Personality and Individual Differences 48 (2010) 527531

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier.com/locate/paid

The ConnorDavidson Resilience Scale (CD-RISC): Testing the invariance

of a uni-dimensional resilience measure that is independent of positive
and negative affect
R.A. Burns *, K.J. Anstey
Ageing Research Unit, Centre for Mental Health Research, Australian National University, Canberra 0200, ACT, Australia

a r t i c l e

i n f o

Article history:
Received 9 April 2009
Received in revised form 11 November 2009
Accepted 19 November 2009
Available online 12 January 2010
Keywords:
Resilience
Factor analysis
Well-being
Epidemiology
Psychometrics

a b s t r a c t
Resilience comprises cognitive and behavioural tendencies that reect dispositional character traits and
patterns of behaviour that develop through life experience. Resilience is associated with positive mental
and physical health outcomes although debate over its function as a predictor and/or outcome of successful stressful life conditions exists. Findings are confounded by a range of operational denitions. This
study tested the factorial structure of the Connor and Davidson Resilience Scale (CD-RISC) and assessed
its independence of two broad affective constructs, positive and negative affect. Participants (n = 1775)
comprised the youngest adult cohort from the PATH study from Canberra, Australia. Results supported
a uni-dimensional CD-RISC measure that was independent of affect at an item level, but supported strong
associations between resilience and affect factors. Comparable Goodness of Fit Indices supported strict
invariance between genders on an oblique 3-factor model of resilience and affect.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Resilience is a multi-dimensional construct that comprises a
network of favourable attitudes and behaviours that enable adaptive coping strategies to acute and chronic stressful life events (Lamond et al., 2008; Rutter, 1985). Resilient people are typically
characterised by internal locus of control, positive self-image and
optimism (e.g. Cederblad, 1996; Werner & Smith, 1992). These
resilient and hardy characteristics are associated with better physical and mental health outcomes (Connor & Davidson, 2003), more
positive adaptive behaviours to negative life events (Aspinall &
MacNamara, 2005) and are protective against the onset of PostTraumatic Stress Disorder (King, King, Fairbank, Keane, & Adams,
1998). In line with proponents of positive psychology (e.g. Ryan
& Deci, 2001), resilience is indicative of positive mental health
(Maddi & Khoshaba, 1994).
Despite the consensus over the characteristics and correlates of
resilience, agreement relating to its temporal stability has been
less unanimous. Evidence that purports a heritable component of
resilience, as indicated by biological markers such as higher levels
of dopamine, neuropeptide Y, testosterone, and increased functionality of 5-HT1A and benzodiazepine receptors (Charney, 2004; Gervai et al., 2005), supports temporal stability and suggests resilience
incorporates a set of temporally-stable psychological resources
* Corresponding author. Tel.: +61 02 6125 3132.
E-mail address: [email protected] (R.A. Burns).
0191-8869/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2009.11.026

available to the individual throughout the lifespan (Vaishnavi, Connor, & Davidson, 2007). This is also supported by longitudinal survey data (e.g. Cederblad, 1996; Werner & Smith, 1992). In contrast,
notions of resilience as behavioural and cognitive responses to a
specic event have been proffered (e.g. Lamond et al., 2008) and
ones capacity for resilience appears shaped by age and life experience (e.g. Gillespie, Chaboyer, & Wallis, in press). Consequently,
resilience is described as either a set of heritable traits, an outcome
of stressful life transactions, or as a process-construct reecting an
interaction between trait attitudes and behaviours with life experiences (Ahern, Kiehl, Sole, & Byers, 2006).
Resilience measures commonly comprise self-report, have not
been extensively validated, nor has their application been widely
documented (Ahern et al., 2006). Yet, growing recognition of the
impact of positive psychological states for individual and social
well-being (Huppert, 2008), suggests an evaluation of available
measures of resilience and their underlying factor structure and
comparability at different stages of the life course, is needed. Similar positive well-being measures are now widely incorporated into
national health surveys including the Health and Retirement Survey and the National Survey of Midlife Development (e.g. Ryff,
Keyes, & Hughes, 2004), and the German Socioeconomic Panel
(e.g. Van Landeghem, 2008).
Well-being research typically delineates between related cognitive and affective psychological constructs (Ryan & Deci, 2001). As
a cognitive psychological resource, resilience may function to optimise subjective well-being by increasing positive and decreasing

528

R.A. Burns, K.J. Anstey / Personality and Individual Differences 48 (2010) 527531

negative affectivity, but since measures of resilience frequently

comprise self-report assessments, indices of resilience may simply
reect affect. However, Burns and Machin (2009) have differentiated between cognitive and affective well-being dimensions at
the item level, whilst still moderately related at a higher-order factor level. It is important to determine a similar nding in respect to
resilience and affect.
A number of resilience measures have been developed and include the Resilience Scale (RS; Wagnild & Young, 1993), which assesses two factors, personal competence and acceptance of self and
life; the Resilience Scale for Adults (RSA; Friborg, Hjemdal, Rosenvinge, & Martinussen, 2003), which measures ve factors: personal
competence, social competence, family coherence, social support
and personal structure; and the ConnorDavidson Resilience Scale
(CD-RISC; Connor & Davidson, 2003). The CD-RISC has been used
with clinical and non-clinical populations (Connor & Davidson,
2003), as a measure of psychological resources in moderating
childhood maltreatment and adult psychiatric symptoms (Campbell-Sills, Cohan, & Stein, 2006), and as an indicator for successful
resilience training (Davidson et al. 2005). Only one large scale validation of the CD-RISC within a general population has been reported (Lamond et al., 2008).
The CD-RISC authors and others typically compute a total CDRISC score (e.g. Campbell-Sills et al., 2006; Lamond et al., 2008),
yet post hoc analysis in the initial review (Connor & Davidson,
2003) indicated a ve-factor structure that reected a multidimensional resilience scale. However, the methods employed in
the initial report (Connor & Davidson, 2003), and subsequent analyses (e.g. Campbell-Sills & Stein, 2007; Vaishnavi et al., 2007) of the
CD-RISCs psychometric properties, reveal several limitations to
the proposed multi-dimensional nature of the CD-RISC. For example, Lamond et al. (2008) reported using a Principal Components
extraction and orthogonal rotation method which is not appropriate for identifying a correlated factor structure (Burns & Machin,
2009). Although using a Principal Axis Factoring method, Connor
and Davidson (2003) also reported an orthogonal rotation method
in their initial analysis which assumes that the resilience components are unrelated when a multi-dimensional resilience model
should comprise oblique factors. Many of the initial factor loadings
(Connor & Davidson, 2003) revealed several moderately crossloading items and suggest that investigation of a more parsimonious structure is warranted. For example, Campbell-Sills and Stein
(2007) used Exploratory Factor Analysis (EFA) with an oblique Promax rotation to identify a rened 10-item uni-dimensional model.
1.1. Aim
Consequently, we seek to test the structural validity of the a priori ve-factor model of the CD-RISC measure (Connor & Davidson,
2003) in a large population-based sample of young-adults and subsequently test its independence from a measure of positive and
negative affect.
2. Method
2.1. Study design and participants
The sample came from the PATH Through Life Project, a large
community survey concerned with the health and well-being of
randomly selected individuals from electoral rolls of Canberra or
Queanbeyan, Australia, and where voting is compulsory (see Anstey, Dear, Christensen, & Jorm, 2005). Results presented here concern the youngest cohort aged 2024 years at wave 1. Of 2404
commencing participants, 1978 were contactable at wave 3 in
2007/8, when data on resilience were collected. As an exploratory

investigation missing data was not imputed and only participants

who provided full data (N = 1775; 45.9% = males) for all measures
were included. Participants had a mean education level of 15 years,
and 88.3% rated their health as good, very good or excellent. Participants were mostly assessed in their homes and asked to complete
a questionnaire under the supervision of a professional interviewer. Basic physical and cognitive tests were also carried out
(e.g. blood pressure, grip strength, reaction time). Participants received a full description of the study and provided informed consent. The study was approved by the Human Research Ethics
Committee at the Australian National University.
2.2. Measures
The ConnorDavidson Resilience Scale (CD-RISC; Connor &
Davidson, 2003) comprises 25 items that measure resilience or
capacity to change and cope with adversity. Respondents indicated
their response on a 5-point Likert scale with higher scores indicating greater resilience. The Positive and Negative Affect Schedule
(PANAS; Watson, Clark, & Tellegen, 1988) is a 20 item self-report
measure of positive and negative affect (10 items per construct),
that reects both trait and state measures of affect, depending on
the time reference included in the item operative. Individuals indicated their response on a 5-point Likert scale, with higher scores
indicating greater well-being on that dimension. Due to an administrative oversight, two negative affect items were excluded from
the questionnaire.
2.3. Statistical analysis
Conrmatory Factor Analysis (CFA) tested the a priori ve-factor CD-RISC model (Connor & Davidson, 2003) in AMOS v17 using
Maximum Likelihood Estimation (MLE). Further model specication was undertaken with Exploratory Factor Analysis (EFA) in SPSS
v17, using Principal Axis Factoring (PAF) and an oblique Direct
Oblimin rotation (delta = 0) since separate CD-RISC factors should
reect related resilience characteristics. We used Parallel Analysis
(PA) to guide our EFA extraction. PA is a sample-based approach to
estimating the number of factors and appears more highly robust
than Kaisers Rule or the Scree plot (Hayton, Allen, & Scarpello,
2004). Goodness of Fit Indices (GFI) of the extracted EFA model
were computed using AMOS v17 and compared to a priori ve-factor model and a revised ten-item uni-dimensional model (Campbell-Sills & Stein, 2007). CFA of the CD-RISC and PANAS tested
the discriminant validity of the CD-RISC measure by evaluating
the independence of resilience items from positive and negative affect. Factorial invariance between gender was tested using the
Multi-Groups Function in AMOS v17.
3. Results
3.1. Testing the structural validity of the RISC
CFA of the a priori ve-factor model (Connor & Davidson, 2003)
revealed a number of large Modication Indices (MI; >20) and
items that did not discriminate the proposed factors. Several very
high correlations indicate considerable overlap between four of
the ve latent variables (Table 1). In addition, GFI revealed poor
t (Table 2) for this model. Consequently, with debate over the extent to which the CD-RISC reects a multidimensional (Connor &
Davidson, 2003) or uni-dimensional model (Campbell-Sills & Stein,
2007), we used EFA to further explore the nature of the CDRISC. Results of a PA did not support an extraction of ve factors, but indicated one factor with a raw eigenvalue greater than 95% of the PA,
with three more eigenvalues greater than their corresponding

529

R.A. Burns, K.J. Anstey / Personality and Individual Differences 48 (2010) 527531

excluded if they loaded P.32 on two or more factors, or failed to

achieve this level on one factor, CD-RISC item 2 was also excluded
from the analysis. All other item loadings and the lack of item misspecication conrmed the discrimination of PANAS and CD-RISC
items and suggests that perceptions of resilience are independent
of affect at the item level for the sample and by gender (Table 3).
The CD-RISC was positively associated with positive affect
(r = .575, p < .001) whilst CD-RISC (r = .261, p < .001) and positive
affect (r = .352, p < .001) were both negatively associated with
negative affect. In comparison with Campbell-Sills and Steins
(2007) modied 10-item uni-variate model, not all of these ten
items (identied in italics type in Table 3) were identied as the
strongest loadings in this sample and this was consistent between
gender.
Despite reporting a signicant chi-square, other GFI revealed
strong support for this three-factor model of resilience and affect
(v2 = 2202.449, df = 667, p = .001; AGFI = .926; CFI = .958;
RMSEA = .036 (95% CI: .034.037)). Signicant covariance paths between error terms were again included within factors that reported
grossly large Modication Indexes (MI P 20). Although item loadings between gender were consistent (Table 3), multi-groups CFA
tested for gender invariance (Table 4). Following Kline (2005), congural and metric invariance was supported with all factor loadings signicant and equivalent between gender. Signicant chisquare differences between an unconstrained model and three
models that constrained, regression weights, error and variance
terms failed to support strong and strict invariance, whilst GFI between models reported comparable moderate good t of the data.
In particular, a comparison of the RMSEA condence intervals reveals t for all models to be comparable across gender and suggests that this model is invariant.

Table 1
Correlations between factors of the a priori ve CD-RISC model.

Factor
Factor
Factor
Factor
Factor

1
2
3
4
5

Factor 1

Factor 2

Factor 3

Factor 4

Factor 5

1
.878
.875
.839
.253

.878
1
.901
.718
.217

.875
.901
1
.767
.248

.839
.718
.767
1
.275

.253
.217
.248
.275
1

mean PA eigenvalues. Therefore we used PAF in SPSS v17 to extract

four factors, explaining 54% of the variance in the CD-RISC. However, most items loaded onto the rst factor with one or two items
cross-loading onto the other three factors, indicating that further
renement of the factor structure was warranted. Extracting three
factors revealed considerable number of cross-loading items and
shared variance between Factors 1 and 3 (r2 = .609) whilst only
items 3 and 9 loaded onto the second factor. With a criterion loading of 6.32 (level at which a factor explains 10% unique item variance), a two-factor structure revealed several moderate crossloadings. Therefore a one-factor model was extracted, but both
items 3 and 9 failed to load >.32 onto this factor. Re-running this
model with items 3 and 9 deleted, revealed that all other items
loaded onto this single factor, accounting for 40% of the explained
variance, and supports Campbell-Sills and Steins (2007) ndings of
a unitary resilience construct.
In comparison to the a priori ve-factor model (Model 1) and
our EFA derived uni-dimensional model (Model 2), GFI and the
v2 test of difference revealed that Model 3, an abridged 10-item
version of the CD-RISC (Campbell-Sills & Stein, 2007), reported better t. However, GFI indicated that signicant improvements were
warranted for both the full-scale and shorter models and signicant covariance paths between error terms were included for
grossly large Modication Indexes (MI P 20). Considerable
improvement in model t was reported for all models, although
comparable t between the models was now reported. Preference
for the uni-dimensional models was chosen on the grounds of parsimony (Table 2).

4. Discussion
Our results failed to support the hypothesis that the CD-RISC
comprises multiple resilience factors (e.g. Connor & Davidson,
2003), but rather supported a one-factor CD-RISC model (Campbell-Sills & Stein, 2007) comprising a revised 22-item CD-RISC
scale. Published results (Connor & Davidson, 2003) revealed that
CD-RISC items 3 and 9 would be problematic in a uni-dimensional
context since they reported cross-factor loadings and item-total
scale correlations that were low. This was supported in this study
suggesting that these items tap a latent construct different to the
other CD-RISC items. The failure of CD-RISC item 2 to load onto
the nal solution was not surprising since its item-total scale correlation reported in Connor and Davidsons (2003) analysis was
also comparatively low.

3.2. Discriminant validity of the CD-RISC with affect between gender

We extended our analyses to explore the relationship between
the larger uni-dimensional CD-RISC model extracted from our EFA
with a measure of affect, the PANAS (Watson et al., 1988). A CFA of
the PANAS and CD-RISC items hypothesised a three-factor extraction excluding CD-RISC items 3 and 9 since these failed to load onto
our earlier unitary factor. Based on our earlier criteria that items be

Table 2
A summary table of Goodness of Fit Indices from CFA of three models of the CD-RISC.

v2

Df

AGFI

CFI

RMSEA

AIC

Model comparisonsd

Five-factor model
Uni-dimensional model
10-Item uni-dimensional model

2530.731
2874.717
200.105

265
230
35

.000
.000
.000

.858
.826
.963

.878
.852
.972

.069 (.067.072)
.081 (.078.083)
.052 (.045.059)

2650.731
2966.717
240.105

M1 > M2

Model 4
Model 5b

Model 1 with signicant covariance paths

Model 2 with signicant covariance paths

154.866
155.385

144
127

.253
.044

.984
.984

.999
.998

.007 (.000.013)
.011 (.002.017)

516.866
458.358

Model 6c

Model 3 with signicant covariance paths

26.646

21

.183

.992

.999

.012 (.000.025)

94.646

Model 1a
Model 2b
Model 3c

M3 > M1
M3 > M2
M4 = M3
M5 = M3
M5 = M4
M6 > M3
M6 = M4
M6 = M5

Model 1 and 4: a priori ve-factor model (Connor and Davidson, 2003).

Model 2 and 5: uni-dimensional model of our EFA derived model.
c
Model 3 and 6: 10-item uni-dimensional model (Campbell-Sills and Stein, 2007).
d
Comparisons of model t undertaken with v2 test of difference (p < .05): (>) indicates better tting model; (<) indicates worse tting model; (=) indicates models t
comparatively.
b

530

R.A. Burns, K.J. Anstey / Personality and Individual Differences 48 (2010) 527531

Table 3
Item loadings from a CFA analysis of the CD-RISC and PANAS items.
Whole sample
RISC items
17
24
16
4
12
5
11
23
18
14
8
15
19
7
21
22
10
6
1
25
20
13
PANAS
Scared
Nervous
Afraid
Distressed
Guilty
Jittery
Upset
Ashamed
Enthusiastic
Interested
Inspired
Excited
Determined
Strong
Proud
Alert
Attentive
Active

CD-RISC

Males
NA

PA

Females

CD-RISC

NA

PA

CD-RISC

NA

PA

.750
.723
.685
.684
.678
.672
.663
.647
.618
.617
.611
.608
.607
.585
.532
.523
.476
.473
.468
.464
.459
.362

.032
.065
.004
.064
.036
.022
.010
.022
.016
.111
.064
.035
.055
.040
.012
.194
.019
.025
.097
.003
.091
.057

.026
.066
.035
.017
.049
.043
.105
.079
.120
.042
.027
.008
.058
.040
.202
.200
.057
.012
.002
.179
.052
.124

.737
.721
.718
.697
.691
.692
.680
.599
.640
.632
.582
.592
.561
.563
.514
.521
.506
.493
.424
.423
.451
.370

.100
.092
.025
.085
.093
.023
.005
.054
.053
.143
.060
.025
.062
.026
.014
.207
.026
.062
.100
.010
.097
.091

.005
.080
.034
.055
.054
.023
.089
.086
.109
.081
.051
.006
.007
.022
.200
.190
.070
.020
.015
.211
.040
.181

.752
.721
.659
.669
.672
.651
.650
.678
.603
.603
.625
.617
.635
.597
.555
.524
.474
.443
.501
.500
.467
.390

.021
.047
.008
.040
.020
.023
.025
.013
.012
.071
.070
.057
.043
.045
.011
.194
.016
.015
.096
.019
.075
.086

.057
.057
.041
.023
.037
.063
.113
.087
.129
.005
.016
.003
.089
.044
.206
.210
.036
.008
.013
.148
.061
.070

.002
.054
.009
.037
.039
.022
.045
.058
.003
.028
.041
.037
.138
.045
.030
.047
.007
.003

.785
.741
.737
.709
.666
.658
.649
.645
.034
.005
.004
.048
.149
.140
.021
.020
.084
.093

.026
.100
.017
.155
.001
.069
.195
.030
.873
.830
.749
.736
.724
.670
.644
.619
.618
.613

.007
.071
.005
.002
.035
.020
.048
.015
.011
.044
.036
.041
.134
.052
.058
.065
.008
.038

.771
.667
.691
.767
.649
.652
.652
.707
.033
.012
.016
.041
.144
.117
.007
.046
.046
.076

.042
.102
.005
.121
.037
.081
.195
.046
.856
.793
.748
.749
.700
.644
.673
.581
.578
.674

.002
.051
.011
.065
.039
.028
.050
.089
.010
.025
.046
.030
.147
.040
.007
.034
.008
.022

.789
.792
.764
.662
.680
.664
.638
.613
.035
.015
.005
.045
.153
.141
.037
.012
.115
.099

.016
.108
.028
.185
.028
.057
.204
.026
.879
.854
.748
.721
.737
.698
.618
.654
.643
.580

Bold and italics type indicate the 10-items identied in Campbell-Sills and Steins (2007) analyses.

Table 4
Multi-groups analysis of a three-factor model of CD-RISC and PANAS testing for invariance between gender.

Unconstrained model
Constrained measurement weights
Constrained structural co-variances
Constrained measurement residuals

v2

df

AGFI

CFI

RMSEA

v2.diff. test

2950.936
3040.188
3057.363
3289.363

1354
1391
1397
1497

.000
.000
.000
.000

.905
.905
.905
.905

.957
.955
.955
.952

.026
.026
.026
.026

p < .001
p < .001
p < .001

(.025.027)
(.025.027)
(.025.027)
(.025.027)

v2 diff. test computed with the unconstrained model as the reference model.

Support for our uni-variate CD-RISC model is reected in the reported GFI and that most of the explained variance was accounted
for by the rst factor. Our revised 22-item model indicated CDRISCs independence of two broad affect states, with moderate
coefcients between affective and cognitive well-being components reported. In comparison to Campbell-Sills and Steins
(2007) abridged 10-item CD-RISC model, the longer uni-dimensional model performed comparatively. Importantly, examination
of the item loadings in our EFA highlighted the limitation of using
an abridged version of the CD-RISC as the 10-items reported in
Campbell-Sills and Steins (2007) analyses were not the strongest
loading items in our representative community sample. However,
whilst other manifest items were more indicative of the resilience

latent construct in our sample, a strong association (r = .950) between the factor scores of the longer and shorter scale forms suggests that these differences are not substantial and that the two
scale forms are comparable. We recognise that a shorter item pool
is more time efcient and user-friendly within the context of a larger battery of survey questionnaires. Further factorial analysis of
all the CD-RISC items may yet indicate consistency in those items
that are identied as common indicators of a uni-dimensional
model of resilience for a short scale format.
Resilience may reect a multi-dimensional structure of self attitudes. As with other models of self-referent beliefs (e.g. Shavelson,
Hubner, & Stanton, 1976) generalised resilience may reect a set of
latent constructs that comprise specic resilience components

R.A. Burns, K.J. Anstey / Personality and Individual Differences 48 (2010) 527531

which relate to an individuals array of coping strategies, self-efcacy beliefs and perceived degree of self-determination. Consequently, measurement of context-specic resilience beliefs may
better indicate the role of resilience in providing a buffer against
the impact of life events. This is clearly an area that needs further
renement as is delineating the role of resilience as an outcome or
predictive construct, however the CD-RISC does not assess such a
model. Importantly though, the CD-RISC is independent of affect
at an item level and supports previous ndings relating to the related yet distinctive nature of cognitive and affective components
of well-being (Burns & Machin, 2009).
A number of limitations are identied. The CD-RISC scale was
only introduced in the third wave of the PATH study and to date,
data collection has only been completed for the younger adult cohort. It is important to test the validity of these ndings with the
other age cohorts as data become available. Furthermore, participants were relatively highly educated and ethnically non-diverse
and so the ndings need to be evaluated in more diverse populations. Age effects and other sampling characteristics may inuence
item response which can be reected in item loadings in EFA. However, this is unlikely to impact on these ndings since participants
were similar in age. In fact, analysis of invariance of the CD-RISC
and the PANAS revealed consistency in factor structure between
gender and supported strict invariance. Although the results of this
study are more appropriate for generalisation to the Australian
population, a further strength is that participants were drawn from
the electoral roll, unlike other analyses (e.g. Campbell-Sills & Stein,
2007) which comprise convenience and clinical samples.

5. Conclusion
This study supported the uni-dimensional structure of the CDRISC and differentiated the CD-RISC items with affect whilst a
strong relationship between these constructs at a higher-order level was reported. Future research should identify either a causal or
reciprocal relationship between resilience and affect. Also, the
independence of resilience with other cognitive well-being constructs, including mastery, has received little empirical
substantiation.
Acknowledgements
We thank Trish Jacomb, Karen Maxwell, Helen Christensen, Andrew MacKinnon, Peter Butterworth and Simon Easteal and the
PATH interviewers. Funding was provided by NHMRC grant #
418039.
References
Ahern, N. R., Kiehl, E. M., Sole, M. L., & Byers, J. (2006). A review of instruments
measuring resilience. Issues in Comprehensive Paediatric Nursing, 29, 103125.
Anstey, K. J., Dear, K., Christensen, H., & Jorm, A. F. (2005). Biomarkers, health,
lifestyle and demographic variables as correlates of reaction time performance
in early, middle, and late adulthood. Quarterly Journal of Experimental
Psychology, 58A, 521.
Aspinall, L. G., & MacNamara, A. (2005). Taking positive changes seriously. Cancer,
104, 25492556.

531

Burns, R. A., & Machin, M. A. (2009). Investigating the structural validity of Ryffs
psychological well-being scales across two samples. Social Indicators Research,
93(2), 359375.
Campbell-Sills, L., Cohan, S. L., & Stein, M. B. (2006). Relationship of resilience to
personality, coping, and psychiatric symptoms in young adults. Behaviour
Research and Therapy, 44, 585599.
Campbell-Sills, L., & Stein, M. B. (2007). Psychometric analysis and renement of the
ConnorDavidson Resilience Scale (CD-RISC): Validation of a 10-item measure
of resilience. Journal of Traumatic Stress, 20(6), 10191028.
Cederblad, M. (1996). Fifty years of epidemiologic studies in child and adolescent
psychiatry in Sweden. Nordic Journal of Psychiatry, 50, 5566.
Charney, D. S. (2004). Psychobiological mechanisms of resilience and vulnerability:
Implications for successful adaptation to stress. American Journal of Psychiatry,
161, 195216.
Connor, K. M., & Davidson, J. R. T. (2003). Development of a new resilience scale: The
ConnorDavidson Resilience Scale (CD-RISC). Depression and Anxiety, 18, 7682.
Davidson, J. R., Payne, V. M., Connor, K. M., Foa, E. B., Rothbaum, B. O., Hertzberg, M.
A., et al. (2005). Trauma, resilience, and saliostatis: Effects of treatment on
posttraumatic stress disorder. International Clinical Psychopharmacology, 20,
4348.
Friborg, O., Hjemdal, O., Rosenvinge, J. H., & Martinussen, M. (2003). A new rating
scale for adult resilience. What are the central protective resources behind
healthy adjustment? International Journal of Methods in Psychiatric Research, 12,
6576.
Gervai, J., Nemoda, Z., Lakatos, K., Ronai, Z., Toth, I., Ney, K., et al. (2005).
Transmission disequilibrium tests conrm the link between DRD4 gene
polymorphism and infant attachment. American Journal of Medical Genetics
Part B (Neuropsychiatric Genetics), 132B, 126130.
Gillespie, B.M., Chaboyer, W., & Wallis, M. (In press). The inuence of personal
characteristics on the resilience of operating room nurses: A predictor study.
International Journal of Nursing Studies, doi: 10.1016/j.ijnurstu.2007.08.006.
Hayton, J. C., Allen, D. G., & Scarpello, V. G. (2004). Factor retention decisions in
exploratory factor analysis: A tutorial on parallel analysis. Organizational
Research Methods, 7(2), 191205.
Huppert, F. (2008) Psychological Wellbeing: Evidence regarding its causes and
consequences State-of-Science Review: SR-X2, Government Ofce for Science
Foresight Project, Mental Capital and Wellbeing: Making the most of ourselves
in the 21st Century, London.
King, L. A., King, D. W., Fairbank, J. A., Keane, T. M., & Adams, G. A. (1998). Resiliencerecovery factors in posttraumatic stress disorder among female and male
Vietnam veterans: Hardiness, postwar social support, and additional stressful
life events. Journal of Personality and Social Psychology, 74, 420434.
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.).
New York: Guilford Press.
Lamond, A. J., Depp, C. A., Allison, M., Langer, R., Reichstadt, J., Moore, D. J., et al.
(2008). Measurement and predictors of resilience among community-dwelling
older women. Journal of Psychiatric Research, 43(2), 148154.
Maddi, S. R., & Khoshaba, D. M. (1994). Hardiness and mental health. Journal of
Personality Assessment, 63, 265274.
Rutter, M. (1985). Resilience in the face of adversity: Protective factors and
resistance to psychiatric disorders. British Journal of Psychiatry, 25, 173180.
Ryan, R., & Deci, E. L. (2001). On happiness and human potentials: A review of
research on hedonic and eudaimonic well-being. Annual Review of Psychology,
52, 141166.
Ryff, C. D., Keyes, C. L. M., & Hughes, D. L. (2004). Psychological well-being in
MIDUS: Proles of ethnic/racial diversity and life course uniformity. In O. G.
Brim, C. D. Ryff, & R. C. Kessler (Eds.), How healthy are we? A national study of
well-being at midlife (pp. 398422). Chicago: University of Chicago Press.
Shavelson, R., Hubner, J., & Stanton, G. (1976). Self concept: Validation of construct
interpretations. Review of Educational Research, 46, 407441.
Vaishnavi, S., Connor, K., & Davidson, J. R. T. (2007). An abbreviated version of the
ConnorDavidson Resilience Scale (CD-RISC), the CD-RISC2: Psychometric
properties and applications in psychopharmacological trials. Psychiatry
Research, 152, 293297.
van Landeghem, B.G.M. (2008) Human Well-Being over the Life Cycle: Longitudinal
Evidence from a 20-Year Panel. LICOS Discussion Papers 21308, LICOS Centre
for Institutions and Economic Performance, K.U. Leuven.
Wagnild, G. M., & Young, H. M. (1993). Development and psychometric validation of
the resilience scale. Journal of Nursing Measurement, 1, 165178.
Watson, D., Clark, A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of
Personality and Social Psychology, 54, 10631070.
Werner, E. E., & Smith, R. S. (1992). Overcoming the odds. High risk children from birth
to adulthood. London: Cornell University Press.