J Happiness Stud (2012) 13:455–467

DOI 10.1007/s10902-011-9273-7

Predicting Well-Being From Personality in Adolescents

and Older Adults

Ana Butkovic • Irma Brkovic • Denis Bratko

Published online: 6 May 2011

Ó Springer Science+Business Media B.V. 2011

Abstract The relationship between personality and well-being is dynamic and therefore
should be examined within aging perspective. In study presented here we compared the
prediction of well-being from personality in two samples–223 adolescents attending high-
school (16–19 years; M = 16.94, SD = 0.51) and 134 older adults (54–90 years;
M = 77.20, SD = 7.39) living in retirement home or at home. Different aspects of well-
being were included in the study. Subjective well-being was measured with Index of Well-
Being, but also self-esteem and loneliness were included as measures of psychological
well-being. Personality (measured with FFPI; Hendriks et al. in Pers Individ Dif
27:307–325, 1999) significantly predicted indicators of well-being (Self-Esteem, Loneli-
ness and Index of Well-Being) in both samples, with emotional stability and extraversion
as most pronounced predictors. Personality explained more variance of well-being in
adolescents than in older adults, and more variance in eudaimonic than hedonic aspects of
well-being.

Keywords Personality
Well-Being
Adolescents
Older adults

1 Introduction

Well-being has been of interest in psychology for some time now, but there are still
differences in opinion on what should be measured as well-being. Well-being is usually
associated with subjective well-being which reflects what people think and feel about their
life (Lucas and Diener 2008). Because it includes both thinking and feeling, subjective

A. Butkovic (&)
I. Brkovic
D. Bratko

Department of Psychology, Faculty of Humanities and Social Sciences, University of Zagreb, Ivana
Lucica 3, 10 000 Zagreb, Croatia
e-mail: [email protected]
I. Brkovic
e-mail: [email protected]
D. Bratko
e-mail: [email protected]

123
456 A. Butkovic et al.

well-being is divided into cognitive and affective component. The cognitive component
refers to individual’s reflective judgment that his or her life is going well and is usually
assessed with measures of life satisfaction. The affective component refers to emotions
experienced in life with the idea that ‘‘life is good’’ if an individual experiences more
positive than negative emotions. This component is measured in different ways that include
experience sampling and measures of positive and negative affect. This tradition of
research on well-being has been referred to as hedonistic tradition, focusing on happiness
(Deci and Ryan 2008). Some authors believe that well-being consists of more than just
happiness and that, if people report of being happy, it does not necessarily mean they are
psychologically well. Second tradition of research on well-being, referred to as eudaimonic
tradition, is concerned with living well or actualizing one’s human potential (Deci and
Ryan 2008). An example of eudaimonic tradition of research on well-being is Ryff’s
(1989a) model and measure of psychological well-being. Ryff’s approach names six
characteristics of psychological well-being—self-acceptance, personal growth, autonomy,
environmental mastery, positive relations with others and purpose in life.
Both views on well-being contribute to its understanding. Integration of these two
traditions could improve our understanding of this concept because only affective com-
ponent of subjective well-being represents hedonic aspect of well-being, while cognitive
component is closer to eudaimonic perspective. Secondly, in a few recent studies that
assessed both hedonic and eudaimonic aspects of well-being substantial overlap between
them was found. Keyes et al. (2002) found the bivariate correlation between the summed
scale of subjective well-being and the summed scale of psychological well-being of .59,
while Waterman et al. (2008) found that measures of hedonic and eudaimonic aspects of
well-being shared between 68 and 76% of common variance.
Two different types of models were used to try to explain what predicts well-being: top-
down models, which state that personality traits predict well-being and bottom-up models,
which emphasize environmental influences and state that contextual factors predict well-
being. Research done so far indicate that both models are needed to explain well-being,
since both traits and sociodemographic variables (e.g. age, gender, marital status, SES,
race/ethnicity) account for variance in well-being outcomes, but to a different degree (Ryff
2008). When it comes to personality traits, the picture is quite clear. Two meta-analyses
(DeNeve and Cooper 1998; Steel et al. 2008) showed that neuroticism is the trait most
strongly associated with negative affect as an indicator of poor well-being, while extra-
version is associated with positive affect indicating higher well-being. As for other traits
from the five-factor model, these meta-analyses have shown that agreeableness and con-
scientiousness are positively associated with well-being, but less strongly than neuroticism
and extraversion, while, for openness to experience, effects were the smallest although also
in direction of positive association with well-being. These two meta-analyses differ
somewhat in effect sizes. While DeNeve and Cooper (1998) found significant but weak
correlations between personality and well being (e.g. for extraversion .17, for neuroticism
-.22), Steel et al. (2008) found much higher correlations using three widely used per-
sonality inventories: the NEO Personality Inventory—Revised (NEO-PI-R; Costa and
McCrae 1992), the Eysenck Personality Inventory (EPI; Eysenck and Eysenck 1964), and
the Eysenck Personality Questionnaire (EPQ; Eysenck and Eysenck 1975). The correla-
tions between neuroticism and negative affect were .54 for the NEO-PI-R, .56 for the EPQ,
and .46 for the EPI, and between extraversion and positive affect .44 for the NEO-PI-R, .35
for the EPQ, and .25 for the EPI. The latter meta-analysis therefore indicates that per-
sonality can explain almost 40% of variance in subjective well-being.

123
Personality and Well-Being 457

As for sociodemographic variables and subjective well-being, Lucas and Diener (2008)
report that different authors found that sociodemographic variables can account at most for
10–15% of variance in subjective well-being. Most studies used only subjective well-
being, but Keyes et al. (2002), for example, used both psychological (eudaimonic) and
subjective (hedonic ? life satisfaction) well-being and showed they are differentially
related to sociodemographic variables and personality. In their study older adults with
lower levels of education showed higher levels of subjective but lower levels of psycho-
logical well-being, whereas younger adults with more education and higher levels of
openness to experience showed higher levels of psychological and lower levels of sub-
jective well-being. Educational level has been shown to be a strong correlate of eudai-
monic well-being. As for personality, they showed, using discriminant analytic strategy,
that individuals who are low on neuroticism and high on extraversion and conscien-
tiousness tend to be the ones with high eudaimonic and hedonic well-being, while indi-
viduals who are high on openness to experience tend to have greater eudaimonic than
hedonic well-being.
The relationship between personality and well-being is not static, and, as Ryff (2008)
points out, studies at the interface of personality, well-being and aging are needed.
Research on personality and well-being separately has indicated that both show stability
and change with aging. For personality, Roberts, and DelVecchio (2000) have shown in
their meta-analysis that there is evidence for rank-order stability of the Big Five traits,
measured by both self-reports and observer ratings, which increases with age. Roberts et al.
(2006) showed in their meta-analysis that there are significant mean-level changes in
personality during adulthood with increases in social dominance (a facet of extraversion),
conscientiousness, and emotional stability especially in young adulthood, increases in
social vitality (another facet of extraversion) and openness in adolescence followed by
decreases in old age, and increase in agreeableness in old age. An awareness of the
different factors that contribute to personality traits and stability comes from behavior-
genetic studies which indicate that heritability of personality is in the 30–50% range and
that genetic factors mainly contribute to stability in personality (Plomin et al. 2008). There
are fewer studies about heritability of well-being than personality, but a recent meta-
analysis indicated that well-being is heritable 39 or 32% after correction for publication
bias (Vukasovic et al. in press). Nes et al. (2006) showed that long-term stability of
subjective well-being was also mainly attributable to stable genetic factors.
Studies of life satisfaction and hedonic well-being using both cross-sectional and lon-
gitudinal designs showed either negligible age differences in well-being or age increments
in life satisfaction and positive affect, concomitant with age decrements in negative affect
(Ryff 2008). For eudaimonic well-being, results have shown somewhat different picture.
Multiple studies have shown little age variation in positive relations with others and self-
acceptance, increases with age for autonomy and environmental mastery, but sharp
decrease from young adulthood to old age for self-rated purpose in life and personal
growth (Ryff 2008). It seems that different aspects of well-being change differently with
age and also that it depends on how these different aspects are measured. For example, for
self-acceptance no change with age was found, but lifespan studies of self-esteem have
shown that self-esteem is relatively high in childhood, it drops in adolescence then
increases in young and middle adulthood, reaches a peak around 60 years and then declines
in old age (Orth et al. 2010; Robins and Trzesniewski 2005; Robins et al. 2002).
In a recent study Gomez et al. (2009) looked at the influence of personality and life
events on subjective well-being from a life span perspective using three age groups—
young, middle-aged, and old adults. For the total sample, only neuroticism together with

123
458 A. Butkovic et al.

life events predicted well-being. When age groups were compared, results showed that
with increasing age neuroticism becomes the dominant personality-related predictor for
well-being, while the influence of extraversion on well-being is only existent in young
adults and disappears with increasing age. Although the level of well-being might be the
same across age, it seems that what contributes to our well-being might be different at
different ages. This would make sense because our life tasks and what is important for our
well-being changes in different life periods. For example, Ryff (1989b) looked at people’s
conceptions of well-being among middle-aged and older adults and found that both groups
endorse good relationships and the pursuit of enjoyable activities as important for well-
being, but middle-aged adults focused more on self-knowledge, competence, and self-
acceptance, and older adults focused more on accepting change. The difference between
adolescents and older adults in what is important for well-being could be even bigger.
In our study we looked at the relationship between personality and well-being in two
samples differing in age with adolescents included in one sample and retirees included in
the other sample. Sociodemographic variables (age, gender, education and marital status)
were also included in our analyses. We tested the hypothesis that personality would be
more important as a predictor of well-being at younger age. In older age we expected that
other variables (e.g. marital status, education) will contribute to well-being together with
personality. Also, different aspects of well-being were included in the study. We looked at
subjective well-being, but also included self-esteem and loneliness as measures of psy-
chological well-being. There are not many studies that looked at both hedonic and eu-
daimonic aspects of well-being. Recently, Grant et al. (2009) looked at Big Five
personality traits as predictors of subjective and psychological well-being in a sample of
managers aged between 20 and 61 years. They found that overall relationship between
personality and well-being was larger for psychological well-being than for subjective
well-being. Also, neuroticism showed similar strong relationship to both subjective and
psychological well-being, while extraversion and conscientiousness were correlated more
with psychological well-being than with subjective well-being. In our study we test if this
relationship can be found on a different sample and in different age groups. We hypoth-
esize that personality will be more strongly associated with psychological well-being than
subjective well-being in both age groups.
Research presented here contributes to the field of well-being research in two ways. On
one hand we applied two approaches to well-being, namely hedonic and eudaimonic
approach, and on the other hand we compared two developmentally distinct age groups in
our study of relationship between personality and well-being.

2 Method

Data used in this paper was collected as part of the larger project Generational differences
in personality and measures of psychological well-being. Two groups of participants were
included in the study—high-school students aged 16–19 years (M = 16.94, SD = 0.51)
and older adults aged 54–90 years (M = 77.20, SD = 7.39). High-school sample included
223 high-school students (82% female) from two high-schools in Zagreb, while older
adults sample included 134 retirees (76% female), of whom 68 (51%) lived in retirement
home and others at home.
All participants filled in a questionnaire booklet that included demographic data, Five
Factor Personality Inventory (FFPI; Hendriks et al. 1999), Index of Well-Being (Campbell
et al. 1976), The Self-Esteem Scale (Rosenberg 1965) and shortened UCLA Loneliness

123
Personality and Well-Being 459

Scale (Russell and Cutrona 1988). The Five-Factor Personality Inventory (FFPI) was
constructed interactively in Dutch, English and German versions to assess the Big Five
factors of personality both for self-ratings and for others ratings. In this study the Croatian
version was used. It consists of 100 brief and concrete statements and gives the following
five factor scores: Extraversion, Agreeableness, Conscientiousness, Emotional stability and
Autonomy. Index of Well-Being is a two-part self-rated measure assessing general affect
and life satisfaction (Campbell et al. 1976). The Index of Well-Being is a sum of the
average score on Index of General Affect, composed of 8 items in semantic differential
format as the opposite poles of 7-point scales, and a single-item assessment of life satis-
faction weighed 1.1 in the combination score. The Self-Esteem Scale (Rosenberg 1965)
was developed as a measure of global self-esteem in adolescents. Originally, the scale was
a ten item Likert type scale with items answered on a four point scale–from 1 (strongly
agree) to 4 (strongly disagree). However, different authors have used different number of
points in the response scale or used different scaling methods (e.g. Guttman scaling or
summing responses across the questions). In this study a 5-point scale was used and a total
score was obtained by summing responses across the questions. The UCLA Loneliness
Scale (Russell and Cutrona 1988) is a 20-item self-report questionnaire designed to assess
loneliness, conceptualized as a unidimensional emotional response to discrepancy between
desired and actual levels of social contact. For this study this scale was shortened to 11
items by keeping those that were clearer, simpler and unambiguous. Each item is
accompanied by a four-point frequency scale: 1 (I never feel this way); 2 (I rarely feel this
way); 3 (I sometimes feel this way); and 4 (I often feel this way). Total score is obtained by
summing responses across items.
Data for high-school students was collected during classes in schools in group settings,
while data for older adults was collected individually. Both high-school students and older
adults were asked first to respond to demographic questions, then to FFPI, and finally to
Index of Well-Being, The Self-Esteem Scale and The UCLA Loneliness Scale. Reliability
coefficients were calculated for all scales and they were in line with previous studies. For
extraversion Cronbach a’s were .92 for adolescents and .83 for older adults, for agree-
ableness .82 for adolescents and .72 for older adults, for conscientiousness .85 for ado-
lescents and .75 for older adults, for emotional stability .89 for adolescents and .88 for
older adults, and for autonomy .85 for adolescents and .76 for older adults. For well-being
they were .88 for adolescents and .82 for older adults, for self-esteem .86 for adolescents
and .71 for older adults, and for loneliness .88 for adolescents and .78 for older adults.

3 Results

Preliminary analyses were run to test normality of distributions and Kolmogorov–Smirnov

test showed that all our results are distributed normally. Descriptive statistics for all our
measures are presented in Table 1.
We compared two age groups in results on three well-being measures and found no
significant differences between groups for well-being (F[1,355] = .02, p = .89), self-
esteem (F[1,355] = .22, p = .64) and loneliness (F[1,355] = 1.14, p = .29). Also, we
calculated correlations between three measures of psychological well-being (Index of
Well-Being, The Self-Esteem Scale and The UCLA Loneliness Scale) in both samples and
they are presented in Table 2. In all our analyses p \ .01 was used as level of significance.
Since in both samples three measures of psychological well-being correlated significantly,
we factor analyzed them in both samples using PCA. In adolescents sample one factor with

123
460 A. Butkovic et al.

Table 1 Descriptive statistics

Adolescents M (SD) Older adults
for well-being, self-esteem,
M (SD)
loneliness and personality in
adolescents and older adults
Well-being 10.45 (2.44) 10.42 (2.09)
Self-esteem 38.20 (7.39) 38.55 (5.78)
Loneliness 21.44 (5.45) 20.84 (4.77)
Extraversion 70.27 (13.57) 64.14 (11.21)
Agreeableness 69.26 (9.09) 74.72 (8.50)
Conscientiousness 63.07 (10.46) 78.25 (7.92)
Emotional stability 66.73 (12.23) 67.84 (12.72)
Autonomy 70.27 (11.34) 64.91 (9.79)

Table 2 Intercorrelations between three measures of psychological well-being

Well-Being Self-esteem Loneliness

Well-being .66 -.60

Self-esteem .46 -.56
Loneliness -.45 -.56
All correlations significant at p \ .01; adolescents above the diagonal, older adults below the diagonal

an Eigenvalue of 2.22 accounted for 73.82% of the variance, and in older adults sample
one factor with an Eigenvalue of 1.98 accounted for 66.08% of the variance. Therefore, in
both samples we retained factor score labeled as psychological well-being and used it as
criteria in our analyses in addition to scores on separate scales.
Next we calculated correlations between personality factors and well-being, self-
esteem, loneliness and psychological well-being in both samples and they are presented in
Table 3. For adolescents, three factors of personality are consistently correlated with our
measures of psychological well-being—extraversion, emotional stability and autonomy.
They correlated positively with well-being, self-esteem and psychological well-being, and
negatively with loneliness. Highest correlations with well-being, self-esteem and psy-
chological well-being were obtained for emotional stability, while with loneliness highest
correlation was for extraversion. For older adults, significant correlations were obtained for
all personality factors and our measures of psychological well-being, with an exception of
nonsignificant correlations for agreeableness with well-being and self-esteem, and between
conscientiousness and well-being. Again, positive correlations were obtained with well-
being, self-esteem and psychological well-being, and negative with loneliness. Although
our measures of well-being correlated in older adults with all personality factors, corre-
lations were higher with the same three factors (extraversion, emotional stability, auton-
omy) as in adolescents.
Final step in our analyses were hierarchical regression analyses to explore predictive-
ness of personality for well-being. Analyses were carried out separately for adolescents and
older adults. In the first step for adolescents we entered gender and age as predictors and
for older adults we entered gender, age, education and marital status as predictors, while in
the second step we entered five personality dimensions. As criteria we used well-being,
self-esteem, loneliness and psychological well-being. Results for adolescents are presented
in Table 4, and for older adults in Table 5.
As can be seen from Table 4, first model with age and gender has not significantly
explained the variance for none of our four criteria. Adding personality traits in the second

123
Personality and Well-Being 461

Table 3 Correlations between

Sample Personality Well- Self- Loneliness Psychological
personality factors and well-
factors being esteem well-being
being, self-esteem, loneliness and
psychological well-being in ado-
lescents and older adults Adolescents
E .40* .49* -.70* .61*
A -.07 -.10 .01 -.07
C .03 .06 .07 .01
ES .61* .68* -.51* .70*
Au .37* .48* -.40* .48*
Older adults
E .43* .38* -.52* .55*
A .15 .18 -.30* .26*
E extraversion, A agreeableness, C .20 .36* -.32* .36*
C conscientiousness, ES ES .50* .45* -.49* .59*
emotional stability, Au autonomy
Au .36* .46* -.40* .50*
* p \ .01

model changed significantly the amount of explained variance in all four criteria. For well-
being we explained 38% of the variance with sociodemographic and personality variables,
with only emotional stability being a significant predictor (b = .56, p \ .01). For self-
esteem, entered predictors explained 50% of the variance, with emotional stability as a
single significant predictor (b = .55, p \ .01). Our model explained 54% of the variance in
loneliness, with extraversion (b = -.66, p \ .01), emotional stability (b = -.24, p \ .01)
and agreeableness (b = -.13, p \ .01) as significant predictors. Finally, 58% of variance
in psychological well-being was explained, with emotional stability (b = .53, p \ .01) and
extraversion (b = .35, p \ .01) as significant predictors.
In older adults, as can be seen from Table 5, first model with age, gender, education and
marital status explained significantly the variance in self-esteem (F[4,129] = 4.39,
p \ .01) and marginally the variance in psychological well-being (F[4,129] = 3.34,
p = 0.012). In the first step education was a significant predictor of self-esteem (b = .27,
p \ .01) and psychological well-being (b = .24, p \ .01), but when personality traits were
entered in the second step it stayed only marginally significant for self-esteem (b = .20,
p = .011). Our final model explained 28% of the variance in well-being with emotional
stability (b = .36, p \ .01) and extraversion (b = .27, p \ .01) as significant predictors.
For self-esteem 33% of the variance was explained with our model and conscientiousness
was an only significant predictor (b = .23, p \ .01). However, both education (b = .20,
p = .011) and emotional stability (b = .21, p = .028) contributed highly to the amount of
explained variance in self-esteem. Sociodemographic and personality variables explained
39% of the variance in loneliness and 49% of the variance in psychological well-being with
the same significant predictors, extraversion (loneliness: b = -.38, p \ .01; psychological
well-being: b = .33, p \ .01) and emotional stability (loneliness: b = -.29, p \ .01;
psychological well-being: b = .35, p \ .01).

4 Discussion

This study had the following goals: to measure both hedonic and eudaimonic aspects of
well-being and to explore predictivness of sociodemographic variables and personality for
different aspects of well-being in two age samples, adolescents and older adults. Obtained

123
462 A. Butkovic et al.

Table 4 Results of hierarchical

Criteria Predictors Beta t Summary of
regression analyses for the model
adolescents
Well-being Model 1
Age .10 1.55 Adjusted R2 = .00
Gender .05 .78 F(2,220) = 1.42
Model 2
Age .08 1.48 Adjusted R2 = .38
Gender .11 2.04 F(7,215) = 20.73*
DR2 = .39
Extraversion .11 1.53 FDR2
Agreeableness -.07 -1.16 (5,215) = 28.10*
Conscientiousness .01 .22
Emotional .56 8.53*
stability
Autonomy .00 .02
Self-esteem Model 1
Age -.01 -.18 Adjusted R2 = .00
Gender -.04 -.60 F(2,220) = .19
Model 2
Age -.05 -.99 Adjusted R2 = .50
Gender .01 .25 F(7,215) = 32.23*
DR2 = .51
Extraversion .15 2.40 FDR2
Agreeableness -.05 -.87 (5,215) = 44.98*
Conscientiousness .07 1.39
Emotional .55 9.20*
stability
Autonomy .12 1.80
Loneliness Model 1
Age -.01 -.18 Adjusted R2 = .00
Gender -.07 -.98 F(2,220) = .48
Model 2
Age .01 .21 Adjusted R2 = .54
Gender -.03 -.55 F(7,215) = 37.81*
DR2 = .55
Extraversion -.66 -10.89* FDR2
Agreeableness -.13 -2.62* (5,215) = 52.51*
Conscientiousness .05 1.02
Emotional -.24 -4.23*
stability
Autonomy .09 1.49
Psychological Model 1
well-being
Age .04 .61 Adjusted R2 = .00
Gender .03 .44 F(2,220) = .27
Model 2
Age .01 .21 Adjusted R2 = .58
Gender .06 1.30 F(7,215) = 44.20*
Beta value of the standardized DR2 = .59
Extraversion .35 6.03* FDR2
regression coefficient, t value of t test,
adjusted R2 adjusted total explained Agreeableness .01 .10 (5,215) = 61.63*
variance, F value of total F test, DR2 Conscientiousness .01 .30
contribution of personality traits for
Emotional .53 9.67*
explaining variance, FDR2 value of
stability
F test for personality traits
Autonomy .01 .19
* p \ .01

123
Personality and Well-Being 463

Table 5 Results of hierarchical regression analyses for older adults

Criteria Predictors Beta t Summary of the model

Well-being Model 1
Age .00 .02 Adjusted R2 = .05
Gender -.12 -1.32 F(4,129) = 2.57
Education .20 2.28
Marital status -.07 -.73
Model 2
Age .04 .50 Adjusted R2 = .28
Gender .00 .01 F(9,124) = 6.70*
DR2 = .25
Education .11 1.33 FDR2 (5,124) = 9.35*
Marital status -.04 -.45
Extraversion .27 3.24*
Agreeableness .01 .07
Conscientiousness .03 .30
Emotional stability .36 3.66*
Autonomy .01 -.12
Self-esteem Model 1
Age -.15 -1.58 Adjusted R2 = .09
Gender -.02 -.22 F(4,129) = 4.39*
Education .27 3.15*
Marital status -.10 -1.08
Model 2
Age -.12 -1.44 Adjusted R2 = .33
Gender .10 1.31 F(9,124) = 8.42*
DR2 = .26
Education .20 2.60 FDR2 (5,124) = 10.38*
Marital status -.07 -.82
Extraversion .16 1.95
Agreeableness -.05 -.58
Conscientiousness .23 2.84*
Emotional stability .21 2.23
Autonomy .15 1.55
Loneliness Model 1
Age .00 -.01 Adjusted R2 = .01
Gender -.05 -.50 F(4,129) = 1.19
Education -.12 -1.37
Marital status .15 1.51
Model 2
Age -.05 -.71 Adjusted R2 = .39
Gender -.17 -2.32 F(9,124) = 10.61*
DR2 = .40
Education -.01 -.20 FDR2 (5,124) = 17.54*
Marital status .09 1.22
Extraversion -.38 -4.91*
Agreeableness -.12 -1.56
Conscientiousness -.10 -1.26
Emotional stability -.29 -3.22*
Autonomy -.04 -.40

123
464 A. Butkovic et al.

Table 5 continued

Criteria Predictors Beta t Summary of the model

Psychological well-being Model 1

Age -.06 -.64 Adjusted R2 = .07
Gender -.04 -.39 F(4,129) = 3.34
Education .24 2.79*
Marital status -.13 -1.39
Model 2
Age -.01 -.14 Adjusted R2 = .49
Gender .12 1.70 F(9,124) = 15.25*
DR2 = .43
Education .13 1.96 FDR2 (5,124) = 22.55*
Marital status -.08 -1.17
Extraversion .33 4.68*
Agreeableness .03 .47
Conscientiousness .15 2.08
Emotional stability .35 4.24*
Autonomy .07 .88

Beta value of the standardized regression coefficient, t value of t test, adjusted R2 adjusted total explained variance,
F value of total F test, DR2 contribution of personality traits for explaining variance, FDR2 value of test for
personality traits
* p \ .01

correlations between our different measures of well-being are in line with previous studies
showing that hedonic and eudaimonic aspects of well-being are significantly correlated
(Keyes et al. 2002; Waterman et al. 2008). Factor analyses in both our samples showed that
one factor explains large amount of variance in well-being measures, 73.82% in adoles-
cents’ sample and 66.08% in older adults’ sample. Since previous studies have shown that
different aspects of well-being have different relations with other variables (e.g. Keyes
et al. 2002), we looked at the relationship between sociodemographic and personality
variables with our three separate measures of well-being, as well as with factor score
labeled psychological well-being.
As expected, sociodemographic variables explained none of the variance in well-being
measures in adolescents. In older adults, only education was a significant predictor of self-
esteem and psychological well-being in the first model, and it stayed marginally significant
predictor of self-esteem in the second model when personality traits were entered. This
result is in line with the finding that education is important predictor of eudaimonic aspects
of well-being (Keyes et al. 2002). In our sample, individuals with higher levels of edu-
cation tend to have higher levels of self-esteem in old age. This finding is expected because
quality of life as well as sense of purpose throughout adulthood is highly determined by
obtained education level and work position related to it. After retirement, when one
reevaluates his or her life based on accomplishments, it does not come as a surprise that
education is important determinant of self-esteem and well-being. Of course, only longi-
tudinal study would allow testing this mediation hypothesis.
Also as expected, personality traits explained large amount of variance in our different
measures of well-being in both samples. The amount of explained variance varied from 38
to 58% in adolescents and from 28 to 49% in older adults, which is in line with the finding
from Steel et al. (2008) meta-analysis that personality can explain around 40% of variance

123
Personality and Well-Being 465

in subjective well-being. In line with our hypotheses, personality explained more variance
in well-being in adolescents than in older adults, and explained more variance in eudai-
monic than hedonic aspects of well-being in both samples. Well-being is determined by
genetic factors, personality, ecological factors such as culture, our experiences etc. Rela-
tive influence of factors contributing to well-being changes over the course of develop-
ment. Our results suggest, in line with our hypothesis, that personality contributes more to
well-being in adolescence than in older age. In older age, more various factors could
influence well-being such as health, education, relationship experiences, and goal fulfill-
ment. Personality is still an important predictor of well-being in older age but to a lesser
degree.
Grant et al. (2009) showed that the relationship between personality factors and psy-
chological well-being was stronger than relationship between personality factors and
subjective well-being. Our results replicate this finding and extend it to adolescents and
older adults. As expected, emotional stability and extraversion were the most important
personality traits for explaining variance in well-being measures. However, in contrast to
Gomez et al. (2009) finding that neuroticism becomes the dominant personality-related
predictor for well-being with increasing age, in our study emotional stability was more
important as predictor of well-being measures in adolescents than in older adults. Also in
Gomez et al. (2009) study the influence of extraversion on well-being was only existent in
young adults and disappeared with increasing age, while in our study extraversion seemed
to be important predictor of well-being in both samples. Gomez et al. (2009) study had
more participants and used structural equation modeling for data analysis, but more studies
are needed in order to explore how the relationship between personality and well-being
changes with aging. In older adults conscientiousness was a significant predictor of self-
esteem, which is in line with the finding from Keyes et al. (2002) that conscientiousness is,
together with neuroticism and extraversion, a personality trait that can discriminate
between people with high and low well-being. It is also in line with the finding from Grant
et al. (2009) that conscientiousness was significantly correlated with psychological well-
being.
Fifth factor of personality is differently conceptualized within different frameworks and
measurement instruments in the field of personality (e.g. intellect, openness to experience,
autonomy). Results about the relationship between this fifth factor and well-being are
therefore inconsistent. The novelty of this study is that it used FFPI as a measure of
personality. It has autonomy as a fifth factor which is closely related to the conception of
personal autonomy (Hendriks et al. 2003), and Perugini and Ercolani (1998) found it is
closely related to measures such as self-awareness, sensitivity to others and generalized
self-efficacy. De Fruyt et al. (2004) correlated the FFPI with the NEO-PI-R in Belgian,
American and Hungarian samples and found that autonomy was not clearly related to
openness, and facet analysis suggested that it might be better interpreted as a dominance
factor. Relationship of autonomy as a personality trait to well-being measures was inter-
esting because of the fact that autonomy as a concept seems to be closely related to
eudaimonia (Deci and Ryan 2008). In Ryff’s (1989a) model the concept of autonomy is
used as one of the six aspects of psychological wellness, defined as self-determination,
independence, and the regulation of behavior from within. In Ryan and Deci’s (2000) self-
determination theory autonomy is one of the three fundamental and universal psycho-
logical needs. From this perspective, living well involves those motives, goals, and
behaviors that are satisfying of the basic psychological needs for competence, relatedness,
and autonomy. In both our samples autonomy, as a personality trait, was correlated sig-
nificantly with all well-being measures, and it should not surprise that autonomy correlated

123
466 A. Butkovic et al.

with measures of well-being in both our samples more than agreeableness and conscien-
tiousness. However, autonomy was not a significant predictor in any of regression analyses.
The reason for this probably lies in the fact that autonomy correlated highly, around .50,
with both extraversion and emotional stability in both samples.
Research presented here extends previous research on personality and well-being in
several ways. First, it showed that personality predicts well-being in adolescence as well as
in very old age. Second, personality was more associated with well-being in adolescence
than in old age. Comparison of these two developmentally distinct groups indicated that
personality remains important base for predicting someone’s well-being throughout life
course. Third, comparison of subjective and psychological well-being measures indicated
that they were related but also distinct concepts throughout life. Personality was associated
more with psychological than with subjective well-being.
There are number of limitations for this study. Both our samples were rather small,
mainly female, and age range in older adults sample was large. Although we measured
different aspects of well-being, our choice of these different aspects was limited and not
driven by any theory. Both personality and well-being measures were self-report measures.
However, despite these limitations, results of this study warrant that more studies on
personality, well-being and aging are needed. It seems clear that personality predicts
significant amount of variance in different aspects of well-being measured in diverse ways.
Also, it seems that contribution of personality for explaining variance in well-being
changes with age, but it is not clear in what way from studies done so far. Another finding
that seems to be rather constant is that neuroticism and extraversion are the most significant
personality traits for well-being. In this study, emotional stability was the most significant
predictor of well-being, self-esteem and psychological well-being in adolescents, while
extraversion was the most significant predictor of loneliness. In older adults both emotional
stability and extraversion were equally significant predictors of well-being, loneliness and
psychological well-being, while conscientiousness was a significant predictor of self-
esteem. Future studies are needed to show if this pattern of results is replicable across
samples and measures.

References

Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality of American life: Perceptions,
evaluations, and satisfactions. New York: Sage.
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO PI-R) and NEO five-factor
inventory professional manual. Odessa, FL: Psychological Assessment Resources.
De Fruyt, F., McCrae, R. R., & Szirmak, Z. (2004). The five-factor personality inventory as a measure of the
five-factor model: Belgian, American, and Hungarian comparisons with the NEO-PI-R. Assessment,
11, 207–215.
Deci, E. L., & Ryan, R. M. (2008). Hedonia, eudaimonia, and well-being: An introduction. Journal of
Happiness Studies, 9, 1–11.
DeNeve, K. M., & Cooper, H. (1998). The happy personality: A meta-analysis of 137 personality traits and
subjective well-being. Psychological Bulletin, 124(2), 197–229.
Eysenck, H. J., & Eysenck, S. B. G. (1964). Manual of the Eysenck personality inventory. San Diego, CA:
Educational and Industrial Testing Service.
Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual for the Eysenck personality questionnaire (junior and
adult). London: Hodder & Stoughton.
Gomez, V., Krings, F., Bangerter, A., & Grob, A. (2009). The influence of personality and life events on
subjective well-being from a life span perspective. Journal of Research in Personality, 43, 345–354.
Grant, S., Langan-Fox, J., & Anglim, J. (2009). The big five traits as predictors of subjective and psy-
chological well-being. Psychological Reports, 105, 205–231.

123
Personality and Well-Being 467

Hendriks, A. A. J., Hofstee, W. K. B., & De Raad, B. (1999). The five-factor personality inventory (FFPI).
Personality and Individual Differences, 27, 307–325.
Hendriks, A. A. J., Perugini, M., Angleitner, A., Ostendorf, F., Johnson, J. A., De Fruyt, F., et al. (2003). The
five-factor personality inventory: Cross-cultural generalizability across 13 countries. European Journal
of Personality, 17, 347–373.
Keyes, C. L. M., Shmotkin, D., & Ryff, C. D. (2002). Optimizing well-being: The empirical encounter of
two traditions. Journal of Personality and Social Psychology, 82(6), 1007–1022.
Lucas, R. E., & Diener, E. (2008). Personality and subjective well-being. In O. P. John, R. W. Robins, & L.
A. Pervin (Eds.), Handbook of personality: Theory and research (3rd ed., pp. 795–815). NY: The
Guilford Press.
Nes, R. B., Røysamb, E., Tambs, K., Harris, J. R., & Reichborn-Kjennerud, T. (2006). Subjective well-
being: Genetic and environmental contributions to stability and change. Psychological Medicine, 36,
1033–1042.
Orth, U., Trzesniewski, K. H., & Robins, R. W. (2010). Self-esteem development from young adulthood to
old age: A cohort-sequential longitudinal study. Journal of Personality and Social Psychology, 98,
645–658.
Perugini, M., & Ercolani, A. P. (1998). Validity of the five factor personality inventory (FFPI): An
investigation in Italy. European Journal of Psychological Assessment, 14, 234–248.
Plomin, R., DeFries, J. C., McClearn, G. E., & McGuffin, P. (2008). Behavioral genetics (5th ed.). New
York: Worth Publishers.
Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality traits from
childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126(1),
3–25.
Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits
across the life course: A meta- analysis of longitudinal studies. Psychological Bulletin, 132(1), 1–25.
Robins, R. W., & Trzesniewski, K. H. (2005). Self-esteem development across the lifespan. Current
Directions in Psychological Science, 14, 158–162.
Robins, R. W., Trzesniewski, K. H., Tracy, J. L., Gosling, S. D., & Potter, J. (2002). Global self- esteem
across the life-span. Psychology and Aging, 17, 423–434.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.
Russell, D., & Cutrona, C. E. (1988). Development and evolution of the UCLA Loneliness Scale. Unpub-
lished manuscript, Center for Health Services Research, College of Medicine, University of Iowa, Iowa
City.
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation,
social development, and well-being. American Psychologist, 55, 68–78.
Ryff, C. D. (1989a). 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. (1989b). In the eye of the beholder: Views of psychological well-being among middle-aged and
older adults. Psychology and Aging, 4, 195–210.
Ryff, C. D. (2008). Challenges and opportunities at the interface of aging, personality and well-being. In O.
P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory and research (3rd ed.,
pp. 399–421). NY: The Guilford Press.
Steel, P., Schmidt, J., & Shultz, J. (2008). Refining the relationship between personality and subjective well-
being. Psychological Bulletin, 134, 138–161.
Vukasovic, T., Bratko, D., & Butkovic, A. (in press). Genetski doprinos individualnim razlikama u sub-
jektivnoj dobrobiti: meta-analiza (Genetic contribution to individual differences in subjective well-
being: meta-analysis). Drustvena istrazivanja.
Waterman, A. S., Schwartz, S. J., & Conti, R. (2008). The implications of two conceptions of happiness
(hedonic enjoyment and eudaimonia) for the understanding of intrinsic motivation. Journal of Hap-
piness Studies, 9, 41–79.

123