Vandierendonck2004 PDF

Personality and Individual Differences 36 (2005) 629–643
www.elsevier.com/locate/paid
The construct validity of Ryff’s Scales of Psychological

Well-being and its extension with spiritual well-being
Dirk van Dierendonck*
Department of Work and Organizational Psychology, University of Amsterdam,
Roetersstraat 15, 1018 WB Amsterdam, The Netherlands
Received 12 August 2002; received in revised form 2 February 2003; accepted 21 February 2003
Abstract
This study examined the factorial and content validity of Ryff’s Scales of Psychological Well-being
(SPWB) in a sample of psychology students (N=233) and a sample of professionals from a diverse occu-
pational background (N=420). The psychometric quality of the SPWB was tested for the versions with
3-items, 9-items and 14-items. It appeared that the factorial validity was only acceptable for the 3-items per
scale version. However, the internal consistency of these 3-items scales was below generally accepted levels.
Therefore, it is suggested to reduce the length of the 14-item scales to 6, 7 or 8 items, depending on the specific
subscale. This resulted in an improved overall psychometric quality. In addition, two new scales were
developed that together refer to spiritual well-being. A second order factor analysis, including vitality, hap-
piness, self-esteem and the Big Five personality dimensions, revealed four underlying dimensions of positive
psychological health: subjective well-being, self-actualization, interpersonal relations and autonomy.
# 2003 Elsevier Ltd. All rights reserved.
Keywords: Well-being; Spirituality; Positive psychological health; Mental health
1. Introduction
The increasing interest in positive psychology, calls for valid and reliable instruments that
measure the essential concepts in this area of research (Keyes & Lopez, 2002). A popular instru-
ment in this field is Ryff’s Scales of Psychological Well-being (SPWB). Ryff (1989a) developed
this instrument on the basis of an extensive literature review and the integration of mental health,
clinical and life span developmental theories. She argued that these perspectives incorporate
similar and complementary criteria of positive psychological health. They have an optimistic
* Tel.: +31-20-525-68-60.
E-mail address: dvandierendonck@fmg.uva.nl (D. van Dierendonck).
0191-8869/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0191-8869(03)00122-3
630 D. van Dierendonck / Personality and Individual Differences 36 (2005) 629–643
outlook on life, emphasizing personal growth and development. The theoretically derived
dimensions of positive psychological health included Self-acceptance, Positive relations with
others, Autonomy, Environmental mastery, Purpose in life, and Personal growth (Ryff, 1989a).
Together, these dimensions can contribute to the assessment of a person’s level of positive func-
tioning and well-being. Nevertheless, despite it’s increased use, studies into the factorial validity
of the SPWB have been scarce. A good factorial validity is one way to establish the construct
validity of a scale. Although earlier studies confirmed the proposed structure (e.g. Ryff, 1989a;
Ryff & Keyes, 1995), a recent study by Kafka and Kozma (2002) showed that the factorial
validity is not unambiguous. This article aims to enhance our insight into the factorial validity of
the SPWB. In addition, the content validity will be addressed. Moreover, it is proposed that a
new dimension, namely spirituality, should be added to the six dimensions of Ryff.
One of the strong points of the SPWB is the way it was constructed (Ryff, 1989b). First, the six
dimensions were theoretically defined. On the basis of these definitions, 80 items were generated
by three persons. Next, these items were evaluated on the basis of criteria such as ambiguity and
fit with the scale definition. The resulting item pool of 32 items per scale (16 positive and 16
negative) was tested in a pilot-study among 321 adults. From each scale, the twelve items that
according to psychometric analysis appeared to be least fitting were removed, thus resulting in a
20-items per scale instrument.
In the following years, three shorter versions of the SPWB were proposed. Ryff, Lee, Essex, and
Schmutte (1994) used a 14-items per scale version. Items from these scales were selected from the
20-items version based on item–total correlation and coherence with the guiding theoretical defi-
nition. Correlations with the original scales ranged from 0.97 to 0.98. A very short version was
formulated as part of a national survey. Ryff and Keyes (1995) selected 3 out of each 20 items
that best captured the conceptual scope of the definitions. These shortened scales still correlated
between 0.70 and 0.89 with their 20-item parent scales, however the internal consistency was low,
ranging from 0.33 to 0.56. Confirmatory factor analysis with the 3-items per scale version
demonstrated that the best fit to the data was achieved by the model that joined the six dimen-
sions together by a single higher order factor model (Ryff & Keyes, 1995). However, the
remarkably high correlation between Self-acceptance and Environmental mastery may point to a
possible five-factor model. The factor analysis of Kafka and Kozma (2002) with the 20-items per
scale version also points in that direction. Recently, Ryff formulated a 9-items per scale version
that is part of an ongoing longitudinal study. Unfortunately, data about its psychometric quality
are as yet not available. Our first research question is whether the six-factor structure of the
SPWB can be confirmed for the versions with 14-, 9- and 3-items per scale.
In her formulation of positive psychological health, Ryff (1989a) criticized the common scales
of well-being (e.g. measures by Fordyce & Bradburn). These scales are considered to only
measure what is now called hedonic well-being (Kahneman, Diener, & Schwarz, 1999). This issue
is related to the construct validity of the SPWB, namely that it measures both similar and differ-
ent aspects of well-being. Only few studies used second order factor analyses with different well-
being and mental health measures to investigate possible underlying dimensions (Compton, 1998;
Compton, Smith, Cornish, & Qualls, 1996; Kafka & Kozma, 2002; Ryff, 1989a). In these studies
subjective well-being and self-actualization or personal growth are suggested as two separate, yet
related, underlying constructs. With regard to the SPWB Scales, these studies show that Self-
acceptance is most strongly related to subjective well-being scales. Both Personal growth and
D. van Dierendonck / Personality and Individual Differences 36 (2005) 629–643 631
Purpose in life consistently load on the self-actualization factor. The results also point in the
direction of a possible third underlying factor, namely Autonomy. The role of Positive relations
with others remains unclear, it loaded once on the autonomy factor and once on the self-
actualization factor. Because of the limited number of studies on this issue and the partly
inconclusive results, future research in this field is called for. The second research question
concerns the different underlying aspects of positive psychological health as measured with the
SPWB.
A related issue is the relation of the SPWB with personality. Previous studies established that
subjective well-being is moderately stable across situations and across life span (Diener &
Lucas, 1999), suggesting that personality is partly related to positive psychological health. As
indicated by Schmutte and Ryff (1997), several dimensions of the SPWB overlap conceptually
with the Big Five personality dimensions. For example, the definitions of agreeableness and
Positive relations with others are quite comparable. Similarly, a high score on Personal
growth indicates openness to new experiences. Their study showed strong but distinctive con-
nections between the personality dimensions and the SPWB Scales. Noteworthy is that not
only the personality dimensions neuroticism and extraversion were related to well-being, as was
earlier reported (Costa & McCrae, 1980; Diener & Lucas, 1999), but that also conscientious-
ness, openness to experiences and agreeableness were linked to different dimensions of the
SPWB. This is in line with the findings in the meta-analysis of DeNeve and Cooper (1998) who
showed that agreeableness predicted positive affectivity. In the present article, the relation
between personality and positive psychological health will be explored as part of the content
validity analysis.
Attention for spirituality has steadily been growing during recent years. More and more peo-
ple are looking for the essence of what it means to be human. What we traditionally experi-
enced as part of our religion has now returned under the denominator spirituality. Spirituality is
often seen as living by your inner truth to produce positive attitudes and relationships in your
life (Hawley, 1993). Such inner resources give a feeling of strength in times of crisis, when
dealing with the uncertainties of life. Spirituality has been noted as a determinant of a better
mental health (Koenig, 1998). According to Ellison (1983), spiritual well-being can be viewed as
an expression of spiritual maturity, but also as the integral experience of a person who is func-
tioning as God intended (Ellison & Smith, 1991). Spirituality as a dimension of well-being
describes, within such a framework, the experience of being in harmony, the integration of inner
and outer.
In the operationalization of spiritual well-being, the basic principle was to measure this aspect
as directly as possible, in terms of behaviour, thoughts and feelings. The basic assumption was to
formulate the items in such a way that they could be filled out by people from different religious
and spiritual background. Overlap with the existing scales of the SPWB was avoided. For exam-
ple, most definitions of spirituality emphasize the search for meaning and purpose. However, this
aspect was already covered by the SPWB. Two aspects were seen as important for the Spirituality
Scale, namely (1) being able to access inner resources and (2) experiencing a satisfying relation-
ship with God or a higher power. The latter aspect was already emphasized as an important ele-
ment of spiritual well-being in studies in the 1970s and the 1980s (Moberg, 1984) and an essential
element of the most widely used scale that specifically focused on spiritual well-being (Ellison,
1983).
To conclude, the purpose of this article is to report on the factorial and content validity of the
SPWB. The following three questions will be addressed: (1) Can the six-factor structure of the
SPWB be confirmed for the 14-, 9- and 3-items per scale versions? (2) To what extent are different
aspects of positive psychological health underlying the SPWB measure? (3) Is it possible to con-
struct a spiritual well-being scale that is conceptually different from the other six scales? These
research questions will be answered with samples from two studies, a sample of 233 first year
psychology students and a sample of 420 professionals from diverse occupational backgrounds.
2. Methods
2.1. Study 1
2.1.1. Subjects
Participants were 233 first year undergraduate psychology students who participated with the
research for course credit. The sample consisted of 156 female (67%) and 77 male (33%) students.
Their mean age was 22 years (S.D.=6).
2.1.2. Measures
Unless otherwise specified, a six-point answering scale was used for all scales, ranging from 1
(totally disagree) to 6 (totally agree).
2.1.2.1. Scales of psychological well-being. This instrument was developed by Ryff (1989a). The
original version consists of six dimensions of 20 items each. In this study the shortened 14-items
per scale version was used. Ryff and Essex (1992) selected items on the basis of fit with the overall
scale and theoretical definition. Two persons with an excellent command of the English language
independently translated the items into Dutch. The translations were then compared, differences
were solved by translating again into English.
2.1.2.2. Spiritual well-being. Spiritual well-being was measured with 14 items. These items were
taken from the religious well-being scale of the Spiritual Well-Being Scale of Ellison (1983) and
from the inner resources and transcendence scales of the Spiritual Assessment Scale developed by
Howden (1992).
2.1.2.3. Vitality. Vitality was measured with the seven items of the Subjective Vitality Scale
(Ryan & Frederick, 1997). The internal consistency was good (a=0.86).
2.1.2.4. Self-esteem. Self-esteem was measured with the Single-Item Self-Esteem Scale (Robins,
Hendin, & Trzesniewski, 2001).
2.1.2.5. Happiness. The happiness measure (Fordyce, 1988) is a two-item measure, including an
11-point Happiness/Unhappiness Scale and a question asking for the time spent in ‘happy’,
‘unhappy’, and ‘neutral’ moods. The primary criterion for happiness is the combination score
that is calculated through the following formula: ½scale score 10 þ happy%=2: This measure is
a simple instrument that demonstrated strong reliability, stability, relative freedom from response
bias, and good validity (Fordyce, 1988).
2.1.2.6. Personality. The Big Five personality dimensions (extraversion, agreeableness, con-
scientiousness, neuroticism and openness to experience) were measured with the 5PFT (Elshout &
Akkerman, 1974). This measure was developed based on Cattell’s personality sphere, condensed form.
This measure consists of 70 items with short descriptions. The respondents answer on a seven-point
scale (ranging from describes me very well to does not describe me at all) to what extent they recognize
themselves in the description. The internal consistencies were good, ranging from 0.80 to 0.88.
2.2. Study 2
2.2.1. Subjects
Participants filled out the survey in an assessment center at a Dutch consultancy firm. For two-
thirds of the group, this session was part of the selection procedure for a new job, for the
remaining one-third it was part of a personal development and coaching programme. Four-
hundred and twenty people from a diverse occupational background participated. Sixty per cent
had a college/university degree. The sample consisted of 130 female (31%) and 290 (69%) male
participants. Their mean age was 36 years (S.D.=8)
2.2.2. Measures
2.2.2.1. Scales of psychological well-being. As in study 1, the Dutch translation of the instrument
developed by Ryff (1989a) was used. Due to the necessity to keep the survey short, the shortened
9-items per scale version was used. This version was suggested by Ryff and is being used in an
ongoing longitudinal study.
2.2.2.2. Spiritual well-being. Spiritual well-being was also shortened to nine items, in line with the
scales of psychological well-being. These items were taken from the inner resources and trans-
cendence scales of the Spiritual Assessment Scale developed by Howden (1992).
2.2.2.3. Vitality. Vitality was measured with the seven items of the Subjective Vitality Scale
(Ryan & Frederick, 1997). The internal consistency was good (a=0.81).
2.2.2.4. Positive and negative affectivity. Positive and negative affectivity was measured with the
PANAS (Watson, Clark, & Tellegen, 1988). On a five-point scale, participants rate (ranging from
very slightly to not at all to extremely) certain feelings they experienced during the last couple of
weeks. This measure consists of 10 positive and 10 negative feelings. The internal consistencies
were good (a=0.85 and 0.68, respectively).
2.2.2.5. Burnout. Burnout was measured with the Dutch version of the MBI-General Survey,
which consists of three subscales: exhaustion, cynicism, and professional efficacy (Schaufeli, Lei-
ter, Maslach, & Jackson, 1996; Schaufeli & Van Dierendonck, 2000). The internal consistency of
all three scales was good: exhaustion (five items) a=0.88, cynicism (four items) a=0.79, profes-
sional efficacy (six items) a=0.74.
3. Results
Our first research question focused on the factorial validity of the Scales of Psychological Well-
being. As described in the introduction, Ryff formulated four different versions of the scales,
ranging from 20 to three items. In this article, we were able to test in two studies the psychometric
quality of the 3- and the 9-items version, and in one study the 14-items version.
Table 1 shows the internal consistencies of the six dimensions. The scales with 14-items all had
good reliabilities, with Cronbach alpha’s ranging from 0.77 to 0.90. The internal consistencies of
the 9-items scales showed a similar pattern. However, in study 1 the internal consistency of Per-
sonal growth was only 0.65. In study 2, the internal consistencies of Purpose in Life and Self-
acceptance appeared to be low, only 0.65 and 0.61, respectively. These are low figures, given the
fact that these are 9-item scales. In both studies the results of the 3-item scales appeared to be
most internally inconsistent. Several scales in both studies showed unacceptably low internal
consistencies, with an absolute low for Purpose in life (0.17 and 0.24!). The latter might come as a
surprise given that the alpha’s of this scale were 0.73 and 0.65 in the 9-item scales. This can be
explained by the fact that, in both studies, these three items showed the lowest item-total corre-
lations within the 9-item scales, ranging between 0.10 and 0.44. In view of the fact that Ryff chose
the 3-item scales on theoretical grounds, and that earlier results also showed low internal con-
sistencies, it is troublesome if the scales are to be used as variables in correlational analysis. Low
reliability diminishes the chance of finding significant correlations (Hunter & Schmidt, 1990).
The factorial validity of the Scales of Psychological Well-being was tested with confirmatory
factor analysis using LISREL 8.5 (Jöriskog & Sörbom, 2001). Covariance matrices were used as
input for the maximum likelihood modelling procedures. Following factor analytic models were
specified: (1) one-factor model which assumes that all items load on a general composite well-
being factor; (2) a two-factor model in which all negatively formulated items load on one factor
and all positively formulated items load on the second factor; (3) the original six-factor structure
as described by Ryff; (4) a five-factor structure where Self-acceptance and Environmental mastery
load on one factor; (5) a six-factor structure and a second-order latent construct called psycho-
logical well-being; (6) the five-factor structure and a second-order latent construct.
The goodness-of-fit of the models was evaluated using relative and absolute indices as recom-
mended by Hu and Bentler (1999). The absolute goodness-of-fit indices calculated were the chi-
square goodness-of-fit index and the standardized root-mean-square residual (SRMR). A value
Table 1
Internal consistency of the Scales of Psychological Well-being
14-items 9-items 3-items
Study 1 Study 1 Study 2 Study 1 Study 2
Self-acceptance 0.90 0.83 0.61 0.68 0.60

Positive relations 0.80 0.77 0.77 0.52 0.40
Autonomy 0.83 0.78 0.69 0.64 0.47
Environmental mastery 0.77 0.77 0.71 0.58 0.51
Purpose in life 0.84 0.73 0.65 0.17 0.24
Personal growth 0.82 0.65 0.69 0.53 0.50
of 0.08 or less is considered as indicating a relatively good fit for the SRMR. The relative good-
ness-of-fit indices computed were the comparative fit index (CFI) and the non-normed fit index
(NNFI or TLI). For both indices, values equal or greater than 0.95 are considered to indicate a
good fit. As a cautionary note, Hu and Bentler (1999) indicate that for smaller samples (i.e.
N4250) these values are harder to reach. Therefore, too strict adherence to these rules might
result in a higher Type II error rate. It should also be noted that the alternative models are nested
only within the stability model, not each other. The Akaike Information Criterion (AIC, Akaike,
1987) was therefore calculated as an additional measure to obtain information for comparing the
fit of alternative models (Cudeck & Browne, 1983). The AIC is always positive; lower values
indicate a better fitting model.
The data of study 1 were used to test the six models for the 14-, the 9- and the 3-item scales (see
Table 2). The results showed that the best fitting model in all versions was the six-factor model
with a single second order factor. However, for the 14- and 9-item versions, even the best com-
parative fit indices were unacceptably low. The fit of the 3-item versions was acceptable only if we
take into account the relative small sample size.
The data of study 2 could be used to test the models for the 9- and the 3-item scales (see
Table 3). As in study 1, the best fitting model was the six-factor model with a single second order
Table 2
Confirmatory factor analyses of study 1
Model w2 df AIC NNFI CFI SRMR
14-Item scales
1. Single factor 7000.53 3402 8802.48 0.53 0.54 0.08
2. Two factors, negative and positive item artifact 6892.41 3401 8528.58 0.54 0.55 0.16
3. Five factors 6706.35 3402 7000.05 0.57 0.58 0.09
4. Six factors 6873.51 3402 7385.28 0.54 0.56 0.19
5. Five factors, single second order factor 6159.61 3398 6535.59 0.64 0.65 0.09
6. Six factors, single second order factor 6097.81 3397 6408.72 0.65 0.65 0.09
9-Item scales
1. Single factor 3000.42 1377 3676.72 0.58 0.59 0.09
3. Five factors 3002.09 1377 3225.47 0.58 0.59 0.18
4. Six factors 3067.05 1377 3458.22 0.56 0.58 0.19
3-Item scales
1. Single factor 318.10 135 392.81 0.71 0.74 0.07
3. Five factors 353.84 135 436.80 0.65 0.69 0.14
4. Six factors 463.79 135 610.38 0.48 0.54 0.17
All w2 : P< 0.001.

Table 3
Confirmatory factor analyses of study 2
Model w2 df AIC NNFI CFI SRMR
9-Item scales
1. Single factor 3466.47 1377 4359.69 0.58 0.59 0.07
3. Five factors 3695.28 1377 4403.77 0.53 0.55 0.14
4. Six factors 3888.39 1377 4919.86 0.49 0.51 0.15
3-Item scales
1. Single factor 362.68 135 462.74 0.73 0.76 0.06
3. Five factors 592.48 135 759.50 0.46 0.53 0.14
4. Six factors 648.49 135 905.40 0.40 0.47 0.15
All w2: P< 0.001.
factor. Again, even for the best fitting model of the 9-items version the comparative fit indices
were unacceptably low, and reasonable for the 3-items per scale version.
To better understand the low fit indices, in both samples possible reasons for this lack of fit
were examined in model six of the 3-items per scale, using the modification indices provided by
LISREL. Although these indices indicated some possibilities to improve the fit, the suggestions
differed between the samples, and were as such probably highly sample specific. In addition, the
highest indices suggested allowing certain items to load on two factors simultaneously. This
points to problems with the measurement model.
Altogether, the conclusions from the reliability analyses and the confirmatory factor analyses
are ambiguous. To reach an acceptable internal consistency, scales should be longer, whereas an
(somewhat) acceptable factorial validity requires the scales to be short. To solve this dilemma,
new short scales were formulated. The purpose was to suggest relatively short scales that would
be internally consistent, comparable to the 9-items per scale as well as having a factorial validity
comparable with the level of the 3-items scales.
A selection of items was made on the basis of the highest item–total correlations (at least 0.30)
on their scale and low cross-loadings on other scales ( <0.40) in the factor analysis (Ferguson &
Cox, 1993). First, items were chosen that were suggested by the results of both datasets in order
to guarantee that the resulting solution would be as sample-unspecific as possible within the
context of this research. A complicating factor was that study 1 included for each scale five items
more than study 2. It proved impossible to come up with a viable solution for the six scales, using
only the 9-items per scale used in studies 1 and 2. By including 1–3 items of the additional items,
short scales were developed that both had good internal consistency and an acceptable factorial
validity.
The resulting short scales ranged from 6 to 8 items (see Table 4). Because study 2 did not
include all items of the short scales, no psychometric analyses were performed for this sample.
Table 4
Descriptive statistics, internal consistency of the short scales and correlation with 14-items scales, study 1
r14-items M S.D. S.E.
Self-acceptance 6-items 0.95 0.81 4.24 0.79 0.05

Positive relations with others 6-items 0.92 0.80 4.67 0.77 0.05
Autonomy 8-items 0.94 0.81 4.01 0.69 0.04
Environmental mastery 6-items 0.91 0.78 4.30 0.69 0.05
Purpose in life 6-items 0.92 0.81 4.28 0.77 0.05
Personal growth 7-items 0.92 0.72 4.67 0.53 0.03
The correlations of the shorter scales with the original 14-item scales were at least 0.91 (for Envir-
onmental mastery). For Self-acceptance this correlation was as high as 0.95. This indicates that the
short scales cover the essence of the longer scales. Also, the internal consistency of the scales was
good, at least 0.72 in study 1 (for Personal growth). A confirmatory factor analysis showed goodness
of fit indices similar to those of the 3-items scales (w2 (696)=1210.44, AIC=1406.82, NNFI=0.84,
CFI=0.85, SRMR =0.07). The modification indices suggested a possible improvement in the fit
of this model by allowing the error variance of three item-pairs to covariate. This resulted in a
reasonable fit (w2 (693)=1110.76, AIC=1302.20, NNFI=0.87, CFI=0.88, SRMR=0.06).
Our third research aim focused on the extension of the instrument with a separate dimension of
well-being, which referred to spirituality. In study 1, an exploratory factor analysis was per-
formed including the 39 items of the shortened scales and the 14 spirituality items. The results
showed seven factors with eigenvalues greater than one. The scree test also suggested seven fac-
tors. After a varimax rotation over the seven-factor solution, the majority of the spirituality items
divided themselves into two factors. This pattern was contrasted with that after an oblimin rota-
tion, which resulted in a similar result. These factors clearly differed from the SPWB dimensions.
As expected, one factor could best be called Inner resources, the other Relationship with a Higher
Power. In study 2, a factor analysis including 30 well-being items and nine spirituality items also
resulted in seven good interpretable factors. Here, spirituality formed one separate scale, similar
to the Inner resources scale in study 1. The items that formed the Relationship with a Higher
Power scale were not included in study 2. Those spirituality items that loaded 0.30 or higher on
one of the other six well-being factors in both studies were excluded, conceptual independence
thus being guaranteed. The resulting Inner Resources Scale consists of six items with an internal
consistency of 0.76 (M=4.04, S.D.=0.84, S.E.=0.06). The Relationship with a Higher Power
Scale consists of four items, with an internal consistency of 0.87 (M=2.33, S.D.=1.20,
S.E.=0.08). Appendix A describes the spirituality items.
To study the content validity of the short scales, a second order factor analytic model was
developed with exploratory factor analysis. In study 1, three other well-being instruments were
included that measured vitality, self-esteem and happiness, respectively. Furthermore, the Big
Five personality dimensions were added. The resulting analysis showed four interpretable factors
(see Table 5). Factor one can be described as the well-being factor, factor two as the self-actua-
lization factor, the third being relationship with others and finally autonomy as factor four.
In study 2, vitality, positive affectivity and negative affectivity and the three dimensions of
burnout were included in addition to the SPWB Scales. For the second order factor analysis, the
Table 5
Second order factor analysis with varimax rotation, study 1
I II III IV
Self-acceptance 0.85
Positive relations with others 0.54 0.47
Autonomy 0.49 0.61
Environmental mastery 0.75
Purpose in life 0.55 0.56
Personal growth 0.54 0.56
Inner resources 0.41 0.67
Relationship with a Higher Power 0.71
Vitality 0.73
Self-esteem 0.81
Happiness 0.77
Extraversion 0.77
Agreeableness 0.73
Conscientiousness 0.61
Neuroticism 0.64
Openness for experiences 0.87
Only factor loadings over 0.40 are depicted. Descriptions of the four factors: (I) well-being, (II) self-actualization, (III)
relationship with others, and (IV) autonomy.
short scales were used. Although in this study these scales did not include all the suggested items,
it was argued that at least in this way, conceptual overlap between scales due to conceptually
unclear items was avoided. Here, a two-factor structure emerged. One factor can be described as
positive affectivity, the other factor as negative affectivity. The positive affectivity factor consisted
of all the SPWB Scales, Inner resources, vitality, positive affectivity and the professional efficacy
dimension of burnout. The negative affectivity factor consisted of negative affectivity, vitality,
two SPWB Scales (i.e. Self-acceptance and Environmental mastery), and the exhaustion and
cynicism dimensions of burnout (Table 6).
4. Discussion
The primary purpose of this article was to examine the factorial and content validity of SPWB
in a sample of first year psychology students and in a sample of professionals from a diverse
occupational background. Three research questions were addressed: (1) can the six-factor struc-
ture of the SPWB be confirmed for the 14-, 9- and 3-items per scale versions? (2) To what extent
does the SPWB measure different underlying aspects of positive psychological health? (3) Can a
Spiritual Well-being Scale be specified that is conceptually different from the other six scales?
The answer to the first question on the factorial validity of the SPWB was unsatisfactory.
Results were only acceptable for the 3-items per scale version, reaching the cuttoff criterion for a
good fitting model for one of the fit indices (i.e. SRMR), but below that of the other two indices (i.e.
NNFI & CFI). Exploration of the modification indices showed that this lack of fit was primarily
due to problems with the measurement model. The fit indices for this version were comparable with
Table 6
Second order factor analysis with varimax rotation, study 2
I II
Self-acceptance 0.45 0.59

Positive relations with others 0.55
Autonomy 0.47
Environmental mastery 0.57 0.55
Purpose in life 0.73
Personal growth 0.69
Inner resources 0.61
Vitality 0.70 0.43
Positive affectivity 0.74
Negative affectivity 0.68
Exhaustion 0.71
Cynicism 0.67
Professional efficacy 0.56
Only factor loadings over 0.40 are depicted. Description of the two factors: (I) positive affectivity, and (II) negative
affectivity.
those reported by Ryff and Keyes (1995). Similarly, Kafka and Kozma (2002) failed to confirm
the six-factor structure for the 20-items per scale version. Regretfully, the internal consistencies of
these 3-items scales were unacceptably low, suggesting that these scales should not be used. The
selection of items on the basis of theoretical arguments is, of course, legitimate. However, for the
SPWB the internal consistency of the resulting scales is so low that it’s use in correlational
research is questionable. A low reliability will lead to an underestimation of the actual correlation
between constructs (Hunter & Schmidt, 1990). In the present article alternative short scales are
therefore suggested (see Appendix B) that showed good internal consistency and reasonable fac-
torial fit indices (with a SRMR value below the 0.08 threshold). According to Hu and Bentler
(1999), the SRMR is the most sensitive index to models with misspecified factor covariances or
latent structures. Four scales (i.e. Autonomy, Environmental Mastery, Personal Growth, & Self-
acceptance) included the three items of the original 3-items scales, that according to Ryff are
essential elements of that particular dimension. Regretfully, this was not the case for Positive
personal relations with others, where only one item could be used and Purpose in life that inclu-
ded none of the items from the 3-items version. The latter should come as no surprise given the
extreme low internal consistencies of this scale in both studies. However, even for the last two
scales the correlation of the short versions with the 14-items version was 0.92, satisfactorily
representing the essential content of that dimension.
With the introduction of the new short scales, the second research question can be answered
more positively. The results of the second order factor analysis dovetail those of previous studies
(Compton, 1998; Compton et al., 1996; Ryff, 1989a). Again, one subjective well-being factor
emerged on which all scales—except for Relationship with a Higher Power—loaded with subjective
well-being scales as vitality, happiness, self-esteem and neuroticism of the Big Five. The second
factor that emerged was also in line with the findings of earlier studies. This self-actualization
factor consisted of Purpose in Life, Personal Growth and both spirituality dimensions. It is
remarkable that among students, a greater striving for self-actualization goes with greater
conscientiousness. The content of this factor seems closely related to one of the three innate
psychological needs in self-determination theory, that is autonomy—‘the organismic desire to
self-organize experience and behavior to have activity be concordant with one’s integrated sense
of self.’ (Deci & Ryan, 2000, p. 231). The third factor referred to as interpersonal relations, con-
sisting of Positive relations with others from the SPWB and extraversion and agreeableness of the
Big Five. Here we can recognize a second innate psychological need according to self-determi-
nation theory, that is relatedness. There is some similarity with the secondary factor analytic
findings of Compton (2001) that also suggest a mature interpersonal relationships factor under-
lying psychological well-being measures. Moreover, our research revealed a possible fourth factor
underlying the SPWB, consisting of the autonomy scale of the SPWB and openness for experi-
ences of the Big Five. It should be noted that autonomy according to Ryff might be different from
autonomy according to self-determination theory. It seems that Ryff’s operationalization most
closely covers independence from others and individualism, something that has been argued by
Deci and Ryan (2000) as incorrectly equated with their view on autonomy. The data in this
research seem to support that position. It also points to the necessity to clearly define autonomy,
if and when we use this term.
It is interesting to compare these four empirically derived factors with the core features of
positive psychological health proposed by Ryff and Singer (1998). They argued that positive
psychological health—The Good Life—involves more than being happy, more than subjective
well-being. They proposed the following core features: leading a meaningful life and having
quality connections to others. The second order factor analysis seems to support this position.
Both dimensions could be identified as separate factors apart from subjective well-being. All in
all, the results suggest that it might be useful to combine these different theoretical positions
(hedonic well-being, The Good Life, and self-determination theory) into one model. It seems that
they have something in common, but also something unique.
The results of the second order factor analysis of study 2 provided insight into the high corre-
lations between Self-acceptance and Environmental mastery that several other studies reported
(Ryff & Keyes, 1995; Kafka & Kozma, 2002). These scales were the only two of the SPWB that
loaded on the negative affectivity factor. It can therefore be speculated that it is the negative
affectivity content in the operationalization of these scales that caused them to cluster together.
Nevertheless, the outcomes of the confirmatory factor analysis supported Ryff and Keyes’s (1995)
conclusion that although these scales may be closely related, they exhibit different elements within
the well-being domain.
This study also made a contribution towards incorporating spiritual well-being into our model
of positive psychological health. The two scales of Spiritual Well-being, Inner resources and
Relationship with a Higher Power, strengthened the self-actualization dimension in the second
order factor analysis in study 1. One might even speculate if these elements put forward the role of
the Daimon, or ‘‘true self’’, as described by Waterman (1993) in his formulation of eudaimonism. At
this point it is noteworthy that the Autonomy Scale of the SPWB (signifying being self-determinant
and independent) is not an element of this dimension. It may well be that being self-determinant is
no guarantee for personal expressiveness or self-actualization, whereas experiencing more inner
resources helps in becoming and living true to oneself. If this is so, it should be considered as an
important element in programmes aimed at enhancing optimal psychological health. A recent study
by Van Dierendonck, Garssen, and Visser (submitted for publication) demonstrated that inner
resources could be enhanced using a programme based on psychosynthesis. This is certainly an

area that requires further study.
The limitations of the present study ask for caution generalizing the results. Although two dif-
ferent samples were employed, there were items that were excluded in one study that turned out
to be essential elements of the short scales. More research in different occupational and age
groups is therefore needed to confirm the psychometric quality of the suggested short scales and
the two Spiritual Well-being Scales. A second limitation is the relatively small sample sizes,
especially in study 1. This limited the possibility of reaching the 0.95 cutoff score for the relative
fit indices (NNFI & CFI; Hu & Bentler, 1999). Another point of caution is the interpretation of
the spirituality scales. In recent years, a variety of new measures have been developed (Slater,
Hall, & Edwards, 2001). It would be useful to specifically study how the new developed scales
relate to other measures that were developed from similar or different theoretical perspectives.
Despite these limitations, the short scales suggested in this article may be considered a viable
alternative for studies that aim to examine positive psychological health from the theoretical
perspective suggested by Ryff (1989a, 1989b; Ryff & Singer, 1998), especially in those situations
where the longer scales (i.e. 9- or 14-items) would take up too much space.
Acknowledgements
The author gratefully acknowledges Lydia Smith, Marieke Rucker and Alec Serlie for their
help with the data collection.
Appendix A. Spiritual Well-being
Inner resources
I can turn to a spiritual dimension within myself for guidance.

I have an inner strength.
I have experienced my own strength in times of struggle.
I have a sense of harmony or inner peace.
My innerness or an inner resource helps me deal with uncertainty in life.
I rely on an inner strength in hard times.
Relationship with a Higher Power
I experience a spiritual dimension that gives me strength and love.

Frequently meditating or praying gives a sense of inner peace.
I get personal strength and support from my God or a Higher Power.
I have a personally meaningful relationship with God or a Higher Power.
Appendix B. Scales of Psychological Well-being, short version
The numbers below refer to the number within the scale in Ryff’s 14-items per scale version.
Self-acceptance: 1, 2, 4, 5, 7, 8.
Positive relations with others: 3, 6, 7, 8, 10, 12.
Autonomy: 2, 4, 6, 9, 10, 11, 13, 14.
Environmental mastery: 1, 2, 4, 6, 13, 14.
Purpose in life: 1, 4, 6, 8, 9, 12.
Personal growth: 2, 4, 5, 6, 9, 11, 13.
References
Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52, 317–332.

Compton, W. C. (1998). Measures of mental health and a five factor theory of personality. Psychological Reports, 83,
371–381.
Compton, W. C. (2001). Towards a tripartite factor structure of mental health: subjective well-being, personal growth,
and religiosity. The Journal of Psychology, 135, 486–500.
Compton, W., Smith, M., Cornish, K., & Qualls, D. (1996). Factor structure of mental health measures. Journal of
Personality and Social Psychology, 71, 406–413.
Costa, P. T., & McCrae, R. R. (1980). Influence of extraversion and neuroticism on subjective well-being: happy and
unhappy people. Journal of Personality and Social Psychology, 38, 668–678.
Cudeck, R., & Browne, M. W. (1983). Cross-validation of covariance structures. Multivariate Behavioral Research, 18,
147–167.
Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and ‘‘why’’ of goal pursuits: human needs and the self-determination of
behavior. Psychological Inquiry, 11, 227–268.
DeNeve, K. M., & Cooper, H. (1998). The happy personality: a meta-analysis of 137 personality traits and subjective
well-being. Psychological Bulletin, 124, 197–229.
Diener, E., & Lucas, R. E. (1999). Personality and Subjective Well-being. In D. Kahneman, E. Diener, & N. Schwarz
(Eds.), Well-being. The foundations of hedonic psychology (pp. 213–229). New York: Russel Sage Foundation.
Ellison, C. W. (1983). Spiritual well-being: conceptualization and measurement. Journal of Psychology and Theology,
11, 330–340.
Ellison, C. W., & Smith, J. (1991). Toward an integrative measure of health and well-being. Journal of Psychology and
Theology, 19, 35–48.
Elshout, J. J., & Akkerman, A. (1974). Vijf persoonlijkheidsfactoren (5PFT). [Five personality factors (5PFT)].
Nijmegen: Berkhout.
Ferguson, E., & Cox, T. (1993). Exploratory factor analysis: a user’s guide. International Journal of Selection and
Assessment, 2, 84–94.
Fordyce, M. W. (1988). A review of research on the happiness measures: a sixty second index of happiness and mental
health. Social Indicators Research, 20, 355–381.
Hawley, J. (1993). Reawaking the spirit in work: the power of dharmic management. San Francisco: Berret-Koehler
Publishers.
Howden, J. W. (1992). Development and psychometric characteristics of the spirituality assessment scale. Unpublished
doctoral dissertation, Texas Woman’s University, Denton, TX.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria
versus new alternatives. Structural Equation Modeling, 6, 1–55.
Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis. Correcting error and bias in research findings.
Newbury Park: Sage Publications.
Jöriskog, K., & Sörbom, D. (2001). LISREL 8: user’s reference guide. Lincolnwood: Scientific Software International.
Kafka, G. J., & Kozma, A. (2002). The construct validity of Ryff’s scales of psychological well-being (SPWB) and their
relationship to measures of subjective well-being. Social Indicators Research, 57, 171–190.
Kahneman, D., Diener, E., Schwarz, N. (Eds.). (1999). Well-being. The foundations of hedonic psychology. New York:
Russel Sage Foundation.
Keyes, C. L. M., & Lopez, S. J. (2002). Toward a science of mental health. In C. R. Snyder, & S. J. Lopez (Eds.),
Handbook of positive psychology (pp. 45–59). Oxford: Oxford University Press.
Koenig, H. G. (1998). Handbook of religion and mental health. San Diego, CA: Academic Press.
Moberg, D. O. (1984). Subjective measures of spiritual well-being. Review of Religious Research, 25, 351–359.
Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001). Measuring global self-esteem: construct validity of a
single-item measure and the Rosenberg self-esteem scale. Personality and Social Psychology Bulletin, 27, 151–161.
Ryan, R. M., & Frederick, C. (1997). On energy, personality, and health: subjective vitality as a dynamic reflection of
well-being. Journal of Personality, 65, 529–565.
Ryff, C. D. (1989a). Beyond Ponce de Leon and life satisfaction: new directions in quest of successful aging. Interna-
tional Journal of Behavioral Development, 12, 35–55.
Ryff, C. D. (1989b). Happiness is everything, or is it? Explorations on the meaning of psychological well-being. Journal
of Personality and Social Psychology, 57, 1069–1081.
Ryff, C. D., & Essex, M. J. (1992). The interpretation of life experience and well-being: The sample case of relocation.
Psychology & Aging, 7, 507–517.
Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological well-being revisited. Journal of Personality and
Social Psychology, 69, 719–727.
Ryff, C. D., Lee, Y. H., Essex, M. J., & Schmutte, P. S. (1994). My children and me: midlife evaluations of grown
children and self. Psychology and Aging, 9, 195–205.
Ryff, C. D., & Singer, B. (1998). The contours of positive human health. Psychological Inquiry, 9, 1–28.
Schaufeli, W. B., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996). The Maslach Burnout Inventory—general survey.
In C. Maslach, S. E. Jackson, & M. P. Leiter (Eds.), Maslach Burnout Inventory manual (3rd ed.) (pp. 19–25). Palo
Alto, CA: Consulting Psychologists Press.
Schaufeli, W. B., & Van Dierendonck, D. (2000). Utrechtse Burnout Schaal (UBOS), handleiding (Utrecht Burnout
Scale, manual). Lisse: Swets & Zeitlinger.
Schmutte, P. S., & Ryff, C. D. (1997). Personality and well-being: reexamining methods and meanings. Journal of
Personality and Social Psychology, 3, 549–559.
Slater, W., Hall, T. W., & Edwards, K. J. (2001). Measuring religion and spirituality: where are we and where are we
going. Journal of Psychology and Theology, 29, 4–21.
Van Dierendonck, D., Garssen, B., & Visser, A. Burnout prevention through personal growth: a study from a transper-
sonal perspective (submitted for publication).
Waterman, A. S. (1993). Two conceptions of Happiness: contrast of personal expressiveness (Eudaimonia) and
Hedonic Enjoyment. Journal of Personality and Social Psychology, 64, 678–691.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and nega-
tive affect: the PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070.

Vandierendonck2004 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Vandierendonck2004 PDF

Uploaded by

Copyright:

Available Formats

Personality and Individual Differences 36 (2005) 629–643

The construct validity of Ryﬀ’s Scales of Psychological

14-items 9-items 3-items

Study 1 Study 1 Study 2 Study 1 Study 2

Self-acceptance 0.90 0.83 0.61 0.68 0.60

Model w2 df AIC NNFI CFI SRMR

All w2 : P< 0.001.

Model w2 df AIC NNFI CFI SRMR

All w2: P< 0.001.

r14-items M S.D. S.E.

Self-acceptance 6-items 0.95 0.81 4.24 0.79 0.05

Self-acceptance 0.45 0.59

resources could be enhanced using a programme based on psychosynthesis. This is certainly an

Appendix A. Spiritual Well-being

I can turn to a spiritual dimension within myself for guidance.

Relationship with a Higher Power

I experience a spiritual dimension that gives me strength and love.

Appendix B. Scales of Psychological Well-being, short version

Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52, 317–332.

You might also like