Psychological Reports, 2003, 93,1217-1222. © Psychological Reports 2003

CONSTRUCT VALIDITY OF THE ADOLESCENTRESILIENCE SCALE I

ATSUSHIOSHIO

Chubu University

SHINJI NAGAMINE

Kanazawa University

HITO~HI KANEKO

Nagoya University

MOTOYUKINAKAYA

Mie University

Summary.-The aim of this study was to assess the construct validity of the Ado­lescent Resilience Scale which measures the psychological features of resilient individ­uals. Research involving this scale, the Negative Life Events Scale, and the GeneralHealth Questionnaire was conducted with a group of 207 Japanese undergraduatestudents (104 men and 103 women; M age=20.2 yr., SD=.9). A cluster analysis forthe Negative Life Events Scale and General Health Questionnaire yielded three clus­ters: (1) mentally healthy with little experience of Negative Life Events, (2) poorermental health with many experiences of Negative Life Events, (3) mentally healthy de­spite many experiences of Negative Life Events. These three groups were defined as(1) Well Adjusted, (2) Vulnerable, and (3) Resilient, respectively. Mean differences inscores on the Adolescent Resilience Scale among the three groups were subjected toone-way analysis of variance. The mean scores of both the Well Adjusted and Resil­ient groups were higher than that of the Vulnerable group, and therefore support theconstruct validity of the Adolescent Resilience Scale.

Individual development is a dynamic process, involving hardships andcrises. This is certainly true in modern society where we meet many difficultchallenges from many factors such as crime and accidents, natural disasters,social unrest, and financial problems. It is therefore essential to cope withthese negative events to attain a state of personal well-being.

Lately in the field of psychology and psychiatry, a concept called "resil­ience" has received increasing attention. Resilience is defined as "the processof, capacity for, or outcome of successful adaptation despite challenging orthreatening circumstances" (Masten, Best, & Garmezy, 1990). This studyfocuses on measurement of resilience during adolescence. Significant changesoccur in the individual's psychological and social development during thisstage. It is reasonable to assume that many difficulties or adversities are asso­ciated with these changes. Resilience is a key factor in the process of over­coming and adapting to negative events. Takahira (1998) indicated thatyoung people such as undergraduate students might experience a number ofevents that are difficult or painful in their daily lives.

'Please address correspondence to Atsushi Oshio, PhD., Department of Psychology, ChubuUniversity, 1200 Matsumoto-cho, Kasugai City, Aichi Prefecture, 487-8501 Japan or e-mail(aoshio@isc.chubu.ac.jp).

Psychological Reports, 2003 , 93,1217-1222. © Psychological Reports 2003

CONSTRUCT VALIDITY OF THE ADOLESCENT RESILIENCE SCALE I

ATSUSHI OSHIO HITO~HI KANEKO

Chubu Universtfy Nagoya University

SHINJ! NAGAMINE MOTOYUKI NAKAYA

Kanazawa University Mie Unzversity

Summary.-The aim of this study was to assess the construct validity of the Ado­lescent Resilience Scale which measures the psychological features of resilient individ­uals. Research involving this scale, the Negative Life Events Scale, and the General Health Questionnaire was conducted with a group of 207 Japanese undergraduate students (104 men and 103 women; M age~20 .2 yr., SD~.9) . A cluster analysis for the Negative Life Events Scale and General Health Questionnaire yielded three clus­ters: (1) mentally healthy with little experience of Negative Life Events, (2) poorer mental health with many experiences of Negative Life Events, (3 ) mentally healthy de­spite many experiences of Negative Life Events. These three groups were defined as (1) Well Adjusted , (2) Vulnerable, and (3 ) Resilient, respectively. Mean differences in scores on the Adolescent Resilience Scale among the three groups were subjected to one-way analysis of variance. The mean scores of both the Well Adjusted and Resil­ient groups were higher than that of the Vulnerable group, and therefore support the construct validity of the Adolescent Resilience Scale.

Individual development is a dynamic process, involving hardships and crises. This is certainly true in modern society where we meet many difficult challenges from many factors such as crime and accidents, natural disasters, social unrest, and financial problems. It is therefore essential to cope with these negative events to attain a state of personal well-being.

Lately in the field of psychology and psychiatry, a concept called "resil­ience" has received increasing attention. Resilience is defined as "the process of, capacity for, or outcome of successful adaptation despite challenging or threatening circumstances" (Masten, Best , & Garmezy, 1990). This study focuses on measurement of resilience during adolescence. Significant changes occur in the individual's psychological and social development during this stage. It is reasonable to assume that many difficulties or adversities are asso­ciated with these changes. Resilience is a key factor in the process of over­coming and adapting to negative events. Takahira (1998) indicated that young people such as undergraduate students might experience a number of events that are difficult or painful in their daily lives.

'Please address correspondence to Atsushi Oshio, PhD. , Department of Psychology, Chubu University, 1200 Matsumoto-cho, Kasugai City, Aichi Prefecture, 487-8501 Japan or e-mail (aoshio@isc.chubu.ac.jp).

1218 A. osmo, ET AL.

Oshio, Nakaya, Kaneko, and Nagamine (2002) reviewed key earlierstudies regarding resilience and internal psychological features (e.g., Rutter& Quinton, 1984; Bandura, 1989; Wolin & Wolin, 1993; Eisenberg, Guth­rie, Fabes, Reiser, Murphy, Holgren, Maszk, & Losoya, 1997; Kumpfer,1999) and developed the Adolescent Resilience Scale to measure the psycho­logical features of resilient individuals. The scale was designed for Japaneseyouth, comprising 21 items and three factors: Novelty Seeking, EmotionalRegulation, and Positive Future Orientation. The present study was design­ed to assess the construct validity of the Adolescent Resilience Scale. Itwould be reasonable to postulate that resilient (young) individuals are ableto maintain mental health, without deterioration, even after experiencingpainful life events. Hence, in this study, scores on the Adolescent ResilienceScale were subjected to cluster analysis. Mean scores of the clusters werecompared along with scores on general health and negative life events.

METHOD

Participants

The target group consisted of 207 undergraduate students in Aichi pre­fecture, Japan. It was composed of 104 men and 103 women. Their agesranged from 19 to 23 years (M = 20.2).

Measures

The Adolescent Resilience Scale, as developed by Oshio, et at. (2002),

was employed. The scale consists of 21 items and three factors: Novelty Seek­ing, Emotional Regulation, and Positive Future Orientation. Respondentswere asked to choose a rating scale using anchors of 5 = Definitely yes and1= Definitely no. According to Oshio, et at. (2002), the coefficients alpha forthe scale and subscales were for the Total score .85, Novelty Seeking .79,Emotional Regulation .77, and Positive Future Orientation .81.

The Negative Life Events Scale was simplified by extracting 30 itemsfrom Takahira's Scale of Life Events in Interpersonal and Achievement Do­mains for Undergraduate Students (1998), and that set was employed. Thisscale taps the negative life events that could be experienced by Japanese un­dergraduate students in their daily lives. Respondents were asked to make aforced choice between 1= Yes, experienced event in the past 3 mo. and0= No, had not been experienced.

The General Health Questionnaire, as simplified by extracting 28 itemsfrom the translated version of the General Health Questionnaire by Naka­gawa and Daibo (1985), was employed. A 4-point Likert scale was used, withhigher points on responses which indicated poorer health.

1218 A. OSHIO, ET AL.

Oshio, Nakaya, Kaneko, and Nagamine (2002) reviewed key earlier studies regarding resilience and internal psychological features (e.g., Rutter & Quinton , 1984; Bandura, 1989; Wolin & Wolin, 1993; Eisenberg, Guth­rie, Fabes, Reiser , Murphy, Holgren , Maszk, & Losoya, 1997; Kumpfer, 1999) and developed the Adolescent Resilience Scale to measure the psycho­logical features of resilient individuals. The scale was designed for Japanese youth, comprising 21 items and three factors: Novelty Seeking, Emotional Regulation, and Positive Future Orientation. The present study was design­ed to assess the construct validity of the Adolescent Resilience Scale. It would be reasonable to postulate that resilient (young) individuals are able to maintain mental health, without deterioration , even after experiencing painful life events. Hence, in this study, scores on the Adolescent Resilience Scale were subjected to cluster analysis. Mean scores of the clusters were compared along with scores on general health and negative life events.

METHOD

Participants

The target group consisted of 207 undergraduate students in Aichi pre­fecture, Japan. It was composed of 104 men and 103 women. Their ages ranged from 19 to 23 years (M =20.2) .

Measures

The Adolescent Resilience Scale, as developed by Oshio, et at. (2002),

was employed. The scale consists of 21 items and three factors: Novelty Seek­ing, Emotional Regulation, and Positive Future Orientation. Respondents were asked to choose a rating scale using anchors of 5 = Definitely yes and 1 = Definitely no. According to Oshio, et at. (2002), the coefficients alpha for the scale and subscales were for the Total score .85, Novelty Seeking .79, Emotional Regulation .77, and Positive Future Orientation .81.

The Negative Life Events Scale was simplified by extracting 30 items from Takahira's Scale of Life Events in Interpersonal and Achievement Do­mains for Undergraduate Students (1998), and that set was employed. This scale taps the negative life events that could be experienced by Japanese un­dergraduate students in their daily lives. Respondents were asked to make a forced choice between 1 = Yes, experienced event in the past 3 mo. and 0= No, had not been experienced.

The General Health Questionnaire, as simplified by extracting 28 items from the translated version of the General Health Questionnaire by Naka­gawa and Daibo (1985), was employed. A 4-point Likert scale was used, with higher points on responses which indicated poorer health.

CONSTRUCT VALIDITY OF ADOLESCENT RESILIENCE 1219

RESULTS AND DISCUSSION

Intercorrelations Among Variables

The Iotal score on the Adolescent Resilience Scale was obtained by cal­culating the mean of the 21-item scores. Similarly, subscale scores were ob­tained by calculating the mean item scores of each factor. The Negative LifeEvents Scale score was calculated as the total number of negative experi­ences within the past 3 mo. The General Health Questionnaire score was ob­tained by calculating the mean of the scores on 28 items.

Table 1 shows the Pearson product-moment correlation coefficientsamong all variables. There were significant positive intercorrelations amongall factors of the Adolescent Resilience Scale. Resilience scores did not showintercorrelations with Negative Life Events. Resilience scores showed a sig­nificant negative intercorrelation with those on the General Health Question­naire. There was a significant positive intercorrelation between Negative LifeEvents and General Health Questionnaire.

TABLE 1INTERCORRELATIONS AMONG Au. VARIABLES

Variable 2 3 4 5 6 M SD

1. Total Resilience 3.35 .522. Novelry Seeking .75* 3.72 .613. Emotional Regulation .79* .35* 3.03 .664. Positive Future Orientation .72* .44* .30* 3.41 .845. Negative Life Events -.09 -.03 -.07 -.10 16.30 7.106. General Health Questionnaire -.49* -.26* -.45* -.38* .28* 2.17 .47

*p< .001.

The Formation 0/Adjustment Groups

Conducting Ward's cluster analysis on respondents using Negative LifeEvents and General Health Questionnaire as variables, three clusters wereidentified with the criterion of 0.1 as the increase in the sum of squares with­in the cluster.

Means of Negative Life Events and General Health Questionnaireamong observed clusters were studied by one-way analysis of variance. Theresults showed scores for Negative Life Events and General Health Question­naire were significantly different between clusters (Fm4 =100.77; F'.204 =176.02, p<.OOl). Effect sizes between clusters were as follows: for NegativeLife Events between the first and second clusters, d= 1.62; between secondand third clusters, d=.49; between first and third clusters, d=3.13; GeneralHealth Questionnaire, between first and second clusters, d=2.63; betweensecond and third clusters, d=2.74; between first and third clusters, d=.15.A multiple comparison test employing Tukey's HSD (p =.05) yielded the fol-

CONSTRUCT VALIDITY OF ADOLESCENT RESILIENCE 1219

RESULTS AND DISCUSSION

Intercorrelations Among Variables

The total score on the Adolescent Resilience Scale was obtained by cal­culating the mean of the 21-item scores. Similarly, subscale scores were ob­tained by calculating the mean item scores of each factor. The Negative Life Events Scale score was calculated as the total number of negative experi­ences within the past 3 mo. The General Health Questionnaire score was ob­tained by calculating the mean of the scores on 28 items.

Table 1 shows the Pearson product-moment correlation coefficients among all variables. There were significant positive inter correlations among all factors of the Adolescent Resilience Scale. Resilience scores did not show intercorrelations with Negative Life Events. Resilience scores showed a sig­nificant negative intercorrelation with those on the General Health Question­naire. There was a significant positive intercorrelation between Negative Life Events and General Health Questionnaire.

TABLE 1 INTERCORREl.A1l0 NS AMONG ALL VARIABLES

Variable 2 3 4 5 6 M SD

1. Total Resilience 3.35 .52 2. Novelty Seeking .75* 3.72 .61 3. Emotional Regulation .79* .35* 3.03 .66 4. Positive Future Orientation .72* .44* .30* 3.41 .84 5. Negative Life Events - .09 -.03 - .07 - .10 16.30 7.10 6. General Health Questionnaire -.49* - .26* -.45* -.38* .28* 2.17 .47

*p< .00 1.

The Formation 0/ Adjustment Groups

Conducting Ward's cluster analysis on respondents using Negative Life Events and General Health Questionnaire as variables, three clusters were identified with the criterion of 0.1 as the increase in the sum of squares with­in the cluster.

Means of Negative Life Events and General Health Questionnaire among observed clusters were studied by one-way analysis of variance. The results showed scores for Negative Life Events and General Health Question­naire were significantly different between clusters (F2.204 = 100.77; F2.204 = 176.02 , p < .OOl ). Effect sizes between clusters were as follows: for Negative Life Events between the first and second clusters, d= 1.62; between second and third clusters, d = .49; between first and third clusters, d = 3.13; General Health Questionnaire, between first and second clusters, d=2.63 ; between second and third clusters, d=2 .74; between first and third clusters, d= .15. A multiple comparison test employing Tukey's HSD (p = .05) yielded the fol-

1220 A. OSHIO, BY AL.

lowing results: Negative Life Events, Third cluster> Second cluster> Firstcluster; General Health Questionnaire, Second cluster> Third cluster - Firstcluster. Therefore, the clusters (Adjustment Groups) were named Well Ad­justed (First cluster), Vulnerable (Second cluster), and Resilient (Third clus­ter) (see Table 2).

TABLE 2GROUP DIFFERENCES ON RESILIENCE, NEGATIVE LIFE EVENTS, AND GENERAL

HEALTH QUESTIONNAIRE: MEANS AND STANDARD DEVIATIONS

Measure Adjustment Group F2~04

Cluster 1 Cluster 2 Cluster 3Well-adjusted Vulnerable Resilient

(n=86) (n=76) (n=45)

M SD M SD M SD

Total Resilience 3.49'.47 3.l1b .49 3.48' 53 14.60tNovelty Seeking 3.83' 59 357b .62 3.78'b .61 3.94*Emotional Regulation 3.20' .66 2.72b 57 3.23' .60 1538tPositive Future Orientation 3.56' .70 3.17b .88 3.52ab .93 5.06tNegative Life Events 1051' 3.87 19.37b 6.78 22.20' 339 100.77tGeneral Health Questionnaire 1.87b .31 2.66' .29 1.91b .24 176.02t

Note.-Means with different superscript letters were significantly different by Tukey's HSDtest (p<.05). *p<.05. tp<.01.

The Adolescent Resilience Scale Score Among Groups

The mean differences in the Resilience scores between AdjustmentGroups were studied by a one-way analysis of variance. There were signifi­cant differences for all Resilience means between groups (see Table 2). Ef­fect sizes between clusters were as follows: ARS total score, between WellAdjusted and Vulnerable, d= .80; between Vulnerable and Resilient, d= .74;between Well Adjusted and Resilient, d = .02; Novelty Seeking, betweenWell Adjusted and Vulnerable, d= .43; between Vulnerable and Resilient,d= .34; between Well Adjusted and Resilient, d= .07; Emotional Regulation,between Well Adjusted and Vulnerable, d= .77; between Vulnerable and Re­silient, d= .88; between Well Adjusted and Resilient, d= .05; Positive FutureOrientation, between Well Adjusted and Vulnerable, d=.49; between Vul­nerable and Resilient, d=.39; between Well Adjusted and Resilient, d=.04.Multiple comparison tests employing the Tukey HSD (p =.05) gave the fol­lowing order of differences: ARS total score, Well Adjusted-Resilient> Vul­nerable; Novelty Seeking, Well Adjusted> Vulnerable; Emotional Regulation,Well Adjusted-Resilient> Vulnerable; Positive Future Orientation, Well Ad­justed > Vulnerable.

ConclusionThe results showed that the Adolescent Resilience Scale score for the

1220 A. OSHIO, EY AL.

lowing results: Negative Life Events, Third cluster> Second cluster> First cluster; General Health Questionnaire, Second cluster > Third cluster - First cluster. Therefore, the clusters (Adjustment Groups) were named Well Ad­justed (First cluster) , Vulnerable (Second cluster), and Resilient (Third clus­ter) (see Table 2).

TABLE 2 GROUP DIFFERENCES ON RESILIENCE, NEGATIVE LIFE EVENTS. AND GENERAL

HEALTH QUESTIONNAIRE: MEANS AND STANDARD DEVIATIONS

Measure Adjustment Group Cluster 1 Cluster 2 Cluster 3

Well-adjusted Vulnerable Resilient (n = 86) (n = 76) (n = 45)

M SD M SD M SD

Total Resilience 3.49' .47 3. 11b .49 3.48' .53 Novelty Seeking 3.83' .59 3.57b .62 3.78,b .61 Emotional Regulation 3.20' .66 2.72b .57 3.23 ' .60 Positive Future Orientation 3.56' .70 3.17b .88 3.52,b .93 Negative Life Events 10.51' 3.87 19,37b 6.78 22.20' 339 General Health Questionnaire 1.87b .31 2.66' .29 1.91 b .24

F2~04

14 .60t 3.94*

1538t 5.06t

100.77t 176.02t

Note.-Means with different superscript letters were significantly different by Tukey's HSD test (p<.05). *p<.05. tp<.01.

The Adolescent Resilience Scale Score Among Groups

The mean differences in the Resilience scores between Adjustment Groups were studied by a one-way analysis of variance. There were signifi­cant differences for all Resilience means between groups (see Table 2). Ef­fect sizes between clusters were as follows: ARS total score, between Well Adjusted and Vulnerable, d= .80; between Vulnerable and Resilient, d= .74; between Well Adjusted and Resilient, d = .02; Novelry Seeking, between Well Adjusted and Vulnerable, d= .43; between Vulnerable and Resilient, d=.34; between Well Adjusted and Resilient, d=.07; Emotional Regulation, between Well Adjusted and Vulnerable, d=.77; between Vulnerable and Re­silient, d= .88; between Well Adjusted and Resilient, d= .05; Positive Future Orientation, between Well Adjusted and Vulnerable, d=.49; between Vul­nerable and Resilient, d=.39; between Well Adjusted and Resilient, d=.04. Multiple comparison tests employing the Tukey HSD (p = .05) gave the fol­lowing order of differences: ARS total score, Well Adjusted-Resilient> Vul­nerable; Novelry Seeking, Well Adjusted> Vulnerable; Emotional Regulation, Well Adjusted-Resilient> Vulnerable; Positive Future Orientation, Well Ad­justed > Vulnerable.

Conclusion

The results showed that the Adolescent Resilience Scale score for the

"

CONSTRUCT VALIDITY OF ADOLESCENT RESILIENCE 1221

Well Adjusted group and the Resilient group were higher than those of the Vulnerable group. This indicates that the scale correctly reflects the psycho­logical features of individuals who show resilience after facing Negative Life Events. Therefore, the construct validity is supported. Further study would be necessary to identify the psychological recovery process of the individuals who have the psychological features indicated by the Adolescent Resilience Scale.

REFERENCES

BANDURA, A. (1989) Human agen cy in social cognitive theory. American Psychologist , 44 , 1175-11 84 .

EISENBERG, N., GUTH RI E, !. K., FABES, R. A. , REISER, M., M uRPHY, B. c., H OLGREN, R., MASZK, 1'. , & LosOYA, S. (1997) The relations of regulation and emotionality to resiliency and com ­petent social fun ctioning in elementary school children . Child Development, 68, 295 -3 11.

KUMPFER, K. L. (1999) Factors and processes contributing to resilience: the resilience frame­work . In M. D. G lantz & J. L. J ohnson (Ed s.), Resilie/tce and development: positive /zfe adaptatzons. New York: Plenum. Pp. 179-224 .

MASTEN, A. 5., BEST, K. , & GARM EZY, N . (1990) Resilience and development: contrib utions from the study of child ren who overcame adversity. Development and Psychopathology, 2, 425-444 .

NAKAGAWA, Y, & DAlBO, !. (1985) [The General Health Questionnaire]. Tokyo: Nihon Bunka Kagakusha. [in J apanese]

OSHlO, A., NAKAYA, M., KANEKO, H ., & NAGAM INE, S. (2002) [Development and validation of an Adolescent Resilience Scale]. [japanese Journal o/Counseling Science], 35 , 57-65. [in Jap a­nese with English abst rac t]

RUTTER, M., & QUINTON, D. (1984) Long-term follow-up of women institutionalized in child­hood : factors promoting good fun ctioning in adult life. British Journal 0/ Developmental Psychology, 18, 225-234 .

TAKAHIRA, M. (1998) [Construction of a scale of life events in interpersonal and achievement domains for undergraduate studen ts]' [japanese Journal 0/ Social Psychology], 14, 12-24 . [in J apanese with English abstract]

WO LIN, S. ] ., &WOLlN, S. (1993) Bound and determined: growing up resilient in a troubled /am­ily. New York: Villard P ress.

Accepted October 24, 2003.

1222 A. OSHIO, EY AL.

APPENDIXITEM CONTENT OF THE ADOLESCENT RESILIENCE SCALE

Novdty SeekingI seek new challenges.I like new or inrriguillg rhings.I think I have a high levd of interest and curiosity.I like to find out about things.I think difficulties form a part of life's valuable experiences.I don't like to do unfamiliar things.I find it bothersome to start new activities.

Emotional RegulationI think I can control my emotions.I can stay caIrn in tough circumstances.I make an effort to always stay caIrn.I think I have perseverance.I find it difficult not to dwell on a negative experience.*I cannot endure adversity.*My behavior varies with my daily moods.*I lose interest quickly.*I have difficulty in controlling my anger.*

Positive Future OrientationI am sure that good things will happen in the future.I think I have a bright future.I fed positive about my future.I have a clear goal for the future.I am striving towards my future goal.

*Reverse-scored items.

1222 A. OSHIO, ET AL.

APPENDIX hEM CONTENT OF THE ADOLESCENT RESILIENCE SCALE

Novdty Seeking I seek new challenges. I like new or intriguing things. I think I have a high levd of interest and curiosity. I like to find out about things. I think difficulties form a part of life's valuable experiences. I don't like to do unfamiliar things. I find it bothersome to start new activities.

Emotional Regulation I think I can control my emotions. I can stay calm in tough circumstances. I make an effort to always stay calm. I think I have perseverance. I find it difficult not to dwell on a negative experience.' I cannot endure adversity.' My behavior varies with my daily moods.' I lose interest quickly.' I have difficulty in controlling my anger.'

Positive Future Orientation I am sure that good things will happen in the future. I think I have a bright future. I fed positive about my future. I have a clear goal for the future. I am striving towards my future goal.

' Reverse-scored items.