Available online at www.sciencedirect.com

ScienceDirect Comprehensive Psychiatry 55 (2014) 388 – 395 www.elsevier.com/locate/comppsych

Psychometric properties of the Creole Haitian version of the Resilience Scale amongst child and adolescent survivors of the 2010 earthquake Jude Mary Cénat⁎, Daniel Derivois Center of Research in Psychopathology and Clinical Psychology (CRPPC), Psychology Institute, University Lumière Lyon 2, France

Abstract Resilience is defined as the capacity of human beings to deal with and adapt to adversity, suffering, tragedy or other traumatic event. This study aims to investigate psychometric properties and the underlying structure of the Creole version of the RS among children and adolescents survivors to the 2010 Haitian earthquake. A total of 872 children and adolescents exposed to the earthquake with an average age of 14.91 (SD = 1.94) completed the Creole version of RS, the Impact Event Scale-Revised, the Children Depression Inventory and the Social Support Questionnaire-6. The current validity of RS and the internal consistency were investigated; sex, age, religion and others sociodemographic variables differences were also analysed. Cronbach’s alpha coefficient for the RS was .77; the split-half coefficient was .72. The goodness-of-fit for the 5-factor model presents the best adjusted indices. The total resilience score was correlated positively with social support (r = .42, p b .01). Mean score of the RS was 131.46 (SD = 21.01). No significant differences were observed about sex, age and residential municipality. The results showed that the Haitian Creole version of RS is a valid and reliable measure in assessing resilience for the children and adolescent survivors to the 2010’s earthquake. © 2014 Elsevier Inc. All rights reserved.

1. Introduction At 16:52 local time on January 12th, 2010, a violent seism of 7.0 magnitude on the Richter scale hit the towns of Portau-Prince, Jacmel and Léogâne in Haiti, the largest since its Independence after the one of May 7th, 1842. Following the earthquake, the American Geological Institute recorded 52 aftershocks of a magnitude equal to, or greater than 4.5. The earthquake affected about half of the Haitian population (5 million people), made more than 600,000 people move and shortly after the seism at least 1.3 million people lived on makeshift camps [1]. Losses are estimated at 7804 billion American dollars, the equivalent to just over the Haitian national GDP in 2009. Over 105,000 private and public buildings have been completely destroyed, and more than 208,000 houses have suffered important damage. The 2010 report of the United Nations Development Programme (UNDP) [2] confirmed more than 222,000 died, over ⁎ Corresponding author. Center of Research in Psychopathology and Clinical Psychology (CRPPC), Psychology institute, University of Lyon 2, France. Tel.: +33 787 69 84 17. E-mail addresses: [email protected] (J.M. Cénat), [email protected] (D. Derivois). 0010-440X/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.comppsych.2013.09.008

300,000 were wounded and there was between 4000 and 7000 amputees. This earthquake not only caused the deaths of around 2.5% of the national population, but also engendered major changes in the lives of the country's inhabitants — raising levels of poverty and instability. Numerous children and adolescents were injured, others are now disabled; many have lost one or both of their parents; some of them witnessed the death of their parents; large numbers have been living in camps since 12 January 2010; many of them no longer go to school. Moreover, they are now exposed to the constant acts of violence which are rife in these camps. This is the grim picture of the problems that these children and adolescents are facing. The harsh realities which they must overcome and from which they must find the necessary psychological strength to bounce back, in a socially and physically hostile environment, are dramatic. Numerous studies on children who have suffered traumatic events and survived in extremely difficult conditions have shown that a good number of them were able to develop well and live positive adult lives [3,4]. However, as stated by Rutter, when several adversities and stresses coexist, the risk of mental health problems in these children increases, even if some of them prove resilient.

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

Resilience is defined by Newman [5] as being the capacity of human beings to deal with and adapt to adversity, suffering, tragedy, all manner of trauma and other stress factors in their daily lives. Various studies on resilience [6–8] stress that this goes beyond dealing with and adapting to traumatic experiences, in the sense of being able to absorb or resist them; but that it is, above all, the capacity to bounce back and develop in a positive way in the face of adverse conditions. Rutter [4,9,10] having carried out longitudinal studies on children who have suffered abuse during childhood, defines resilience as an interactive, dynamic process between the risk factors and the factors of internal and external protection of the individual that enable him/her to deal with stressful life events. This reflects work that we have previously undertaken, in which we define resilience as a psycho-socio-environmental process, i.e., an interaction between the risk and protective factors stemming from the psychological reality and the personality of the individual and the domestic, social and environmental factors of the subject’s development [11]. Wagnild & Young [12] define the five principle components of resilience as: a balanced perspective on life and previous experiences (equanimity); the capacity to carry on regardless of events and adverse conditions (perseverance); the awareness of our own strengths and weaknesses, (self-belief or self-reliance); having a goal in life (meaningfulness); recognition of our individuality; i.e. a sense of uniqueness and acceptance of life (existential aloneness). Numerous instruments are used to measure the various characteristics of resilience using children, adolescents, adults and as a group. Békaert, Masclet & Caron [13] list the following: the Resilience Scale for Adolescents (READ); the Connor-Davidson Resilience Scale (CDRISC); the Adolescent Resilient Scale (ARS); the Resiliency Scale and the Resilience Scale (RS) as instruments for evaluating resilience in adolescents who have experienced traumatic events. 1.1. Resilience Scale (RS) The Resilience Scale has been translated into more languages than any other resilience evaluation instrument [13]: German, Swedish, Russian, Spanish, Dutch, Italian, Japanese, French and others languages — and many studies have evaluated the psychometric properties of this scale [14–20]. The Resilience Scale was developed by Wagnild & Young [21]. It was based on work carried out with 24 women who had experienced traumatic events and who had managed, over time, to overcome these adversities. Although the five components of resilience (given above) were included in the construction of this instrument, the exploratory factor analysis gave rise to two factors (Personal competence, 17 items and Acceptance of self and life, 8 items). In studies carried out prior to this, the numbers of identified factors vary from 2 to 5 depending on the sample.

389

Studies carried out on the internal consistency of the Resilience Scale have shown a Cronbach alpha (α) ranging from .76 to .94 [16,20]. Very few studies have evaluated the test–retest reliability. The Nygren and collaborators [14] study in Sweden showed a reliability coefficient of .78 whereas that of Lei and collaborators [20], amongst student survivors of the Wenchuan earthquake in China in 2008, showed a test–retest reliability coefficient of .82. The Haitian Creole version was produced using the WHO (World Health Organization) standard process, supplied by the Scale’s authors. A first translation was submitted for review by a panel of five experts from the Haitian State University where Creole is the mother-tongue, this was then subject to a back-translation by an independent expert, of English mother-tongue, who had no knowledge of the RS. Finally, pre-testing (with 24 participants), cognitive interviewing (with 8 participants) and discussion with the 12 investigators through which we found significant change two words of two different items in order to make easier the understanding of the subjects. The steps described above have been carried out with our Haitian partners to produce the final creole version of the RS used in this research carried two and a half years after the 2010 earthquake in Haiti. This study is part of the Lyon 2 University project ANR10-HAIT-002 RECREAHVI, led by D. Derivois. It follows on from research work carried out on traumatic events, post-traumatic stress disorder, resilience and the creative processes of Haitian child and adolescent victims of natural disasters, and in particular, of the earthquake on 12 January 2010. The aim of this paper is to evaluate the psychometric properties of the Resilience Scale among the Haitian child and adolescent survivors of the earthquake on 12 January 2010, and to clarify the underlying structure of the Creole version of the Resilience Scale. This will enable us to provide an instrument specifically designed for Haitian professionals working in the field of mental health — who currently use those validated by other countries. It will also enable us to evaluate levels of resilience by age, gender and the link between resilience and the range of social support available to the children and adolescents who survived the 2010 quake in Haiti, and who have experienced multiple traumatic events. 2. Method 2.1. Participants The participants of this study conducted two years and a half after the earthquake of 2010 are 872 children and adolescents; including 491 are girls (56.30%) with an average age of 14.91 (SD = 1.94). They were chosen from 12 schools scattered in all Port-au-Prince. The study includes also street children who do not attend school (6.42%). The inclusion criteria in the group were as follows: (i) they were aged between 7 and 17; (ii) they were present in one of the

390

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

most seriously affected zones of Port-au-Prince during the quake, (iii) they had not received any psychological treatment, and (iv) the informed consent form had been signed by one of their parents. 2.2. Procedure The study was conducted in accordance with a protocol approved by the Institutional Review Boards of Ministry of Public Health and Population, Ministry of National Education and Professional Formation, Ministry of Social Affairs and State University of Haiti. The study was conducted between June and July 2012, at 12 schools and 2 homeless children’s centres in the Haitian capital and its surrounding areas. The children and adolescents we studied came from several municipalities in the capital that were badly damaged during the quake. Other questionnaires were administered by door-to-door canvassing in several districts of the capital. The parents of each child were asked to sign an informed consent form guaranteeing the confidentiality of this study. For the street children in the homeless centres, however, these were signed by the centre directors. Of 872 participants, 202 (23.16%) come from Port-au-Prince downtown municipality, 173 (19.84%) from Carrefour, 151 (17.32) from Delmas, 106 (12.15) from Petion-Ville, 105 (12.04%) from Tabarre, 79 (9.06%) from Cité Soleil and 56 (6.42%) street children who often walk through several municipalities. Twelve graduating psychology students from the Human Science Faculty of the Haiti State University (one of the project partners) administered the questionnaires. These students were all experienced in the administration of questionnaires and had been given a day’s training on preparing the study. 2.3. Instruments All the instruments used for this research are self-reported questionnaires. But, about the filling of the questionnaires, the investigators read themselves the items for all participants and completed to facilitate their understanding. Our first step was to collect socio-demographic data relating to age, gender, level of education, family situation and parental occupation; to the damage inflicted by the earthquake on the child’s family; to family members who were killed during the quake and to other factors. French versions of these questionnaires were used following a cross-cultural analysis in order to identify any words liable to shock or cause offence in a given cultural context. The Resilience Scale was the only instrument translated into Haitian Creole. 2.3.1. The Resilience Scale (RS) The Resilience Scale contains 25 items and respondents are asked to indicate their level of agreement with each one; ranging from 1 (strongly disagree) to 7 (strongly agree). The total score varies from 25 to 175 with 6 stages going from

very low (25 to 100) to a very high resilience score (161 to 175). As stated above, work carried out on the psychometric properties of this scale has demonstrated the sound reliability of the instrument. Developed for adults, this instrument has been used with children and adolescents [22]. Békaert collaborators [13] and Waaktaar & Torgersen [23] consider it to be one of the most widely used for the evaluation of resilience — in particular with children and adolescents, and for which the psychometric properties are the most valid. 2.3.2. The Social Support Questionnaire (SSQ) Social support was evaluated using the Social Support Questionnaire (SSQ-6) developed by Sarason and colleagues [24] — which is an abridged version of 6 items of the Social Support Questionnaire [25]. Two scores are calculated based on six situations presented to the subject. There are two parts to each situation: the first part evaluates the number of people who are available to the individual, these are the odd numbers (1, 3, 5, 7, 9 and 11) and it scores the perceived availability of support. The second part evaluates the level of satisfaction with each perceived availability of support using a 6-point Likert scale going from very dissatisfied to very satisfied. The level of satisfaction is evaluated by the even numbers (2, 4, 6, 8, 10 and 12). This scale presents sound internal consistence (Cronbach alpha = .90 to .93), or from .83 to .89 for the level of satisfaction and from .95 to .98 for the perceived availability of support. For the purposes of this study, we used the French version [26,27], which seemed appropriate for Haitian culture. 2.3.3. Child Depression Inventory Inspired by the Beck Depression Inventory [28], the CDI (Child Depression Inventory) scale [29,30] is used with children and adolescents aged between 7 and 17 [31]. It contains 27 items which evaluate several factors of depression (negativity, relationship problems, ineffectiveness, anhedonia, low self-esteem). For our research, we used the French version [32].

2.3.4. Impact of Event Scale-Revised (IES-R) Following discussions with the research group, we used the IES-R rather than the Children’s revised impact of event scale (CRIES-13), which we deemed more culturally appropriate to our population of children and adolescents. The IES-R contains 3 sub-scales and 22 items [33]: Avoidance (8 items), Intrusion (8 items) and Hyperarousal (6 items). Each item presents a 5-point Likert scale (not at all, a little, moderately, a lot, enormously) scored from 0 to 4. The seriousness of PTSD symptoms is evaluated by a global score which varies from 0 to 88. This questionnaire is widely used and has demonstrated sound reliability and internal consistency with alpha coefficients ranging from .78 to .89 [33–36]. The French language version that we used for this research presents alpha coefficients from .81 to .93 [35].

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

2.4. Data analyses All statistical analysis was carried out using the Statistical Package for Social Science (SPSS) — version 19 and version 6 of Statistica. We carried out t tests to examine the difference between age and gender for the resilience scale. First, we established descriptive statistics item by item (medium, median and interquartile - IQR) to calculate the prevalence of resilience characteristics amongst the children and adolescents in our sample. Bivariate correlation analyses were used to identify the links between the resilience scale and other scales. Exploratory and confirmatory factoral analyses were carried out. We used the scree test method [36] in order to explore the various dimensions of the RS. We applied an extraction of principle components and a varimax rotation. We considered saturations above .30. Finally, internal coherence (Cronbach alpha) was tested.

3. Results For the 25 items in the Resilience Scale, the average score was 131.46 (SD = 21.01). The scores obtained by partici-

391

pants range between 47 and 175. For the Social Support Questionnaire (SSQ), scores for the sub-scale for availability of social support (N), vary from 0 to 34 with an average score of 10.11 (SD = 4.82); for the satisfaction sub-scale, the average score is 23.27 (SD = 8.90) and scores varied from 0 to 36. The following score distribution was found amongst the Haitian child and adolescent earthquake survivors: very low: 7.2%; low: 15.7%; mod. low: 24.9%; mod. high: 22.7%; high: 21.8% and very high: 7.6%. The descriptive analyses carried out item by item are presented in Table 1. The Creole version for each item is presented in the same table. The t-test for independent samples did not show significant differences according to gender. The average RS score was 132.56 (SD = 20.25) for boys and 130.60 for girls, t (870 = 1.37, p = 0,171). Similar observation for age: no significant difference was observed between children and adolescents aged 7 to 13 (group 1) and those aged 14 to 17 (group 2). The average resilience score for group 1 is 131.79 (SD = 20.95) whereas for group 2, this is 131.36 (SD = 21.04), t (870) = 0.26, p = .80. A t-test analysis of independent samples was also carried out factor by factor which did not reveal any significant differences. There were no differences observed based on whether the child had been

Table 1 Median, inter-quartile range (IQR) and distribution in % for single items for the total group (N = 872). The items are given in English and Creole. Items

Median 1 (IQR) (%)

2 3 4 (%) (%) (%)

5 6 (%) (%)

1. When I make plans, I follow through with them. Lèm gen yon pwojè, mfè tout sam kapap poum reyalize l. 2. I usually manage one way or another. Mwen toujou jwenn yon fason poum rezoud yon pwoblèm. 3. I am able to depend on myself more than anyone else. M pran abitid plis konte sou tèt mwen pase sou lòt moun. 4. Keeping interested in things is important to me. Toujou rete kirye, sa enpòtan anpil pou mwen. 5. I can be on my own if I have to. Mwen ka debrouye m pou kont mwen si m ta gen bezwen. 6. I feel proud that I have accomplished things in life. Mwen santi m fyè de sa m reyalize nan lavim. 7. I usually take things in stride. Mwen pran lavi a jan li vini. 8. I am friends with myself. Mwen santi m byen ak tèt mwen. 9. I feel that I can handle many things at a time. Mwen santi m ka mennen plizyè aktivite an menm tan. 10. I am determined. Mwen se yon moun ki konn kote li vle ale. 11. I seldom wonder what the point of it all is. Se ra pou m mande: « Sa sa ap sèvi ? » 12. I take things one day at a time. Jou mwen wè a se li ki pa m. 13. I can get through difficult times because I've experienced difficulty before. Mwen pase maladi, mwen konn remèd. 14. I have self-discipline. Mwen konn kijan poum kontwole pwòp tèt mwen. 15. I keep interested in things. Mwen toujou enterese ak sa kap pase nan antouraj. 16. I can usually find something to laugh about. Mwen toujou jwen yon bagay ki pou fèm ri. 17. My belief in myself gets me through hard times. Konfyans mwen gen nan pwòp tèt mwen pèmèt mwen travèse moman difisil yo. 18. In an emergency, I'm someone people can generally rely on. Lè gen ijans, m se moun yo ka konte sou li. 19. I can usually look at a situation in a number of ways. Mwen toujou dispoze gade yon pwoblem sou tout fas li yo. 20. Sometimes I make myself do things whether I want to or not. Sa konn rive mwen pouse m fè sa mwen anvi fè osinon sa m pa anvi fè. 21. My life has meaning. Lavi m gen sans. 22. I do not dwell on things that I can't do anything about. Mwen pa bay tèt mwen pwoblèm pou bagay mwen pa ka chanje. 23. When I'm in a difficult situation, I can usually find my way out of it. Lè m nan sityasyon difisil, mwen toujou jwenn yon solisyon. 24. I have enough energy to do what I have to do. Mwen gen anpil enèji poum fè sa mwen dwe fè. 25. It's okay if there are people who don't like me. Sa pa fè anyen si gen moun ki pa renmen m.

7 (1.0) 7 (2.0) 7 (3.0)

6.9 1.0 1.8 7.7 3.7 7.8 71.1 6.8 0.1 2.8 14.2 7.2 11.6 57.3 12.2 1.3 1.1 10.9 3.9 8.4 62.3

4 4 7 5 7 4 7 4 7 4

(6) (5) (3) (5) (2.5) (5) (3) (4) (3) (4)

29.5 22.2 8.1 22.4 6.9 23.5 7.8 20.5 52.4 20.4

9.1 4.9 1.8 3.1 1.0 3.7 2.8 4.4 7.0 4.1

4.8 4.4 2.6 3.7 2.8 2.9 1.1 2.6 4.9 3.7

22.4 22.6 14.4 17.5 14.3 23.3 15.1 23.7 19.3 24.1

3.2 3.9 27.2 6.4 6.4 33.0 5.4 7.8 59.7 4.6 6.2 42.5 4.9 8.3 61.8 5.8 10.4 30.4 4.6 5.5 63.1 8.0 6.4 34.3 2.9 0.7 12.8 4.8 5.3 37.6

7 6 7 7

(3) (3) (2) (2)

8.5 17.0 6.1 8.8

1.0 2.9 1.4 1.8

3.2 1.8 1.8 1.5

15.1 19.2 13.3 11.0

5.3 9.1 6.0 4.5

7.5 6.0 9.4 8.6

7 (%)

59.4 44.2 62.0 63.8

6 (3) 5 (3)

11.6 2.2 1.8 22.2 7.3 11.8 43.0 15.7 1.7 3.7 24.0 7.8 10.2 36.9

4 (4)

20.4 3.9 3.8 24.3 7.8

8.7 31.1

7 (2) 7 (3)

5.5 .2 1.4 13.9 5.3 16.5 3.3 3.1 14.0 5.4

9.1 64.7 7.2 50.5

7 (2)

6.5 1.6 3.0 13.5 7.9 11.9 55.5

7 (3) 7 (3)

8.8 1.6 2.9 14.9 6.1 10.3 55.4 15.8 2.9 0.9 13.1 3.8 5.3 58.3

392

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

orphaned, religion or losses caused by the quake. We made the same observation about the residential municipality (Table 2).

index). In addition, a comparison of the two models shows a Δx 2(Δdf) = 56,08(7), p b 0,001. It was for this reason that we chose to retain the 5-factor model in Table 3. This table also presents factorial saturations ranging from .34 to .68.

3.1. Exploratory and confirmatory factoral analysis of the Creole version of the RS 25

3.2. Internal consistence of the RS

Principal axes factorisation with a varimax rotation was carried out in order to establish a feasible RS structure for our sample. The Kaiser–Meyer–Olkin measure of sampling adequacy is .93 and the Bartelett (X 2) sphericity test is 2960.34 (df = 300; p b .0001). This is very satisfactory for the purposes of performing factoral analysis with our data. The results showed 8 factors with an eigenvalue higher than 1. However, considering it relevant, we used the scree test method [36] which enabled us to retain two models; the first having three factors and the second having five. Table 3 presents the two models that we will now compare. The 3factor model explains 29.81% of the variance. The 5-factor model presented in Table 2 explains 39.29% of the variance. The goodness-of-fit presented in Table 3 confirms the two models and presents significant X 2’s (p = .000001). However, the goodness-of-fit for the 5-factor model presents the best adjusted indices (Akaike Information Criterion, Bayesian Information Criterion and expected cross-validation

All the correlation coefficients are significant at .0001 between the different factors and vary from .18 and .71. The Cronbach alpha is estimated at .77, the split-half coefficient is .72. The internal consistencies for the 5 dimensions, on the other hand, are below .70. For other instruments, Cronbach's alpha in our sample is also adequate: for the IES-R, Cronbach’s α of .88; for the CDI, Cronbach’s α of.77; for the SSQ-6, Cronbach α is .77 (SSQ-6-N) and .81 (SSQ-6S). Analyses of bivariate correlations between the RS and the other scales used for the study do not really show conclusive correlations, except for the SSQ evaluating social support which correlates very significantly with the RS (see Table 4). 4. Discussion The aim of this study was to evaluate the psychometric properties of the Resilience Scale in order to clarify the underlying structure of the Haitian Creole version amongst

Table 2 Factor analysis with factor loadings and explained variance. Items Equanimity I am determined. (10) I am friends with myself. (8) My life has meaning. (21) I take things one day at a time. (12) When I'm in a difficult situation, I can usually find my way out of it. (23) I have self-discipline. (14) I have enough energy to do what I have to do. (24) Autonomy I feel that I can handle many things at a time. (9) I seldom wonder what the point of it all is. (11) Keeping interested in things is important to me. (4) I can be on my own if I have to. (5) Perseverance When I make plans, I follow through with them. (1) I usually manage one way or another. (2) I feel proud that I have accomplished things in life. (6) Resistance I keep interested in things. (15) I do not dwell on things that I can't do anything about. (22) I usually take things in stride. (7) I can usually find something to laugh about. (16) Sometimes I make myself do things whether I want to or not. (20) I can get through difficult times because I've experienced difficulty before. (13) Self-reliance It's okay if there are people who don't like me. (25) I am able to depend on myself more than anyone else. (3) My belief in myself gets me through hard times. (17) In an emergency, I'm someone people can generally rely on. (18) I can usually look at a situation in a number of ways. (19) Explained variance in percentage

Factor 1

Factor 2

Factor 3

Factor 4

Factor 5

0.675 0.562 0.498 0.474 0.434 0.426 0.380 0.683 0.554 0.551 0.492 0.681 0.672 0.572 0.616 0.542 0.448 0.440 0.389 0.371

17.20

7.25

5.36

4.96

0.654 0.602 0.476 0.357 0.341 4.53

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

393

Table 3 Goodness-of-fit statistics for the two models (n = 872). X2

Models 3-Factor model 5-Factor model

973.61 916.53

df

P

RMSEA

NNFI

CFI

GFI

AIC

BIC

ECVI

272 265

.00⁎ .00⁎

0.05 0.05

0.72 0.73

0.75 0.76

0.92 0.92

1.24 1.19

1.53 1.51

1.24 1.19

⁎ p = .000000.

child and adolescent survivors of the earthquake on 12 January 2010 in Haiti. It was also used to evaluate their resilience scores by age, gender and social support. The results of this study have shown that there are no significant differences between our two age groups. Neither any significant difference between the girls and boys included in our sample. The great majority of studies carried out on resilience have not shown significant differences between genders as a result of natural disasters. This is true in spite of the fact that female gender is considered as a risk factor for the development of PTSD. Moreover, a recent article by Wagnild [37] concerning studies using the Resilience Scale has shown that, of the 12 studies used, only two revealed differences according to age, gender, race and other socio-demographic factors. However, the recent study on adolescents in Wenchuan, China, produced lower RS scores amongst girls than amongst boys. This was previously concluded by Hunter & Chandler [22] in a study on adolescents in secondary education in New England. Child and adolescent survivors of the Haiti earthquake on 12 January scored a higher average on the resilience scale compared to those of the Wenchuan earthquake in China (mean score = 118.77; SD = 20.83) [21]: 131.46 (SD = 21.01). We do not possess enough information to explain this difference, since some factors, such as social support, were not evaluated in the Chinese study. The Resilience Scale has been used in several countries amongst numerous populations. For the purposes of this study, we decided to produce a culturally adapted version of the scale — in spite of its sound reliability and validity. Concerning the structure of the scale; previous studies have presented various possibilities. Although the Wagnild & Young [12] study presented a 2-factor model, confirmed by the first Swedish study with 142 participants [14], studies on a population of 997 students aged 12 to 19 in Brazil,

Table 4 Bivariate correlations between the RS and the other scales.

RS F1 F2 F3 F4 F5

Social Support number

Social Support satisfaction

Depression

PTSD

0.101** 0.084** 0.022 0.098** 0.066 0.073**

0.424** 0.336** 0.197** 0.326** 0.248** 0.341**

−0.045 −0.52 0.005 −0.76** −0. 017 −0.028

−0.087** −0.73* −0.13 −0.64 −0.97** 0.058

*p b .05; **p b .05.

however, presented a 3-factor structure [15]. The Lundman and colleagues [16] study on a larger Swedish population (n = 1,719) used a 5-factor structure — whereas the Chinese study used a 4-factor structure [21]. For the purposes of this study, the confirmatory factorial analysis enabled us to opt for a 5-factor structure. This choice was motivated by the adjustment indices that were better for the 5-factor model in comparison to the 3-factor model (Table 3). The five factors include the 25 items in the questionnaire and three can be labelled according to the basic theory of resilience [4]: factor 1, equanimity; factor 3, perseverance; factor 5, Self-reliance. The two others factors can be labelled: factor 2, autonomy, capacity of selfgovernment, to be independent to decide his life; and factor 4, resistance, ability to put up with adversity and unfortunate event. The reliability coefficients (Cronbach alpha and Splithalf) all exceed .70 for the entire scale. However, this was not the case for the other factors. The reduction in the number of items could explain this reduction in the Cronbach alpha for some factors. The administration of the questionnaire demonstrated sound reliability in the understanding of the items. During a debriefing session, the students administering the questionnaire identified an item which could be better stated using an alternative synonym — one which was more accessible to the population; this was subsequently included in the final questionnaire. Concerning the correlation coefficients between the RS and the other scales used in this research; these are often significant but very low, apart from social support (where correlation is good). This is in contrast with several of the 12 studies cited in works by Wagnild [12,38] as well as that of Lei and colleagues [21]. However, it should be stressed that resilience does not mean the absence of PTSD, as stated by Almedom & Glandon [39]. In their review of 500 articles dealing with resilience and PTSD, they demonstrated that resilience rarely has a negative correlation with PTSD. It strongly implies that resilience is primarily based on psychological, domestic, economic and environmental factors. This work has also enabled us to identify the important role which social support plays in the construction of an individual’s resilience. The correlation is good between the RS and the satisfaction sub-scale in the SSQ. It reflects the works of Peng and colleagues [40] and Piertzak and colleagues [41] and shows that: “it is not simply the availability of friends and relations that counts, but more the quality of the relationships an individual can have with other

394

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395

people and the use made of these relationships” [4, p. 126]. This confirms the role of mutual assistance and social support in the construction of resilience by children. In a recent work, we have already demonstrated the mentoring role played by schools and other social institutions in the development of resilience amongst the Haitian child and adolescent victims of natural disasters [11]. This study was also subject to some constraints, the first being the lack of data concerning the development of resilience amongst child and adolescent survivors of the quake. It would have been useful to see how this had evolved during the period spanning one to 30 months after the quake. This could have shed light onto the mental processes, the strategies for coping and the social factors used for the reconstruction of these children and adolescents facing and living in adversity. However, in addition to the points of interest mentioned above, it should be stressed that the standard procedure employed for the translation and cultural adaptation of the Resilience Scale, with all the experts involved in the process, has made it possible to produce an adapted version with valid content — which did not present any problems with the administration of the questionnaire. This meant that we were able to provide the professionals and native researchers working with children and adolescents in Haiti with the first resilience evaluation questionnaire to demonstrate reliable and valid psychometric properties. We also observed that the child and adolescent survivors of the 12 January 2010 earthquake in Haiti have the necessary mental and social resources to cope with the multiple traumas caused by this quake. In addition to the high average score, we note that 52.1% of them presented a score going from mod. high to very high (131 to 175). This would appear high considering the massive damage caused by the quake and that the majority of children and adolescents in our sample saw dead bodies lying in the streets, lost family members or even watched them die and had friends killed during the quake. This research has therefore made it possible to explore the psychometric properties of the RS, which would appear to be reliable and valid. The five factors emerging from the factoral analysis correspond well with the questionnaire’s basic hypotheses [12]. This research has also allowed us to assess the capacity of Haitian children and adolescents to cope with traumatic events and bounce back after a natural disaster. Acknowledgment This research and the project ANR-10-HAIT-002 RECREAHVI were funded by the National Research Agency (ANR) of France. We thank members of the research team at University Lyon 2 (Lisbeth Brolles, Gildas Bika, Nathalie Guillier-Pasut, Min-sung Kim, Christo Zafimaharo, Laura Coursol, Amira Karray-Khemiri, Bernard Chouvier), at the State University of Haiti including

investigators (Ronald Jean Jacques, Marjory Mathieu), at Haitian Bible Society (Pierre Michel and Marie Carme Derivois). We wish acknowledge Gail Wagnild, the author of RS for comments on a draft of this manuscript and Nicolas Fieulaine for critics of the statistical analyzes.

References [1] International Organization for Migration (IOM). 2010 report. New York: Author; 2010. [2] United Nations Development Programme (UNDP). 2010 report. New York: Author; 2010. [3] Werner EE. High risk children in young adulthood: a longitudinal study from birth to 32 years. Am J Orthopsychiatry 1989;59:72-81. [4] Rutter M. Resilience in the face of adversity: protective factors and resistance to psychiatric disorders. Br J Psychiatry 1985;147:598-611. [5] Newman R. APA’s resilience initiative. Prof Psychol Res Pract 2005;36:227-9. [6] Cyrulnik B. Le murmure des fantômes. France, Paris: Odile Jacob; 2003. [7] Anaut M. Résiliences familiales ou familles résilientes ? Reliance 2006;19:14-7. [8] Earvolino-Ramirez M. Resilience: a concept analysis. Nurs Forum 2007;42:73-82. [9] Rutter M. Resilience concepts and findings: implications for family therapy. J Fam Ther 1999;21:119-44. [10] Rutter M. Proceeding from observed correlation to causal inference: the use of natural experiments. Perspect Psychol Sci 2007;2: 377-95. [11] Cénat JM, Derivois D, Mérisier GG. Ecole et Résilience chez les enfants et adolescents dans l’Haïti post-séisme (School and resilience among children and adolescents during the post-earthquake of Haiti). Revue Québécoise de Psychologie 2013;34:189-01. [12] Wagnild GM, Young HM. Development and psychometric evaluation of the Resilience Scale. J Nurs Meas 1993;1:165-78. [13] Békaert J, Masclet G, Caron R. Les instruments de mesure de la résilience chez les adolescents ayant été confrontés à un traumatisme: une revue de la littérature. Ann Med Psychol 2011;169:510-6. [14] Nygren B, Randstrom KB, Lejonklou AK, Lundman B. Reliabilityand validity of a Swedish language version of the Resilience Scale. J Nurs Meas 2004;12:169-78. [15] Pesce RP, Assis SG, Avanci JQ, Santos NC, Malaquias JV, Carvalhaes R. Cross-cultural adaptation, reliability and validity of the Resilience Scale. Cad Saude Publica 2005;21:436-48. [16] Lundman B, Strandberg G, Eisemann M, Gustafson Y, Brulin C. Psychometric properties of the Swedish version of the Resilience Scale. Scand J Caring Sci 2007;21:229-37. [17] Nishi D, Uehara R, Kondo M, Matsuoka Y. Reliability and validity of the Japanese version of the Resilience Scale and its short version. BMC Res Notes 2010;3:310, http://dx.doi.org/10.1186/1756-0500-3-310 [http://www.biomedcentral.com/1756-0500/3/310]. [18] Heilemann MV, Lee KA, Kury FS. Psychometric evaluation of Spanish version of the Resilience Scale. J Nurs Meas 2003;11:61-72. [19] Portzky M, Wagnild G, De Bacquer D, Audenaert K. Psychometric evaluation of the Dutch Resilience Scale RS-nl on 3265 healthy participants: a confirmation of the association between age and resilience found with the Swedish version. Scand J Caring Sci 2010;24: 86-92. [20] Lei M, Li C, Xiao X, Qiu J, Dai Y, Zhang Q. Evaluation of the psychometric properties of the Chinese version of the Resilience Scale in Wenchuan earthquake survivors. Compr Psychiatry 2012;53: 616-22. [21] Wagnild G, Young HM. Resilience among older women. J Nurs Sch 1990;22:252-5.

J.M. Cénat, D. Derivois / Comprehensive Psychiatry 55 (2014) 388–395 [22] Hunter AJ, Chandler GE. Adolescent resilience. Image J Nurs Sch 1999;31:243-7. [23] Waaktaar T, Torgersen S. How resilient are resilience scales? The Big Five scales outperform resilience scales in predicting adjustment in adolescents. Scand J Psychol 2010;51:157-63. [24] Sarason IG, Sarason BR, Shearin EN, Pierce RG. A brief mesure of social support: practical and theoretical implications. J Soc Pers Relat 1987;4:497-510. [25] Sarason IG, Levine HM, Basham RB, Sarason BR. Assessing social support: the Social Support Questionnaire. J Pers Soc Psychol 1983;44:127-39. [26] Racle N, Bruchon-Schweitzer M, Sarason IG. The short form of Sarason’s Social Support Questionnaire (SSQ6). A French adaptation and validation. Psychol Rep 2005;97:195-202. [27] Bruchon-Schweitzer M, Racle N, Gélie F, Fortie C, Sifakis Y, Constant A. Le questionnaire de soutien social de Sarason (SSQ6). Une adaptation française. Psychol Fr 2003;48:41-54. [28] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561-71. [29] Kovacs M. The Children’s Depression Inventory (CDI). Psychol Bull 1985;21:995-8. [30] Kovacs M. Manual for the Children’s Depression Inventory. New York, North Tonawanda: Multi-Health Systems; 1992. [31] Kovacs MK. Children ́s Depression Inventory. Canada, Toronto: Multi-Health Inc.; 1981. [32] Mack C, Moor L. Versions françaises d'échelles d'évaluation de la dépression, Les échelles CDI et ISC de Maria Kovacs. Neuropsychiatr Enfance 1982;30:627-52.

395

[33] Weiss DS, Marmar CR. The Impact of Event Scale-Revised. In: Wilson J, & Keane TM, editors. Assessing psychological trauma and PTSD. New York: The Guilford Press; 1997. p. 399-411. [34] Creamer M, Bell R, Failla S. Psychometric properties of the Impact of Event Scale-Revised. Behav Res Ther 2003;41:1489-96. [35] Brunet A, St-Hilaire A, Jehel L, King S. Validation of a French version of the Impact of Event Scale-Revised. Can J Psychiatry 2003; 48:56-61. [36] Jens B, Heidi S, Harald G, Claus S, Karl-Heinz L. The Impact of Event Scale-Revised: evaluation of the subscales and correlations to psychophysiological startle response patterns in survivors of a lifethreatening cardiac event. An analysis of 129 patients with an implanted cardioverter defibrillator. J Affect Disord 2004;82:29-41. [37] Cattell RB. The scree test for the number of factors. Multivar Behav Res 1966;1:245-76. [38] Wagnild GM. A review of the Resilience Scale. J Nurs Meas 2009;17:105-13. [39] Almedom AM, Glandon D. Resilience is not the absence of PTSD anymore than health is the absence of disease. J Loss Trauma 2007;12: 127-43. [40] Peng L, Zhang J, Li M, Li P, Zhang Y, Zuo X, et al. Negative life events and mental health of Chinese medical students: the effect of resilience, personality and social support. Psychiatry Res 2012;196:138-41. [41] Piertzak RH, Johnson DC, Goldstein MB, Malleye JC, Rivers AJ, Morgan CA, et al. Psychosocial buffers of traumatic stress, depressive symptoms, and psychosocial difficulties in veterans of Operations Enduring Freedom and Iraqi Freedom: the role of resilience, unit support, and postdeployment social support. J Affect Disord 2010;120:188-92.