. D. S. (1989). search for protective factors. In J. E. Stevenson
Child maltreatment and attachment theory. In D. (Ed.), Recent research in developmental psycho-
Cicchetti & V. Carlson (Eds.). Child maltreat- pathology. Journal of Child Psychology and Psy-
ment: Theory and research on the causes and con- chiatry Book (Suppl. No. 4, pp. 213-233). Oxford:
sequences of child abuse and neglect (pp. 432- Pergamon Press.
463). Cambridge: Cambridge University Press. Garmezy, N", & Masten, A. S. (in press). The protec-
Curtiss, S. (1977). Genie: A psycholinguists study of tive role of competence indicators in children at
a modern-day "Wild Child." New York: Academic risk. In E. M. Cummings (Ed.), Life-span develop-
Press. mental psychology: Perspectives on stress and cop-
de Vries, M. W. (1984). Temperament and infant ing. Hillsdale, NJ: Erlbaum.
mortality among the Masai of East Africa. Ameri- Garmezy, N., & Phipps-Yonas, S. (1984). An early
can Journal of Psychiatry, 141, 1189-1194. crossroad in research on risk for schizophrenia:
Demo, D. H., & Acock, A. C. (1988). The impact of The Dorado Beach Conference. In N. F. Watt,
divorce on children. Journal of Marriage and the E. J. Anthony, L. C. Wynne, & J. E. Rolf (Eds.),
Family, 50, 619-648. Children at risk for schizophrenia: A longitudinal
Dugan, T. F., & Coles, R. (Eds.). (1989). The child in perspective (pp. 6-18). Cambridge: Cambridge
our times: Studies in the development of resiliency. University Press.
New York: Brunner/Mazel. Garmezy, N., & Rutter, M. (1983). Stress, coping, and
Egeland, B., Jacobvitz, D., & Sroufe, L. A. (1988). development in children. New York: McGraw-Hill.
Breaking the cycle of abuse. Child Development, Garmezy, N., & Rutter, M. (1985). Acute reactions to
59, 1080-1088. stress. In M. Rutter & L. Hersov (Eds.), Child
Elder, G. H. (1974). Children of the Great Depres- psychiatry: Modern approaches (2nd ed., pp. 152-
sion: Social change in life experience. Chicago: 176). Oxford: Blackwell Scientific.
University of Chicago Press. Garmezy, N., & Streitman, S. (1974). Children at risk:
Elder, G. H., Jr., Nguyen, T. V., & Caspi, A. (1985). The search for the antecedents of schizophrenia.
Linking family hardship to children's lives. Child Part I: Conceptual models and research methods.
Development, 56, 361-375. Schizophrenia Bulletin, 8, 14-90.
Emery, R. E. (1988). Marriage, divorce, and children's Gleser, G. C , Green, B. L., & Winget, C. (1981).
adjustment (Vol. 14). Newbury Park, CA: Sage. Prolonged psychosocial effects of disaster: A
Eth, S., & Pynoos, R. S. (1985). Post-traumatic stress study of Buffalo Creek. New York: Academic.
disorder in children. Washington, DC: American Gottesman, 1.1., & Shields, J. (1982). Schizophrenia:
Psychological Association Press. The epigenetic puzzle. New York: Cambridge Uni-
Farber, E. A., & Egeland, B. (1987). Invulnerability versity Press.
442 A. S. Masten, K. M. Best, and TV. Garmezy

Hesse, P., & Cicchetti, D. (1982). Perspectives on an Lane, H. (1976). The wild boy of Aveyron. Cam-
integrated theory of emotional development. New bridge, MA: Harvard University Press.
Directions for Child Development, 16, 3-48. Lempers, J. D., Clark-Lempers, D., & Simons, R. L.
Heston, L. L. (1960). Psychiatric disorders in foster (1989). Economic hardship, parenting, and dis-
home reared children of schizophrenic mothers. tress in adolescence. Child Development, 60, 25-
British Journal of Psychiatry, 112, 819-825. 39.
Hetherington, E. M., Stanley-Hagan, M., & Ander- Long, J. V. F., & Vaillant, G. E. (1984). Natural his-
son, E. R. (1989). Marital transitions: A child's tory of male psychological health, XI: Escape
perspective. American Psychologist, 44, 303-312. from the underclass. American Journal of Psychi-
Hirschi, T., & Hindelang, M. J. (1977). Intelligence atry, 141, 341-346.
and delinquency: A revisionist review. American Maccoby, E. E. (1983). Social-emotional development
Sociological Review, 42, 571-587. and response to stressors. In N. Garmezy & M.
Hodges, J., & Tizard, B. (1989a). IQ and behavioral Rutter (Eds.), Stress, coping, and development in
adjustment of ex-institutional adolescents. Journal children (pp. 217-234). New York: McGraw-Hill.
of Child Psychology and Psychiatry, 30, 53-75. Malmquist, C. P. (1986). Children who witness paren-
Hodges, J., & Tizard, B. (1989b). Social and family tal murder: Posttraumatic aspects. Journal of the
relationships of ex-institutional adolescents. Jour- American Academy of Child Psychiatry, 25, 320-
nal of Child Psychology and Psychiatry, 30, 77- 325.
97. Masten, A. S. (1988). Toward a developmental psy-
Horowitz, F. D. (April, 1989). The concept of risk: A chopathology of early adolescence. In M. D. Lev-
re-evaluation (Invited address). Presented at Soci- ine & E. R. McAnarney (Eds.), Early adolescent
ety for Research in Child Development, Kansas transitions (pp. 261-278). Lexington, MA: D. C.
City, MO. Heath.
Hornik, R., & Gunnar, M. R. (1988). A descriptive Masten, A. S. (1989). Resilience in development: Im-
analysis of infant social referencing. Child Devel- plications of the study of successful adaptation for
opment, 59, 626-634. developmental psychopathology. In D. Cicchetti
Huesmann, L. R., Eron, L. D., & Yarmel, P. W. (Ed.), The emergence of a discipline: Rochester
(1987). Intellectual functioning and aggression. Symposium on Developmental Psychopathology
Journal of Personality and Social Psychology, 52, (Vol. 1, pp. 261-294). Hillsdale, NJ: Erlbaum.
232-240. Masten, A. S., & Braswell, L. (in press). Develop-
Huttunen, M. O., & Nyman, G. (1982). On the conti- mental psychopathology: An integrative frame-
nuity, change and clinical value of infant tempera- work for understanding behavior problems in chil-
ment in a prospective epidemiological study. In dren and adolescents. In P. R. Martin (Ed.),
R. Porter & G. M. Collins (Eds.), Temperamental Handbook of behavior therapy and psychological
differences in infants and young children (pp. 240- science: An integrative approach. New York: Per-
246). London: Pitman. gamon.
Ianni, F. A. J. (1989). The search for structure. New Masten, A. S., & Garmezy, N. (1985). Risk, vulnera-
York: The Free Press. bility, and protective factors in developmental psy-
Kagan, J. (1983). Stress and coping in early develop- chopathology. In B. B. Lahey & A. E. Kazdin
ment. In N. Garmezy & M. Rutter (Eds.), Stress, (Eds.), Advances in clinical child psychology (Vol.
coping, and development in children (pp. 191— 8, pp. 1-51). New York: Plenum.
216). New York: McGraw-Hill. Masten, A. S., Morison, P., Pellegrini, D., & Tel-
Kalter, N., Riemer, B., Brickman, A., & Chen, J. W. legen, A. (1990). Competence under stress: Risk
(1985). Implications of parental divorce for female and protective factors. In J. Rolf, A. S. Masten,
development. Journal of the American Academy D. Cicchetti, K. H. Neuchterlein, & S. Weintraub
of Child Psychiatry, 24, 538-544. (Eds.), Risk and protective factors in the develop-
Kandel, E., Mednick, S. A., Kirkegaard-Sorensen, L., ment of psychopathology (pp. 236-256). New
Hutchings, B., Knop, J., Rosenberg, R., & Schul- York: Cambridge University Press.
singer, F. (1988). IQ as a protective factor for sub- Masten, A. S., & O'Connor, M. J. (1989). Vulnerabil-
jects at high risk for antisocial behavior. Journal ity, stress, and resilience in early development of a
of Consulting and Clinical Psychology, 56, 224- high risk child. Journal of the American Academy
226. of Child and Adolescent Psychiatry, 28, 274-278.
Kaufmam, J., & Zigler, E. (1989). The intergenera- McFarlane, A. C. (1987). Posttraumatic phenomena
tional transmission of child abuse. In D. Cicchetti in a longitudinal study of children following a nat-
& V. Carlson (Eds.), Child maltreatment: Theory ural disaster. Journal of the American Academy
and research on the causes and consequences of of Child and Adolescent Psychiatry, 26, 764-769.
child abuse and neglect {pp. 129-150). Cambridge: Moskovitz, S. (1985). Longitudinal follow-up of child
Cambridge University Press. survivors of the Holocaust. Journal of the Ameri-
Kinzie, J. D., Sack, W. H., Angell, R. H., Manson, can Academy of Child Psychiatry, 24, 401-407.
S., & Rath, B. (1986). The psychiatric effects of Murphy, L. B., & Moriarty, A. E. (1976). Vulnerabil-
massive trauma on Cambodian children: I. The ity, coping, and growth: From infancy to adoles-
children. Journal of the American Academy of cence. New Haven, CT: Yale University Press.
Child Psychiatry, 25, 370-376. Nowicki, S., Jr., & Strickland, B. R. (1973). A locus
Kopp, C. B., & Krakow, J. B. (1983). The develop- of control scale for children. Journal of Consult-
mentalist and the study of biological risk: A view ing and Clinical Psychology, 40, 148-154.
of the past with an eye toward the future. Child O'Dougherty, M., Wright, F. S., Garmezy, N.,
Development, 54, 1086-1108. Loewenson, R. B., & Torres, F. (1983). Later com-
Resilience and development 443

petence and adaptation in infants who survive se- tive mechanisms. In J. Rolf, A. S. Masten, D.
vere heart defects. Child Development, 54, 1129- Cicchetti, K. H. Neuchterlein, & S. Weintraub
1142. (Eds.), Risk and protective factors in the develop-
O'Grady, D., & Metz, J. R. (1987). Resilience in chil- ment of psychopathology (pp. 181-214). New
dren at high risk for psychological disorder. Jour- York: Cambridge University Press.
nal of Pediatric Psychology, 12, 3-23. Rutter, M., Maughan, B., Mortimore, P., Ouston,
Pargament, K. I. (August, 1987). God help me: To- J., & Smith, A. (1979). Fifteen thousand hours:
wards a theoretical framework of coping for the Secondary school and their effects on children.
psychology of religion. Presented at the American Cambridge, MA: Harvard University Press.
Psychological Association, New York. Rutter, M., & Quinton, D. (1984). Long term follow-
Pasamanick, B., & Knobloch, H. (1960). Brain dam- up of women institutionalized in childhood: Fac-
age and reproductive casualty. American Journal tors promoting good functioning in adult life. Brit-
ofOrthopsychiatry, 30, 298-305. ish Journal of Developmental Psychology, 18,
Pasamanick, B., & Knobloch, H. (1961). Epidemio- 225-234.
logical studies on the complications of pregnancy Sameroff, A. J., & Chandler, M. J. (1975). Reproduc-
and the birth process. In G. Caplan (Ed.), Preven- tive risk and the continuum of caretaking casualty.
tion of mental disorders in children: Initial explor- In F. D. Horowitz, E. M. Hetherington, S. Scarr-
ations (pp. 74-94). New York: Basic Books. Salapatek, & G. M. Siegel, (Eds.), Review of child
Peterson, J. L., & Zill, N. (1986). Marital disruption, development research (Vol. 4, pp. 187-243). Chi-
parent-child relationships, and behavior problems cago: University of Chicago Press.
in children. Journal of Marriage and the Family, Sameroff, A. J., & Seifer, R. (1990). Early contribu-
48, 295-307. tors to developmental risk. In J. Rolf, A. S. Mas-
Pinkney, A. (1987). Black Americans (3rd ed.). New ten, D. Cicchetti, K. H. Nuechterlein, & S. Wein-
York: Prentice Hall. traub (Eds.), Risk and protective factors in the
Putnam, F. W. (1989). Diagnosis and treatment of development of psychopathology (pp. 52-66).
multiple personality disorder. New York: Guil- New York: Cambridge University Press.
ford. Scarr, S., & McCartney, K. (1983). How people make
Pynoos, R. S., Frederick, C , Nader, K., Arroyo, W., their own environments: A theory of genotype
Steinberg, A., Eth, S., Nunez, F., & Fairbanks, "RA" environmental effects. Child Development,
L. (1987). Life threat and posttraumatic stress in 54, 424-435.
school-age children. Archives of General Psychia- Silber, E., Perry, S., & Bloch, D. (1958). Patterns of
try, 44, 1057-1063. parent-child interaction in a disaster. Psychiatry,
Rachman, S. J. (1979). The concept of required help- 21, 159-167.
fulness. Behavior Research and Therapy, 17, 1-6. Skuse, D. H. (1984a). Extreme deprivation in early
Radke-Yarrow, M., & Sherman, T. (1990). Hard childhood —I. Diverse outcomes for three siblings
growing: Children who survive. In J. Rolf, A. S. from an extraordinary family. Journal of Child
Masten, D. Cicchetti, K. H. Nuechterlein, & S. Psychology and Psychiatry, 25, 523-541.
Weintraub (Eds.), Risk and protective factors in Skuse, D. (1984b). Extreme deprivation in early child-
the development of psychopathology (pp. 97-119). hood-II. Theoretical issues and a comparative re-
New York: Cambridge University Press. view . Journal of Child Psychology and Psychiatry,
Rodnick, E. H., & Goldstein, M. J. (1974). Premorbid 25, 543-572.
adjustment and the recovery of the mothering Sroufe, L. A. (1979). The coherence of individual de-
function in acute schizophrenic women. Journal velopment: Early care, attachment, and subse-
of Abnormal Psychology, 83, 623-628. quent developmental issues. American Psycholo-
Rosenblatt, R. (1983). Children of war. Garden City, gist, 34, 834-841.
NY: Anchor Press/Doubleday. Sroufe, L. A., & Rutter, M. (1984). The domain of
Rutter, M. (1979a). Maternal deprivation, 1972-1978: developmental psychopathology. Child Develop-
New findings, new concepts, new approaches. ment, 55, 17-29.
Child Development, 50, 283-305. Sternberg, R. J. (1985). BeyondIQ: A triarchic theory
Rutter, M. (1979b). Protective factors in children's of human intelligence. Cambridge: Cambridge
responses to stress and disadvantage. In M. W. University Press.
Kent & J. E. Rolf (Eds.), Primary prevention of Suomi, S. J. (1987). Genetic and maternal contribu-
psychopathology, Vol. 3: Social competence in tions to individual differences in Rhesus monkey
children (pp. 49-74). Hanover, NH: University biobehavioral development. In N. A. Krasnegor,
Press of New England. E. M. Blass, M. A. Hofer, & W. P. Smotherman,
Rutter, M. (1985). Resilience in the face of adversity: (Eds.), Perinatal development: A psychobiological
Protective factors and resistance to psychiatric perspective (pp. 397-417). New York: Academic
disorder. British Journal of Psychiatry, 147, 598- Press.
611. Terr, L. C. (1979). Children of Chowchilla: A study
Rutter, M. (1987). Psychosocial resilience and protec- of psychic trauma. Psychoanalytical Study of the
tive mechanisms. American Journal of Orthopsy- Child, 34, 552-623.
chiatry, 57, 316-331. Terr, L. C. (1983). Chowchilla revisited: The effects
Rutter, M. (1989). Temperament: Conceptual issues of psychic trauma four years after a school-bus
and clinical implications. In G. A. Kohnstamm, kidnapping. American Journal of Psychiatry, 140,
J. E. Bates,&M. K. Rothbart (Eds.), Temperament 1543-1550.
in childhood (pp. 463-479). New York: Wiley. Tizard, B., & Hodges, J. (1978). The effect of early
Rutter, M. (1990). Psychosocial resilience and protec- institutional rearing on the development of eight
444 A. S. Masten, K. M. Best, and N. Garmezy

year old children. Journal of Child Psychology J. E. (Eds.). (1984). Children at risk for schizo-
and Psychiatry, 19, 99-118. phrenia: A longitudinal perspective. Cambridge:
Tizard, B., & Rees, J. (1974). A comparison of the Cambridge University Press.
effects of adoption, restoration to the natural Werner, E. E. (1989). Children of the garden island.
mother, and continued institutionalization on the Scientific American, April, 106-111.
cognitive development of four-year-old children. Werner, E. E. (1990). Protective factors and individ-
Child Development, 45, 92-99. ual resilience. In S. J. Meisels & M. Shonkoff
van der Kolk, B. A. (1987). Psychological trauma. (Eds.), Handbook of early intervention. New
Washington, DC: American Psychological Associ- York: Cambridge University Press.
ation Press. Werner, E. E., & Smith, R. S. (1982). Vulnerable but
Waddington, C. H. (1966). Principles of development invincible: A study of resilient children. New
and differentiation. New York: Macmillan. York: McGraw-Hill.
Wallerstein, J. S. (1984). Children of divorce: Prelimi- West, D. J., & Farrington, D. P. (1973). Who be-
nary report of a ten-year follow-up of young chil- comes delinquent? London: Heinemann.
dren. American Journal of Orthopsychiatry, 54, White, J. L., Moffitt, T. E., & Silva, P. A. (1989). A
444-458. prospective replication of the protective effects of
Wallerstein, J. S. (1985). Children of divorce: Prelimi- 1Q in subjects at high risk for juvenile delinquency.
nary report of a ten-year follow-up of older chil- Journal of Consulting and Clinical Psychology,
dren and adolescents. Journal of the American 57,719-724.
Academy of Child Psychiatry, 24, 545-553. Wolfe, D. A. (1987). Child abuse. Newbury Park,
Wallerstein, J. S., Corbin, S. B., & Lewis, J. M. CA: Sage.
(1988). Children of divorce: A 10-year study. In Wolfer, J. A., & Visintainer, M. A. (1979). Prehospi-
E. M. Hetherington & J. D. Arasteh (Eds.), Im- tal preparation for tonsillectomy patients: Effects
pact of divorce, single parenting, and stepparent- on children's and parents' adjustment. Pediatrics,
ing on children (pp. 197-214). Hillsdale, NJ: Erl- 64, 646-655.
baum. Yule, W., & Williams, R. M. (1990). Post-traumatic
Wallerstein, J. S., & Kelly, J. B. (1980). Surviving stress reactions in children. Journal of Traumatic
the breakup: How children and parents cope with Stress, 3, 279-295.
divorce. New York: Basic Books. Zimrin, H. (1986). A profile of survival. Child Abuse
Watt, N. F., Anthony, E. J., Wynne, L. C , & Rolf, and Neglect, 10, 339-349.