Discussion 2: Dispositional Versus Situational Factors and Job Attitudes

Contents lists available at ScienceDirect

Personality and Individual Differences

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

The relationship between personality and job satisfaction across
occupations☆

Maria Törnroosa,⁎, Markus Jokelab, Christian Hakulinenb
a Department of Management and Organisation, Hanken School of Economics, Helsinki, Finland
b Department of and Logopedics, Faculty of Medicine, University of Helsinki, Helsinki, Finland

A R T I C L E I N F O

Keywords:
Job satisfaction
Personality
Person-environment fit
Person-occupation fit

A B S T R A C T

Research shows that people select themselves and are selected into occupations, partly because of their per-
sonality, and this has implications for their person-environment fit. Although it has been shown that personality
congruence between the individual and the environment is important to job satisfaction, the effect of personality
congruence in occupations on job satisfaction is not well understood. In a sample of 22,787 individuals, nested
within 25 occupational groups from the British Household Panel Survey and the UK Household Longitudinal
Study, we examined (1) whether average levels of personality vary across occupational groups, and (2) whether
there is a cross-level interaction between the occupational mean personality and the individual’s personality,
with job satisfaction. We found there were modest differences across occupational groups in all FFM traits.
Neuroticism and openness interacted with the corresponding mean personality, showing that for these traits the
fit between an individual’s personality and the average personality of the occupation makes a difference for job
satisfaction.

1. Introduction

Vocational choices are not random. Individuals often seek out, and
are selected into, experiences and environments compatible with their
personality, creating person-environment fit (Kristof-Brown,
Zimmerman, & Johnson, 2005; Schneider, 1987). Person-environment
fit is suggested as a key tool in personnel selection and retention
(Ostroff & Zhan, 2012), and better fit has been associated with higher
employee job satisfaction and wellbeing, as well as lower turnover rates
(Kristof-Brown & Guay, 2011; Verquer, Beehr, & Wagner, 2003). The
general theme of person-environment fit is sometimes discussed in
more specific terms of person-organization, person-job, person-group or
person-supervisor fit (Kristof-Brown et al., 2005). However, individuals
might not be able to select the specific organizations in which they
work, the job they do, or their supervisors. People have more oppor-
tunities to select which broader occupational pathways to follow, as
vocational choice precedes organizational choice (Bradley-Geist &
Landis, 2011; King et al., 2016). Person-occupation fit can therefore be
viewed as an early instance of person-environment fit processes that
may eventually develop into more specific fits between persons, orga-
nizations, and jobs (Bradley-Geist & Landis, 2011). Therefore, the
present study examines whether individuals cluster in occupations

based on their personality, and whether the fit between the individual’s
personality and the mean personality of the occupation is associated
with job satisfaction.

1.1. Occupational choices based on personality traits

Individuals with similar personalities develop similar interests and
drift toward occupations that match those interests, creating person-
ality homogeneity in occupations (Holland, 1997). A personality model
that has frequently been used to study vocational interests is the Five-
Factor Model of personality (FFM). The FFM is one of the most used
taxonomies of personality and defines five broad personality traits:
neuroticism (e.g., anxious, hostile, nervous), extraversion (e.g., soci-
able, fun loving, optimistic), conscientiousness (e.g., well organized,
reliable, persevering), agreeableness (e.g., trusting, flexible, sympa-
thetic), and openness to experience (e.g., originality, independence,
intellectual curiosity) (Costa & McCrae, 1992; Digman, 1990). The FFM
traits have been shown to relate to important outcomes of personal
wellbeing (Hakulinen et al., 2015; Strickhouser, Zell, & Krizan, 2017),
and job-related criteria such as job satisfaction (Avery, Smillie, & Fife-
Schaw, 2015; Cohrs, Abele, & Dette, 2006; Judge, Heller, & Mount,
2002).

https://doi.org/10.1016/j.paid.2019.03.027
Received 14 January 2019; Received in revised form 17 March 2019; Accepted 19 March 2019

☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Declarations of interest: none.
⁎ Corresponding author.
E-mail address: [email protected] (M. Törnroos).

Personality and Individual Differences 145 (2019) 82–88

Available online 26 March 2019
0191-8869/ © 2019 Elsevier Ltd. All rights reserved.

T

http://www.sciencedirect.com/science/journal/01918869

https://www.elsevier.com/locate/paid

https://doi.org/10.1016/j.paid.2019.03.027

https://doi.org/10.1016/j.paid.2019.03.027

mailto:[email protected]

https://doi.org/10.1016/j.paid.2019.03.027

http://crossmark.crossref.org/dialog/?doi=10.1016/j.paid.2019.03.027&domain=pdf

Previous studies examining the association between occupational
clustering of the FFM personality traits have focused on Holland’s
Vocational Type theory (Holland, 1997) to classify occupations. Hol-
land’s framework identifies six vocational interest dimensions, Realistic,
Investigative, Artistic, Social, Enterprising, and Conventional (RIASEC).
The six vocational types characterize both occupational interests and
individual differences in interests (Holland, 1997). For example, Rea-
listic individuals tend to prefer working with things and dislike helping
activities, and typical Realistic jobs include building construction and
auto mechanics (see Holland, 1997, for a detailed description of all
RIASEC types, and Rounds et al., 1999, for a description on which oc-
cupations match the RIASEC).

Early meta-analytic evidence showed that four of the five person-
ality traits were associated with the RIASEC types: openness with
Artistic and Investigative interests, extraversion with Enterprising and
Social interests, conscientiousness with Conventional and Enterprising
interests, and agreeableness with Social interests (Barrick, Mount, &
Gupta, 2003; Larson, Rottinghaus, & Borgen, 2002). More recent stu-
dies have shown that Enterprising and Social occupations are more
likely to be occupied by extravert employees, whereas Conventional
and Realistic occupations are more common for highly conscientious
individuals (King et al., 2016). Furthermore, neuroticism has been
shown to be inversely related with Realistic interests (McKay & Tokar,
2012) and Enterprising interests (Wille & De Fruyt, 2014).

Taken together, although FFM personality traits have been asso-
ciated with RIASEC interest types, it remains unclear how personality
traits are distributed across occupations. By assessing mean levels of
FFM personality traits across 25 occupational groups, we expand the
current knowledge on the extent of occupational homogeneity. This
delivers important information on the contribution of personality to
occupational sorting, which has implications for the composition of
employee pools in organizations.

1.2. Person-occupation fit

The attraction-selection-attrition (ASA) model states that people are
attracted to, and selected by, specific environments according to their
individual dispositions (Schneider, 1987). This selection effect leads to
the environments becoming more homogeneous over time, which re-
sults in an increasingly strong person-environment fit between in-
dividuals and specific environments (Schneider, 1987). Similarly, ac-
cording to Holland’s (1997) theory, people select themselves into work
environments that match their vocational interests, and this congruence
leads to greater satisfaction and better performance. Person-environ-
ment fit is thus the degree to which individual and environmental at-
tributes are compatible (Kristof-Brown et al., 2005).

There is some evidence linking the fit between the person and the
occupation to job satisfaction, although the findings are mixed. Early
meta-analytic work showed that fit matters for some vocational types
but not for all (Tranberg, Slane, & Ekeberg, 1993). A recent study
showed that congruence of RIASEC interest types with occupation was
not associated with job satisfaction (Wille, Tracey, Feys, & De Fruyt,
2014). However, to the best of our knowledge, previous studies have
not examined whether the congruence of personality traits and mean
personality of occupations is associated with job satisfaction. This is
quite surprising, given the importance of person-environment fit for
career counselling (Hartung, 2010) – by showing individuals the oc-
cupation where they might be most satisfied, they can pursue a career
in that occupation but not bound by a certain job. Managers should also
benefit from knowing which occupation is congruent with certain
personality traits, as they can use this information to match the in-
dividual to the right job inside the organization in order to increase
satisfaction at work.

1.3. Present study

The present study has two aims: (1) to examine the average levels of
personality by occupational group, and (2) to examine the association
of the cross-level interaction between the occupational mean person-
ality and the individual’s personality, with job satisfaction.

Regarding the first aim, we expect differences in the mean levels of
the personality traits across occupations. Based on evidence from pre-
vious studies on the association between the RIASEC types and FFM
personality traits (King et al., 2016; Larson et al., 2002; McKay & Tokar,
2012; Wille & De Fruyt, 2014), we expect the mean level of neuroticism
to be lower in Realistic and Enterprising occupations, such as science
and technology, skilled occupations, and managerial occupations. For
extraversion, the mean level will be higher in occupations characterized
by Enterprising and Social interests, such as managerial, personal ser-
vices, and sales. The mean level of conscientiousness will be higher in
Conventional and Enterprising occupations, such as administrative and
managerial. Agreeableness will be highest in Social occupations, such
as education and personal services. Finally, the highest mean level for
openness will be found in occupations characterized by Investigative
and Artistic interests, such as culture, health, and research occupations.

With regard to our second aim, we examine whether the match or
mismatch between an individual’s personality trait and the occupa-
tional mean of that personality trait differentially relates to job sa-
tisfaction. Thus, we examine whether also occupation could be im-
portant for person-environment fit, as occupational categories are
important in guiding people’s career prospects. We base our aim on
person-environment fit theory, which states that it is the congruence
between the individual and the environment that defines fit, and the
better the fit the higher the satisfaction (Kristof-Brown et al., 2005).
Thus, we expect an environment with a corresponding level regarding a
personality trait should increase person-environment fit, and conse-
quently job satisfaction. For example, individuals with low levels of
neuroticism will be more satisfied in Realistic occupations, where the
mean level of neuroticism is lower (fit between the individual’s per-
sonality and the mean personality of the occupation).

2. Method

2.1. Participants and procedure

The data for this study were drawn from the British Household
Panel Survey (Taylor, Brice, Buck, & Prentice-Lane, 2010) and the
Understanding Society – UK Household Longitudinal Study (University
of Essex, 2014). We used data extracted from two waves (collected in
2005 and 2012) for our analyses.

Inclusion criteria were: (a) employed by someone other than your-
self (self-employed participants were not included), (b) data from either
BHPS or UKHLS (participants who had continued from BHPS to UKHLS
contributed data only to BHPS), (c) employable age range (16–65 years;
614 participants were over 65), and (d) complete data on job satisfac-
tion, personality, demographics (sex and age), and occupational group.
We combined the BHPS and UKHLS datasets to form one pooled da-
taset. Based on the criteria outlined above, 22,787 participants were
included in our analyses – 7372 from BHPS and 15,415 from UKHLS. Of
the participants, 12,307 (54%) were female and the mean age was 41
(range 16–65).

2.2. Measures

2.2.1. Job satisfaction
Job satisfaction was measured by a question on overall satisfaction

with the current job (“All things considered, how satisfied or dissatisfied
are you with your present job overall?”) on a 7-point scale from 1 (not
satisfied) to 7 (completely satisfied).

M. Törnroos, et al. Personality and Individual Differences 145 (2019) 82–88

83

2.2.2. Personality
Personality was measured using a 15-item questionnaire on the

Five-Factor Model traits (Gerlitz & Schupp, 2005). Each personality
dimension was assessed with three questions coded on a 7-point scale,
where 0 indicated “does not apply to me at all”, and 6 indicated “applies
to me perfectly”. The value range for each trait’s total score was 0–18.
The internal consistencies (Cronbach α) across our sample (for BHPS
wave 15 and UKHLS wave 3) were: neuroticism = 0.69, extraver-
sion = 0.63, conscientiousness = 0.54, agreeableness = 0.58, open-
ness = 0.65. Although the internal consistency estimates were modest,
the short BFI measure has been shown to have good reliability in terms
of test-retest correlations, and convergent and discriminant validity
against longer personality inventories (Hahn, Gottschling, & Spinath,
2012; Lang, John, Lüdtke, Schupp, & Wagner, 2011; Soto & John, 2017)

2.2.3. Occupation
Occupation was coded according to SOC2000 (Office of National

, 2000) with 25 sub-major occupational groups. SOC2000-
coding is a universal occupational classification system, developed to
allow government agencies and private industry to produce comparable
data. More information on SOC2000 can be found at https://www.bls.
gov/soc/2000/socguide.htm. The sub-major groups of the occupational
coding provide information on the kind of work performed and the skill
level needed for the job. The 25 sub-major groups are listed in the
Appendix (Table A.1).

2.2.4. Control variables
Previous research shows that both age and sex are associated with

job satisfaction (Dobrow Riza, Ganzach, & Liu, 2018; Zou, 2015). Thus,
we included them as control variables. Due to differences in data col-
lection, we additionally controlled for the data source.

2.3. Analyses

In our data, individuals were nested within occupational groups.
The control variables for all analyses were: grand-mean centred age,
sex, and study (0 = BHPS, 1 = UKHLS). The FFM personality traits and
job satisfaction were standardized.

Thus, the model for one personality trait, including a random in-
tercept for job satisfaction, a random slope for personality trait, and a
cross-level interaction between an individual’s personality trait and the
average personality trait of the individual’s occupational group, took
the form: = + + + + +y u P u P P Pij j ij j ij ij j ij00 0 1 1 2 , where γ00 is the
overall mean job satisfaction, yij is job satisfaction of individual i of
occupational group j, Pij is the personality score of person i of occupa-
tional group j, Pj is the average personality score of occupational group
j, u0j is the random intercept of job satisfaction across occupational
groups, u1j is the random slope of personality trait across occupational
groups, β1 and β2 are fixed regression coefficients, and εij is the error
term. In addition, sex, age, and study were included as individual-level
fixed-effect covariates. We first modelled all the fixed effects of the
personality traits together (model 1), then added random slopes for
personality traits for which the random slopes improved model fit
(model 2), and finally added interaction effects between the individual’s
and occupation’s personality traits for the personality traits for which
random slopes had been included (model 3).

We calculated R2 at each of the two levels using the formula re-
commended by Snijders and Bosker (Snijders & Bosker, 1994), which
indicates the explained variance at each level as a proportion of the
total variance. All statistical analyses were conducted using Stata ver-
sion 13 statistical software.

3. Results

Descriptive statistics and correlations of the study variables are
presented in Table 1.

The frequencies of occupational groups are shown in the Appendix
(Table A.1). The largest group was corporate managers (2520 in-
dividuals; 11.06% of the sample) and the smallest was skilled agri-
cultural trades (156; 0.68%).

We calculated the intraclass correlations (ICC1) and the group mean
reliability estimates (ICC2) of our study variables. The ICC1 is an esti-
mate of the extent to which individual ratings are attributable to oc-
cupation, whereas the ICC2 is an estimate of the reliability of the group
means (Bliese, 2000). In our study, ICC2 value was 0.90 for job sa-
tisfaction (ICC1 = 0.010), indicating a highly reliable group mean,
even though only 1% of the variance in job satisfaction was explained
by occupational grouping. ICC2 estimates of the personality traits were
0.97 for neuroticism (ICC1 = 0.031), 0.91 for extraversion
(ICC1 = 0.011), 0.90 for conscientiousness (ICC1 = 0.010), 0.95 for
agreeableness (ICC1 = 0.022), and 0.98 for openness to experience
(ICC1 = 0.043). The ICC2 estimates thus indicated that the occupation
means of the study variables could be reliably estimated with the large
sample, despite the ICC1 estimates being modest (Bliese, 2000).

3.1. Occupational differences in personality traits

There was significant random variance in the intercept of all five
personality traits between occupations (neuroticism: σ2 = 0.09, stan-
dard error [s.e.] = 0.03; extraversion: σ2 = 0.10, s.e. = 0.03, con-
scientiousness: σ2 = 0.06, s.e. = 0.02; agreeableness: σ2 = 0.04, s.e. =
0.01; openness: σ2 = 0.59, s.e. = 0.17), suggesting that the mean level
of the FFM personality traits varies across occupations. The estimated
means and standard errors for the five personality traits are shown in
the Appendix (Table A.2). The standard deviation (SD) for the random
intercept of job satisfaction was 0.12 (square root of 0.0135), which
represented around one-tenth of the overall standard deviation in job
satisfaction across individuals (SD = 1.4).

3.2. Interaction between personality and mean occupational personality
explaining job satisfaction

Results for the association between the FFM personality traits and
job satisfaction across occupations are shown in Table 2. The results of
Model 1 showed significant (p < .05) fixed effects for all personality traits; lower neuroticism and higher extraversion, conscientiousness, agreeableness, and openness were all associated with higher job sa- tisfaction. To test if random slopes were needed, we individually added the slope variance of each personality trait (random-coefficients model; Model 2), and performed a likelihood ratio test to examine which slopes improved the model. The likelihood ratio tests showed that the slopes of neuroticism (p < .001) and openness (p < .001) improved the model. Model 2 results showed there is variance in the slopes of neuroticism and openness, suggesting that for these traits some occupations are a better fit than others. Model 3 in Table 2 included the interaction term between the per- sonality traits with significant slope variance and the corresponding occupation mean personality. Adding the interaction term increased explained variance in level 2, indicating that occupational mean per- sonality and the cross-level interaction explained 4.5% of the variance of job satisfaction at the occupational level. Occupational mean neu- roticism and openness moderated the association between job sa- tisfaction and individual-level neuroticism, and openness, respectively. Examination of the simple slopes revealed there was a negative asso- ciation between neuroticism and job satisfaction when occupational mean neuroticism was high (i.e. +1 SD; γ = −0.11, p < .001), and when occupational mean neuroticism was low (i.e. -1 SD; γ = −0.16, p < .001). There was also a negative association between openness and job satisfaction when occupational mean openness was low (i.e. -1 SD; γ = −0.05, p < .001). The association between openness and job satisfaction was non-significant when occupational mean openness was high (i.e. +1 SD; γ = 0.01, p = .291). These simple slopes are plotted M. Törnroos, et al. Personality and Individual Differences 145 (2019) 82–88 84 https://www.bls.gov/soc/2000/socguide.htm https://www.bls.gov/soc/2000/socguide.htm in Fig. 1. The interaction terms explained 62.5% and 72.7% of the variance in the individual-level neuroticism and openness slope, re- spectively (calculated as the proportion of change between the slope variance in Model 2 and Model 3; neuroticism = [0.0008–0.0003]/ 0.0008, openness = [0.0011–0.0003]/0.0011). To illustrate how the associations of neuroticism and openness with job satisfaction differed across occupations depending on the corre- sponding mean personality of the occupation, we calculated the best linear unbiased predictions (BLUPs), using empirical Bayes predictions for each occupation, producing 25 regression slopes for each person- ality trait. Fig. 2 shows the predicted estimate of each occupation's regression slope. The steepest slopes (indicating the strongest effect of the trait) for neuroticism were found in occupations with the lowest average levels of neuroticism, and the flattest slopes (indicating the weakest effect of the trait) in occupations with the highest average levels of neuroticism. A similar trend was visible also for openness, where occupations with low average levels of openness had a steeper negative slope, and those with high average levels of openness had a flatter slope. 4. Discussion In order to extend person–environment fit theory and offer a better understanding of person-occupation fit, the aim of this study was to examine personality-based fit in occupations. The ASA-model (Schneider, 1987) and Holland's (Holland, 1997) model, on vocational interests, suggest that individuals choose their environment based on their personality, and that some environments then become homo- genous and provide a better fit than others. Drawing on these theories, we proposed that (1) personality steers people into certain occupations, so that different occupations have different mean levels of personality traits, and (2) the variability in how personality is associated with sa- tisfaction in different occupations is partly related to the average level of the personality trait in that occupation. A large sample of almost 23,000 participants from 25 different occupational groups enabled us to model variation in the intercept and slopes of personality traits and job Table 1 Descriptive statistics for the study variables. Variable (range) Mean SD 1. 2. 3. 4. 5. 6. 7. 1. Sex (0 = male; 1 = female) 0.54 0.50 1 2. Age (16–65) 41.46 12.11 −0.01 1 3. Neuroticism (0–18) 7.58 3.85 0.22⁎⁎⁎ −0.09⁎⁎⁎ 1 4. Extraversion (0–18) 10.71 3.59 0.12⁎⁎⁎ −0.09⁎⁎⁎ −0.18⁎⁎⁎ 1 5. Conscientiousness (0–18) 13.43 3.03 0.13⁎⁎⁎ 0.17⁎⁎⁎ −0.15⁎⁎⁎ 0.19⁎⁎⁎ 1 6. Agreeableness (0–18) 13.59 3.01 0.18⁎⁎⁎ 0.08⁎⁎⁎ −0.07⁎⁎⁎ 0.16⁎⁎⁎ 0.37⁎⁎⁎ 1 7. Openness (0–18) 10.84 3.51 −0.06⁎⁎⁎ −0.06⁎⁎⁎ −0.10⁎⁎⁎ 0.26⁎⁎⁎ 0.18⁎⁎⁎ 0.19⁎⁎⁎ 1 8. Job satisfaction (1–7) 5.30 1.38 0.06⁎⁎⁎ 0.03⁎⁎⁎ −0.13⁎⁎⁎ 0.07⁎⁎⁎ 0.12⁎⁎⁎ 0.12⁎⁎⁎ 0.02⁎⁎⁎ ⁎⁎⁎ Significant at the 0.001 level (two-tailed). Table 2 Estimates and standard errors in the multilevel analyses on the associations between personality traits and job satisfaction. Job satisfaction Null model Random-intercept (Model 1) Random-coefficient (Model 2) Cross-level interaction (Model 3) Estimate s.e. Estimate s.e. Estimate s.e. Estimate s.e. Level 1 fixed effects Intercept 0.00 0.02 0.01 0.02 0.01 0.02 −0.00 0.02 Covariates Sex (0 = male, 1 = female) 0.12⁎⁎⁎ 0.02 0.13⁎⁎⁎ 0.02 0.13⁎⁎⁎ 0.02 Age 0.00 0.00 0.00 0.00 0.00 0.00 Data id (0 = BHPS, 1 = UKHLS) −0.16⁎⁎⁎ 0.01 −0.16⁎⁎⁎ 0.01 −0.16⁎⁎⁎ 0.01 Independent variables Neuroticism (N) −0.13⁎⁎⁎ 0.01 −0.14⁎⁎⁎ 0.01 −0.14⁎⁎⁎ 0.01 Extraversion (E) 0.02⁎⁎ 0.01 0.02⁎⁎ 0.01 0.02⁎⁎ 0.01 Conscientiousness (C) 0.06⁎⁎⁎ 0.01 0.06⁎⁎⁎ 0.01 0.06⁎⁎⁎ 0.01 Agreeableness (A) 0.07⁎⁎⁎ 0.01 0.07⁎⁎⁎ 0.01 0.07⁎⁎⁎ 0.01 Openness (O) −0.02⁎⁎ 0.01 −0.02 0.01 −0.02⁎ 0.01 Level 2 fixed effects Occupational mean N (OmN) −0.07 0.13 Occupational mean O (OmO) 0.09 0.10 Cross-level interactions N x OmN 0.12⁎⁎ 0.05 O x OmO 0.15⁎⁎⁎ 0.04 Random effects σ2 0.0135 0.0045 0.0112 0.0038 0.0112 0.0038 0.0105 0.0036 σ2(N) 0.0008 0.0005 0.0003 0.0004 σ2(O) 0.0011 0.0006 0.0003 0.0003 R2level 1 0.043 0.043 0.046 R2level 2 0.157 0.157 0.202 Akaike's information criterion 64,513.62 63,557.82 63,541.25 63,531.11 Bayesian information criterion 64,537.72 63,646.19 63,645.69 63,667.69 σ2 = intercept variance; σ2(N) = slope variance of neuroticism; s.e. = standard error. ⁎ p < .05. ⁎⁎ p < .01. ⁎⁎⁎ p < .001. M. Törnroos, et al. Personality and Individual Differences 145 (2019) 82–88 85 satisfaction across the occupational groups. Regarding our first aim, all FFM traits showed occupational varia- tion, with the greatest found in openness to experience. These findings are in line with the previous studies, where people in the same occu- pation exhibit personality traits more similar to each other than to people in other occupations (King et al., 2016; Woods & Hampson, 2010). Our specific findings were in line with our expectations. The lowest average levels of neuroticism were found in skilled construction and building, and in managerial and health occupations. For extraver- sion, the highest mean levels were in managerial occupations and in business and public service associate professionals. Conscientiousness clustered around skilled trades and managerial occupations, and agreeableness around personal care, leisure, and teaching occupations. The highest mean levels of openness were found in culture, media and sports, and teaching and research professionals. With regard to our second aim, we found that for neuroticism and openness, the fit between an individual's own personality trait and the mean level in the occupation of the corresponding trait, matters for job satisfaction. When occupational mean neuroticism was low, individuals displaying high neuroticism had lower job satisfaction than when they were in an occupation with correspondingly high average levels of neuroticism. Conversely, individuals low in neuroticism had greater job satisfaction in occupations where the average level of neuroticism is low, than in occupations where the average level of neuroticism is high. This finding is in line with research and theories on person-environment fit and personality congruence (Kristof-Brown et al., 2005), and shows that neuroticism appears to be important for person-occupation fit. Furthermore, we found that when occupational mean openness was low, individuals displaying high openness to experience had lower job satisfaction than individuals low on openness. Perhaps those in- dividuals are not happy with the occupation's relatively low level of novelty, cognitive challenge, and opportunities for personal innovation. This may be analogous to a mismatch between a person's and occupa- tion's educational or skill requirements; comparatively over-educated and over-skilled individuals are less likely to be happy with their job (Flisi, Goglio, Meroni, Rodrigues, & Vera-Toscano, 2016). The current study has some practical implications. First, our results suggest some personality dimensions may be …