Authenticity Scale PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

Journal of Counseling Psychology Copyright 2008 by the American Psychological Association

2008, Vol. 55, No. 3, 385399 0022-0167/08/$12.00 DOI: 10.1037/0022-0167.55.3.385

The Authentic Personality: A Theoretical and Empirical Conceptualization


and the Development of the Authenticity Scale

Alex M. Wood P. Alex Linley


University of Manchester Centre for Applied Positive Psychology

John Maltby and Michael Baliousis Stephen Joseph


University of Leicester University of Nottingham

This article describes the development of a measure of dispositional authenticity and tests whether
authenticity is related to well-being, as predicted by several counseling psychology perspectives. Scales
were designed to measure a tripartite conception of authenticity, comprising self-alienation, authentic
living, and accepting external influence, which was supported with exploratory factor analysis. Multi-
group confirmatory factor analysis showed that the factor loadings were invariant across sample,
ethnicity, and gender. The scale showed substantial discriminant validity from the Big Five personality
traits, nonsignificant correlations with social desirability, and 2- and 4-week testretest correlations
ranging from r .78 to .91. Each subscale was strongly related to self-esteem and aspects of both
subjective and psychological well-being. This article provides the first direct test of several theoretical
models that view authenticity as integral to well-being.

Keywords: authenticity, well-being, positive psychology, person centered, self-determination

To thine own self be true, and it must follow, as the night the day, thou human strengths, but the psychological database is spotty (p.
canst not be false to any man. 205); and Harter (2002) concluded that there is no single, coher-
Shakespeare, Hamlet ent body of literature on authentic self-behavior, no bedrock of
To know yourself and to act accordingly has been seen as a knowledge (p. 382). The recent positive-psychology movement
moral imperative throughout history (Harter, 2002). Within hu- (see Linley, Joseph, Harrington, & Wood, 2006) has encouraged a
manistic and existential psychology, individual differences in au- resurgence of interest in authenticity. This is partly through high-
thenticity have been considered critically important to understand- lighting understudied areas of research (Gable & Haidt, 2005) and
ing well-being and freedom from psychopathology (May, 1981; partly through promoting an increased dialogue between human-
Rogers, 1959, 1964, 1980; Yalom, 1980), with the importance of istic and empirical psychologists, involving the rigorous scientific
authenticity also stressed by psychodynamic writers (Horney, testing of ideas with humanistic and counseling psychology lin-
1951; Winnicott, 1965). However, the study of authenticity has eage (Joseph & Linley, 2006; Linley, 2006; Patterson & Joseph,
largely been neglected in empirical psychology, and there have 2007). For the study of authenticity to progress, there seems to be
been no direct and psychometrically valid measures of trait au- a need to identify and quantify authenticity as an individual dif-
thenticity yet developed (Sheldon, 2004). Lopez and Rice (2006) ference variable. There is also a need for a direct test of the
lamented the virtual absence of available measures of the con- theoretically driven hypothesis that the authentic personality is
struct (p. 362); Peterson and Seligman (2004) noted that most related to well-being.
[people] agree that integrity, authenticity, and honesty are basic In the empirical study of authenticity, there has been definitional
confusion regarding the construct (Harter, 2002). As a result,
previous research has either asked people to rate themselves on a
false-self to true-self continuum (e.g., Harter, Marold, Whitesell,
Alex M. Wood, Department of Psychology, University of Manchester,
Manchester, England; P. Alex Linley, Centre for Applied Positive Psy- & Cobbs, 1996) or used less direct measures, such as the extent to
chology, Coventry, England; John Maltby and Michael Baliousis, School which peoples behavior varies across social roles (e.g., Sheldon,
of Psychology, University of Leicester, Leicester, England; Stephen Jo- Ryan, Rawsthorne, & Ilardi, 1997). As noted by Harter (2002), this
seph, School of Sociology and Social Policy, University of Nottingham, has led to a diffuse body of literature, which at times is difficult to
Nottingham, England. interpret. In developing a scale of dispositional authenticity, there
This research was supported in part by a University of Warwick Re- is a need for a clear definition of the construct, both for item
search Fellowship awarded to Alex M. Wood. This research was conducted
development and to interpret the existing literature. Fortunately,
while Alex M. Wood was at the University of Warwick. We are grateful to
Anna Cunningham for her invaluable comments at all stages of the project. such a definition emerges from person-centered psychology, where
Correspondence concerning this article should be addressed to Alex M. substantial debate and conceptualization has led to a clear expla-
Wood, Department of Psychology, University of Warwick, Coventry, nation of the construct, with consensus on the content and bound-
England CV4 7AL. E-mail: alex.wood@warwick.ac.uk aries of authenticity (see Wyatt, 2001). The person-centered model

385
386 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

(Line 3). Taken together, self-alienation, authentic living, and


A: Actual physiological
states/emotions/deep accepting external influence compose the tripartite person-
level cognitions centered view of authenticity.
Although the foregoing discussion has focused on the person-
3: Accepting external centered conception, the concept of authenticity is considered
1: Self- influence essential to understanding the human condition in psychodynamic
alienation (e.g., Horney, 1951; Winnicott, 1965), existential (e.g., May, 1981;
Yalom, 1980), developmental (e.g., Harter et al., 1996), social
psychological (e.g., Kernis & Goldman, 2005; Lopez & Rice,
B: Conscious awareness Social-
of physiological states/ Environment 2006), positive psychological (e.g., Sheldon, 2004), and clinical
emotions/cognitions perspectives (e.g., Ehlers, Maercker, & Boos, 2000; Joseph &
Linley, 2005). We have focused on the person-centered definition
of authenticity simply because it appears to provide the widest and
2: Authentic most comprehensive explanation of the construct. Authenticity
living 3: Accepting external appears to represent an area of agreement between various coun-
influence seling, clinical, and empirical perspectives, with each conception
of authenticity mapping on one or more of the lines in Figure 1.
C: Behavior and
The integrative nature of our definition is evident in treatments of
emotional expression
the relationship between authenticity and well-being.

Figure 1. The person-centered conception of authenticity. Authenticity and Well-Being


In many mainstream counseling psychology perspectives, au-
thenticity is seen as the most fundamental aspect of well-being
is presented in Figure 1 (based substantially on the theory of (Horney, 1951; May, 1981; Rogers, 1961; Winnicott, 1965;
Rogers, 1959, 1961). Yalom, 1980). These researchers see authenticity not simply as an
In the person-centered conception, authenticity is a tripartite aspect or precursor to well-being but rather the very essence of
construct defined by Barrett-Lennard (1998, p. 82) as involving well-being and healthy functioning. As such, departures from
consistency between the three levels of (a) a persons primary authenticity are seen as involving increasing psychopathology.
experience, (b) their symbolized awareness, and (c) their outward However, many of these approaches have not been subjected to
behavior and communication (corresponding to Lines 1, 2, and 3, empirical verification, and the empirical evidence that does exist
in Figure 1). This account begins by contrasting actual experience regarding the relationship between authenticity and well-being is
(the true self, including actual physiological states, emotions, and mostly indirect and focuses primarily on one or another of the
schematic beliefs; Box A of Figure 1) with the aspects of experi- three facets of authenticity.
ence that are represented in cognitive awareness (Box B). The first From a psychodynamic perspective, both Winnicott (1965) and
aspect of authenticity involves the inevitable mismatch between Horney (1951) focused on how internalizing external influence,
the conscious awareness and actual experience. Perfect congruence particularly during childhood, leads to self-alienation. Self-
between these aspects of experience is never possible, and the alienation was in turn seen to be the cause of psychopathology.
extent to which the person experiences self-alienation between From the existential perspective, Yalom (1980) and May (1981)
conscious awareness and actual experience (the true self) com- focused particularly on self-alienation, again viewing this as the
poses the first aspect of authenticity (Line 1 in Figure 1) and leads core of authenticity and the cause of mental distress. Both of these
to psychopathology. The subjective experience of not knowing existential perspectives and also more recent humanistic accounts
oneself, or feeling out of touch with the true self, is indicative of (Joseph, 2004; Joseph & Linley, 2005) have conceptualized post-
this aspect of authenticity. traumatic stress disorder (PTSD) in terms of a shattered, inauthen-
The second aspect of authenticity involves the congruence be- tic self, and linked the distress element of PTSD with bringing
tween experience as consciously perceived (Box B) and behavior self-alienation to awareness. Joseph and Linley (2005) presented a
(Box C; Rogers, 1959, 1961). Authentic living involves behaving purely theoretical account, however a qualitative study (with a
and expressing emotions in such a way that is consistent with the priori coding) by Ehlers et al. (2000) found that both self-
conscious awareness of physiological states, emotions, beliefs, and alienation and completely accepting external influence were re-
cognitions (Line 2). In other words, authentic living involves being lated to more intense PTSD symptoms. Using a similar method-
true to oneself in most situations and living in accordance with ology, Dunmore, Clark, and Ehlers (2001) found that accepting
ones values and beliefs. external influence led to the worsening of symptoms over time.
The third aspect of authenticity involves the extent to which one The role of self-alienation was also examined empirically by
accepts the influence of other people and the belief that one has to Harter et al. (1996), who found that greater self-alienation was
conform to the expectations of others. Humans are fundamentally related to lower levels of hope in children.
social beings, and both self-alienation and authentic living are In the only study to examine dispositional authenticity, Gold-
affected by the social environment (Schmid, 2005). Introjecting man and Kernis (2002) asked 60 questions designed to measure
the views of others and accepting external influence affects both authenticity and found strong correlations between authenticity
feelings of self-alienation and the experience of authentic living and both self-esteem and a composite, subjective well-being
THE AUTHENTIC PERSONALITY 387

(SWB; although this should be considered preliminary given that to be at a particular premium, and we wanted to reduce participant
internal consistencies of their authenticity scale were as low as burden as much as possible. Preliminary evidence is also presented
.32, and the study used only 79 college students). regarding the relationship between authenticity and SWB.
Neff and Harter (2002) examined people who subordinated their
needs in close relationships to avoid confrontation, accepting ex-
Method
ternal influence. Providing they subjectively felt inauthentic, they
reported lower levels of self-esteem and more depression. Lopez Development of the Item Pool
and Rice (2006) rigorously developed a measure of authentic
living and accepting external influence with respect to romantic Items were developed to measure the a priori three-factor def-
relationships and found correlations with self-esteem, depression, inition of authenticity described in the introduction and illustrated
anxiety, and satisfaction with life. Lopez and Rice also found in Figure 1. Specifically, items were designed to measure self-
correlations between authenticity and relationship satisfaction, alienation, authentic living, and accepting external influence. As
even after controlling for gender, self-esteem, commitment level, noted above, this definition is derived from the person-centered
avoidance, and anxiety. However, Lopez and Rice were very clear literature (see Wyatt, 2001) and encompasses the focus of exis-
that they were measuring the process of authenticity in relation- tential and psychodynamic approaches, as well as empirical work
ships rather than authenticity as a disposition, and it is not clear from a variety of perspectives. The exact items were developed by
whether the results will generalize to individual differences on a P. Alex Linley (an expert in existential psychology and positive
personality level. This study also did not examine the self- psychology) and Stephen Joseph (an expert in positive psychology
alienation dimension, probably appropriately given the focus was and a psychotherapist specializing in person-centered practice). P.
on the relationship rather than on the individual. Alex Linley took the lead in reviewing the literature, initially
Social psychological research has demonstrated that the extent conducting electronic searches on the PsycINFO database (http://
to which people feel their personality varies between roles is psycnet.apa.org/) using the term authenticity and subsequently
related to their levels of well-being, with less role variation being working through the reference sections of the articles that
correlated with higher well-being (e.g., Roberts & Donahue, emerged. This revealed several empirical studies that researched
1994). Sheldon et al. (1997) specifically related this to authentic authenticity (e.g., Bettencourt & Sheldon, 2001; Goldman & Ker-
living by showing that people who reported more variability be- nis, 2002; Harter et al., 1996; McGregor & Little, 1998; Neff &
tween roles saw themselves as less authentic. Greater feelings of Harter, 2002; Sheldon et al., 1997) and several theoretical litera-
authenticity were negatively correlated with anxiety, stress, and tures from humanistic psychology (Rogers, 1959, 1964, 1980),
depression, and positively correlated with self-esteem, and this psychodynamic theory (Winnicott, 1965), existential psychology
partially mediated the relationship between role variability and (May, 1958/1994; Yalom, 1980), and positive psychology (Harter,
well-being. In a related study, Bettencourt and Sheldon (2001) 2002; Peterson & Seligman, 2004; Sheldon, 2004).
showed that subjective authenticity in different roles was related to P. Alex Linley and Stephen Joseph met weekly during the
both SWB and group connectedness, and this correlation persisted development phase and developed the items together; there were
when these variables were measured via the peer report of a group no disagreements regarding which items to include. Through a
member. consensual process, it was agreed that 7 items parsimoniously and
There is an increasing body of empirical evidence that supports accurately represented self-alienation (e.g., I feel out of touch
counseling psychology perspectives on authenticity. We suggest a with the real me), 11 items represented authentic living (e.g., I
tripartite definition of authenticity, grounded in a well-accepted always stand by what I believe in), and 7 items represented
definition of person-centered psychology, which sees authenticity accepting external influence (e.g., Other people influence me
as being comprised of self-alienation, accepting external influence, greatly). The two of them agreed that these 25 items accurately
and authentic living. This definition provides a framework in covered the construct of authenticity and measured each of the
which to interpret the existing empirical work, answering Harters three factors (see Table 1 for a full list of items). Each item was
(2002) call for such an integration. We developed a measure to expressed as a statement (e.g., I am true to myself in most
assess this tripartite conception, to directly test whether disposi- situations), with which participants expressed their agreement on
tional authenticity was related to well-being, and to provide a new a 1 (does not describe me at all) to 7 (describes me very well)
tool for counseling psychology research. Likert-type scale; intermediate scale points were not anchored.

Study 1 Sample and Procedure

Introduction Two hundred undergraduate students (79 men, 121 women)


participated in Study 1. Ages ranged from 18 to 54 years, with
The aim of Study 1 was the initial development of the Authen- 90% of participants aged below 26 years. Participants were pre-
ticity Scale, through standard psychometric procedures (Clark & dominantly of a White ethnicity (64%), with the next highest
Watson, 1995), to measure the tripartite conception of authenticity represented ethnicities being Indian (11.5%) and Chinese (9%).
described in the introduction. An initial item pool was generated Most participants were single (86%), with a minority either mar-
and analyzed with exploratory factor analysis (EFA) in order to ried (6.5%) or in other forms of relationships (7.5%). Participants
check whether the expected three-factor structure emerged. We were presented with a study information sheet in the course of
aimed to develop a short scale, as the scale is likely to be used in academic lectures and invited to take part in the study, being
counseling psychology settings. In these settings, time is expected advised that they were free to withdraw at any time. Upon com-
388 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

Table 1
Communalities and Factor Loadings From the Exploratory Factor Analysis (Study 1)

Factor Communalities

Item 1 2 3 Initial Extracted


*
7. I feel as if I dont know myself very well. .79 .17 .01 .51 .54
*
18. I feel out of touch with the real me. .74 .03 .07 .47 .49
*
20. I feel alienated from myself. .70 .05 .14 .54 .51
*
23. I dont know how I really feel inside. .69 .04 .02 .51 .49
25. I feel cut off from who I really am. .63 .08 .02 .42 .37
3. I have to hide the way I feel inside. .39 .06 .01 .29 .16
12. I am in touch with the real me. .42 .37 .04 .57 .44
*
16. I always stand by what I believe in. .27 .73 .06 .47 .45
*
17. I am true to myself in most situations. .08 .76 .06 .65 .67
*
8. I think it is better to be yourself, than to be popular. .17 .52 .06 .40 .35
*
19. I live in accordance with my values and beliefs. .08 .52 .09 .35 .28
15. I find it easier to get on with people when Im being myself. .19 .50 .03 .22 .20
21. My daily behavior reflects the real me. .18 .44 .00 .46 .30
4. I can be myself in my day-to-day activities. .27 .43 .02 .45 .36
22. I am in touch with all of my feelings. .18 .40 .14 .38 .21
9. I feel free to express my emotions to others. .07 .38 .09 .32 .15
6. I feel that I am doing the things that are right for me. .17 .38 .04 .40 .24
2. I dislike people who pretend to be what they are not. .18 .26 .04 .17 .06
*
5. I usually do what other people tell me to do. .10 .04 .73 .48 .51
*
10. Other people influence me greatly. .02 .01 .74 .53 .54
*
13. I am strongly influenced by the opinions of others. .12 .04 .69 .46 .44
*
24. I always feel I need to do what others expect me to do. .16 .07 .64 .47 .48
14. I feel pressured to behave in certain ways. .18 .12 .61 .54 .44
11. I usually laugh because other people are laughing. .03 .02 .35 .21 .13
1. I make my own choices in life. .01 .26 .29 .42 .25

Note. N 200. Principal axis exploratory factor analysis with oblique rotation; loadings over .35 in bold type.
Factor 1 represents self-alienation, Factor 2 represents authentic living, and Factor 3 represents accepting
external influence.
*
Item included in final 12-item scale.

pletion and return of the paper-and-pencil survey, participants person) to 7 (a very happy person) scale. The Subjective Happi-
were debriefed on the nature of the research by P. Alex Linley, and ness Scale has high testretest validity over periods varying from
any questions were answered. 1 month (r .90) to 1 year (r .55) and convergent validity with
measures of depression and life satisfaction (Lyubomirsky & Lep-
Measures per, 1999). In the current study, .82.

Authenticity item pool. All participants completed the full item


pool of 25 items. Results
Anxiety. The Tension subscale of the Profile of Mood States Factor Analysis of the Initial Item Pool
(McNair, Lorr, & Droppleman, 1971) was used to measure anxi-
ety. Participants rate how they have been feeling over the last week A first step in scale construction involves identifying the under-
on nine adjectives (e.g., anxious, tense, shaky, on edge) on a 0 (not lying dimensions that exist in the item pool (Clark & Watson,
at all) to 4 (extremely frequently) scale. The subscale is one of the 1995). Using Sample 1 (n 200), we submitted the whole pool of
most commonly used measures of anxiety (McNair et al., 1971). In 25 items to principal axis EFA, with initial communalities gener-
the current study, .92. ated using squared multiple correlations. Bartletts test suggested
Stress. The Perceived Stress Scale was used to measure sub- that the data were suitable for an EFA, 2(300) 1696.95, p
jective stress (Cohen & Williamson, 1988). Participants rate 10 .001, and the KaiserMeyerOlkin measure indicated that there
items regarding how often in the last month they have found their was an adequate sample size for this specific analysis (.823). The
lives unpredictable (e.g., been upset because of something that first 10 factors had eigenvalues of 6.20, 2.36, 2.25, 1.38, 1.30,
happened unexpectedly), uncontrollable (e.g., been unable to 1.14, 0.97, 0.93, 0.89, and 0.81, respectively, and respectively
control irritations in your life), and overwhelming (e.g., felt that accounted for 24.78%, 9.43%, 8.10%, 5.51%, 5.19%, 4.54%,
you were on top of things). Items are rated on a 0 (never) to 4 3.87%, 3.70%, 3.56%, and 3.23% of the variance.
(very often) scale. In the current study, .83. The decision on the number of factors to retain was based on
Happiness. Happiness was measured with the Subjective Hap- parallel analysis and the minimum average partial (MAP) method.
piness Scale (Lyubomirsky & Lepper, 1999). Five items assess the In studies using simulated data, Velicer, Eaton, and Fava (2000)
participants perception of their happiness (e.g., In general, I and Zwick and Velicer (1986) demonstrated that parallel analysis
consider myself . . .), which are rated on a 1 (not a very happy and the MAP method produced more accurate decisions regarding
THE AUTHENTIC PERSONALITY 389

the number of factors to retain than did examination of the scree develop a short scale for use in counseling psychology settings.
plot or the Kaiser eigenvalue 1 criterion. We did not form subscales with less than four items, as Saucier and
Parallel analysis involves the generation of random data corre- Goldberg (2002) have demonstrated that scales normally have low
lation matrices with the same number of variables and participants internal consistency and poor psychometric properties with less
and calculation of the average eigenvalues for each factor in the than four items. The results from the EFA (see Table 1) showed
data sets. Any factor in the real data set with eigenvalues exceed- that factor loadings dropped off markedly after the fifth item for
ing the randomly generated values is considered substantive. Using each subscale, so we considered forming subscales each comprised
SPSS syntax developed by OConnor (2000), we calculated 1,000 of either four or five items. The four-item subscales had internal
randomly generated data sets with 200 cases and 25 variables, for consistency of .69 for Authentic Living, .78 for Accepting External
which the first five mean eigenvalues were 1.71, 1.59, 1.50, 1.43, Influence, and .78 for Self-Alienation. We examined the change in
and 1.36. These values were exceeded by the first three eigenval- alpha that would occur if we included the fifth highest loading item
ues in our actual data set, indicating an optimal three-factor struc- in each factor. For each of the subscales, adding a fifth item
ture. increased alpha by between .03 and .04. We did not feel that such
The MAP method involves separating common and unique marginal changes in alpha justified burdening the participant with
variance and only retaining factors comprising common variance an additional item, particularly given the aim of developing a short
(see OConnor, 2000; Velicer et al., 2000). In the current data set, scale. As such, we used the four highest loading items on each
the average squared partial correlations associated with the first factor to measure self-alienation, authentic living, and accepting
five components were .024, .022, .015, .016, and .018, with the external influence. The final 12 items used in the Authenticity
smallest average squared partial correlation being associated with Scale are indicated in the Appendix, along with revised item
the third component, again suggesting a three-factor solution. numbers.
Based on parallel analysis and the MAP method, we extracted
three factors that were rotated with an oblique rotation. An oblique
Authenticity and SWB
rotation is the most appropriate when the components are theoret-
ically or empirically related (Fabrigar, Wegener, MacCallum, & Table 2 shows preliminary correlations between the Authentic-
Strahan, 1999). ity Scale and SWB. Each of the subscales was also correlated with
Table 1 shows the initial and extracted communalities and all happiness. Authentic Living and Accepting External Influence
factor loadings. The initial communalities were considerably lower were correlated with anxiety and stress. The correlations of Self-
than 1, supporting the use of EFA over principal component Alienation with anxiety and stress were particularly notable (r
analysis, as principal component analysis assumes all variance is .43 and .54, respectively).
shared variance (Tinsley & Tinsley, 1987). There were few dif-
ferences between the initial and extracted communalities, again
Discussion
suggesting that a sufficient number of factors had been extracted.
Most of the items loaded strongly and uniquely on one factor. The In Study 1, the Authenticity Scale was developed, and initial
factors were readily interpretable. evidence supported the existence of the expected factor structure
Factor 1 comprised negatively worded statements, such as I of self-alienation, accepting external influence, and authentic liv-
feel as if I dont know myself very well and I feel out of touch ing. This suggests that the items are indeed assessing the a priori
with the real me, and corresponded to the self-alienation factor of tripartite conception of authenticity. Based on these three factors,
authenticity. The highest loading items of Factor 2 were I always a 12-item Authenticity Scale was created.
stand by what I believe in and I am true to myself in most Each of the subscales was correlated with happiness. Addition-
situations and corresponded to the authentic living factor of ally, anxiety and stress were positively correlated with Authentic
authenticity. Factor 3 was defined by such items as I usually do Living and negatively correlated with Accepting External Influ-
what other people tell me to do and Other people influence me ence. Given that authenticity was theoretically expected to be
greatly and represented accepting external influence. Thus, the associated with SWB (e.g., Horney, 1951; Rogers, 1964; Winni-
factor analysis supported the structure we expected to find based cott, 1965), this provides preliminary evidence for the validity of
on the person-centered definition of authenticity (Rogers, 1961; the scale.
Wyatt, 2001) and suggested that the items we had developed
mapped onto this conception as desired. The three factors were
intercorrelated. Self-alienation correlated with authentic living at
r .44 and with accepting external influence at r .40. Table 2
Authentic living was correlated with accepting external influence Preliminary Correlations Between the Authenticity Scale and
at r .38. Subjective Well-Being (Study 1)

Subscale M SD Anxiety Stress Happiness


Development of the Authenticity Scale
Authentic Living 22.05 3.72 .07 .11 .26**
The Authenticity Scale was developed from the results of the Accepting External
EFA. Three subscales were created to represent each of the factors. Influence 13.34 4.95 .16* .22** .16*
Self-Alienation 10.84 4.91 .43** .54** .55**
We hypothesized each of the three factors to be equally important
and therefore purposefully selected an equal number of items for Note. N 200.
each of the subscales. As noted in the introduction, we aimed to *
p .05. ** p .01.
390 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

Study 2 Authenticity and Well-Being

Introduction Convergent validity is provided with SWB, PWB, and self-


esteem. SWB involves high positive affect, low negative affect
Study 2 aimed to (a) confirm the factor structure with a new (including low anxiety and stress), and high satisfaction with life.
sample, (b) investigate whether a higher order factor structure PWB involves fulfilling human potential existential challenges of
might best represent the data, (c) test whether the factor structure life (Keyes, Shmotkin, & Ryff, 2002, p. 1008). Ryff (1989)
is invariant across sample, gender, and ethnic group, (d) compare operationalized PWB as comprising autonomy, environmental
the a priori three-factor model with an alternate one-factor model, mastery, positive relations with others, personal growth, purpose in
(e) investigate the temporal stability of the subscales with test life, and self-acceptance. SWB and PWB are theoretically and
retest correlations, (f) present discriminant validity from social empirically distinct concepts, with about 45% of people off-
desirability and the Big Five (Goldberg, 1993), and (g) present diagonal, that is, high on SWB and low on PWB or vice versa
convergent validity with self-esteem and a greater number of (Keyes et al., 2002). The literature on authenticity predicts that
subjective and psychological well-being (PWB) characteristics. authenticity will be related to both aspects of well-being, with
authentic people both experiencing positive emotional experience
Multigroup Confirmatory Factor Analysis (CFA) and also engaging in the existential challenges of living. Self-
esteem was also predicted to be related to authenticity, as self-
The factor structure was tested with multigroup CFA. CFA is esteem is a proxy for unconditional self-regard, which person-
commonly used in scale development to test a factor structure that centered conceptions (e.g., Rogers, 1959, 1961) strongly link with
has emerged through EFA (Clark & Watson, 1995). Multigroup authenticity.
CFA builds on a conventional CFA by additionally testing whether
the factor structure is invariant across samples and demographic
groups (Byrne, 2004). The multigroup approach both provides Method
several replications of the CFA and supports the generalizability of
the measure across the samples and demographic groups included
Participants and Procedure
in the analysis. Sample 1 (ethnically diverse sample). Sample 1 was com-
In addition to testing the factor structure suggested by Study 1, prised of 180 people (94 men, 86 women) from the local commu-
we also used the multigroup CFA to test whether authentic living, nity. Ages ranged from 24 years to 70 years (M 38.6, SD 9.0),
self-alienation, and accepting external influence exist under a and ethnicity was equally balanced between Asian (60 people),
higher order factor. If these three factors are indeed aspects of Black (60 people), and White (60 people) participants. Participants
authenticity, then they would be expected to load highly on a were married (45.6%), cohabiting (21.7%), single (17.8%), dating
higher order authenticity factor. Low loadings would suggest that (8.9%), separated (3.3%), divorced (2.2%), or widowed (0.6%).
the scales are actually measuring fundamentally different concepts. Most participants were employed (95.6%), with a diverse range of
occupations represented, the most common of which were com-
Discriminant Validity puter operations (20.6%), education (12.6%), and sales (6.7%).
Sampling was designed to obtain an ethnically diverse sample of
Study 2 presents discriminant validity from socially desirable working adults. The participants in the sample were originally
responding and the Big Five personality traits. Discriminant va- contacted via five workplaces and four community groups in
lidity from the Big Five would be provided if the Authenticity Northern England. Respondents were sought four at a time, look-
Scale could not be reduced to a linear combination of one or more ing to fill a quota of equal numbers of three broadly defined
Big Five traits. Correlations between the Big Five and authenticity ethnicities. Once identified, participants were given a paper-and-
may be expected as authenticity is conceptualized as a variable pencil questionnaire packet, which they completed and returned to
related to well-being and social life, domains with which the Big the researcher. The procedure was repeated until the target number
Five are correlated. The Authenticity Scale was especially ex- of 180 participants was reached. Prior to completing the question-
pected to correlate with extraversion, neuroticism, and agreeable- naire, each individual was told about the broad nature of the
ness, as these traits respectively include positive affect, negative research, their right to withdraw, and were assured of confidenti-
affect, and prosocial tendencies (Costa & McCrae, 1995). ality. They were also asked to provide personal contact details,
However, the discriminant validity of the Authenticity Scale should they be willing to complete a second questionnaire at a
would be undermined if the majority of variance in the scale could future time point. Informed consent was achieved by the signing of
be predicted by one or more Big Five traits. If this were the case, a document. Each participant was recontacted either 2 or 4 weeks
then authenticity research may still be useful, as it would explain after they first completed the questionnaire (whether they were
how people with certain Big Five trait configurations see their contacted in 2 or 4 weeks was determined by random assignment).
world. However, it would seem that the Authenticity Scale has At the first time point, participants completed the 12-item Authen-
greater potential to make a contribution to the literature if it ticity Scale and measures of SWB (satisfaction with life, positive
encompasses more than simply a linear combination of the Big and negative affect), PWB (autonomy, environmental mastery,
Five. As there has been much recent interest in a sixth (humility) positive relations with others, personal growth, purpose in life, and
factor of personality, which is not represented in the Big Five (Lee self-acceptance), socially desirable responding, and the sixth factor
& Ashton, 2004), we also provide correlations between the Au- of personality (humility). At the second time point, participants
thenticity Scale and a scale measuring this factor. only completed the 12-item Authenticity Scale (to establish test
THE AUTHENTIC PERSONALITY 391

retest reliability). Participants were debriefed on all aspects of the Measures


study following completion of the second time point measure.
Sample 2 (College Student Sample A). Sample 2 comprised Socially desirable responding. The full 40-item Balanced In-
158 undergraduate students (21 men, 137 women) who were ventory of Desirable Responding (Paulhus, 1984) was used to
recruited at two university campuses. Ages ranged from 18 years measure socially desirable responding. The widely used inventory
to 50 years, with 96.2% of participants being aged below 26 years. provides two orthogonal scales measuring deliberate misreporting
Most participants were of White (79.1%) or Indian (9.5%) ethnic- of items to create a positive effect (Impression Management [IM]
ity and described their relationship status as single (88.5%) or subscale) and characteristic positivity bias (Self-Deception [SED]
married (3.8%). Participants were presented with a study informa- subscale). In the current study, alphas were .90 for both subscales.
tion sheet in the course of academic lectures and invited to take Big Five. The Big Five personality traits of extraversion, neu-
part in the study, being advised that they were free to withdraw at roticism, agreeableness, openness, and conscientiousness were as-
any time. Participants completed a paper-and-pencil questionnaire sessed with the Big Five Inventory (John & Srivastava, 1999).
packet including the 12-item Authenticity Scale, as well as mea- Each trait is measured with between 8 and 10 items and contains
sures of self-esteem and SWB (stress, satisfaction with life, and a mixture of positively and negatively coded items. The Big Five
positive and negative affect). Following completion, participants Inventory is one of the mostly widely used measures of the Big
were debriefed on the nature of the research by either P. Alex Five; for each trait, Cronbachs alpha and testretest reliability
Linley or Stephen Joseph (depending on location) and any ques- have been shown to range from .79 to .90, and each subscale
tions were answered. correlates with the corresponding scale of the NEO Personality
Sample 3 (College Student Sample B). Sample 3 included 213 InventoryRevised (Costa & McCrae, 1992) and the Trait Descrip-
second-year psychology students who participated in return for tive Adjectives (Goldberg, 1992) at between r .83 and r .99
course credit. Participants included 43 men and 170 women, with (mean r .94; John & Srivastava, 1999). In the current study,
a mean age of 19.45 years (SD 2.45 years). Participants were alphas ranged from .81 to .86.
primarily of a White (79.9%) or Indian (8.1%) ethnic background, Sixth factor of personality. Lee and Ashtons (2004) Honesty/
and single/never married (94.7%). Participants were presented Humility subscale of the HEXACO Personality Inventory was
with a study information sheet as part of an academic course and used to represent the proposed sixth primary factor of personality.
invited to take part in the study, being advised that they were free Sixteen items assess self-perceptions of honesty (e.g., If I knew
to withdraw at any time. Participation in the study was part of a that I could never get caught, I would be willing to steal a million
course in personality psychology, although several other alternate dollars) and humility (e.g., I am an ordinary person who is no
options were available to obtain course credit. Students who better than others). Lee and Ashton demonstrated that these items
agreed to participate in the study were given a questionnaire operationalized the proposed sixth factor of personality and have
packet. All participants completed the 12-item Authenticity Scale incremental validity above the Big Five. In the current study, alpha
and a measure of self-esteem. Taking advantage of the large was .86.
participant pool, we then asked participants to complete either a Self-esteem. Rosenbergs (1965) 10-item Self-Esteem Scale
measure of the Big Five (n 97) or measures of PWB (autonomy, assessed global self-esteem. Five items are oriented in a positive
environmental mastery, positive relationships with others, grati- direction (e.g., I feel that I am a person of worth, at least on an
tude, and emotional intelligence; n 115). equal plane with others), and five in a negative direction (e.g., At
Sample 4 (community sample). Sample 4 was recruited from a times I think I am no good at all). Participants rate statements on
participant panel run by P. Alex Linleys university. The initial a 1 (strongly disagree) to 4 (strongly agree) scale. The Rosenberg
sample was comprised of 117 people (18 men, 99 women), aged Self-Esteem Scale is one of the most widely used measures of
between 14- and 76-years old (M 32.23, SD 15.93). Partic- self-esteem. In the current study, alphas ranged from .87 to .90.
ipants were predominantly White (82.1%) or Chinese (6.8%) and Life satisfaction. In the Satisfaction with Life Scale (Diener,
single (45.3%) or married (35.9%), with a minority divorced Emmons, Larsen, & Griffin, 1985), participants rate their agree-
(5.1%) or with other relationship status. Due to potential develop- ment with five statements regarding how satisfied they are with
mental differences in authenticity, we excluded 13 participants their life (e.g., I am satisfied with my life) on a 1 (strongly
aged below 18, leaving a final sample of 104. disagree) to 7 (strongly agree) scale. The Satisfaction with Life
Information about the study was provided via the internet for Scale is the most commonly used measure of the evaluative
potentially interested parties. Having read the study information, component of SWB and has a high degree of temporal stability
participants then indicated their agreement with the study protocol (ranging from r .89 over 2 weeks to .54 over 4 years) while still
and procedure by signifying their consent online. At a secure being sensitive to the effects of therapy (Pavot & Diener, 1993). In
university website, participants completed the 12-item Authentic- the current study, alphas ranged from .83 to .87.
ity Scale and measures of self-esteem and SWB (anxiety, satisfac- Affect. The frequency of positive and negative affect was
tion with life, and positive and negative affect). Upon completion measured with the 20-item Positive and Negative Affect Schedule
of the study, participants were sent a debrief to their nominated (PANAS; Watson, Clark, & Tellegen, 1988). Ten items form a
email address. Due to an error in the production of the electronic Positive Affect subscale assessing participants positive affect
questionnaire, participants completed the Authenticity Scale on a 1 (e.g., interested, excited, and enthusiastic), and 10 items form a
(does not describe me at all) to 5 (describes me very well) Likert- second subscale assessing negative affect (e.g., guilty, scared, and
type scale rather than the usual 7-point scale. Due to this anomaly, hostile). Consistent with research on the independence of positive
Sample 4 was not used in the psychometric analysis and was used and negative affect, the two subscales are minimally correlated.
only as a cross-validation sample for the correlational analyses. The PANAS is an extremely widely used measure of affect, as the
392 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

independence of positive and negative affect is better operation- largely independently of the Big Five, social desirability, and
alized than it is in many other similar measures (Watson et al., coping styles (McCullough et al., 2002; Wood, Joseph, & Linley,
1988). In the current study, alphas ranged from .83 to .88. 2007; Wood, Maltby, Stewart, & Joseph, 2008), and to have high
Scales of PWB. The short versions of the six subscales of testretest reliability (Wood, Maltby, Gillette, Linley, & Joseph, in
Ryffs (1989) scales of PWB were used to measure aspects of press). In the current study, alpha was .86.
PWB. Each of the subscales contains three items, including a
balance of positively and negatively worded items, all of which are
Results
rated on a 1 (strongly disagree) to 6 (strongly agree) scale. The
Autonomy subscale measures independence and self- Descriptive Statistics
determination (e.g., I have confidence in my opinions, even if
they are contrary to the general consensus), Environmental Mas- Table 3 shows the descriptive statistics and subscale intercorre-
tery measures a persons sense of mastery and competence in lations for all of the samples. Of note are the relatively low
managing the environment (e.g., In general, I feel I am in charge intercorrelations between the subscales, supporting their discrimi-
of the situation in which I live), Positive Relations With Others nant validity. Internal consistencies ranged from .70 to .86.
measures the participants impression of the quality of their close
personal relationships (e.g., I have not experienced many warm Multigroup CFA
and trusting relationships with others [reverse coded]), Personal
Growth measures an orientation toward self-improvement and Multigroup covariance structural equation modeling was per-
actualization (e.g., For me, life has been a continuous process of formed with the AMOS software (see Byrne, 2004), using the
learning, changing, and growth), Purpose in Life measures beliefs maximum likelihood model of estimation. As the scales showed
regarding purpose and meaningfulness in life (e.g., Some people some negative skew, we applied the Satorra-Bentler (Satorra &
wander aimlessly through life, but I am not one of them) and Bentler, 2001) correction for nonnormality.
Self-Acceptance measures positive attitudes about the self (e.g., I Multigroup CFA involves two steps. In Step 1, separate CFAs
like most aspects of my personality). Extensive studies have are performed for each of the groups. In Step 2, two models are
previously used these scales and have shown that the scales are compared for difference in fit. The fit of the first model (the
independent from SWB (e.g., Keyes et al., 2002; Ryff & Keyes, unconstrained model) is simply the sum of the chi-squared statis-
1995). In the current study, alphas for the three-item subscales tics from the separate CFAs in Step 1. In this model, the values of
ranged from .54 to .79. factor loadings have been free to vary between groups. The second
Gratitude. The Gratitude Questionnaire-6 (McCullough, Em- model (the constrained model) is a single CFA that constrains the
mons, & Tsang, 2002) was used to assess trait gratitude, which was factor loadings to be equal across the groups. Invariance of the
included as an additional well-being variable. Six items measure measure across groups is inferred if the fit of the constrained
grateful affect in terms of intensity (e.g., I feel thankful for what model is not significantly worse than is the fit of the unconstrained
I have received in life), frequency (e.g., Long amounts of time model. As the models are nested, the difference in the fit between
can go by before I feel grateful to something or someone), and the chi-squared values of the two models is itself chi-squared
density, reflecting the number of events or people that can elicit the distributed, with number of degrees of freedom equal to the dif-
emotion (e.g., I am grateful to a wide variety of people). Items ference between the degrees of freedom of the competing models
(two reverse coded) are rated on a 7-point scale ranging from 1 (see Byrne, 2004).
(strongly disagree) to 7 (strongly agree). The Gratitude The first multigroup analysis was performed to test the fit and
Questionnaire-6 has been shown to correlate with well-being sample invariance of the three-factor model suggested in Study 1.

Table 3
Descriptive Statistics and Scale Intercorrelations (Study 2)

Central tendency Intercorrelations

Subscale M SD Authentic Living Accepting External Influence Self-Alienation

Sample 1 (n 180)
Authentic Living .82 19.52 5.20 .40** .32**
Accepting External Influence .84 13.11 5.77 .42**
Self-Alienation .82 13.03 5.16
Sample 2 (n 158)
Authentic Living .70 21.75 3.42 .21** .16*
Accepting External Influence .77 13.91 4.71 .32**
Self-Alienation .84 9.95 4.79
Sample 3 (n 210)
Authentic Living .79 22.41 3.07 .27** .32**
Accepting External Influence .77 13.14 4.18 .20**
Self-Alienation .82 9.49 4.15
*
p .05. **
p .01.
THE AUTHENTIC PERSONALITY 393

Three latent factors were specified, corresponding to self- both genders exhibited a good fit for the three-factor model. The
alienation, authentic living, and accepting external influence. Each multigroup CFA showed that again the constrained model,
of these latent factors was defined by the items of the subscales. 2(111) 146.85, CFI .98, RMSEA .03 (90% CI .01.04),
We also specified that the latent factors of self-alienation, authen- provided an equally good fit as the unconstrained model did,
tic living, and accepting external influence existed under a higher 2(102) 135.39, CFI .98, RMSEA .03 (90% CI .01.04;
order authenticity factor. No error variances were allowed to 2 11.46, df 9, p .25), suggesting the gender invariance
correlate. of the measure.
The individual fit from the separate CFAs for Samples 1, 2, and Finally, we recombined the samples and split the sample be-
3 are presented in Table 4. Model fit was tested with the chi- tween three ethnic groups. Participants were classed as either
squared test of the difference between the implied and reproduced White (n 283), Asian (n 109), or Black (n 65). Finer
correlation matrices, the standardized root-mean-square residual grained comparisons between ethnic groups (e.g., Indian or Chi-
(SRMR), the comparative fit index (CFI), and the root-mean- nese) were not possible due to an insufficient sample size for CFA.
square error of approximation (RMSEA). As the chi-squared test is Table 4 shows a model fit for each of the ethnic groups. The
highly sensitive to sample size, Hu and Bentler (1999) recommend multigroup CFA showed the constrained model, 2(189)
basing model fit assessments on the CFI and SRMR. Based on 273.31, CFI .96, RMSEA .03 (90% CI .02.04), provided
their Monte Carlo analyses, they suggested that good model fit is an equally good fit as the unconstrained model did, 2(180)
individually indicated with approximate values of SRMR .08, 261.82, CFI .96, RMSEA .03 (90% CI .02.04; 2
CFI .95, and RMSEA .06; conventional values for accepting 11.49, df 9, p .09), suggesting that the measure is invariant
good models are substantially more lenient than these values are. across ethnic groups.
Based on these values, individually any of the samples provided a Factor loadings. The multigroup CFAs suggested that the
good fit for the three-factor model. In the multigroup comparison, model is invariant across sample, gender, and ethnic groups (the
the constrained model, 2(171) 312.93, CFI .94, RMSEA factor loadings are equal for each of these groups). Given the
.04 (90% CI .03.05), provided an equally good fit as the factor loadings are equal across each group, Figure 2 presents the
unconstrained model did, 2(153) 285.69, CFI .94, RM- factor loadings based on a combination of all three samples. Visual
SEA .04 (90% CI .03.05; 2 27.25; df 18, p .08), inspection of the factor loadings from the separate CFAs from each
indicating that the factor loadings were equal between the groups, group confirmed the statistical finding that the loadings were near
and the measure was sample invariant. This is important, as the identical. Inspection of Figure 2 shows reasonable factor loadings
samples differed in terms of sampling technique and comparison (between .60 and .78). The latent factors also load highly on a
(two student groups and one ethnically diverse occupational sam- higher order authenticity factor (between .58 and .63).
ple). Comparing one- and three-factor models. The results of the
As the measure showed sample invariance, it is acceptable to multigroup CFAs suggested that the three-factor model provides a
combine the samples and create new groups based on demographic good fit for the data. However, the three latent factors loaded
groups (Byrne, 2004). We combined the samples and split them highly on a higher order authenticity factor. Although this is
according to gender (144 men, 325 women). As shown in Table 4, consistent with each factor being representative of authenticity, it

Table 4
Results From the Multigroup Confirmatory Factor Analysis

Model
Model 1: Three factors Model 2: One factor comparisons

RMSEA RMSEA
Group n 2(51) SRMR CFI (90% CI) 2(54) SRMR CFI (90% CI) df 2

Between sample comparisons


1. Sample 1 180 55.55 .04 .99 .02 (.00, .05) 353.45*** .13 .63 .18 (.16, .19) 3 297.90***
2. Sample 2 158 79.25** .08 .94 .06 (.03, .08) 260.76*** .16 .53 .16 (.14, .18) 3 181.51***
3. Sample 3 213 90.06*** .08 .94 .06 (.04, .08) 365.87*** .16 .50 .16 (.15, .18) 3 275.81***

Between gender group comparisons


4. Men only 144 48.44 .05 .99 .00 (.00, .05) 284.16*** .14 .62 .17 (.15, .19) 3 235.72***
5. Women only 325 83.45** .04 .98 .04 (.03, .06) 714.70*** .15 .52 .19 (.18, .20) 3 631.25***

Between ethnic group comparisons


6. White only 283 61.98 .04 .99 .03 (.00, .05) 449.07*** .14 .51 .16 (.15, .18) 3 387.09***
7. Asian only 109 74.30* .06 .96 .07 (.03, .10) 297.80*** .16 .54 .20 (.18, .23) 3 223.50***
8. Black only 65 53.43 .07 .99 .03 (.00, .09) 148.22*** .14 .60 .17 (.13, .20) 3 94.79***

Note. SRMR standardized root-mean-square residual; CFI comparative fit index; RMSEA root-mean-square error of approximation; CI
confidence interval.
*
p .05. ** p .01. *** p .001.
394 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

Authenticity

.60 -.58 -.63

Self-Alienation Accepting External Authentic Living


Influence

.77 .74 .64 .78 .74 .70 .73 .65 .68 .70 .78 .60

Item 10 Item 7 Item 2 Item 12 Item 6 Item 4 Item 3 Item 5 Item 11 Item 9 Item 8 Item 1

Figure 2. Diagram of the confirmatory factor analysis, with item and latent variable loadings. Error variances
omitted for clarity.

raises the question of whether a one-factor model would provide Discriminant validity from the Big Five. Table 6 presents
an equally good fit for the data. To test this, for each of the zero-order correlations between the Authenticity Scale and the Big
samples, genders, and ethnicities, we compared the three-factor Five, along with three multiple regressions where the Big Five are
model with a second one-factor model where all items loaded on sequentially regressed on Authentic Living, Accepting External
a single factor. As can be seen in Table 4, for each group, the Influence, and Self-Alienation. There were different patterns of
one-factor model provided a poor fit of the data. As also reported zero-order correlations for each of the subscales, but a consistent
in Table 4, for every group, direct nested comparisons of chi- overall picture emerges where all the significant correlations are in
squared values showed that the three-factor model provided a the same direction. As a construct, authenticity appears to be
significantly better fit than the one-factor model did. On this basis, positively correlated with extraversion, agreeableness, conscien-
the one-factor model was rejected. tiousness, and openness, and negatively correlated with neuroti-
cism. However, the results of the multiple regression suggest that
Reliability and Validity authenticity cannot be reduced to a linear combination of Big Five
traits, with the Big Five only accounting for a small but significant
Testretest reliability. Table 5 provides 2- and 4-week test 11%13% of the variance in Authentic Living, Accepting External
retest reliability. For each of the subscales, responses at Time 1 Influence, and Self-Alienation. We also correlated the Authenticity
were correlated with responses at Time 2 at between r .78 and Scale with the recently conceptualized sixth factor of personality.
r .91. In each case, the 4-week testretest correlation differed None of the subscales was significantly correlated with the
from the corresponding 4-week correlation by only r .01. HEXACO measure of this factor (largest absolute r .11, p
Additionally, at both time intervals each of the subscales showed .14), ruling out strong correlations with the sixth factor of person-
group-level stability, with small and nonsignificant mean level
ality as an explanation for the discriminant validity of the Authen-
differences between the two time points.
ticity Scale from the Big Five.
Discriminant validity from social desirability. Both the IM and
SED subscales of social desirability showed very low and nonsignif-
icant correlations with the Authenticity Scale. Neither scale was Correlations With Self-Esteem and SWB
significantly correlated with accepting external influence (IM r
.09; SED r .08), self-alienation (IM r .08; SED r .08), To test whether authenticity is related to self-esteem and SWB,
or authentic living (IM r .05; SED r .06), with the smallest p we correlated the Authenticity Scale with measures of self-esteem,
.19. The results suggest that responding to the Authenticity Scale is anxiety, stress, happiness, satisfaction with life, and positive and
not influenced by socially desirable responding. negative affect. These results are presented in Table 7.

Table 5
Test-Retest Reliabilities at 2 Weeks and 4 Weeks

Time 1 Time 2 Mean change


Stability
Subscale M SD M SD t p r

2 weeks
Authentic Living 19.02 5.26 19.27 5.01 0.69 .49 .79
Accepting External Influence 13.67 5.50 13.91 5.12 1.08 .28 .84
Self-Alienation 12.46 5.15 12.41 4.78 0.13 .90 .78
4 weeks
Authentic Living 20.02 5.12 19.63 5.25 1.09 .28 .78
Accepting External Influence 12.66 6.02 12.84 5.47 0.51 .61 .81
Self-Alienation 13.61 5.12 13.47 5.14 0.38 .71 .79

Note. All participants from Sample 1. Two-week n 90, 4-week n 90, all dfs 89. For each r, p .001.
THE AUTHENTIC PERSONALITY 395

Self-esteem. As predicted, all subscales of the Authenticity

.03 (.01) .23*


.01 (.01) .09
.01 (.01) .06

Scale were correlated with self-esteem in four samples. Self-
Alienation had large correlations with self-esteem (range of r

Openness
.45 to .59). Authentic Living and Accepting External Influence

B (SE)
were also correlated with self-esteem in each of the samples. The
size of the correlations of self-esteem with Authentic Living and
Accepting External Influence were generally of a medium magni-

*
.23 .21
.06 .10
.13 .04
r
tude (range of absolute r .20 to .36). The hypothesis that

*
authenticity would be linked to self-esteem was supported in four

samples across all of the subscales.


SWB. As can be seen from Table 7, each of the authenticity
Neuroticism

.06 .23 .04 (.02)


.01 (.02)
.02 (.02) subscales was correlated with the SWB variables. Self-Alienation
B (SE)

was particularly strongly correlated with each of the SWB vari-


ables (absolute rs ranged from .21 to .50). Accepting External
Influence showed the same pattern of correlations, but the corre-
*

.25*
.19

lations between satisfaction with life and positive affect were not
r

stable across all samples. Authentic Living was correlated with


Summary of Zero-Order Correlations and Four Multiple Regressions of the Big Five on the Authenticity Scale and Subscales

.03 (.02) .18


.19

each of the well-being variables except anxiety, although the


correlation with negative affect seems less stable. With a few


Conscientiousness

exceptions, there was a remarkable level of consistency and rep-


.01 (.02)

.03 (.02)
B (SE)

lication across samples and strong support for the conception of


authenticity as a variable related to SWB.
PWB. Table 8 shows the correlations between the Authenticity
.02 (.02) .11 .28**

Scale and PWB. Each of the subscales was significantly correlated


.17
.18 .16
r

with almost all of the seven aspects of PWB (Accepting External


Influence was not correlated with gratitude, and Authentic Living
.19

was not correlated with purpose in life). Additionally, results were


replicated in a second sample for the three variables that were


Agreeableness

.04 (.02)
.04 (.02)

represented in both samples. The results support the hypothesis


B (SE)

that authenticity is related to PWB.

Discussion and General Discussion


**

.04 (.02) .21* .24*


.27
.07 (.02) .30** .07
r

Two studies reported the development and testing of the Au-


thenticity Scale. Study 1 developed the Authenticity Scale based
.00 (.02) .01

on a tripartite conception of authenticity. Study 2 confirmed the


factor structure, presented reliability and validity information, and


Extraversion

presented the first stringent test of whether trait authenticity is


B (SE)

related to SWB and PWB.


Note. Participants are from Sample 3 (n 94). Adj. Adjusted.

The factor structure of the scale appears very robust. EFA in


Study 1 showed that the factor structure measured the intended
.33**
.29**

three-factor conception, which was supported through multigroup


.11
r

CFA in Study 2. Of particular note was the factor invariance across


each sample, between both genders and broad ethnic grouping.
F(5, 89) Adj. R2

This provides early indication that the Authenticity Scale behaves


**

.13**
.11*
.12

consistently across diverse demographic groups.


The Authenticity Scale appears to have good psychometric
properties. The 2-week and 4-week testretest reliabilities ranged
3.46
3.27
3.92

from r .78 to .91, suggesting that responses on the scale are


stable across short intervals, as would be expected for a trait
Accepting External Influence

measure. Correlations with social desirability were all small and


nonsignificant. It appears that responding to the scale is neither
p .05. ** p .01.

confounded with responses designed to manage impressions nor


Subscale

does it represent an overly positive impression of the self. The


Authentic Living

Authenticity Scale also seems to have distinct variance from the


Self-Alienation

Big Five traits. The scale was meaningfully related to the Big Five,
with more authentic people being more extraverted, agreeable,
Table 6

conscientious, open, and less neurotic. This pattern of correlations


is consistent with the conceptualization of authenticity as an aspect
*
396 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

Table 7
Authenticity and Subjective Well-Being

Subscale Self-Esteem Anxiety Stress SWLS Positive Affect Negative Affect

Sample 1 (n 180)
Authentic Living .22** .23** .07
Accepting External Influence .35** .23** .20**
Self-Alienation .34** .21** .21**
Sample 2 (n 158)
Authentic Living .24** .20** .22** .17* .10
Accepting External Influence .23** .26** .13 .15 .18*
Self-Alienation .57** .47** .50** .35** .49**
Sample 3 (n 213)
Authentic Living .23**
Accepting External Influence .27**
Self-Alienation .45**
Sample 4 (n 104)
Authentic Living .36** .18 .21* .20* .27*
Accepting External Influence .20* .20* .06 .06 .21*
Self-Alienation .59** .39** .34** .31** .48**

Note. SWLS Satisfaction With Life Scale.


*
p .05. ** p .01.

of positive emotional and social functioning. However, a linear thenticity and the well-being variables. The strong relationship
combination of the Big Five only explained a maximum of 13% of between the Authenticity Scale and well-being is a good example
the variance in the subscales of the Authenticity Scale, suggesting of how classical perspectives in counseling psychology can inform
that the scale is more than just a reflection of a configuration of the direction of current empirical work in personality psychology
Big Five traits. (cf. Linley, 2006).
The Authenticity Scale also was correlated with self-esteem, It is also notable that authenticity was correlated with both SWB
SWB, and PWB. As well as providing convergent validity for the and PWB. SWB and PWB are separate concepts, with different
scale, this provides the first test using a validated scale of whether theoretical positions, causes, correlates, and consequences (Keyes
trait authenticity is related to well-being. This is important, as et al., 2002; Ryan & Deci, 2001; Ryff & Keyes, 1995). However,
authenticity is considered central to well-being in several counsel- as predicted, authenticity is related to both conceptions of well-
ing psychology conceptions (Horney, 1951; May, 1981; Rogers, being. The Authenticity Scale was also strongly and robustly
1959; Winnicott, 1965; Yalom, 1980). Indeed, some of the corre- related to self-esteem. This is in keeping with Rogers (1959)
lations of authenticity and well-being were particularly high. For linking of authenticity and unconditional positive regard and Ker-
example, the correlation between self-alienation and satisfaction nis (2003) association of authenticity and secure self-esteem.
with life ranged between r .34 and .50. In Park, Peterson, The field of authenticity research has been hampered by the lack
and Seligmans (2004) assessment of the relationship between 24 of a valid personality measure. The development of the Authen-
character strengths and satisfaction with life, values of |.34| would ticity Scale allows for further tests of the theoretical positions, as
be higher than all but six strengths, and values of .50 are higher well as the several questions that emerge from this article.
than all 24 strengths except for hope. It appears that authenticity is First, longitudinal research could address the order of causality
one of the strongest predictors of well-being. This is particularly between authenticity and well-being and the developmental anteced-
notable as there is no item overlap between the measure of au- ents of authenticity. For example, authenticity could lead to well-

Table 8
Authenticity and Psychological Well-Being

Environmental Personal Purpose Self-


Subscale Autonomy Mastery PRWO Growth in Life Acceptance Gratitude

Sample 1 (n 180)
Authentic Living .18* .17* .18* .25** .08 .28**
Accepting External Influence .25** .21** .27** .30** .22** .41**
Self-Alienation .17* .21** .23** .28** .15* .39**
Sample 3 (n 119)
Authentic Living .45** .40** .34** .37**
Accepting External Influence .59** .27** .24* .15
Self-Alienation .33** .52** .44** .35**

Note. PRWO Positive Relations With Others subscale.


*
p .05. ** p .01.
THE AUTHENTIC PERSONALITY 397

being as Rogers (1959) has suggested, well-being could lead to people References
having the courage to be authentic, or the two could operate in a spiral
in a broaden-and-build fashion (Fredrickson & Joiner, 2002). Barrett-Lennard, G. T. (1998). Carl Rogers helping system: Journey and
substance. London: Sage.
Second, from a developmental perspective, it would be interest-
Bettencourt, B. A., & Sheldon, K. (2001). Social roles as mechanisms for
ing to examine both mean level authenticity across different age psychological need satisfaction within social groups. Journal of Person-
groups and what kind of environments lead to dispositional au- ality and Social Psychology, 81, 11311143.
thenticity. Rogers (1959) suggested that people were naturally Byrne, B. M. (2004). Testing for multigroup invariance using AMOS
authentic at an early age but that this authenticity decreased later graphics: A road less traveled. Structural Equation Modeling, 11, 272
in life due to the imposition of conditions of worth. Similarly, 300.
Harter et al. (1996) and Neff and Harter (2002) found that people Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in
were more authentic when their self was being accepted by other objective scale development. Psychological Assessment, 7, 309 319.
people. It would be pertinent to see whether this equated to Cohen, S., & Williamson, G. (1988). Perceived stress in a probability
different levels of dispositional authenticity. sample of the United States. In S. Spacapam & S. Oskamp (Eds.), The
social psychology of health (pp. 31 67). Newbury Park, CA: Sage.
Third, authenticity could also illuminate differences between
Costa, P. T., & McCrae, R. R. (1992). Revised NEO personality inventory
groups. In addition to the disadvantages suffered by all stigmatized (NEO-PI-R) and the NEO five-factor inventory (NEO-FFI): Profes-
groups, certain group members may have a potential identity sional manual. Odessa, FL: Psychological Assessment Resources.
which is not visually clear (such as Jewish people, lesbian, gay, Costa, P. T., & McCrae, R. R. (1995). Domains and facets: Hierarchical
bisexual, and transsexual people, and people with unseen disabil- personality assessment using the revised NEO personality inventory.
ities, such as epilepsy). For such people, they have the additional Journal of Personality Assessment, 64, 2150.
strain of not knowing whether people would treat them differently Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert,
if their true group membership was known (Crocker, Major, & S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology
Steele, 1998). Issues of authenticity may be particularly important (Vol. 2, pp. 505553). New York: Oxford University Press.
for such groups. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The
Satisfaction With Life scale. Journal of Personality Assessment, 49,
Fourth, several conceptions have seen increased authenticity as
7175.
sometimes arising in people who have undergone trauma (Joseph, Dunmore, E., Clark, D. M., & Ehlers, A. (2001). A prospective investiga-
2004; Joseph & Linley, 2005; May, 1981). This may be one of the tion of the role of cognitive factors in persistent posttraumatic stress
benefits that people often report after the trauma, in addition to disorder (PTSD) after physical or sexual assault. Behaviour Research
their intense suffering. The Authenticity Scale could be used to test and Therapy, 39, 10631084.
not only whether this was the case but also test more complex Ehlers, A., Maercker, A., & Boos, A. (2000). Posttraumatic stress disorder
models, such as whether authenticity only arises as a form of following political imprisonment: The role of mental defeat, alienation,
trauma-related growth when unconditionally accepting relation- and perceived permanent change. Journal of Abnormal Psychology, 109,
ships are present (Joseph, 2004; Joseph & Linley, 2005). 4555.
Fifth, each of the counseling and existential psychology per- Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J.
(1999). Evaluating the use of exploratory factor analysis in psycholog-
spectives on authenticity (Horney, 1951; May, 1981; Rogers,
ical research. Psychological Methods, 4, 272299.
1959; Winnicott, 1965; Yalom, 1980) saw the authentic disposi- Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger upward
tion as being increased through psychotherapy. This could be spirals toward emotional well-being. Psychological Science, 13, 172
tested with the Authenticity Scale, such as by comparing longitu- 175.
dinal change scores between those undergoing therapy and a Gable, S. L., & Haidt, J. (2005). What (and why) is positive psychology?
control group. This would be in keeping with an increasing focus Review of General Psychology, 9, 103110.
on the efficacy of counseling and a drive to evaluate therapy by Goldberg, L. R. (1992). The development of markers for the Big-Five
other criteria than those based on the medical model of psycho- factor structure. Psychological Assessment, 4, 26 42.
pathology (Joseph & Worsley, 2005). Goldberg, L. R. (1993). The structure of phenotypic personality traits.
Sixth, future research could also widen understanding of how American Psychologist, 48, 26 34.
Goldman, B. M., & Kernis, M. H. (2002). The role of authenticity in
authenticity fits in with other personality traits. In particular, it is
healthy psychological functioning and subjective well-being. Annals of
not clear how authenticity is related to its nonfelicitous opposites. the American Psychotherapy Association, 5, 18 20.
For example, Peterson and Seligman (2004) pointed out that ant- Greenwald, A. G., & Farnham, S. D. (2000). Using the implicit association
onyms of authenticity include deceitfulness, insincerity, preten- test to measure self-esteem and self-concept. Journal of Personality and
tiousness, and falseness. It would be informative to see whether Social Psychology, 79, 10221038.
these were part of the same higher order factor as authenticity. Harter, S. (2002). Authenticity. In C. R. Snyder & S. J. Lopez (Eds.),
Kernis (2003) has posited that authenticity should be related more Handbook of positive psychology (pp. 382394). Oxford, England: Ox-
to secure self-esteem than to insecure self-esteem, and this could ford University Press.
now be tested directly. Although such traits as insincerity and Harter, S., Marold, D. B., Whitesell, N. R., & Cobbs, G. (1996). A model
secure self-esteem are difficult to measure, considerable advances of the effects of perceived parent and peer support on adolescent false
self behavior. Child Development, 67, 360 374.
have been made into measuring these traits through implicit mea-
Horney, K. (1951). Neurosis and human growth. London: Routledge.
sures (Greenwald & Farnham, 2000). Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indices in
There are multiple new areas of research for authenticity in both covariance structure analysis: Conventional criteria versus new alterna-
counseling psychology and personality psychology research. It is tives. Structural Equation Modeling, 6, 155.
hoped that the development of the Authenticity Scale will aid these John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: History,
research endeavors and support therapeutic applications. measurement, and theoretical perspectives. In L. A. Pervin & O. P. John
398 WOOD, LINLEY, MALTBY, BALIOUSIS, AND JOSEPH

(Eds.), Handbook of personality: Theory and research (2nd ed., pp. Paulhus, D. L. (1984). Two-component models of socially desirable re-
102138). New York: Guilford. sponding. Journal of Personality and Social Psychology, 46, 598 609.
Joseph, S. (2004). Client-centered therapy, post-traumatic stress disorder Pavot, W., & Diener, E. (1993). Review of the Satisfaction With Life scale.
and post-traumatic growth: Theoretical perspectives and practical impli- Psychological Assessment, 5, 164 172.
cations. Psychology and Psychotherapy: Theory, Research and Practice, Peterson, C., & Seligman, M. E. P. (2004). Introduction: Strengths of
77, 101119. courage. In C. Peterson & M. E. P. Seligman (Eds.), Character strengths
Joseph, S., & Linley, P. A. (2005). Positive adjustment to threatening and virtues: A handbook and classification (pp. 197212). Oxford,
events: An organismic valuing theory of growth through adversity. England: Oxford University Press.
Review of General Psychology, 9, 262280. Roberts, B. W., & Donahue, E. M. (1994). One personality, multiple selves
Joseph, S., & Linley, P. A. (2006). Positive therapy: A meta-theory for - integrating personality and social roles. Journal of Personality, 62,
positive psychological practice. New York: Routledge. 199 218.
Joseph, S., & Worsley, R. (Eds.). (2005). Person-centred psychopathology: Rogers, C. R. (1959). A theory of therapy, personality and interpersonal
A positive psychology of mental health. Ross-on-Wye, England: PCCS
relationships as developed in the client-centered framework. In S. Koch
Books.
(Ed.), Psychology: A study of a science. Vol. 3: Formulations of the
Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem.
person and the social context (pp. 181256). New York: McGraw-Hill.
Psychological Inquiry, 14, 126.
Rogers, C. R. (1961). On becoming a person: A therapists view of
Kernis, M. H., & Goldman, B. M. (2005). From thought and experience to
psychotherapy. London: Constable.
behavior and interpersonal relationships: A multicomponent conceptu-
Rogers, C. R. (1964). Toward a modern approach to values: The valuing
alization of authenticity. In A. Tesser, J. V. Wood, & D. Stapel (Eds.),
process in the mature person. Journal of Abnormal and Social Psychol-
On building, defending, and regulating the self: A psychological per-
spective (pp. 3152). New York: Psychology Press. ogy, 68, 160 167.
Keyes, C. L. M., Shmotkin, D., & Ryff, C. D. (2002). Optimizing well- Rogers, C. R. (1980). A way of being. Boston: Houghton Mifflin.
being: The empirical encounter of two traditions. Journal of Personality Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
and Social Psychology, 82, 10071022. NJ: Princeton University Press.
Lee, K., & Ashton, M. C. (2004). Psychometric properties of the HEXACO Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A
personality inventory. Multivariate Behavioral Research, 39, 329 358. review of research on hedonic and eudaimonic well-being. Annual
Linley, P. A. (2006). Counseling psychologys positive psychological Review of Psychology, 52, 141166.
agenda: A model for integration and inspiration. Counseling Psycholo- Ryff, C. D. (1989). Happiness is everything, or is itExplorations on the
gist, 34, 313322. meaning of psychological well-being. Journal of Personality and Social
Linley, P. A., Joseph, S., Harrington, S., & Wood, A. M. (2006). Positive Psychology, 57, 1069 1081.
psychology: Past, present, and (possible) future. The Journal of Positive Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological
Psychology, 1, 316. well-being revisited. Journal of Personality and Social Psychology, 69,
Lopez, F. G., & Rice, K. G. (2006). Preliminary development and valida- 719 727.
tion of a measure of relationship authenticity. Journal of Counseling Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test
Psychology, 53, 362371. statistic for moment structure analysis. Psychometrika, 66, 507514.
Lyubomirsky, S., & Lepper, H. S. (1999). A measure of subjective hap- Saucier, G., & Goldberg, L. R. (2002). Assessing the Big Five: Applica-
piness: Preliminary reliability and construct validation. Social Indicators tions of ten psychometric criteria to the development of marker scales.
Research, 46, 137155. In B. De Raad & M. Perugini (Eds.), Big Five assessment (pp. 29 58).
May, R. (1981). Freedom and destiny. New York: Basic Books. Seattle, WA: Hogrefe and Huber.
May, R. (1994). Contributions of existential psychotherapy. In R. May, E. Schmid, P. F. (2005). Authenticity and alienation: Towards an understand-
Angel, & H. F. Ellenberger (Eds.), Existence (pp. 3791). Northvale, NJ: ing of the person beyond the categories of order and disorder. In S.
Jason Aronson. (Original work published 1958) Joseph & R. Worsley (Eds.), Person-centred psychopathology (pp. 75
McCullough, M. E., Emmons, R. A., & Tsang, J. A. (2002). The grateful 90). Ross-on-Wye, England: PCCS Books.
disposition: A conceptual and empirical topography. Journal of Person- Sheldon, K. M. (2004). Integrity (honesty/authenticity). In C. Peterson &
ality and Social Psychology, 82, 112127.
M. E. P. Seligman (Eds.), Character strengths and virtues (pp. 249
McGregor, I., & Little, B. R. (1998). Personal projects, happiness, and
272). New York: Oxford University Press.
meaning: On doing well and being yourself. Journal of Personality and
Sheldon, K. M., Ryan, R. M., Rawsthorne, L. J., & Ilardi, B. (1997). Trait
Social Psychology, 74, 494 512.
self and true self: Cross-role variation in the big-five personality traits
McNair, D. M., Lorr, M., & Droppleman, L. F. (1971). Manual for the
and its relations with psychological authenticity and subjective well-
profile of mood states. San Diego, CA: Educational and Industrial
being. Journal of Personality and Social Psychology, 73, 1380 1393.
Testing Services.
Tinsley, H. E. A., & Tinsley, D. J. (1987). Uses of factor-analysis in
Neff, K. D., & Harter, S. (2002). The authenticity of conflict resolutions
among adult couples: Does womens other-oriented behavior reflect counseling psychology research. Journal of Counseling Psychology, 34,
their true selves? Sex Roles, 47, 403 417. 414 424.
OConnor, B. P. (2000). SPSS and SAS programs for determining the Velicer, W. F., Eaton, C. A., & Fava, J. L. (2000). Construct explication
number of components using parallel analysis and Velicers MAP test. through factor or component analysis: A review and evaluation of
Behavior Research Methods, Instrumentation, and Computers, 32, 396 alternative procedures for determining the number of factors or compo-
402. nents. In R. D. Goffin & E. Helmes (Eds.), Problems and solutions in
Park, N., Peterson, C., & Seligman, M. E. P. (2004). Strengths of character human assessment: Honoring Douglas N. Jackson at seventy. Norwell,
and well-being. Journal of Social and Clinical Psychology, 23, 603 MA: Kluwer Academic.
619. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
Patterson, T. G., & Joseph, S. (2007). Person-centered personality theory: validation of brief measures of positive and negative affectthe
Support from self-determination theory and positive psychology. Jour- PANAS scales. Journal of Personality and Social Psychology, 54,
nal of Humanistic Psychology, 47, 117139. 10631070.
THE AUTHENTIC PERSONALITY 399

Winnicott, D. W. (1965). The maturational processes and the facilitating Wood, A. M., Maltby, J., Stewart, N., & Joseph, S. (2008). Conceptualiz-
environment. New York: International Universities Press. ing gratitude and appreciation as a unitary personality trait. Personality
Wood, A. M., Joseph, S., & Linley, P. A. (2007). Coping style as a and Individual Differences, 44, 619 630.
psychological resource of grateful people. Journal of Social and Clinical Wyatt, G. (2001). Rogers therapeutic conditions: Evolution, theory and
Psychology, 26, 1108 1125. practice. Vol. 1: Congruence. Ross-on-Wye, England: PCCS Books.
Wood, A. M., Maltby, J., Gillette, R., Linley, P. A., & Joseph, S. (in press). Yalom, I. D. (1980). Existential psychotherapy. New York: Basic Books.
The role of gratitude in the development of social support, stress, and Zwick, W. R., & Velicer, W. F. (1986). Comparison of five rules for
depression: Two longitudinal studies. Journal of Research in Personal- determining the number of components to retain. Psychological Bulletin,
ity. doi:10.1016/j.jrp. 2007.1011.1003 99, 432 442.

Appendix

Items of the Final Authenticity Scale

1. I think it is better to be yourself, than to be 9. I am true to myself in most situations.


popular.
10. I feel out of touch with the real me.
2. I dont know how I really feel inside.
11. I live in accordance with my values and beliefs.
3. I am strongly influenced by the opinions of oth-
12. I feel alienated from myself.
ers.

4. I usually do what other people tell me to do. Scoring Instructions

5. I always feel I need to do what others expect me All items are presented on a 1 (does not describe me at all) to
7 (describes me very well) scale. Total Items 1, 8, 9, and 11 for
to do.
Authentic Living; Items 3, 4, 5, and 6 for Accepting External
6. Other people influence me greatly. Influence; and Items 2, 7, 10, and 12 for Self-Alienation.

7. I feel as if I dont know myself very well. Received January 24, 2007
Revision received March 6, 2008
8. I always stand by what I believe in. Accepted March 9, 2008

E-Mail Notification of Your Latest Issue Online!


Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!

You might also like