Changing jobs like changing clothes: the hobo syndrome among career starters

The hobo syndrome (i.e., the wanderlust someone posits to frequently change employers) has a behavioural (i.e., frequent job-quitting behaviour) and an attitudinal dimension (i.e., attitudes towards frequent job-quitting). Across two studies, we examine both dimensions across 348 career starters. By doing so, we expand our understanding of Ghiselli’s hobo syndrome in two ways: (a) we explore the effect of both ‘bright’- and ‘dark’-side personality traits on each dimension of the hobo syndrome, and (b) using longitudinal research, we shed light on the role of each dimension of the hobo syndrome in predicting actual job-quitting behaviour. Data for both studies were gathered through a survey and LinkedIn. Results of regression analyses show that psychopathy is associated with both dimensions of the hobo syndrome. Conversely, openness to experience is only associated with the attitudinal dimension, while agreeableness and extraversion are only associated with the behavioural dimension. Finally, we find that only the behavioural dimension is associated with the length of tenure with the first employer, suggesting that one’s intentions to frequent job-quitting are more important in predicting one’s actual job-quitting behaviour than one’s attitudes towards frequent job-quitting.


Introduction
Employees quitting their job are costly for organizations (see e.g., [1][2][3][4]). Therefore, a vast amount of research has focused on why people leave their jobs (e.g., [5][6][7]). The present study focusses on the 'hobo syndrome' , which refers to "the periodic itch to move from a job in one place to some other job in some other place" (Ghiselli [8], p. 81). In other words, the hobo syndrome refers to the wanderlust someone posits to frequently change employers [9,10]. However, recruiters try to avoid hiring hobos (i.e., individuals who exhibit prototypical characteristics of the hobo syndrome) as they are likely to create an unstable workforce [11]. Therefore, increasing our understanding of the personality of hobos is useful, especially because little is known about what exactly makes someone a hobo [10]. To do so, we explore in our study the role of personality traits in the hobo syndrome, which we measure on both a behavioural and an attitudinal dimension. Additionally, we analyse whether the hobo syndrome influences actual job-quitting behaviour. In particular, we investigate whether the hobo syndrome already manifests itself in career starters' first job.
Our study contributes to the research area of the hobo syndrome in several ways. First, to the best of our knowledge, this is the first paper to measure the hobo syndrome before entering the workforce and to investigate its potential 2 Theoretical background and hypotheses development

The hobo syndrome
The hobo syndrome, which refers to the wanderlust to flutter between employers [9,10], should not be confused with the "advancement motive" for changing jobs (i.e., changing jobs for career advancement) because hobos do not necessarily want to have a successful career [10]. Rather, the hobo syndrome is linked to the "escape motive" (i.e., changing jobs for escaping disliked work environments) [14]. Importantly, as argued by Woo [10], both one's past and one's future job-quitting are, at least in part, explained by the hobo syndrome. Past job switching is therefore an insufficient measure of the hobo syndrome, which is a broader construct of underlying psychological characteristics. Hence, she suggests that the hobo syndrome needs to be measured along two distinct dimensions: (1) frequent job-quitting behaviour, and (2) positive attitudes towards quitting. In doing so, she states that we should describe hobos as "people who frequently quit their jobs for the sake of quitting itself" (p. 7). So, in conclusion, hobos are not simply frequent job-quitters, but are people who also really enjoy the quitting itself.

Personality and the hobo syndrome
Research on the relation between personality and the hobo syndrome is scant. Judge and Watanabe [9], nonetheless, suggested that there needs to be a better understanding of the psychology behind the hobo syndrome in order to validate it. To the best of our knowledge, Woo [10] is the only study so far that has specifically examined the relation between personality traits (openness to experience and impulsivity) and the hobo syndrome, documenting a positive association between openness to experience and the hobo syndrome. Woo [10] claims that hobos score higher on openness to experience because they believe in the goodness of frequent job-quitting, and therefore behave in line with those beliefs as well. Further increasing our understanding of the personality of hobos is useful, as recruiters tend to eliminate job applicants with indications of the hobo syndrome in fear of creating an unstable workforce [11].
While the hobo syndrome has been scarcely researched in personality research, personality does have a long history in the research on vocational behaviour. A large literature shows that personality is associated with frequent job movement [15]. More specifically, 'bright' 1 personality traits have been found to be associated with frequent job-quitting. For instance, several studies have shown that openness to experience and extraversion are positively associated with spontaneous quitting [15][16][17]. Conversely, honesty-humility, agreeableness and conscientiousness are negatively associated with (frequent) job-quitting [17][18][19][20]. The evidence on emotionality is mixed, as one stream of research has found it to be 1 3 a good predictor of intentions to quit and actual turnover [17,[20][21][22], while another stream of research suggests that in order to avoid turnover, organizations should recruit employees with high emotional stability [18,23,24]. Given the documented, multidirectional associations of 'bright' personality traits with spontaneous (and frequent) job-quitting (i.e., indicators of the behavioural dimension of the hobo syndrome), we expect that 'bright' personality traits are associated with the hobo syndrome. Because of mixed prior findings, we do not specify any direction for this association. Specifically, we propose the following explorative hypotheses: Hypothesis 1a 'bright' personality traits are associated with frequent job-quitting intentions.
Hypothesis 1b 'bright' personality traits are associated with positive attitudes towards frequent job-quitting.
Literature on the role of 'dark' personality traits on job mobility is scarcer (e.g., [13,25]), but does show that such personality traits are at least as informative as 'bright' personality traits in predicting job-quitting behaviour [13]. For instance, prior research has found that employees scoring high on narcissism are more likely to be dissatisfied with their job [26], which might increase their intention to quit [27] and to eventually leave their organization [7]. Narcissists might switch organizations to find "narcissistic gratification" [28] as frequently quitting jobs might provide ego satisfaction [29]. Narcissists also show some signs of impulsivity [30][31][32]. Taken together, this suggests that narcissists are more likely to exhibit the hobo syndrome. Further, also psychopathy has been linked to the intention to leave the organization [33]. Especially, its link to impulsivity [30,31,[34][35][36] suggests a potential relation with the hobo syndrome. At last, the evidence on Machiavellianism is mixed. One stream of research suggests that Machiavellians are rather strategic and plan ahead instead of being impulsive, the key element distinguishing them from psychopaths [30]. Further, they cannot deal well with change as it creates uncertainty and stress [37]. Another stream of research, however, suggests that Machiavellians experience lower levels of attachment and dedication to their workplace [38]. In summary, this discussion leads us to expect an association between 'dark' personality traits and the hobo syndrome. Because of mixed prior findings, we do not specify any direction for this association. Specifically, we propose the following explorative hypotheses:

Overview of the current study
The aim of our study is to deepen the existing knowledge of Ghiselli's hobo syndrome. To do so, we divide this study in Study 1 and Study 2. First, we investigate in both studies the relation between both 'bright' and 'dark' personality traits and the hobo syndrome. 2 Second, we examine the role of the hobo syndrome in predicting future job-quitting behaviour. Specifically, in Study 1, we first examine the role of both 'bright' (Hypothesis 1a) and 'dark' (Hypothesis 2a) personality traits on career starters' intentions to perform frequent job-quitting behaviour. Using longitudinal data, we then analyse the link between these intentions and actual job-quitting behaviour (Hypothesis 3). In Study 2, we investigate the role of both 'bright' (Hypothesis 1b) and 'dark' (Hypothesis 2b) personality traits on career starters' attitudes towards frequent job-quitting behaviour. Using longitudinal data, we then examine the influence of these attitudes on actual job-quitting behaviour (Hypothesis 4). Figure 1 gives an overview of the paths that are studied.

Participants
The sample includes 458 career starters (i.e., students in their final semester before receiving their masters' degree) from four Flemish universities. They are aged between 24 and 46 and more than half are men (63%). Participants were first recruited online and then in class in order to increase the response rate. They were not compensated for their participation. Ethics approval was obtained from the university ethics committee before the data collection began.

Materials
Personality traits were measured with two scales. First, we used the Brief HEXACO Inventory (BHI; [45]) in order to measure 'bright' personality traits (i.e., honesty-humility (HON), emotionality (EMO), extraversion (EXTR), agreeableness (AGR ), conscientiousness (CONSC), and openness to experience (OPEN)) [46]. The BHI consists of 24 items. Responses to items were measured on scales ranging from 1 (strongly disagree) to 7 (strongly agree). Second, we used the Short Dark Triad (SD3; [47]) in order to measure 'dark' personality traits (i.e., narcissism (NARC ), Machiavellianism (MACH), and psychopathy (PSY)) [36]. The SD3 scale consists of 27 items. Responses to items were measured on scales ranging from 1 (strongly disagree) to 7 (strongly agree). Table 1 gives an overview of the Cronbach's alpha of each scale. 3 Frequent job-quitting intentions (INT_JOBQUIT) refer to the behavioural dimension of the hobo syndrome and is assessed by measuring how many times the respondent intents to switch between employers during the next ten years. This variable is self-constructed based on Woo [10] and other papers in the field of psychology focussing on job-quitting (see e.g., [48,49]).  2 We focus on the most common studied personality traits, distinguishing between 'bright' personality traits (i.e., honesty-humility, emotionality, extraversion, agreeableness, conscientiousness, and openness to experience) [46] and 'dark' personality traits (i.e., psychopathy, Machiavellianism, and narcissism) [47]. 3 Cronbach's alpha for all subscales of the SD3 are above or close to 0.7, indicating high internal consistency of each scale [63]. Cronbach's alpha for the HEXACO subscales are relatively low, but comparable to those reported by other studies ( [68], e.g., [45,[64][65][66]) and justified by de Vries [45]. The BHI has also been shown to strongly correlate with longer measures of HEXACO personality [45].

5.
1 Actual job-quitting behaviour (ACT_JOBQUIT) measures the length of tenure with the respondent's first employer (see e.g., [50][51][52]). More specifically, this variable measures the duration of employment in months with the first employer.
Demographic variables were included as control variables since prior research shows that these demographic variables are related to job-quitting. We control for sex (SEX) and age (AGE) following Dougherty et al. [53] who studied the job changes of business graduates. We include sex as women tend to switch jobs more often compared to men [54]. Age is included as prior research shows that age is negatively related to frequent job-quitting behaviour [55]. Additionally, we collected information on when the career starter entered the labour market to measure how many months the participant was active in the labour market as of February 2020 (ACTIVE). This allows us to make a more precise comparison, as not all participants started working at the same time. Given that some students take a sabbath year or started another study, we argue that this is a better measure than the one used by Dougherty et al. [53] which controlled for the number of years since graduation.

Procedure
All data were retrieved from two sources. First, we administered a survey during the spring semester of 2016, which included measures capturing personality traits and the hobo syndrome as well as demographic information. The survey was composed in the original English version. Participants were first recruited online and then in class in order to increase the response rate. Second, for our longitudinal research, we gathered data on actual behaviour (i.e., length of tenure with the first employer) through LinkedIn during the month of February 2020.
We had to exclude 110 participants due to more than 50 percent missing data for the variables needed. This yields a total sample of 348 respondents for the first analysis. We conduct a missing value analysis on all items of the scales for these 348 respondents. The value for Little's MCAR test is not significant (χ 2 = 155,488, df = 197, p = 0.987), allowing us to assume that the data which are missing are completely random. For the longitudinal analysis, another 88 respondents had to be excluded because of the impossibility to follow them up four years later. 4 This generates a total sample of 260 respondents for the second analysis.
To test our hypotheses, we apply linear regression analysis using the statistical program SPSS. We estimate the following models to evaluate Hypotheses 1a and 2a (1), and Hypothesis 3 (2): where i represents the regression error term. 5  (2)]. Panel A shows that our respondents intend to switch, on average, 2.13 times from employer in the next ten years (INT_JOBQUIT). Most of the respondents are men (63%) (SEX) and the average age is 24.24 years (AGE). Except for the psychopathy personality trait, respondents' score for personality traits lies on average higher than the personalities' scale midpoint. This is in line with prior personality research ( [47], e.g., [56,57]). Panel B reveals that the respondents' length of tenure with their first employer is, on average, 26.97 months (i.e., 2.25 years) (ACT_JOBQUIT). In accordance with Panel A, the respondents intended to switch, on average, 2.14 times from employer in their first 10 years of employment (INT_JOBQUIT). The

Descriptive statistics and univariate results
majority of respondents are men (58%) (SEX) and the average age is 28.04 years (AGE). The respondents have on average been active in the labour market for 38.85 months (i.e., 3.07 years) (ACTIVE).
Further, Table 1 presents the Pearson correlation matrix for the variables in Study 1. The highest pairwise correlation in panel A is -0.467 between PSY and HON, indicating that career starters who score high on psychopathy generally score low on honesty-humility. The highest pairwise correlation in panel B is 0.600 between ACT_JOBQUIT and ACTIVE, indicating that career starters being active in the labour market for a longer period of time, tend to stay longer with their first employer. Multicollinearity is not a limiting factor in any of the regressions as none of the variance inflation factors exceed 1.7 [58].

Personality and frequent job-quitting intentions
To investigate Hypotheses 1a and 2a, we regress career starters' job-quitting intentions against personality traits and our two control variables SEX and AGE. Table 2 presents the results of this regression analysis. The results show that extraversion (EXTR: B = 0.107, p = 0.077) and agreeableness (AGR : B = − 0.095, p = 0.085) are significantly associated with INT_JOBQUIT, indicating that career starters with higher scores on extraversion or lower scores on agreeableness are more likely to have high job-quitting intentions. These findings are in line with Hypothesis 1a, although extraversion and agreeableness appear to be the only 'bright' personality traits to have an influence on career starters' job-quitting intentions. The results also show that psychopathy (PSY: B = 0.132, p = 0.052) is significantly and positively associated with INT_JOBQUIT, indicating that career starters with higher scores on psychopathy are more likely to possess high job-quitting intentions. This finding is in line with Hypothesis 2a, although psychopathy appears to be the only 'dark' personality trait to have an influence on career starters' job-quitting intentions. As for the control variables, AGE (B = 0.166, p < 0.010) is positively and significantly associated with INT_JOBQUIT. This implies that older career starters are more likely to have higher job-quitting intentions.

Frequent job-quitting intentions and actual behaviour
To investigate Hypothesis 3, we regress career starters' tenure with their first employer against job-quitting intentions and the control variables SEX, AGE, and ACTIVE. Table 3 presents the results of this regression analysis. The results show that job-quitting intentions (INT_JOBQUIT: B = −0.120, p = 0.018) are significantly and negatively associated with ACT_ JOBQUIT, indicating that length of tenure with the first employer is likely to be shorter for career starters with higher job-quitting intentions. These findings are in line with Hypothesis 3. Further, we also find ACT_JOBQUIT to be positively and significantly associated with AGE (B = 0.137, p = 0.010) and ACTIVE (B = 0.541, p < 0.010), indicating that older career starters and these who are in the labour market for a longer period of time tend to stay longer with their first employer.

Participants
This section corresponds to Sect. 3.1.1.

Materials
In this study, we merely use the same materials as in Study 1. Therefore, we refer to Sect. 3.1.2 for more information on the materials. However, Study 2 contributes to Study 1 by focussing on the second measure of the hobo syndrome. For that reason, the new variable in this study is ATTITUDE. Attitudes towards frequent job-quitting behaviour (ATTITUDE) are measured with two questions suggested by Woo [10]. 6 That is, 'I feel positive about changing jobs regularly'; and 'I  6 Originally, the scale of Woo [10] contains a third item [i.e., 'I believe that quitting is bad and that persistence is a virtue' (reverse-coded)]. We had to exclude this item due to its low item reliability. It is not uncommon to delete a reverse-coded item as these kinds of items usually have a lower item reliability. Some research even suggests to avoid reverse-coded items altogether (e.g., [67]). The results using the threeitem scale are qualitatively and quantitatively similar to those reported with the two-item scale.

3
believe that staying at one place too long leads to stagnation' , r(133) = 0.45, p < 0.001). 7 Responses were measured on a scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Procedure
The data collecting process is the same as elaborated in Sect. 3.1.3. Further, we had to exclude 86 participants due to more than 50% missing data for the variables needed. This generates a total sample of 135 respondents. We conduct a missing value analysis on all items of the scales for these 135 respondents. The value for Little's MCAR test is not significant (χ 2 = 14.567, df = 15, p = 0.483). This allows us to assume that the data which are missing are completely random. For the longitudinal analysis, another 43 respondents had to be excluded because of the impossibility to follow them up four years later. 8 This yields a total sample of 92 respondents.
To test our hypotheses, we apply linear regression analysis using the statistical program SPSS. We estimate the following models to evaluate Hypotheses 1b and 2b (3), and Hypothesis 4 (4): where i represents the regression error term. 9 (4)]. Panel A shows that the respondents have, on average, a rather neutral attitude towards frequent job-quitting (M = 4.11) (ATTITUDE). Most of the respondents are men (67%) (SEX) and the average age is 23.82 years (AGE). Except for the psychopathy personality trait, respondents' score for personality traits lies higher on average than the personalities' scale midpoint. This is in line with prior personality research ( [47], e.g., [56,57]). Finally, the respondents have on average a length of tenure with the first employer of 26.97 months (i.e., 2.25 years) (ACT_JOBQUIT).

Table 4 presents the descriptive statistics for the variables used in Study 2. Panel A contains the observations of the first analysis [Model (3)], while Panel B reports on the observations of the longitudinal analysis [Model
Panel B presents the descriptive statistics for the variables used in the longitudinal analysis. Our respondents have, on average, a length of tenure of 24.11 months (i.e., 2.01 years) with their first employer (ACT_JOBQUIT). Again, most of the respondents are men (62%) (SEX) and the average age is 27.77 years (AGE).
Further, Table 4 also presents the Pearson correlation matrix for the variables in Study 2. The highest pairwise correlation in Panel A is − 0.523 between HON and MACH, indicating that career starters scoring high on Machiavellianism generally score low on honesty-humility. The highest pairwise correlation in Panel B is 0.655 between ACT_JOBQUIT and ACTIVE, indicating that career starters who are already longer in the labour market, also stay significantly longer with their first employer. Multicollinearity is not a limiting factor in any of the regressions as none of the variance inflation factors exceed 1.9 (Menard 1995).

Personality and attitudes towards frequent job-quitting behaviour
To investigate Hypotheses 1b and 2b, we regress career starters' attitudes towards frequent job-quitting behaviour against personality traits and our two control variables SEX and AGE.  indicating that career starters with higher scores on openness to experience, tend to possess positive attitudes towards frequent job-quitting behaviour. These findings are in line with Hypothesis 1b, although openness to experience appears to be the only 'bright' personality trait to have an influence on career starters' attitudes towards frequent job-quitting. Further, the results show that the standardized coefficient of psychopathy (PSY: B = 0.241, p = 0.020) is positively and significantly associated with ATTITUDE, indicating that career starters with higher scores on psychopathy tend to possess positive attitudes towards frequent job-quitting behaviour. These findings are in line with Hypothesis 2b, although psychopathy appears to be the only 'dark' personality trait to have an influence on career starters' attitudes towards frequent job-quitting behaviour. Finally, for the control variables, AGE (B = 0.178, p = 0.038) is positively and significantly associated with ATTITUDE. This implies that older career starters are more likely to have positive attitudes towards frequent job-quitting behaviour.

Attitudes towards frequent job-quitting behaviour and actual behaviour
To investigate Hypothesis 4, we regress career starters' tenure with their first employer against attitudes towards frequent job-quitting behaviour and the control variables SEX, AGE, and ACTIVE. Table 6

Discussion: study 2
The objective of Study 2 is to find out whether: (a) personality is linked to attitudes towards frequent job-quitting behaviour; and (b) these attitudes are linked to actual job-quitting behaviour. We measure the former through a survey conducted by career starters from four Flemish universities, and the latter trough follow-up data collected from LinkedIn. Our results show a positive association between the personality traits openness to experience and psychopathy, and positive attitudes towards frequent job-quitting behaviour. The result of openness to experience corroborates Woo's [10] research. Further, the result of psychopathy provides support for the statement by Woo et al. [13] that 'dark' personality traits are at least as informative as 'bright' personality traits in predicting job-quitting behaviour. In particular, we find that the standardized loading of psychopathy is the highest of all personality traits. Further, our results show that older respondents have a higher intention to perform job-quitting behaviour. This seems to contradict previous work suggesting that younger individuals tend to switch employer more often [55]. However, this is likely to be explained by the fact that all our respondents are career starters (i.e., there is limited variation in age in our sample). Additionally, our results show that women tend to have a longer length of tenure with their first employer. We do however not have enough evidence to conclude that people with positive attitudes towards frequent job-quitting behaviour have shorter length of tenure with their first employer. Our results do thus not provide support for Woo's [10] assertion that people with positive attitudes towards frequently quitting jobs, also quit their job more frequently.

General discussion
The aim of this study was twofold. First, we examined whether both 'bright' and 'dark' personality traits are associated with the hobo syndrome. More specifically, this study aspired to found out whether both kinds of personality traits are linked to each dimension of the hobo syndrome, i.e. the behavioural and the attitudinal dimension. Second, we tested whether both dimensions of the hobo syndrome are associated with career starters' tenure with their first employer (as a proxy for actual job-quitting behaviour). We found support for all associations, with one exceptions; the attitudinal dimension of the hobo syndrome is not associated with career starters' tenure with their first employer.

Theoretical implications
Our study has several implications for research. Across two studies, the current paper extends our understanding of the hobo syndrome. First, we find that openness to experience, agreeableness, extraversion, and psychopathy are associated with the hobo syndrome. While prior research by Woo [10] already suggested an association between the hobo syndrome and openness to experience, the current paper also shows the association with two other 'bright' personality traits (agreeableness and extraversion). However, both agreeableness and extraversion have been linked to job-quitting in prior research [15][16][17][18][19]. This suggests that hobos frequently quit their jobs not only because they are curious about what is around the corner (openness to experience), but also because they might be unforgiving and critical for previous experiences during their past jobs (agreeableness) or because they enjoy meeting new people and feel confident about themselves, which might lower the barriers towards quitting their job and start somewhere else (extraversion). Further, to the best of our knowledge, this is the first paper to conjointly investigate the Dark Triad personality traits, which are unexplored in this area [13]. We find that the hobo syndrome is positively associated with psychopathy. This finding aligns with our results on 'bright' personality traits as prior research found that psychopathy is positively associated with openness to experience and extraversion, and negatively with agreeableness [36], the three 'bright' personality traits we found to be linked to the hobo syndrome. Further, in prior research, psychopathy has been linked to the intention to leave the organization [33]. Given that psychopaths exhibit high levels of impulsivity and low levels of anxiety [30,31,[34][35][36], our results suggest that hobos might frequently quit their jobs without thinking it through and fearless of the consequences. Importantly, this adds to Woo's [10] study who did not find a significant association between the hobo syndrome and impulsivity, which is a core feature of psychopathy but is not equal to it [59]. In conclusion, this paper provides thus a more complete picture of the relation between personality traits and the hobo syndrome. Second, as suggested by Woo [10], this study examines the hobo syndrome by measuring it through two dimensions, thereby capturing both behavioural and attitudinal aspects. In doing so, we find that psychopathy is associated with both the behavioural and the attitudinal dimension of the hobo syndrome. Conversely, openness to experience is only associated with the attitudinal dimension of the hobo syndrome, while agreeableness and extraversion are only associated with the behavioural dimension of the hobo syndrome.
Finally, in this study, we not only delve into the underlying roots of the hobo syndrome, but we also seek to explore its influence on actual vocational behaviour such as career starters' tenure with their first employer. While prior research on personality and job mobility primarily focused on job movement during one's career, our study measured one's hobo-ness before entering the labour market and investigates the effect of it on actual vocational behaviour with the first employer. Taking this approach leads us to conclude that only the behavioural dimension of the hobo syndrome predicts one's length of tenure with the first employer, again confirming the link in prior research between intentions and behaviour [39][40][41][42]. No such association is found for the attitudinal dimension of the hobo syndrome, although this was suggested by prior research [10,29]. These results suggest that one's intentions to frequent job-quitting are more important in predicting one's actual job-quitting behaviour than one's attitudes towards frequent job-quitting.

Practical implications
Our study has implications for recruiters and employers as well. Expanding our understanding of the hobo syndrome's construct, could help recruiters before hiring in recognizing individuals' characteristics that are associated with the hobo syndrome. Particularly, our study showed that the behavioural dimension of the hobo syndrome remains the most important dimension in predicting actual job-quitting behaviour. Therefore, besides asking potential recruits about how they feel about working for multiple employers during the course of their career, it could be beneficial for recruiters to also consider their curriculum vitae or inquire about their intentions towards frequently changing jobs. In addition, personality appears to be related to the hobo syndrome as well. Further, even when certain employers do not mind hiring an employee who does not plan to stay for the long term, it might be useful for recruiters to have a secure insight into the personality of hobos in order to find an appropriate job match. Specifically, our study indicates that the hobo syndrome is linked to both 'bright' personality traits (i.e., openness to experience, extraversion, and agreeableness), that have a rather positive notion, and 'dark' personality traits (i.e., psychopathy), that have a rather negative notion. Indeed, dark traits are considered undesirable and are regularly associated with counterproductive work behaviour such as dominance, manipulativeness, impulsiveness, or aggressiveness [36]. Nevertheless, there are also 'bright' aspects to 'dark' traits in particular occupations. Specifically, psychopaths can remarkably well deal with stress [60], making them suitable employees for stressful jobs where rotation is not as harmful for the employer.

Future directions and limitations
There are a few potential weaknesses in the current paper that need to be taken into account. First, we were not able to distinguish between voluntary or involuntary job-quitting behaviour. As the hobo syndrome is referred to as some kind of wanderlust [9,10], indicating that the job has been quit voluntarily, future research could benefit from explicitly examining voluntary and involuntary quitting behaviour separately. Second, we did not examine for which new job participants quit their first job. That is, we did not investigate the direction of the job movement (i.e., upward, downward, or lateral). We are therefore unable to make a clear distinction between the "advancement motive" for changing jobs and the "escape motive" [14]. Third, we acknowledge that our study is based on data from career starters. The role of personality in job-quitting may increase over one's career [15]. Fourth, we cannot exactly determine which subdivision of the personality traits openness to experience, extraversion, agreeableness, and psychopathy are associated with the hobo syndrome. Future research could take a more detailed look at the role of these personality traits in explaining the hobo syndrome. Fifth, examining strategies for employers on how to prevent a hobo quitting the firm goes beyond the scope of our study. Future research could investigate whether hobos leave their employer less frequently when they are given a job with a lot of variation, are repositioned in the firm, or work in a team whose composition changes regularly. Sixt, given the relative short time period of our study, we measure length of tenure with the first employer. Future research could benefit from examining a longer time span and measure actual quitting behaviour across different employers. Finally, we had contradictory results on age in comparison with prior research [55]. Future research should not only focus more on this variable and how it manifests itself with the hobo syndrome, but also on other demographical variables like for example age, education, profession, and nationality to find out whether and how these variables play a role to increase our understanding of the hobo syndrome.
Authors' contributions LK: conceptualization; data curation; formal analysis; investigation; methodology; project administration; software; visualization; roles/writing-original draft. KH: conceptualization; data curation; resources; supervision; validation; writing-review and editing. M-LV: conceptualization; investigation; supervision; validation; roles/writing original draft; writing-review and editing. All authors read and approved the final manuscript.
Funding Not applicable.

Data availability
The data that support the findings of this study are available from the corresponding author upon request.

Data transparency
The study and hypotheses are not based on a pre-registration and the hypotheses are exploratory. Some data used in this research are already, although in part, used in a previously published paper [68]. Specifically, the data that were used in this paper concern both the bright and dark personality traits, gender and age. We are however convinced that the current paper differs extensively from the already published paper. The already published paper focusses on the role of personality career choice of graduating students with a focus on the career choice for accounting. Moreover, the paper continues with a focus on career choice within the accounting profession (Big 4 vs. non-Big 4 accounting firms) and tests whether the Theory of Planned Behaviour mediates the association between personality traits and career choice. The current paper does not focus on which graduation students start in which profession (based on the personality), but focusses on which graduation students already show signs of being a hobo (based on personality), even before starting their career. Therefore, not only the used theory (Theory of Planned Behaviour vs. the hobo syndrome) and hypotheses differ broadly, also the analyses (Structural Equation Modelling vs. linear regression) are distant. The survey used for both papers was set up especially for making these two different papers out of it. Due to ethical reasons, the data cannot be made openly available.
Code availability SPSS syntax can be made available upon request.

Declarations
Competing interests The authors declare that they have no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.