Early research into resilience began by examining the types of factors that enabled children and young people to achieve positive adaptation and positive social adjustment despite being considered ‘high–risk’ or ‘at risk’ (Garmezy and Streitman, 1974; Garmezy et al., 1984; Luthar et al., 2000; Werner and Smith, 1977, 1992). Werner and Smith (1977), for example, found the resilient children in their study differed from their non–resilient peers in a number of ways.
They found that most of the resilient children grew up in a family with a maximum of three other siblings (four children in total) and that there were at least two years between the resilient child and any other siblings (Werner and Smith, 1992: 56). The resilient children had not experienced any prolonged separation fom a primary caregiver in their first year of life, and they formed a close bond with one or more caregivers who could be either parental or ‘substitute parents’, for example a grandparent or older sibling (Werner and Smith, 1992: 56).
As infants they had temperaments that “elicited positive attention”, were considered active, affectionate (females), and good-natured (males), and also had “fewer eating and sleeping habits that distressed their parents” (Werner and Smith, 1992: 56).
In difference to their low–risk peers, the high–risk resilient group tended to withdraw from troubled relationships with parents, but it could be argued the low–risk group did not need to as they did not experience the same problems with their relationships with their parents. Many of the children, by nature of being from high–risk families, had parents who divorced, had illnesses, or lived in households with ‘chronic family discord’ (Werner and Smith, 1992: 65). The resilient children tended to cope by
their high–risk peers who continued to be involved (Werner and Smith, 1992: 65).
Despite this the resilient cohort’s family played an important role in the positive outcomes of many of the resilient children, despite contributing to or causing their high–risk status. Educational level of an opposite–sex parent was strongly associated with positive adaptation in adulthood (Werner and Smith, 1992: 177). The males tended to have older fathers, had more positive interactions with caregivers, and had higher ratings of family stability (Werner and Smith, 1992: 179).
In addition to family, community also played an important role in the resilient cohort’s positive adaption. Resilience was associated with having additional caring adults, including grandparents, uncles and aunts, neigh- bours, parents of partners and boy– or girlfriends, youth leaders, church leaders, and, in adolescence, teachers (Garmezy and Streitman, 1974: 64; Werner and Smith, 1992: 178).
These factors all suggest that, despite their adverse beginnings, these resilient children learned how to establish and maintain important social relationships with family, peers, and elders who they are able to draw on for support and encouragement. This is in contrast to their peers who were high–risk but not resilient, who struggled to achieve positive outcomes and who may have lacked the social skills necessary to form such bonds.
In cases where children had mental health problems the ability of the family to obtain psychological and psychiatric support from professional and community services—and the knowledge of the existence of such support—helped the children to manage their condition (Garmezy and Streitman, 1974: 63; Werner and Smith, 1977: 216).
Perhaps one of the most powerful protective factors among the resilient young people was “. . . faith that life made sense, [and] that the odds could be overcome” (Werner and Smith, 1992: 177), an ‘internal locus of control’. Locus of control and competence had a positive effect even on
high–risk youths with more severe needs, such as learning disability or long–term mental health problems:
The degree to which youth had faith in the effectiveness of their own actions was related not only to the effectiveness with which they used their intellectual resources in scholas- tic achievement but also to positive change in behaviour in adolescence. An internal locus of control was a significant correlate of improvement (Werner and Smith, 1977: 220)
Work ethic (‘hard work’ and ‘persistence’) was mentioned by young people with more severe mental health issues in childhood who later improved (Werner and Smith, 1977: 221).
Werner and Smith argued that one of the biggest differences separating the high–achieving and low–achieving high–risk individuals was their goal setting and aspirations. Career and employment success was the most important goal for the resilient cohort but the lowest priority for their non–resilient peers (Werner and Smith, 1992: 69).
The resilient cohort faced many of the same difficulties as their high–risk peers, but took opportunities when presented with them to recover from these difficulties. These opportunities presented at ‘major life transitiions’, and included marriage or entering a long–term committed relationship, the birth of a child, employment and establishment of a career, graduating from high school, going to and graduating from college (university), joining the military, and becoming an active member of a church group (Werner and Smith, 1992: 178). The authors contrasted the resilient cohort’s focus on taking opportunities with their high–risk peers who instead discussed life events that limited opportunities, including divorce or the break–up of a long–term relationship, the death of a parent (women), and moving away from home (men) (Werner and Smith, 1992: 178).
The resilient cohort overall had greater work satisfaction, measured by self–rated satisfaction with work or school achievement at age 31 or 32 when asked in a structured interview or self–completion questionnaire
(Werner and Smith, 1992: 179). The resilient cohort had higher self–rated satisfaction with their state of life (Werner and Smith, 1992: 181). In adulthood the resilient cohort tended to have positive relationships with their parents–in–law or the parents of their long–term partner, and many resilient women in particular sought emotional support from their parents–in–law (Werner and Smith, 1992: 66). The resilient cohort had more ‘satisfying’ relationships with their siblings as adults based on self– rated responses, and this was most notable among siblings who had alcoholic or mentally ill parents (Werner and Smith, 1992: 67). They also had more satisfying relationships with parents, spouses or partners, and children at age 31 or 32 (Werner and Smith, 1992: 180).
Their relationships with friends were more complex; they shared a similar number of friends with their high–risk peers and had similar satisfaction with their relationships, but tended to be more self–reliant and rely on friends less for financial support and counsel than their high–risk peers (Werner and Smith, 1992: 68–69). As outlined above these resilient individuals tended to have an internal locus of control and be more self– confident, so it is perhaps not surprising that they relied on themselves more to address problems. In addition, they may simply have had greater financial resources as a result of their employment—which tended to be of a higher grade—or simply be better at managing their own money.
2.4.1
Sex and gender
The authors noted that resilient girls tended to have increased autonomy and responsibility in households where the mother worked and the father was absent, for example by providing care to younger siblings (Werner and Smith, 1992: 57). For women, having a mother who had steady employment also had positive results (Werner and Smith, 1992: 177). Resilient women were significantly more likely to have had “regular household chores and domestic responsiblity during adolescence” while
resilient men had higher self–rated temperament and activity scores (Werner and Smith, 1992: 177). It seems unlikely that such a result is genetic given what is known about gendered role profiles but, nevertheless, perhaps the resilient young people found their lives somehow easier if they conformed to these gendered expectations.
Resilient boys tended to have a positive male role model, although this was not necessarily the child’s father (Werner and Smith, 1992: 57). Both resilient boys and resilient girls had additional role models outside of the family, including close friends, teachers, neighbours, youth leaders, ministers or faith leaders, or elders (Werner and Smith, 1992: 57). In females, internal protective factors—self–esteem, for example—had the biggest effect on resilience. For males, outside sources of support—for example from caregivers, friends, and family—had the biggest effect on resilience.
2.4.2
Development stage
Because of the longitudinal nature of these studies the authors were able to explore which factors affected the participants’s resilience at any given developmental stage. By comparing groups of children and young people with different socio-economic and familial circumstances in early life and comparing their trajectories into adulthood these studies were able to explore which factors led to positive outcomes.
In the 1972 follow–up Werner and Smith were successful in tracking down 88% of the original Kauai cohort. This included an ‘at–risk’ and a control group of young people, matched for age, sex, socio-economic status, and ethnicity. In the 1985 follow–up the researchers managed to obtain responses from 82% of the original cohort, for which data was available at birth, infancy, age two, age 10, and age 18 (Werner and Smith, 1992: 34). All participants were surveyed for education and health outcomes, ability, achievement, and personality using standard instruments contemporary
for the time (Werner and Smith, 1977: 24).
Werner and Smith (1992) suggest that during infancy and early childhood, constitutional factors such as health and temperament played the biggest part in effecting resilience. This changed as the resilient children matured, and by middle school their verbal and reasoning skills played a bigger part in their positive development. By late adolescence and adulthood their personality characteristics—self–esteem and internal locus of control— most helped to reinforce their resilience and positive adaptation (Werner and Smith, 1992: 57).
At age two the resilient cohort displayed alertness and autonomy, sought out experiences, had a ‘positive social orientation’, and had better com- munication, locomotion, and self–help skills than their high–risk but non–resilient peers (Werner and Smith, 1992: 56). The coping style of the primary caregiver at age two, observed by psychologists and paedia- tricians, was linked with positive adaptation, as was the presence of rules and structure in the household.
In infancy the resilient group had good sleeping and eating habits, and were considered ‘affectionate and cuddly’ (girls) or ‘very active’ (boys) (Werner and Smith, 1992: 173). At age two assessments by paediatricians and psychologists found the resilient children to be more agreeable, relaxed, responsive, self–confident, and sociable. In comparison, their high–risk non–resilient peers were characterised by anxiety, fearfulness and suspicion and were more frequently withdrawn (Werner and Smith, 1992: 176). At age ten, teachers (for boys) and parents (for girls) noted fewer be- havioural problems, and at age 17 and 18 the resilient cohort enjoyed greater popularity among their peers (Werner and Smith, 1992: 176). In grade four (approximate age nine to ten) the resilient children had higher reading achievement scores, especially among the boys (Werner and Smith, 1992: 176). In elementary school the resilient children got along well with classmates, had better reasoning skills, better reading skills, and had many interests including “activities and hobbies that were
not narrowly sex–typed” (Werner and Smith, 1992: 56).
Between age 10 to follow–up at age 17–18 ‘perception of parental under- standing’, peer support, the young person’s belief in their own abilities, hard work, persistence, and ability to communicate in the first language of the island (‘standard English’) were associated with improvement and positive change (Werner and Smith, 1977: 216).
The authors compared characteristics of these high–risk children who had positive outcomes with high–risk children who did not fare so well by ages 10 and 18, matched for age and sex. These young people by contrast had learning problems, mental health problems, and ‘serious delinquencies’ (Werner and Smith, 1992: 56).
In senior year of high school (approximate age 18) the resilient group considered their school experience to be more positive and had higher— and more realistic—expectations for their future (Werner and Smith, 1992: 176). In addition, their interviewers considered them to have higher self–esteem.
In adulthood (ages 31 and 32) the resilient group had lower self–rated distress and emotionality using the EAS temperament survey instrument, and women had higher self–rated sociability and lower anger (Werner and Smith, 1992: 176). The authors also found a significant association between the resilient groups’ problem solving skills (PMA IQ) at age ten and successful adaptation in adulthood (Werner and Smith, 1992: 176).
2.4.3
Validation
Using discriminant function analysis, Werner and Smith were able to enter these protective factors chronologically into their model. In 94.4% of cases they were able to identify the correct, resilient, individuals by entering all protective factors. The authors were still able to identify the majority—87.5%—of individuals correctly by entering protective factors the cohort were exposed to between birth and age two (Werner and Smith,
1992: 182–183). Crucially the success rate was similar, even higher, for high–risk non–resilient peers (96.8%, all protective factors) suggesting validity in the measures (Werner and Smith, 1992: 183).