Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Journal of Occupational Health Psychology © 2010 American Psychological Association

2010, Vol. 15, No. 4, 421– 433 1076-8998/10/$12.00 DOI: 10.1037/a0020720

Feeling Good and Doing Great: The Relationship Between


Psychological Capital and Well-Being
Satoris S. Culbertson and Maura J. Mills
Clive J. Fullagar Hofstra University
Kansas State University

This study seeks to determine the relationship between psychological capital and an employee’s
eudaimonic and hedonic well-being. Panel data were collected from 102 extension agents over a
2-week interval. In addition, daily surveys were collected from 67 of the participants. Results
from the panel data indicated that the relation between psychological capital and hedonic
well-being, measured two weeks later, is mediated by eudaimonic well-being. Results from the
daily surveys found that daily eudaimonic work well-being was significantly associated with both
daily positive mood and daily life satisfaction and that variance in eudaimonic work well-being
was predicted by one’s psychological capital.

Keywords: psychological capital, positive psychology, hedonic well-being, eudaimonic well-being

In recent years, there has been a growing interest in essary effort to succeed at challenging tasks; (b)
adopting a positive approach when examining orga- making a positive attribution (optimism) about suc-
nizations and organizational behavior (e.g., Luthans, ceeding now and in the future; (c) persevering toward
2002a). Rather than focus on ways to predict negative goals and, when necessary, redirecting paths to goals
outcomes for individuals and organizations (e.g., (hope) in order to succeed; and (d) when beset by
turnover, burnout), researchers have begun to exam- problems and adversity, sustaining and bouncing
ine ways in which positive outcomes can be realized back and even beyond (resilience) to attain success”
and to identify factors that facilitate individual flour- (pp. 3). Despite the main impetus behind POB being
ishing and wellness. More specifically, positive or- to understand those factors that enable individuals to
ganizational behavior (POB) has been defined as “the thrive, little research has investigated the relation
study and application of positively oriented human between PsyCap and well-being. Studies that have
resource strengths and psychological capacities” examined this relation (e.g., Avey, Luthans, Smith, &
(Luthans, 2002b, p. 59). Palmer, 2010) have used a limited conceptualization
An important construct that has emerged from the of well-being.
POB movement is psychological capital (PsyCap), a Well-being has been defined as “optimal psycho-
higher-order constellation of positive psychological logical functioning and experience” (Ryan & Deci,
components that consists of efficacy, optimism, hope,
2001, p. 142) and can be differentiated into hedonic
and resilience (Luthans, Avolio, Avey, & Norman,
well-being and eudaimonic well-being (Ryff &
2007). As described by Luthans, Youssef, and Avolio
Keyes, 1995; Ryan & Deci, 2001). Hedonic well-
(2007), PsyCap is “characterized by (a) having con-
being is the more typical operationalization of well-
fidence (self-efficacy) to take on and put in the nec-
being, consisting of subjective happiness and the
experience of pleasure (Ryan & Deci, 2001). It is
typically assessed using subjective measures consist-
Satoris S. Culbertson and Clive J. Fullagar, Depart-
ment of Psychology, Kansas State University; and Maura J. ing of life satisfaction, the presence of positive mood,
Mills, Department of Psychology, Hofstra University. and the absence of negative mood (Waterman, 2008).
A version of this paper was presented at the 24th Annual Together, these components are often summarized as
Conference of the Society for Industrial and Organizational
happiness (Diener, 1984). Thus, in the interest of
Psychology, New Orleans, LA in April 2009. This research
was funded in part by the Center for Engagement and parsimony, we refer to hedonic well-being as happi-
Community Development, Kansas State University. ness throughout this article.
Correspondence concerning this article should be ad- More recently, there has been a move to broaden
dressed to Satoris S. Culbertson, Kansas State University,
the notion of well-being to include motivational and
Department of Psychology, 492 Bluemont Hall, 1100 Mid-
Campus Drive, Manhattan, KS 66506-5302. E-mail: behavioral components (Ryff, 1989; Ryff & Keyes,
satoris@ksu.edu 1995). From a motivational perspective, eudaimonic

421
422 CULBERTSON, FULLAGAR, AND MILLS

well-being focuses on striving for self-realization eudaimonic well-being as positive functioning for the
(Waterman, 2008). Behaviorally, it includes optimal remainder of this article.
positive functioning and the act of striving (Ryff, Research has demonstrated that the various con-
1989; Ryff & Singer, 1998). Thus, eudaimonic well- stituent elements of PsyCap are related to well-being.
being involves a sense of fulfillment of one’s poten- This research supports the contention that positive
tial, aspects not subsumed in the conceptualization of cognitive resources inherent in the PsyCap dimen-
happiness. Additionally, eudaimonic well-being is sions are associated with the positive affective expe-
more cognition-based: An individual is arguably mo- rience of well-being. However, such research has
tivated to take actions because of an underlying cog- focused on hedonic, at the expense of eudaimonic,
nition regarding what would be of greatest benefit to well-being. For example, Luthans, Avolio, et al.
the individual and his or her overall positive feelings (2007) found that a composite operationalization of
and thoughts regarding him or herself. PsyCap was positively related to job satisfaction.
Building on positive psychological theories of However, the conceptualization of job satisfaction as
Erickson (1959) and Maslow (1954); Ryff (1989) indicative of well-being is limited, and a more com-
developed a multidimensional conceptualization of prehensive assessment is necessary.
eudaimonic well-being that includes (a) self- POB researchers have developed a set of criteria
acceptance, understanding and accepting one’s for including constructs (Luthans 2002a, 2002b;
strengths and weaknesses; (b) purpose in life, having Luthans, Avolio, et al., 2007). One of these is that
objectives that give life meaning and direction; (c) constructs should be predominantly state-like versus
personal growth, a belief that one’s skills and poten- trait-like. State-like constructs vary situationally and
tial are being realized and enhanced; (d) positive are open to development. To understand these con-
relations with others, having close and valued inter- structs, it is insufficient to rely on between-
actions with others; (e) environmental mastery, feel- individual, cross-sectional designs. Within-individual
ing in control of life and able to manage its demands; variation can only be captured longitudinally. For
and (f) autonomy, being self-determined and owning example, multilevel modeling (e.g., hierarchical lin-
one’s actions. Confirmatory factor analysis has em- ear modeling, or HLM) allows researchers to distin-
pirically confirmed this model and its indication of guish between-individual and within-individual com-
the higher order construct of eudaimonic well-being ponents of variance and to ascertain whether
(Ryff & Keyes, 1995). variables are behaving more like states or traits. In
It is important to distinguish between PsyCap and addition, multilevel modeling allows for the study of
eudaimonic well-being. Despite some overlap (e.g., the impact of higher-level, individual dispositions on
PsyCap efficacy is similar to Ryff’s environmental lower-level, day-to-day well-being.
mastery), these two constructs are theoretically dis- In terms of well-being, there is evidence that hap-
tinct. PsyCap can be conceptualized as personal psy- piness and positive functioning have state-like and
chological capacities and resources. As such, PsyCap trait-like properties. Research suggests individuals
may serve to support or guide individuals in how they have an affective “set point” (Heady & Wearing,
conceptualize life experiences. Conversely, eudai- 1992) or a dispositional mood or attitude level around
monic well-being is best thought of as optimal pos- which reactions occur. Thus, although daily fluctua-
itive functioning—an outcome we hypothesize is in- tions in well-being are likely, there may also be
fluenced by psychological resources (i.e., PsyCap). convergence toward a more stable level. Individuals
Although Ryff’s (1989) model of eudaimonic also likely vary in their optimal functioning. Ilies,
well-being was developed to be context free, Van Morgeson, and Nahrgang (2005) have likened posi-
Horn, Taris, Schaufeli, and Schreurs (2004) have tive functioning to Csikszentmihalyi’s (1990) con-
drawn extensively on Ryff’s research to create a cept of flow—the feeling of complete involvement in
comprehensive model of well-being specific to work. an activity such that individuals seemingly forget
These authors note that Ryff’s approach is more everything but the activity itself—which recent re-
detailed and inclusive than other approaches to work search has shown to have daily fluctuations (Fullagar
well-being in that it incorporates a behavioral com- & Kelloway, 2009). Whether positive functioning
ponent in addition to affective and motivational as- has a similar set-point as happiness remains unclear.
pects. As such, we felt justified in adapting Ryff’s The current research sought to answer three pri-
operationalization to the work context. In the contin- mary research questions. First, we addressed the
ued interest of parsimony and given its focus on question: To what extent does PsyCap contribute to
optimal positive striving and functioning, we refer to an employee’s positive functioning and happiness at
PSYCHOLOGICAL CAPITAL AND WELL-BEING 423

a later time? In line with other researchers (e.g., Ryff suffer deleterious effects of having “too much of a
& Singer, 1998) who have proposed that affect can be good thing.”
considered an outcome of positive functioning, we According to COR theory (Hobfoll, 1989), PsyCap
proposed that positive functioning contributes to hap- optimism would serve as a resource positively im-
piness and that the relation between PsyCap and pacting happiness and positive functioning. For ex-
happiness is mediated through positive functioning. ample, in terms of happiness, research on optimism
Second, we wanted to determine the extent to which has demonstrated a positive relation with mental
daily positive functioning and happiness varied to- well-being (Scheier & Carver, 1992) as well as life
gether. Third, we sought to ascertain the relation satisfaction (Seligman, 2002). Regarding positive
between PsyCap and daily positive functioning. We functioning, optimism is linked to behaviors like
proposed that positive functioning is related to daily coping (Scheier & Carver, 1985). Research has dem-
mood and life satisfaction (measures of happiness), onstrated that people with superior coping styles are
both across and within individuals. Additionally, we better able to handle adversity, benefit from chal-
posited that individual variance in positive function- lenge, and flourish (Park, 1998). Thus, such behav-
ing is associated with one’s PsyCap. iors can be deemed optimal functioning behaviors,
likened to positive functioning. Thus, individuals
Psychological Capital with high levels of PsyCap optimism are likely to
report greater happiness and positive functioning.
According to Hobfoll’s (1989) Conservation of The third PsyCap dimension is hope. Based on
Resource (COR) theory, individuals seek to acquire Snyder’s (2000) work, hope involves persevering and
and maintain resources. When these resources are redirecting paths toward goals. Similar to optimism,
gained, the result is an increase in well-being. Ac- hope is likely a resource that influences happiness,
cording to Hobfoll, there are four kinds of resources with research demonstrating a link between hope and
that can lead to greater well-being, including physical subjective well-being (Kato & Snyder, 2005). Addi-
objects (e.g., one’s home), personal characteristics tionally, hope is related to such optimal functioning
(e.g., self-esteem), conditions (e.g., marital status), or behaviors as coping (Chang & DeSimone, 2001) and,
energies (e.g., time, knowledge). We propose that as such, is likely related to positive functioning.
PsyCap serves as a personal characteristic resource The final PsyCap dimension is resiliency, defined
resulting in increased well-being when at high levels. as “the capacity to rebound or bounce back from
PsyCap is comprised of four constituent elements: adversity, conflict, failure, or even positive events,
efficacy, optimism, hope, and resiliency. Together, progress, and increased responsibility” (Luthans,
these elements serve as an overarching personal char- 2002a, p. 702). There is evidence that resilience is
acteristic that positively influences well-being. For related to positive effects on well-being at work. For
example, PsyCap efficacy is based on Bandura’s example, Maddi (1987) reported a positive relation
(1997) work concerning beliefs about one’s abilities between resilience and happiness among employees
to successfully perform a given task. In terms of COR undergoing downsizing. Additionally, researchers
theory, efficacy can serve as a cognitive resource. For have demonstrated a link between PsyCap resiliency
example, highly efficacious individuals are less im- and job satisfaction (Larson & Luthans, 2006) as well
pacted by self-doubt, negative feedback, setbacks, as happiness (Youssef & Luthans, 2007). Thus, we
and criticism (Bandura & Locke, 2003). From this would expect a high level of PsyCap resiliency is
perspective, individuals with a high level of efficacy related to increased happiness as well as greater
would more likely report greater happiness as well as positive functioning.
greater positive functioning.
The second PsyCap dimension is optimism, which Hypotheses
is based on Seligman’s (1998) work and involves an
individual’s positive attributional style about success. A primary purpose of our study was to ascertain
According to Seligman, optimists tend to take credit the relation between PsyCap and a more compre-
for positive occurrences in their lives while providing hensive assessment of well-being (see Figure 1).
external, temporary, situation-specific explanations Avey et al. (2010) recently demonstrated that Psy-
to negative occurrences. PsyCap optimism differs Cap was related to well-being at a later time. Their
from traditional optimism, however, in that it has the focus, however, was exclusively on psychological
caveats of being both realistic and flexible. In this well-being (i.e., affective or emotional well-being,
manner, it remains a resource that is not likely to akin to hedonic well-being). Previous researchers
424 CULBERTSON, FULLAGAR, AND MILLS

e11
Res 2

Hedonic .96
e1 Efficacy Well-being Positive
R2=.61 Affectivity
.75

e2 Hope
.77
Psychological .78
Capital
.83
Resiliency .75
e3
.70 Res 1

e4 Optimism Eudaimonic
Well-being
R2=.56
.81
.74 .57
.50 .55 .67

Autonomy Environmental Personal Work Purpose Self


Mastery Growth Relations In Work Acceptance

e5 e6 e7 e8 e9 e10

Figure 1. Standardized estimates for the structural model.

(e.g., Ryff & Singer, 1998) have noted that affect responsive to environmental conditions” (Csik-
can be considered an outcome of positive function- szentmihalyi & Hunter, 2003, p. 186). Further-
ing, suggesting that positive functioning may be an more, positive functioning conceptually overlaps
intermediate variable between PsyCap and positive with the construct of flow; both tap engagement in
affectivity. Thus, we build on Avey et al.’s finding work, a sense of control and purpose, and mastery
that PsyCap is related to happiness, and posit that over the work environment. Flow has been found
the relation can be explained by positive function- to have daily fluctuations and to be predominantly
ing. state-like (Fullagar & Kelloway, 2009). Conse-
quently we expect positive functioning to also fluc-
Hypothesis 1a: PsyCap is positively related to
tuate daily. Thus, we expected considerable with-
positive functioning at a later time.
in-individual variance in both daily positive
Hypothesis 1b: Positive functioning mediates functioning and happiness.
the relation between PsyCap and happiness.
Hypothesis 2a: Daily positive functioning ex-
A second purpose of our study was to examine hibits both within-individual variance and be-
fluctuations in daily positive functioning and hap- tween-individual variance with the former ac-
piness and PsyCap’s influence on these daily fluc- counting for a substantial proportion of the
tuations. As described earlier, although there is variance in ratings and indicating state-like
evidence that individuals have an affective “set properties.
point” (Heady & Wearing, 1992), individuals are
likely to experience daily fluctuations in their hap- Hypothesis 2b: Daily happiness exhibits both
piness. Indeed, the emotional components of well- within-individual variance and between-individ-
being have been argued to be “quite volatile and ual variance with the former accounting for a
PSYCHOLOGICAL CAPITAL AND WELL-BEING 425

substantial proportion of the variance in ratings bias, we administered two surveys at two different
and indicating state-like properties. points in time (Podsakoff, MacKenzie, Lee, & Pod-
sakoff, 2003). The first survey assessed PsyCap and
We also predicted that daily positive functioning demographic variables. The second survey was ad-
would relate to daily measures of happiness, both ministered two weeks later and assessed positive
across and within individuals. Both happiness and functioning and current positive affectivity. The tim-
positive functioning are concerned with optimal ex- ing of the surveys also set up a temporal logic that
periences (Ryan & Deci, 2001). We propose that PsyCap would predict later positive functioning,
daily positive functioning will predict happiness. which in turn would be associated with present pos-
That is, if an individual is functioning optimally, he itive emotions.
or she is also likely to experience higher levels of In addition to the Time 1 and Time 2 surveys,
satisfaction and overall happiness. Of course, the participants had the option of participating in a daily
reverse may also be true, that happy individuals func- survey study over the two weeks. Two weeks was
tion at a higher level at work. selected as the data collection time period primarily
for practical reasons. First, Reis and Wheeler (1991)
Hypothesis 3: Daily positive functioning is pos- argued that “the 2-week record-keeping period is
itively associated with daily happiness both assumed to represent a stable and generalizable esti-
across and within individuals. mate of social life” (p. 287). Second, we believed that
a longer time period may be too demanding for
Our final hypothesis was based on the rationale
participants and result in either lower response rate or
regarding the link between PsyCap, a state-like trait
increased participant attrition.
(Luthans, Avolio, et al., 2007), and positive function-
As an incentive to complete the daily surveys,
ing, a momentary state. Specifically, based on Hob-
volunteers received a $20 gift card conditional upon
foll’s (1989) COR theory, PsyCap serves as a per-
completion of at least five daily surveys. Two days
sonal characteristic resource that, when at higher
after completing the Time 1 survey, agents who had
levels, would be expected to lead to greater levels of
agreed to participate in the daily survey (N ⫽ 84)
well-being. Building on our first hypothesis that Psy-
were sent two daily e-mails including links to the
Cap is a direct predictor of positive functioning,
online home and work surveys for a period of 10
which in turn predicted happiness, we predicted that
working days. Respondents were asked to complete
PsyCap at the individual level would explain a sig-
the work survey immediately after finishing work and
nificant proportion of within-individual variance in
to complete the home survey just prior to retiring for
daily positive functioning.
the day. The average time to respond was 5:51 p.m.
Hypothesis 4: PsyCap is positively related to (SD ⫽ 43 minutes) for the work survey and 8:33 p.m.
daily positive functioning. (SD ⫽ 72 minutes) for the home survey. The average
response frequency was 8.84 daily work and home
surveys.
Methods A total of 102 agents returned completed the Time
1 and Time 2 surveys, yielding a response rate of
Participants and Procedure 42%. Of these participants, 34% were male and the
average age was 41.06 years (SD ⫽ 12.18). Partici-
The sample consisted of county extension agents pants reported working an average of 48.81 hours per
from a Midwestern state in the United States. Exten- week (SD ⫽ 5.13). The average tenure in their cur-
sion agents work in service to their communities and rent position was 11.06 years (SD ⫽ 9.64). Data from
in event planning. Areas of service include Family these 102 participants were used to Test Hypotheses
and Consumer Sciences, Agriculture and Natural Re- 1a and 1b.
sources, and 4H/Youth Development, and include The remaining hypotheses were tested using only
such activities as teaching healthy eating habits, ed- participants who completed both work and home
ucating about conservation, and teaching young peo- surveys on at least five days. A total of 67 (79.67%)
ple leadership, citizenship, and life skills. All 245 individuals (68.7% female) met these criteria. The
county extension agents in the state received an ini- average age of this subsample was 41.67 years (SD ⫽
tial e-mail describing the research. Agents opting to 12.50), with an average tenure of 10.21 years (SD ⫽
participate were then sent an email with a link to an 9.70). Their average work week was 48.98 hours
online survey. In order to reduce common method (SD ⫽ 5.41).
426 CULBERTSON, FULLAGAR, AND MILLS

Measures second-order, latent construct indicated by the six


facets of well-being. Goodness-of-fit indices for this
Time 1 survey. Apart from collecting demo- second-order measurement model were acceptable:
graphic information (e.g., age, gender, tenure, mar- ␹ 2 (9) ⫽ 16.54, CFI ⫽ .96, RMSEA ⫽ .08,
ital status) the main purpose of the Time 1 survey RMR ⫽ .01.
was to assess PsyCap. We measured PsyCap using Happiness was also assessed at Time 2, using the
Luthans, Youssef, and Avolio’s (2007) 24-item positive affectivity subscale from Watson, Clark, and
measure. Six items each assessed the four facets: Tellegen’s (1988) Positive and Negative Affect
efficacy (e.g., “I feel confident analyzing a long-term Schedule (PANAS). The PANAS includes 10 de-
problem to find a solution”), optimism (e.g., “I al- scriptors of positive affect (interested, excited,
ways look on the bright side of things regarding my strong, proud, alert, enthusiastic, active, attentive,
job”), hope (e.g., “There are lots of ways around any determined, and inspired). For each item, participants
problem”), and resiliency (e.g., “I feel I can handle were asked to describe the extent to which they
many things at a time at this job”). Response choices generally feel that way. Response choices ranged
ranged from 1 (strongly disagree) to 6 (strongly from 1 (very slightly/not at all) to 5 (extremely).
agree). We used the four facets as indicators of a Coefficient alpha for this scale was .93.
second order latent construct labeled “PsyCap” (see Daily work survey. The daily work survey as-
Figure 1). Confirmatory factor analysis on data from sessed positive functioning at the end of workday. It
the full sample of 102 participants indicated that each consisted of six items adapted from Ryff’s (1989)
of the four facets had high loadings on the latent Psychological Well-Being Scale. The six items were
PsyCap construct (Optimism ⫽ .74; Resiliency ⫽ chosen because they had consistently high loadings
.77; Hope ⫽ .83; Efficacy ⫽ .79). Goodness-of-fit on the psychological well-being dimensions in pre-
indices were well within Hu and Bentler’s (1999) vious factor analytic studies (e.g., Ryff & Keyes,
recommended values: ␹2(2) ⫽ 6.10, comparative fit 1995; Springer & Hauser, 2005) and had high face
index (CFI) ⫽ .98, root mean square error of approx- validity to the world of work. Each item reflected one
imation (RMSEA) ⫽ .02, root mean residual of the facets of Ryff’s concept of well-being applied
(RMR) ⫽ .01. The reliabilities for the subscales were to the work setting (e.g., Self-Acceptance, “I feel
satisfactory for both the full sample (N ⫽ 102; Effi- positive about myself and the events that happened at
cacy, ␣ ⫽ .86; Optimism, ␣ ⫽ .86; Hope, ␣ ⫽ .79; work today”; Positive Relations, “I had positive and
Resiliency, ␣ ⫽ .63) as well as the subsample (N ⫽ satisfying relations with those at work today”; Au-
67; Efficacy, ␣ ⫽ .88; Optimism, ␣ ⫽ .86; Hope, ␣ ⫽ tonomy, “Social pressures and the expectations of
.79; Resiliency, ␣ ⫽ .67). others made me act and think in certain ways at work
Time 2 survey. Positive functioning and happi- today” (reverse coded); Environmental Mastery, “I
ness were measured at Time 2 (two weeks after Time had difficulty managing my daily affairs and control-
1). Positive functioning was measured with Ryff’s ling events at work today” (reverse coded); Purpose
(1989) Psychological Well-Being Scale, which con- in Work, “I did not have a sense of purpose and
sists of 42 items divided evenly into six dimensions. meaning in my work today” (reverse coded); Per-
Items were reworded for the work context. Sample sonal Growth, “My work today challenged me and
items include self-acceptance (“In general, I feel con- made me grow as a person.”). Response options
fident and positive about myself;” ␣ ⫽ .73), purpose ranged from 1 (strongly disagree) to 4 (strongly
in work/life (“I am an active person in carrying out agree). A total score was computed by summing the
the plans I set for myself at work;” ␣ ⫽ .65), personal scores on these six items. Coefficient alpha of the
growth (“I have the sense that I have developed a lot positive functioning scale (computed on within-
as a person through my work;” ␣ ⫽ .76), positive individual mean item ratings) was .82.
work relations with others (“I enjoy personal and Daily home survey. The daily home survey was
mutual conversations with the people I work with;” completed at the end of the day and assessed two
␣ ⫽ .74), environmental mastery (“I am quite good at aspects of happiness; daily mood and life satisfaction.
managing the many responsibilities of my daily work Mood was measured with 10 positive emotions (alert,
life;” ␣ ⫽ .68), autonomy (“My work decisions are happy, cheerful, strong, active, sociable, involved,
not usually influenced by what everyone else at work excited, clear, relaxed) and 10 negative emotions
is doing;” ␣ ⫽ .70). Participants responded from 1 (drowsy, sad, irritable, weak, passive, lonely, de-
(strongly disagree) to 4 (strongly agree). Again, pos- tached, bored, confused, tense) arranged as bipolar
itive functioning was included in the model as a items (e.g., alert— drowsy). These emotions have
PSYCHOLOGICAL CAPITAL AND WELL-BEING 427

been used to assess happiness in previous research single factor model did not fit the data well: ␹2(45,
(e.g., Diener, Smith, & Fujita, 1995). Participants N ⫽ 102) ⫽ 202.41, p ⫽ .000; goodness-of-fit index
indicated on a 7-point scale their mood “at this mo- (GFI) ⫽ .69; CFI ⫽ .67; normed fit index (NFI) ⫽
ment” for each bipolar pair. Higher scores indicated .62; RMSEA ⫽ .18. These results suggest mono-
greater positive mood. Coefficient alpha for this scale method bias may not be of great concern and is likely
was .93. To assess life satisfaction, we used a single to have minimal impact on interpretation of the re-
item measure (“All things considered, I am satisfied sults. The proposed measurement model indicated
with my life at this moment”). Participants responded satisfactory goodness-of-fit and the respecified model
on a scale from 1 (strongly disagree) to 7 (strongly was less parsimonious than the original measurement
agree). Our decision to use a single-item measure model (see Table 2). Consequently the unrevised
was based on the desire for the daily survey to not be measurement model was used in subsequent analy-
overly cumbersome for participants. This decision, ses.
however, should not have resulted in sacrificed mea-
surement quality. That is, single-item measures of
The Structural Model
happiness have been shown to have good convergent
and divergent validity as well as high correlations The indices of fit for the proposed recursive model
with multiple item scales of overall happiness and are shown in Table 2. All fit indices indicated a
facet satisfaction (Abdel-Khalek, 2006; Nagy, 2002). satisfactory fit. The estimated standardized coeffi-
cients for PsyCap à positive functioning (␤ ⫽ .75,
Results p ⬍ .001) and positive functioning à happiness (␤ ⫽
.78, p ⬍ .001) were significant (see Figure 1). These
To Test Hypotheses 1a and 1b, following Ander- relationships accounted for 56% of the variance in
son and Gerbing (1988), a two-step structural equa- positive functioning and 61% of the variance in hap-
tion modeling approach was used. In the first step, a piness. The proposed model provided a significantly
confirmatory second-order factor analysis was under- better fit than the null/independence model that spec-
taken to determine the goodness-of-fit of the mea- ified no correlations between the latent constructs,
surement model. As noted, PsyCap was indicated by ⌬␹2(13) ⫽ 466.70, p ⬍ .001. We also assessed the
its four constituent factors. Positive functioning was goodness-of-fit for a direct effects model where
indicated by the six factors outlined by Ryff (1989). PsyCap and positive functioning were correlated and
Happiness was indicated by a single observed vari- had a direct effect on happiness. This model indicated
able, positive affectivity. In order to avoid the as- a weak improvement in fit, ⌬␹2(1) ⫽ 4.39, .01 ⬍ p ⬍
sumption that happiness was measured without error, .05. It also indicated that the standardized coefficient
we estimated the error variance based on the reliabil- for the direct path between PsyCap and happiness
ity of positive affectivity (␣ ⫽ .93).1 In the second was significant (␤ ⫽ .28, p ⫽ .03) but weak. This
step, the structural model (see Figure 1) was assessed suggests that the impact of PsyCap on happiness is
by comparing the goodness-of-fit of the hypothesized primarily mediated by positive functioning, thus par-
model with an independence/null model. tially supporting Hypotheses 1a and 1b.
Descriptive statistics for the full sample of 102 For the remaining hypotheses in which we inves-
participants are shown in Table 1. The data were tigate the relationships between PsyCap, positive
screened to check for multivariate outliers and any functioning, and happiness, we used HLM (Bryk &
violations of normality. No outliers were detected Raudenbush, 1992). HLM is an extension of multiple
and no violations were revealed. regression that enables the analysis of relationships
between variables manifested at two levels of analy-
The Measurement Model sis. In this study, we took a series of repeated daily
measures of positive functioning, mood, and life sat-
We used AMOS 16 to calculate goodness-of-fit isfaction for each participant. These constituted the
indices and estimate coefficients in both the measure- lowest level (Level 1) of data. The individual level
ment and structural models. The proposed measure- data (PsyCap) assessed in the Time 1 survey were the
ment model was compared to a single-factor model second level data (Level 2). Although HLM is a
where all observed variables loaded on a single latent correlative analysis, and thus cannot impute causal-
construct. This single-factor model (often referred to
as Harman’s single factor test; Podsakoff et al., 2003)
1
is an estimate of common method variance. This Estimated error variance ⫽ (1 – ␣)(Variance).
428 CULBERTSON, FULLAGAR, AND MILLS

Table 1
Descriptive Statistics: Means, Standard Deviations, Reliabilities, and Correlationsⴱ for Study Variables
Variable M SD 1 2 3 4 5 6 7 8 9 10 11
Psychological capital
1. Efficacy 4.86 .59 (.86)
2. Hope 4.71 .55 .64 (.79)
3. Resiliency 4.66 .46 .49 .63 (.63)
4. Optimism 4.45 .68 .52 .58 .63 (.86)
Eudaimonic well-being
5. Autonomy 2.83 .42 .38 .45 .43 .27 (.70)
6. Environmental mastery 3.04 .32 .17 .39 .38 .40 .34 (.68)
7. Personal growth 3.12 .39 .28 .37 .39 .24 .28 .33 (.76)
8. Work relations 3.03 .40 .16 .27 .32 .30 .34 .52 .14 (.74)
9. Purpose in work 3.09 .31 .32 .45 .37 .35 .40 .53 .49 .40 (.65)
10. Self acceptance 3.07 .37 .42 .47 .40 .49 .50 .56 .45 .39 .57 (.73)
Hedonic well-being
11. Positive affectivity 3.70 .64 .41 .39 .53 .49 .56 .36 .46 .43 .29 .56 (.63)
Note. Coefficient alphas are along the diagonal.

Critical values of Pearson r (df ⫽ N-2 ⫽ 102), r ⬎ .20, p ⬍ .05, r ⬎ .25, p ⬍ .01.

ity, the temporal logic of the study strengthened ied 67 individuals who had an average of 8.84 ob-
causal inferences. servations each. Thus, there was sufficient power to
To study relationships between positive function- detect effects.
ing, mood and life satisfaction, mood and life satis-
faction were independently regressed onto positive Between- and Within-Individual Variation
functioning for each of the 67 participants who com-
pleted both daily surveys on at least five days. At the Means, standard deviations, and intercorrelations
second level, the parameters estimated at Level 1 across the 67 participants for all study variables are
(intercepts and slopes) for positive functioning were presented in Table 3. Before proceeding, systematic
regressed on the Level 2 variable, PsyCap. HLM 5 within- and between-individual variance in the rat-
was used (Bryk, Raudenbush, & Congdon, 2000) to ings of positive functioning, mood, and life satis-
test the hierarchical models. faction were investigated. These statistical as-
One design issue in HLM is having a sufficiently sumptions were tested using three null models
large sample size. Maas and Hox (2005) demon- (oneway analysis of variance, or ANOVA, models)
strated that when Level 2 sample sizes are larger than for each variable. Since there are no Level 1 or
50, the estimation of regression coefficients, variance Level 2 predictors in these null models, the total
components, and standard errors are accurate and within-individual and between-individual variance
relatively unbiased. In the current research, we stud- can be calculated.

Table 2
Summary of Fit Indices for the Measurement and Structural Models (N ⫽ 102)
Model ␹2 df GFI NFI CFI PGFI RMSEA
Measurement models
Single-factor model 202.41 45 .69 .62 .67 .47 .18
Measurement model 64.98 42 .90 .88 .95 .57 .07
Structural models
Null/Independence model 531.68 55 .35 .00 .00 .29 .29
Hypothesized model 69.37 43 .90 .87 .95 .58 .07
Direct effects model 64.98 42 .90 .88 .95 .57 .07
Note. GFI ⫽ goodness-of-fit index; NFI ⫽ normed fit index; CFI ⫽ comparative fit index; PGFI ⫽ parsimony
goodness-of-fit index; RMSEA ⫽ root mean square error of approximation.
PSYCHOLOGICAL CAPITAL AND WELL-BEING 429

Table 3
Means (M), Standard Deviations (SD), and Intercorrelations Across Individuals for All Study Variables
(N ⫽ 67)
M SD 1 2 3
1. Life satisfaction 4.88 0.59
2. Positive mood 4.29 0.58 0.30ⴱⴱ
3. Eudaimonic work well-being 3.03 0.24 0.16 0.39ⴱⴱ
4. Psychological capital 4.70 0.51 0.17 0.19 0.26
ⴱ ⴱⴱ
Correlation is significant at the .05 significance level (one-tailed). Correlation is significant at the .01 significance level
(one-tailed).

Results revealed substantial within-individual vari- sion of positive functioning onto mood (␥10) indi-
ance (␴2) and between-individual variance (␶00) for cates that higher levels of positive functioning are
all three variables (Positive Functioning: ␴2 ⫽ 0.09, associated with more positive mood. The magni-
␶00 ⫽ 0.04; Positive Mood: ␴2 ⫽ 0.50, ␶00 ⫽ 0.27; tude of this relation was calculated by comparing
Life Satisfaction: ␴2 ⫽ 0.25, ␶00 ⫽ 0.35). Chi-square within-individual variance in mood from the first
tests indicated that the between-individual variance null model with within-individual variance after
was significant in all three cases (Positive Function- controlling for positive functioning. 3 Positive
ing: ␶00 ⫽ 0.04, ␹2(66) ⫽ 346.10, p ⬍ .01; Positive functioning accounted for 22% of within-individ-
Mood: ␶00 ⫽ 0.27, ␹2(66) ⫽ 374.57, p ⬍ .01; Life ual variance in mood. Similarly, comparing Level
Satisfaction: ␶00 ⫽ 0.35, ␹2(66) ⫽ 934.81, p ⬍ .01). 2 variance components of random regression
Intraclass correlations2 indicated that between- model and the null model indicated that positive
individual variance accounted for 31%, 35%, and functioning explained 26% of between-individual
59% of the total variance in positive functioning, variance in mood.4 Thus, results suggest that daily
positive mood, and life satisfaction, respectively. positive functioning is associated with daily posi-
This suggests that most of the variance in positive tive mood.
functioning (69%) and positive mood (65%) was The positive functioning—life satisfaction rela-
attributable to within-individual variation, indicating tion was also significant, ␥10 ⫽ 0.19, t(66) ⫽ 2.37,
they vary considerably on a daily basis and are asso- p ⬍ .05; see Table 4. However, the effect was
ciated with state or situational factors. The variation weak and mainly attributable to positive function-
in life satisfaction, however, was predominantly be- ing explaining 5% of the between-individual vari-
tween-individual rather than within-individual. Thus, ance in life satisfaction. Positive functioning
Hypothesis 2a was supported whereas Hypothesis 2b explained zero percent of the variance in within-
was partially supported. individual life satisfaction. The within-individual
variance in life satisfaction, however, was re-
Eudaimonic and Hedonic Well-Being stricted (M ⫽ 4.85, SD ⫽ 0.77). Our data suggest
that overall life satisfaction does not vary much
We hypothesized that daily positive functioning within individuals on a day-to-day basis. Life sat-
relates to daily mood and life satisfaction, both isfaction seems to vary more considerably between
across and within individuals. Positive functioning individuals. Thus Hypothesis 3 was partially sup-
represents the deviation of the individual’s score ported in that positive functioning was associated
from the grand mean of the sample. Given this, the
with individual level life satisfaction but was not
variance in the intercept term equals the between-
predictive of daily life satisfaction.
individual variance in either positive mood or life
satisfaction after controlling for positive function-
ing, or the adjusted between group variance. The 2
ICC ⫽ (␶00/(␴2 ⫹ ␶00).
3
results of two random coefficients regression mod- R2 for level-1 model ⫽ (␴2(oneway ANOVA model) ⫺
els revealed that, for positive mood, the pooled ␴ 2 (random regression model))/␴ 2 (Oneway ANOVA
model).
slope parameter was significantly different from 4
R 2 for level-2 model ⫽ (␶ 00 (oneway ANOVA
zero, ␥10 ⫽ 1.15, t(66) ⫽ 15.43, p ⬍ .01; see Table model) ⫺ ␶00(random regression model))/␶00(oneway
4. The direction of the pooled slope of the regres- ANOVA model).
430 CULBERTSON, FULLAGAR, AND MILLS

Table 4
Parameter Estimates and Variance Components of the HLM Models Testing the Relationship Between
Work Well-Being, Psychological Capital, Positive Mood, Life Satisfaction, and Psychological Capital
(N ⫽ 592)
Model equations ␥00 ␥10 ␥01 ␴2 ␶00
a
Level 1: Random coefficient regression models
Moodij ⫽ ␤0j ⫹ ␤1j (Eudaimonic work well-beingij) ⫹ rij 4.27 1.15ⴱⴱ — 0.39 0.20
Life satisfactionij ⫽ ␤0j ⫹ ␤1j (Eudaimonic work well-beingij) ⫹ rij 4.87 0.19ⴱ — 0.24 0.33
Level 1: Oneway analysis of variance model
Eudaimonic work well-beingij ⫽ ␤0j ⫹ rij 3.01 — — 0.09 0.04
Intercepts-as-outcomes model
L1: Eudaimonic work well-beingij ⫽ ␤0j ⫹ rij
L2: ␤1j ⫽ ␥10 ⫹ ␥01(Psychological capitalj) ⫹ U1j 2.44 — 0.12ⴱ 0.09 0.04
a
For each model the Level 2 equations are: ␤0j ⫽ ␥00 ⫹ U0j and ␤1j ⫽ ␥10 ⫹ U1j where ␥00 is the mean of the intercepts
across groups; ␥10 is the mean of the slopes across groups; ␴2 (rij) is the level 1 residual variance; ␶00 (U0j) is the variance
in intercepts; and ␶11 (U1j) is the variance in slopes.

p ⬍ .05. ⴱⴱ p ⬍ .01.

Psychological Capital and investigating the relation between PsyCap and two
Positive Functioning different operationalizations of well-being (hedonic
and eudaimonic).
We hypothesized that individual variance in posi- Our results indicated that PsyCap was related to
tive functioning was associated with PsyCap. To test both types of well-being both across two weeks and
this, two models were calculated: a oneway ANOVA on a daily basis. Furthermore, individuals doing work
or null model and an intercepts-as-outcomes model that was eudaimonic (i.e., reflective of one’s abilities
(see Table 4). Tests were undertaken to ascertain the and strengths) were more likely to experience posi-
proportion of variance in the intercept terms ex- tive affectivity and high life satisfaction. This finding
plained by PsyCap at Level 2. The t tests for the fixed supports research that suggests positive functioning
effects in the intercepts-as-outcomes model assessed is a better predictor of life satisfaction than pleasure
the significance of the relationship between PsyCap alone (Peterson, Park, & Seligman, 2005).
and daily positive functioning. These tests indicated Organizational researchers often operationalize oc-
PsyCap, ␥11 ⫽ 0.12, t(65) ⫽ 2.16, p ⬍ .05, was a cupational well-being as job satisfaction. Although
significant predictor of between-groups variance in PsyCap is associated with job satisfaction (e.g.,
positive functioning. Luthans, Avolio, et al., 2007), our results suggest it is
To determine the magnitude of the PsyCap— daily also reliably related to a much broader definition of
positive functioning relationship, we looked at resid- well-being that is more indicative of human happi-
ual variance in the intercepts and found PsyCap ac- ness, flourishing, and thriving. This is an important
counted for 17% of between-individual variance in association if researchers are to establish PsyCap as a
positive functioning (␶00 ⫽ 0.04, ␹2(65) ⫽ 320.79, core construct in the application of positive psychol-
p ⬍ .01).5 Hypothesis 4 was supported. ogy to organizational behavior. Research has indi-
cated a significant link between psychological well-
Discussion being and job performance ratings (e.g., Cropanzano
& Wright, 2001), suggesting organizations can im-
PsyCap has been identified as an important under-
prove their effectiveness by increasing employee
lying construct in the field of positive organizational
well-being. However, relatively little guidance exists
behavior (Luthans, Avey, & Patera, 2008; Luthans,
as to how to achieve this. Our findings suggest that
Avolio, et al., 2007). Despite the fact that the over-
one way to do this would be to improve employees’
arching aim of positive psychological approaches to
PsyCap.
organizational behavior is an understanding of opti-
mal human flourishing, no research has investigated
the association between PsyCap and individual well- 5 2
R ⫽ (␶00(oneway analysis of variance) ⫺ ␶00(inter-
being. The current research aimed to redress this by cepts-as-outcomes))/␶00(oneway analysis of variance).
PSYCHOLOGICAL CAPITAL AND WELL-BEING 431

The PsyCap constituent elements are considered to ways. First, we presented an examination of a rela-
be “state-like” in that they can change and be devel- tively new construct—PsyCap—and its role as a pre-
oped (Luthans, Avolio, et al., 2007). As such, orga- dictor of well-being. While this literature is still in its
nizations interested in developing their employees’ infancy, any research that helps to broaden the no-
well-being are encouraged to direct effort toward mological network of PsyCap is beneficial and
enhancing employee hope, resiliency, efficacy, and timely. Second, we expanded the work on well-being
optimism. Recent research has shown that interven- within the workplace to include positive functioning.
tions can improve PsyCap (e.g., Luthans, Avey, Although eudaimonia dates back to the time of
Avolio, Norman, & Combs, 2006; Luthans et al., Aristotle (Ryan & Deci, 2001), it has been relatively
2008). Using a 1-hr microintervention Luthans et al. unexplored in the workplace literature (see Ilies et al.,
(2006) were able to improve PsyCap among both 2005 for an exception). Thus, whereas much research
management students and practicing managers from exists on happiness at work, our research provides
a variety of organizations. Luthans et al. (2008) one of the few examinations of positive functioning
found similarly encouraging results with their Inter- at work.
net-based intervention. Such programs would be Another strength of our research is the methodology.
more time- and cost-effective in improving well- Our longitudinal approach and use of experiential sam-
being than more traditional interventions, such as job pling allowed us to establish that positive functioning
redesign and organizational behavior modification. and happiness fluctuate daily. These findings suggest
Findings from the current study also provide sup- that exploring well-being using a cross-sectional design
port for the notion that one’s work life can spillover may be insufficient and potentially misleading, de-
into one’s personal life. Specifically, we found that pending on the research question. Additionally, by
positive functioning was related to mood and life
assessing individuals over time, at different points in
satisfaction at home. Given the temporal logic of the
the day, we provided a temporal logic that allowed us
current study, it is safe to say that it was positive
to make some inferences regarding the likely causal
functioning influencing mood and life satisfaction at
order of events.
home, rather than the other way around. That said, it
Although our longitudinal approach and use of
would be interesting to examine the influence of
experiential sampling are strengths of our research,
flourishing and optimal functioning at home on mood
we limited our focus to a 2-week time period for
and satisfaction levels at work.
practical reasons. Although this time frame was con-
Such spillover from one’s work life to one’s per-
sistent with other longitudinal research examining
sonal life is supported by Fredrickson’s (2001)
daily within-individual variation in work-related
broaden-and-build theory of positive emotions,
which posits that positive thoughts and well-being variables (e.g., Judge, Ilies, & Scott, 2006; Sonnentag
create a positive spiral generating further positive & Zijlstra, 2006), it may have been too short to allow
experiences, thoughts, and feelings, which is benefi- for an adequate sampling of workplace experiences.
cial to optimal functioning. This theory lends support Nevertheless, even within the limited time frame, we
to the implementation of interventions to improve found clear variation in employee positive function-
PsyCap and, ultimately, well-being. The findings of ing and happiness. Despite this, sampling from a
our research would empirically support the notion longer time period would help better understand the
that enabling individuals to experience positive func- experiences of individuals over time, as well as pro-
tioning at work leads to positive emotional states. vide greater understanding into the predictive poten-
Future research needs to identify further antecedent tial of PsyCap over greater time spans.
conditions that elicit happiness, particularly as posi- Related to this, we limited our assessment of
tive emotional well-being is an important mediating PsyCap to Time 1, using it to predict later positive
variable to optimal individual and organizational functioning and happiness. In doing so, we treated
functioning. PsyCap as more dispositional than it has been pro-
posed to be. Luthans, Avolio, et al. (2007) have
described PsyCap as being “state-like,” which sug-
Strengths, Limitations, and Directions for gests that it, too, may yield daily fluctuations. The
Future Research extent to which this is the case, however, remains
unknown. Furthermore, both positive functioning and
Our research contributes to the growing literature happiness were measured simultaneously at Time 2
on positive organizational scholarship in several raising the question of whether one could cause or
432 CULBERTSON, FULLAGAR, AND MILLS

mediate the other. Future research is needed to con- two weeks later, was mediated by positive function-
firm our causal ordering. ing. Additionally, results showed that daily positive
In our test of Hypothesis 1b, we used the PANAS functioning was significantly associated with both
as our measure of happiness, which may not be ideal daily positive mood and daily life satisfaction and
given its focus on activation-based (vs. pleasantness- that PsyCap predicted between-groups variance in
based) descriptors of emotions (vs. moods). Never- positive functioning. These findings suggest that or-
theless, although the PANAS has activation-based ganizations interested in enhancing employee well-
descriptors, they are valenced descriptors, in a sense being would be well-advised to target employee
making them pleasantness-based and activation- PsyCap.
based. In addition, the PANAS is measuring an
emotional state, thus falling under the umbrella of References
hedonic well-being. That is, moods are simply less-
specific and more enduring emotional states. Thus, Abdel-Khalek, A. M. (2006). Measuring happiness with a
the PANAS, which measures emotional states, does single-item scale. Social Behavior and Personality, 34,
139 –150.
indeed measure at least a part of hedonic well-being, Anderson, J. C., & Gerbing, D. W. (1988). Structural equa-
albeit potentially a less-stable indicator. Neverthe- tion modeling in practice: A review and recommended
less, we recognize that more emotionally laden de- two-step approach. Psychological Bulletin, 103, 411–
scriptors may be more accurately indicative of hedo- 423.
nic well-being, and thus, we would suggest that Avey, J. B., Luthans, F., Smith, R. M., & Palmer, N. F.
(2010). Impact of positive psychological capital on em-
future researchers designing similar studies consider ployee well-being over time. Journal of Occupational
using a measure of mood (e.g., Diener et al., 1995) as Health Psychology, 15, 17–28.
perhaps a more precise measure of hedonic well- Bandura, A. (1997). Self-efficacy: The exercise of control.
being. New York, NY: Freeman.
Bandura, A., & Locke, E. (2003). Negative self-efficacy and
Another concern is that four of our 11 measures goal effects revisited. Journal of Applied Psychology, 88,
had reliability coefficients less than .70. Such low 87–99.
alphas are indicative of high measurement error. Al- Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical
though HLM makes adjustments for measurement linear models: Applications and data analysis methods.
error, such low internal consistencies suggests future Thousand Oaks, CA: Sage.
Bryk, A. S., Raudenbush, S. W., & Congdon, R. T., Jr.
researchers find more reliable assessments of resil- (2000). HLM 5 for Windows. Chicago, IL: Scientific
iency and certain components of positive function- Software International.
ing. Chang, E. C., & DeSimone, S. L. (2001). The influence of
Another limitation concerns our focus on employ- hope on appraisals, coping, and dysphoria: A test of hope
theory. Journal of Social & Clinical Psychology, 20,
ees from a single organization. Although this focus 117–129.
helps rule out alternative explanations for findings Cropanzano, R., & Wright, T. A. (2001). When a “happy”
related to contextual differences across organiza- worker is really a “productive” worker: A review and
tions, the experiences of agents may be unique com- further refinement of the happy-productive worker thesis.
pared to experiences of individuals in other occupa- Consulting Psychology Journal: Practice and Research,
53, 182–199.
tions. Nevertheless, the process-related nature of the Csikszentmihalyi, M. (1990). Flow: The psychology of op-
questions we examined may remain relatively con- timal experience. New York, NY: Harper Collins.
sistent across occupations and organizations. Regard- Csikszentmihalyi, M., & Hunter, J. (2003). Happiness in
less, further research on this topic using diverse oc- everyday life: The uses of experience sampling. Journal
of Happiness Studies, 4, 185–199.
cupations and organizations is needed. Diener, E. (1984). Subjective well-being. Psychological
Finally, a potential limitation of the present study Bulletin, 95, 542–575.
involves the single-source nature of our data. How- Diener, E., Smith, H., & Fujita, F. (1995). The personality
ever, many of our variables were collected at differ- structure of affect. Journal of Personality and Social
ent times, reducing the potential bias of self-report Psychology, 69, 130 –141.
Erickson, E. (1959). Identity and the life cycle. Psycholog-
data. Additionally, results from Harman’s single- ical Issues, 1, 18 –164.
factor test suggested common method bias was not a Fredrickson, B. L. (2001). The role of positive emotions in
concern. Finally, Spector (2006) has provided empir- positive psychology: The broaden-and-build theory of
ical evidence suggesting common method variance positive emotions. American Psychologist, 56, 218 –226.
Fullagar, C., & Kelloway, E. K. (2009). ‘Flow” at work: An
does not significantly inflate correlations. experience sampling approach. Journal of Occupational
In conclusion, our results demonstrated that the and Organizational Psychology, 81, 595– 615.
relation between PsyCap and happiness, measured Heady, B., & Wearing, A. (1992). Understanding happi-
PSYCHOLOGICAL CAPITAL AND WELL-BEING 433

ness: A theory of subjective well-being. Melbourne, Aus- search: A critical review of the literature and recommended
tralia: Longman Chesire. remedies. Journal of Applied Psychology, 88, 879 –903.
Hobfoll, S. E. (1989). Conservation of resources: A new Reis, H. T., & Wheeler, L. (1991). Studying social interaction
attempt at conceptualizing stress. American Psychologist, with the Rochester Interaction Record. In M. P. Zanna
44, 513–524. (Ed.), Advances in experimental social psychology (Vol. 24)
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit (pp. 270 –318). San Diego: Academic Press.
indexes in covariance structure analysis: Conventional Ryan, R. M., & Deci, E. L. (2001). On happiness and human
criteria versus new alternatives. Structural Equation potentials: A review of research on hedonic and eudaimonic
Modeling, 6, 1–55. well-being. Annual Review of Psychology, 52, 141–166.
Ilies, R., Morgeson, F. P., & Nahrgang, J. D. (2005). Au- Ryff, C. D. (1989). Happiness is everything, or is it? Ex-
thentic leadership and eudaemonic well-being: Under- plorations on the meaning of psychological well-being.
standing leader-follower outcomes. Leadership Quar- Journal of Personality and Social Psychology, 57, 1069 –
terly, 16, 373–394. 1081.
Judge, T. A., Ilies, R., & Scott, B. A. (2006). Work-family Ryff, C. D., & Keyes, C. L. M. (1995). The structure of
conflict and emotions: Effects at work and at home. psychological well-being revisited. Journal of Personal-
Personnel Psychology, 59, 779 – 814. ity and Social Psychology, 69, 719 –727.
Kato, T., & Snyder, C. R. (2005). The relationship between Ryff, C. D., & Singer, B. (1998). The contours of positive
hope and subjective well-being: Reliability and validity human health. Psychological Inquiry, 9, 1–28.
of the dispositional hope scale, Japanese version. Japa- Scheier, M., & Carver, C. (1985). Optimism, coping, and
nese Journal of Psychology, 76, 227–234. health: Assessment and implications of generalized out-
Larson, M., & Luthans, F. (2006). Potential added value of come expectancies. Health Psychology, 4, 219 –247.
psychological capital in predicting work attitudes. Journal
Scheier, M., & Carver, C. (1992). Effects of optimism on
of Leadership and Organizational Studies, 13, 44 – 61.
psychological and physical well-being: Theoretical over-
Luthans, F. (2002a). The need for and meaning of positive
view and empirical update. Cognitive Therapy and Re-
organizational behavior. Journal of Organizational Be-
search, 16, 201–228.
havior, 23, 695–706.
Seligman, M. (1998). Learned optimism. New York, NY:
Luthans, F. (2002b). Positive organizational behavior: De-
veloping and managing psychological strengths. Acad- Pocket Books.
emy of Management Executive, 16, 57–72. Seligman, M. (2002). Authentic happiness. New York, NY:
Luthans, F., Avey, J. B., Avolio, B. J., Norman, S. M., & Free Press.
Combs, G. M. (2006). Psychological capital develop- Snyder, C. R. (2000). Handbook of hope. San Diego, CA:
ment: Toward a micro-intervention. Journal of Organi- Academic Press.
zational Behavior, 27, 387–393. Sonnentag, S., & Zijlstra, F. R. H. (2006). Job characteristic
Luthans, F., Avey, J. B., & Patera, J. L. (2008). Experimen- and off-job activities as predictors of need for recovery,
tal analysis of a web-based training intervention to de- well-being, and fatigue. Journal of Applied Psychology,
velop positive psychological capital. Academy of Man- 91, 330 –350.
agement Learning & Education, 7, 209 –221. Spector, P. E. (2006). Method variance in organizational
Luthans, F., Avolio, B. J., Avey, J. B., & Norman, S. M. research: Truth or urban legend? Organizational Re-
(2007). Positive psychological capital: Measurement and search Methods, 9, 221–232.
relationship with performance and satisfaction. Person- Springer, K. W., & Hauser, R. M. (2005). An assessment of
nel Psychology, 60, 541–572. the construct validity of Ryff’s scales of psychological
Luthans, F., Youssef, C. M., & Avolio, B. J. (2007). Psy- well-being: Method, mode and measurement effects. So-
chological capital: Developing the human competitive cial Science Research, 35, 1079 –1101.
edge. New York, NY: Oxford University Press. Van Horn, J. E., Taris, T. W., Shaufeli, W. B., & Schreurs,
Maas, C. J. M., & Hox, J. J. (2005). Sufficient sample sizes P. J. (2004). The structure of occupational well-being: A
for multilevel modeling. Methodology, 1, 86 –92. study among Dutch teachers. Journal of Occupational
Maddi, S. R. (1987). Hardiness training at Illinois Bell and Organizational Psychology, 75, 365–375.
Telephone. In P. Opatz (Ed.), Health promotion evalua- Waterman, A. S. (2008). Reconsidering happiness: A eu-
tion (pp. 101–115). Stevens Point, WI: National Well- daimonist’s perspective. The Journal of Positive Psy-
ness Institute. chology, 3, 234 –252.
Maslow, A. H. (1954). Motivation and personality. New Watson, D., Clark, L. A., & Tellegen, A. (1988). Develop-
York, NY: Harper. ment and validation of brief measures of positive and
Nagy, M. S. (2002). Using a single-item approach to mea- negative affect: The PANAS scales. Journal of Person-
sure job satisfaction. Journal of Occupational and Orga- ality and Social Psychology, 54, 1063–1070.
nizational Psychology, 75, 77– 86. Youssef, C. M., & Luthans, F. (2007). Positive organiza-
Park, C. (1998). Stress-related growth and thriving through tional behavior in the workplace: The impact of hope,
coping: The roles of personality and cognitive processes. optimism and resiliency. Journal of Management, 33,
Journal of Social Issues, 54, 267–277. 774 – 800.
Peterson, C., Park, N., & Seligman, M. E. P. (2005). Orienta-
tions to happiness and life satisfaction: The full life versus
the empty life. Journal of Happiness Studies, 6, 25– 41.
Received May 4, 2009
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, Revision received March 24, 2010
N. P. (2003). Common method biases in behavioral re- Accepted June 7, 2010 y

You might also like