Subjective career success has been widely researched by academics and researchers as it provides job and career satisfaction that can lead to the perceived life satisfaction of employees, as well as their engagement in organizations. This study demonstrates that subjective career success depends not merely on career adaptability but also on the connections people build throughout their professional lives.
The study was conducted in the socioeconomic context of Latvia with a sample size of 390 respondents. Interpersonal behavioral factors from the perception of career success measure and the influence of the Career Adapt-Abilities Scale (CAAS) on subjective career success (two statements from Gaile et al., 2020) were used. The constructed research model was tested using the SPSS 28 and WarpPLS 8.0 software tools. The primary data analysis method used was partial least squares structural equation modeling (PLS-SEM). Then 12 moderators and their effects on the main relationships of the model were reviewed.
The study confirms that relationships at work have the most significant effect on subjective career success, followed by control behaviors and curiosity behaviors. Moreover, a list of significant and insightful moderation effects was found, most significantly the relationship between connections and subjective career success.
Until now, the CAAS was not integrated with the behaviors and attitudes that depict the social relationships of individuals at work. This study aims to narrow this gap by exploring whether (and, if so, how) career adaptability and interpersonal relationships in the workplace (i.e. professional connections) contribute to subjective career success.
學者和研究人員一直對主觀的事業成功課題進行廣泛的討論和研究, 這類研究會給予僱員工作和事業的滿足感, 繼而使他們感到生活圓滿, 並促進他們對組織的參與。本研究展示了主觀的事業成功不但取決於生涯調適力, 同時也取決於僱員在整個職業生涯裡人際聯繫的建立。
研究人員在拉脫維亞的社會經濟背景下進行這個研究; 樣本為390名回應者。研究人員使用了衡量事業成功概念內的人際行為因素, 以及職業適應能力量表對主觀事業成功的影響 (來自 (Gaile 等, 2020) 的兩個聲明) 。研究人員採用 SPSS 28和 WarpPLS 8.0兩個軟件工具, 來測試他們構建的研究模型。主要分析數據的方法為基於偏最小平方法的結構方程模型, 研究人員仔細審核12個調節因素和它們對模型的主要關聯的影響。
研究確認了工作方面的關係對主觀的事業成功影響最為顯著, 其次則為控制行為和好奇行為; 而且, 研究人員發現了一系列重要的、富有洞察力的調節效果; 更具意義的是, 他們發現了建立聯繫與主觀事業成功之間的關係。
職業適應能力量表至今仍未融合於可描繪在工作上各個個體的社會關係的行為和態度。本研究探索了職業適應能力和在工作場所的人際關係, 如何能促進主觀的事業成功; 就此而言, 本研究縮窄了有關的研究缺口。
Introduction
Career success and the role of proactive behaviors in shaping job-related development have been central topics in vocational behavior research (Spurk, 2021). The key prerequisites for maintaining a successful career include, among other factors, one’s strengths and core interests (Brown and Lent, 2016), career competencies (Presti et al., 2022; Ayoobzadeh, 2022; Francis-Smythe et al., 2012; Kong, 2010; Kuijpers and Scheerens, 2006), one’s ability to overcome career hurdles (Ng and Feldman, 2014), the application of a boundaryless career approach (Bartel, 2021; Arthur and Rousseau, 1996), opportunities to follow one’s professional calling (Bloom et al., 2021; Praskova et al., 2014) and career adaptability (Savickas and Porfeli, 2012).
The widely used Career Adapt-Abilities Scale (CAAS), grounded in career construction theory by Savickas and Porfeli (2012), portrays individuals as active creators of their work environments. This theory highlights the pivotal role of everyday behaviors and choices in crafting careers. The CAAS centers around four key behaviors, known as the 4 C’s – curiosity, concern, confidence and control – that are established antecedents of career success, supported by substantial empirical evidence (Rudolph et al., 2017).
This study bridges a gap by proposing the inclusion of workplace social relationships, labeled as “Connections,” as the fifth C in the CAAS. In contrast to prior research, it highlights the overlooked role of moderators in individuals' career construction efforts. Categorizing these moderators into family life, job attributes, company size and demographic characteristics enhances our understanding of the diverse factors influencing career success, complementing the recognized importance of social capital in previous studies (Lin, 1999; Seibert et al., 2001; Chow, 2002; Lo Presti et al., 2019; Sou et al., 2022; Kauffeld and Spurk, 2022; Boat et al., 2022).
The goal of this study is to examine the influence of the CAAS and interpersonal relationships on subjective career success.
The study aims to develop an expanded Subjective Career Success Model by integrating interpersonal behavior with career adaptabilities. Data were collected from 390 individuals through social media platforms in Latvia and subjected to rigorous reliability and validity testing using SPSS 28 and WarpPLS 8.0 (specialized software for structural equation modeling). The findings of the study demonstrate the effectiveness of the newly developed model.
The study concludes that interpersonal relationships significantly impact subjective career success, with sociodemographic factors like age, education and gender acting as moderators. Suggesting an expansion of the CAAS to include a fifth factor, “Connections,” the study advocates its use in future career success studies. These nuanced findings serve as a roadmap for individuals and career counselors, emphasizing the suitability of different behaviors in diverse life-cycle situations, considering various sociodemographic and family background factors.
This paper is structured as follows. The literature review depicts the theories of vocational behavior, in particular Savickas and Profeli’s theory of career adaptability. Next, the role of interpersonal relationships and connections in career success is explored. The methodology addresses the measurement explanations and research model descriptions of this study. Following this are a discussion and the authors' conclusions.
Literature review
Career Adapt-Ability Scale within theories of vocational behavior
According to Brown and Lent (2016), past vocational behavior research can be classified into three broad areas: (1) agency in career development, (2) equity in the workforce and (3) well-being in work and educational settings. Within the agency category, there is an emphasis on career success and career adaptability. To date, much attention has been paid to human capital within the following topics: career success (Ng et al., 2005; Järlström et al., 2020), psychological and demographic factors (Lyons et al., 2015; Haenggli and Hirschi, 2020; Zacher, 2014), social capital (Järlström et al., 2020; Barthauer et al., 2016), organizational support, organization–person fit (Ng and Feldman, 2014), leader–member exchange (e.g. Restubog et al., 2011; Spurk et al., 2014), member–member exchange (Drabe et al., 2015), performance (Dries et al., 2008) and career adaptability behaviors (Haenggli and Hirschi, 2020; Zacher, 2014) as factors that interplay in the development of a successful career.
Savickas and Porfeli’s (2012) CAAS, a recognized antecedent for career success (Rudolph et al., 2017; Zacher, 2014), aligns with the focus on continuous individual activity in career development and success studies. Lee et al. (2021) note a lack of widespread criticism for the current CAAS version, establishing its positive impact on individuals and organizations. However, this paper critically examines the CAAS, highlighting its task-focused nature and proposing the inclusion of a fifth “C”: connections, acknowledging the undervalued importance of social skills and interpersonal relationships (Rigby and Sanchis, 2006).
Interpersonal relationships (connections) also play a significant role when individuals report their own subjective career success (e.g. Gattiker and Larwood, 1986; Shockley et al., 2016; Ng and Feldman, 2014; Poona et al., 2015). Thus, connections should be added as a separate element to the CAAS to build a more profound and coherent picture of the main factors that lead to career success (which are directly affected by each individual’s behavior).
Social capital and interpersonal connections as predictors of career success
The workplace’s social context is pivotal for success, especially in customer-centric industries like hospitality, retail, tourism and sales. Relationships stand as one of the three pillars of career success in the pursuit of a boundaryless career (Arthur and Defillippi, 2001), intertwined with motivation (Defillippi and Arthur, 1994).
Interpersonal connections form the nucleus of career success predictors, encompassing trust (Wang, 2014), leader–member exchanges (Restubog et al., 2011; Spurk et al., 2014), member–member exchanges (Drabe et al., 2015), developmental networks (Chollet et al., 2021; Cheung et al., 2016; Cotton et al., 2011; Dobrow et al., 2012), mentoring (Ng and Feldman, 2014; Higgins and Kram, 2001; Lancau and Scandura, 2002; Kong, 2010; Defillippi and Arthur, 1994) and sponsored mobility (Maurer and Chapman, 2013).
Numerous studies underscore how social capital impacts executive compensation (Belliveau et al., 1996; Burt, 1997), reduces turnover rates (Krackhardt and Hanson, 1993) and influences career orientations, especially for women (Rodrigues et al., 2019; Choi, 2019). Recent research extends this impact to the career and development outcomes of HR professionals (Gubbins and Garavan, 2016).
Subjective career success
Career success, encompassing job aspects, finances, interpersonal relationships, personal lives, learning and development, is influenced by an individual’s raised life standard and educational level, leading to varied needs (Harrington and Hall, 2007). This shift in understanding subjective career success emphasizes overall job and career satisfaction. Vocational behavior plays a crucial role in achieving the desired outcome of subjective career success.
Possible moderators of subjective career success
Recognizing the complexity of career behavior, the authors employ moderators to assess intervention effects across diverse research sample groups (MacKinnon, 2011). Previous studies highlight demographic factors, such as age (Van der Heijden et al., 2022), gender (Fernandez et al., 2023) and marital status (Agrawal and Singh, 2022), influencing subjective career success. To address this, employers are adopting work–family enrichment options (Awan et al., 2021) and age-adjusted human resource development policies (Van der Heijden et al., 2022).
Job satisfaction has always been related to subjective career success, but it remains debated whether job satisfaction leads to perceived career success (Drabe et al., 2015; Schwormal et al., 2017) or subjective career success leads to job satisfaction (Lehtonen et al., 2022; Sou et al., 2022). Education is the foundation for the individual’s profession and perceived self-worth in the labor market (Duta et al., 2021; Hildenbrand, 2015; Sönmez et al., 2021). Education makes job crafting possible, which lately has been introduced as an antecedent of subjective career success (Kundi et al., 2022; Lo Presti et al., 2023).
Microenvironments and macroenvironments impact individuals' career satisfaction. Family influence, while potentially contributing to career success, may also hinder it if career-building overshadows family commitments (Liu and Yu, 2021). Organizations, as a microenvironment, offer career resources like networking and mentoring, fostering career success through enhanced learning and professional growth opportunities (Agrawal and Singh, 2022; Lehtonen et al., 2022). Larger organizations, with increased resources, are positioned to provide better support for successful careers.
Considering the macroenvironment is crucial in evaluating career success. Socioeconomic status, global income inequality and dependence on institutional capital like cities are pivotal factors (Awan et al., 2021; Bagdadli et al., 2021; Fernandez et al., 2023; Guo and Baruch, 2021). Unemployment experience, tied to macroeconomic shifts, significantly impacts individuals' employability and career success (Manzoni and Mooi-Reci, 2020; Borgen et al., 2021).
The study’s moderators fall into four groups:
Family-related factors (marital status, number of children, proportional contribution to family budget and previous unemployment experience).
Job-related factors (total work experience, years in a current position, current monthly salary level, liking a current job from the start and education).
Organizational attribute (company size).
Demographic characteristics (gender and age).
The methodology applied for the research is outlined in the next section, and the application of these moderators is designed to increase the practicality of the study.
Methodology
Subjective Career Success model has been created. It includes the Savickas and Porfeli (2012) CAAS behaviours: concern, control, curiosity and confidence, and interpersonal behavior (connection) behaviours by Gattiker and Larwood (1986). All statements were measured using the ten-point Likert scale, which has a higher validity and explanatory power than the five-point Likert scale suggested by Coelho and Esteves (2007).
Finally, subjective career success was measured in accordance with Gaile et al. (2020) by using two statements: “To what extent are you satisfied with your job?” from Colakoglu (2011), Converse et al. (2014) and Verbruggen et al. (2015), and “I am satisfied with the success of my career” from Greenhaus et al. (1990). A calculated Cronbach’s alpha with a value of 0.768 indicates good internal consistency for the construct of this factor.
The study data were collected in April 2020 by inviting people from different backgrounds via social media platforms, such as LinkedIn and Facebook. The questionnaire was written in Latvian, suggesting that the results can be attributed to Latvia’s socioeconomic context. A total of 390 valid responses were obtained. Table 1 details the study sample descriptions.
Study’s sample descriptives
| Total number of respondents: 390 | # | % | |
|---|---|---|---|
| Gender | Female | 283 | 72.56 |
| Male | 107 | 27.44 | |
| Age | Range | 23–67 | |
| Mean | 40.07 | ||
| S.D. | 9.65 | ||
| Education | Primary school | 2 | 0.51 |
| Secondary school | 40 | 10.26 | |
| Bachelor’s degree | 104 | 26.67 | |
| Master’s degree | 229 | 58.72 | |
| Doctoral degree | 15 | 3.85 | |
| Marital status | Not married | 174 | 44.62 |
| Married | 216 | 55.38 | |
| Children | Range | 0–5 | |
| Mean | 1.04 | ||
| S.D. | 1.1 | ||
| Contribution to family budget (%) | Range | 10–100 | |
| Mean | 64.09 | ||
| S.D. | 25.30 | ||
| Salary level (EUR) | 500 or less | 0 | 0.00 |
| 501–1,000 | 101 | 25.90 | |
| 1,001–2,000 | 177 | 45.38 | |
| 2,001–5,000 | 85 | 21.79 | |
| More than 5,000 | 27 | 6.92 | |
| Company size | Solo | 25 | 6.41 |
| Small | 106 | 27.18 | |
| Midsized | 112 | 28.72 | |
| Large | 147 | 37.69 | |
| Liking current job at its start (Likert-10) | Range | 0–10 | |
| Mean | 7.82 | ||
| S.D. | 1.76 | ||
| Years in current position | Range | 0–44 | |
| Mean | 5.63 | ||
| S.D. | 6.14 | ||
| Total work experience (years) | Range | 0–50 | |
| Mean | 18.93 | ||
| S.D. | 9.75 | ||
| Unemployment experience | Yes | 214 | 54.87 |
| No | 176 | 45.13 | |
Source(s): Authors
The constructs' reliability and validity were assessed using SPSS 28 (Statistical Package for the Social Sciences). Subsequently, a mathematical model was implemented in WarpPLS 8.0, a user-friendly software package for variance- and factor-based structural equation modeling (SEM) employing the partial least squares method (PLS), as recommended by Hair et al. (2014). PLS-SEM is a widely accepted method for exploratory research in various fields, including management and organizational development (Al-Emran et al., 2018; Kock and Hadaya, 2018).
WarpPLS, a software program developed by ScriptWarp Systems, is a powerful tool for predictive PLS-SEM cases rooted in established theories (Hair et al., 2014). Notably, for exploratory research, PLS-SEM is favored (Kock and Hadaya, 2018), and WarpPLS stands out by allowing the explicit identification of nonlinear functions connecting latent variables in SEM models and calculating associated multivariate coefficients of association, a capability unique to this software (Kock, 2010). Unlike other tools that offer solely linear functions, WarpPLS is the first to provide classic PLS algorithms alongside factor-based PLS algorithms for SEM (Kock, 2019).
Research model
Figure 1 presents the initial research model, which contains the interpersonal behavior construct (connections) as well as the four career adaptabilities constructs (control, curiosity, confidence and concern). This reflects the starting point for the research meta-model in this paper and the key constructs for further PLS-SEM analysis.
The general PLS-SEM analysis results include model fit and quality indexes: average path coefficient (APC), average R-squared (ARS), average adjusted R-squared (AARS), average block variance inflation factor (AVIF) and average full collinearity VIF (AFVIF). It is recommended that the p values (significance) for APC, ARS and AARS should all be equal to or lower than 0.05; this was the case for the main research model (APC = 0.220, p < 0.001; ARS = 0.410, p < 0.001; AARS = 0.402, and p < 0.001). Ideally, both AVIF and AFVIF should be equal to or lower than 3.3 (particularly in models where most of the constructs are measured through two or more indicators); this was true for the model used (AVIF = 1.699 and AFVIF = 1.863).
Measurement model: reliability and validity
To ensure the validity and reliability of the reflective measurement model, various tests were conducted using the most frequent techniques according to Ringle et al. (2012). Thus, the internal consistency reliability with Cronbach’s alpha (CA) and composite reliability (CR), the convergent validity with the average extracted variance (AVE) and the discriminant validity with the Fornell–Larcker criterion (Ringle et al., 2012) were tested. Validity indicates the degree to which a measurement model can predict what will be measured. By contrast, reliability checks the degree to which the same measured values lead to the same results; this represents the failure rate (Burns and Burns, 2008; Weiber and Mühlhaus, 2014).
In the beginning, the internal consistency reliability was tested (for which the average correlation of all of the individual items of the same construct are compared). Thus, this shows the accuracy of a group of variables or items measuring a latent variable. Internal consistency reliability is mostly measured with CA and CR (Ringle et al., 2012). The higher the values of CA and CR, the more congruent the items (ergo, the higher the internal reliability). In the model, the CA of all the constructs was considerably greater than 0.7 (Table 2), which is mentioned as the threshold. The CR should have a value of at least 0.6, which is reflected in the current model values in Table 2 (ranging from 0.855 to 0.903) (Weiber and Mühlhaus, 2014). Thus, this model is internally reliable.
Reliability and validity measures
| CR | CA | AVE | PCS | CONN | CONT | CURI | CONF | CONC | |
|---|---|---|---|---|---|---|---|---|---|
| PCS | 0.896 | 0.768 | 0.811 | 0.901 | 0.511 | 0.389 | 0.114 | 0.268 | 0.251 |
| CONN | 0.881 | 0.797 | 0.712 | 0.511 | 0.844 | 0.353 | 0.140 | 0.234 | 0.248 |
| CONT | 0.860 | 0.803 | 0.506 | 0.389 | 0.353 | 0.712 | 0.511 | 0.562 | 0.468 |
| CURI | 0.855 | 0.788 | 0.543 | 0.114 | 0.140 | 0.511 | 0.737 | 0.666 | 0.634 |
| CONF | 0.903 | 0.871 | 0.609 | 0.268 | 0.234 | 0.562 | 0.666 | 0.780 | 0.461 |
| CONC | 0.890 | 0.844 | 0.619 | 0.251 | 0.248 | 0.468 | 0.634 | 0.461 | 0.787 |
Note(s): (CR: composite reliability; CA: Cronbach’s alpha; AVE: average variance extracted) as well as inter-construct correlation matrix (with square roots of AVEs shown in italic on diagonal)
Source(s): Authors
Regarding testing the convergent validity, AVE was used; this determined the average percentage of the items that explain the dispersion of the latent construct. In the literature, a threshold of 0.5 is mentioned for this factor (Fornell and Larcker, 1981; Bagozzi and Yi, 1988); this was the case for all the constructs in the model (Table 2). Lastly, the discriminant validity with the Fornell–Larcker criterion (Fornell and Larcker, 1981) was checked. Therefore, the square root of each AVE in a diagonal must be compared with the correlation coefficients for each construct. Table 2 shows that the AVE was higher in each case in the model, so the discriminant validity is accepted. Table 3 provides structure loadings and cross-loadings for deeper insights into the construct’s reliability and validity.
Structure loadings and cross-loadings
| PCS | CONN | CONT | CURI | CONF | CONC | |
|---|---|---|---|---|---|---|
| PCS1 | 0.901 | 0.466 | 0.367 | 0.113 | 0.245 | 0.223 |
| PCS2 | 0.901 | 0.456 | 0.333 | 0.092 | 0.238 | 0.23 |
| CONN1 | 0.447 | 0.844 | 0.355 | 0.146 | 0.251 | 0.239 |
| CONN2 | 0.458 | 0.865 | 0.317 | 0.108 | 0.18 | 0.212 |
| CONN4 | 0.387 | 0.821 | 0.217 | 0.1 | 0.16 | 0.175 |
| CONT1 | 0.493 | 0.362 | 0.632 | 0.231 | 0.321 | 0.423 |
| CONT2 | 0.292 | 0.281 | 0.767 | 0.41 | 0.384 | 0.393 |
| CONT3 | 0.267 | 0.273 | 0.773 | 0.465 | 0.539 | 0.37 |
| CONT4 | 0.124 | 0.146 | 0.699 | 0.339 | 0.327 | 0.236 |
| CONT5 | 0.267 | 0.281 | 0.733 | 0.386 | 0.492 | 0.305 |
| CONT6 | 0.24 | 0.164 | 0.653 | 0.326 | 0.312 | 0.273 |
| CURI1 | 0.025 | 0.003 | 0.337 | 0.759 | 0.365 | 0.458 |
| CURI2 | 0.116 | 0.115 | 0.38 | 0.809 | 0.48 | 0.583 |
| CURI3 | 0.107 | 0.163 | 0.441 | 0.67 | 0.445 | 0.403 |
| CURI4 | 0.253 | 0.234 | 0.46 | 0.713 | 0.669 | 0.407 |
| CURI6 | −0.077 | 0.011 | 0.277 | 0.726 | 0.508 | 0.472 |
| CONF1 | 0.229 | 0.222 | 0.403 | 0.436 | 0.739 | 0.28 |
| CONF2 | 0.216 | 0.211 | 0.42 | 0.454 | 0.73 | 0.242 |
| CONF3 | 0.173 | 0.128 | 0.429 | 0.635 | 0.788 | 0.487 |
| CONF4 | 0.197 | 0.114 | 0.384 | 0.563 | 0.779 | 0.428 |
| CONF5 | 0.162 | 0.203 | 0.461 | 0.527 | 0.837 | 0.35 |
| CONF6 | 0.283 | 0.22 | 0.529 | 0.499 | 0.806 | 0.363 |
| CONC1 | −0.007 | 0.069 | 0.135 | 0.43 | 0.212 | 0.734 |
| CONC2 | 0.143 | 0.148 | 0.276 | 0.441 | 0.3 | 0.753 |
| COCN3 | 0.234 | 0.206 | 0.437 | 0.52 | 0.43 | 0.842 |
| CONC4 | 0.259 | 0.218 | 0.458 | 0.487 | 0.353 | 0.744 |
| CONC5 | 0.336 | 0.315 | 0.508 | 0.604 | 0.494 | 0.851 |
Source(s): Authors
Results
The study results in this section are presented in three sub-sections. The first describes the main model, as well as the key relationships between the dependent variable of the subjective career success and the independent variables of connections (control, curiosity, confidence and concern). The second outlines the statistically significant moderating effects of the 12 variables (Table 4) on the main relationships in the research model. The third explores how the significant moderating effects altered the relational strengths among the independent and dependent constructs.
Overview of moderating effects in main research model
| CONN → PCS | CONT → PCS | CURI → PCS | CONF → PCS | CONC → PCS | |
|---|---|---|---|---|---|
| Marital | β = 0.160***p < 0.001 | β = −0.090* p = 0.036 | |||
| Children | β = −0.147**p = 0.002 | β = −0.099* p = 0.024 | |||
| FamilyBudget | β = 0.100* p = 0.023 | ||||
| Unemployment | β = 0.195***p < 0.001 | ||||
| YearsTotal | β = −0.094* p = 0.030 | ||||
| YearsCurrent | – | – | – | – | – |
| Salary | – | – | – | – | – |
| LikeStart | β = −0.193***p < 0.001 | ||||
| Education | β = −0.085* p = 0.045 | β = −0.104* p = 0.019 | |||
| CompanySize | β = −0.107* p = 0.016 | ||||
| Gender | β = 0.087* p = 0.042 | ||||
| Age | β = 0.095* p = 0.028 |
Source(s): Authors
Main model
The main results of the structural model are presented in Figure 2. The β values that are noted next to each arrow demonstrate the strength of the relationships among the constructs, and the asterisks mark their statistical significance (the R2 contributions are presented in brackets). All the paths in the model are statistically significant. The model visibly demonstrates how significantly strong the effect of the connections is on explaining subjective career success (β = 0.412, p < 0.001). This is especially true when comparing it with the path coefficients of the other four constructs that represent the career adaptability factors (β = 0.236–0.117, p < 0.001–0.050).
The total effects and effect sizes are also provided in Figure 2. The effect sizes (f2) determine whether the effects indicated by the path coefficients are small (0.02), medium (0.15) or large (0.35). This study reveals that relationships at work (connections) dominate in the size of their effect on subjective career success (f2 = 0.217) as compared to the effect sizes of the other four constructs (f2 = 0.096–0.023) related to the CAAS. The current model in Figure 2 portrays the differences by changing the arrow sizes and thicknesses accordingly.
Of the four career adaptability factors, control had the strongest predictive power on subjective career success (β = 0.236, p < 0.001), with a medium to small effect size (f2 = 0.096). The next largest effect on subjective career success (β = 0.198, p < 0.001) came from curiosity, with a small effect size (f2 = 0.023). The effect of confidence (f2 = 0.039) and concern (f2 = 0.036) on subjective career success is minimal. The strengths of these relationships, however, were comparatively the weakest in the model, with confidence (β = 0.139, p < 0.01) being slightly stronger than concern (β = 0.117, p < 0.05) in predicting subjective career success for the study sample.
The coefficient of the determination value (R2) indicates the model’s ability to explain and predict the constructs (Ringle et al., 2012). The R2 of the overall model was 41%, which shows quite a good predictive accuracy according to Hair et al. (2014). As the model’s value is around 50%, this indicates that the measurements fit well to the model and that the independent variables explain the variance of the dependent variable quite effectively.
Moderating effects
In a further analysis, the potential effects of the moderators (Table 4) on the main relationships of the research model were investigated. In PLS-SEM analysis, moderating effects provide deeper and richer insights into how various factors might influence the strengths of model relationships. Typically, there are three ways they do this: a moderator increases, decreases or has no significant effect on a relationship in a model. Figure 3 presents a conceptual version of the research model by portraying an example of a moderator and its relational effect on the main paths of the research model.
Each moderator was analyzed separately to gain insights into its effects on the main model relationships. The 12 moderators of this study (Table 4) were as follows: marital status (Marital), number of children (Children), proportional contribution to family budget (FamilyBudget), previous unemployment experience (Unemployment), total work experience in years (YearsTotal), years in current position (YearsCurrent), current monthly salary level (Salary), liking one’s current job from the start (LikeStart), education level (Education), company size (CompanySize), gender of the individual (Gender) and age (Age). After running the analysis with the 12 moderators, all the significant moderating effects were summarized. Table 4 shows this summary, along with the moderators’ strengths (β values) and levels of significance (p values).
Table 4 provides clear evidence that the relationship between connections and subjective career success had the widest variety of significant moderating effects (both positive and negative). Not only did connections have the greatest effect on explaining individuals’ subjective career success, but they also had the highest sensitivity regarding many personal and contextual parameters. The relationships that control and concern behaviors have with subjective career success had no significant moderating effects according to this study. The importance of curiosity behaviors for subjective career success was negatively moderated by both the marital status (Marital) and educational level (Education) of an individual. Similarly, the importance of confidence behaviors in determining subjective career success was also negatively moderated by two factors: one’s number of children (Children) and total work experience in years (YearsTotal).
Deeper exploration of moderating effects
In this exploratory research, the WarpPLS software offered a unique opportunity to investigate the nonlinear functions of moderating effects on the connecting pairs of latent variables. Therefore, deeper and richer insights into some of the most interesting moderating effects are shared in this section.
Positive moderating effects on CONN > PCS relationship. First, Figure 4 provides a closer look into the moderating effect of age on the CONN > PCS relationship. The 3D graph visually portrays how the effect of connections on subjective career success increases with age. The turning point is somewhere around 50 years of age. In the focused graph, a difference between the younger and older generations of the respondents is depicted.
Positive moderating effect of age on relationship between connections and subjective career success
Positive moderating effect of age on relationship between connections and subjective career success
Figure 5 provides deeper insights into the moderating effect of the proportional contribution to the family budget on the relationship between connections and subjective career success. A significant increase happens at around the levels of 60 and 80% of the proportional contributions. The focused graph depicts the difference between the groups of respondents depending on low vs high degrees of proportional contributions to the family budget.
Positive moderating effect of proportional contribution to family budget on relationship between connections and subjective career success
Positive moderating effect of proportional contribution to family budget on relationship between connections and subjective career success
Discussion
In the career-management domain, the predominant focus revolves around choosing a profession aligned with personal aspirations. This study emphasizes a crucial question: “What kind of people am I willing to work with?” This adds a significant theoretical dimension to subjective career success. Notably, relationships at work wield greater influence (β = 0.412, p < 0.001) than career adaptability factors (β = 0.236–0.117, p < 0.001–0.05) on individuals' perceived career success. Moderators like years in the current position and current monthly salary show no significant effect, possibly due to their short-term nature, aligning with counterintuitive findings on pay level and job satisfaction from previous research (Judge et al., 2010).
A noteworthy discovery is that control and concern behaviors' relational effects on subjective career success remained unaffected by any moderator. The study indicates a general trend where career adaptability behaviors' relational effects on subjective career success were less influenced by moderators than the effects on subjective career success. This suggests that adaptability, focusing on internal self-regulation strategies, differs from relationships at work, which reflect external (social) behaviors.
The study indicates that higher education fosters independence, reducing reliance on work relationships. Conversely, work connections hold significant importance for less-educated individuals in perceiving career success. Additionally, a negative moderating effect among more-educated respondents was observed in the relationship between curiosity and subjective career success. This suggests that higher education may lead to elevated career expectations, potentially resulting in lower career satisfaction due to a misalignment with market realities.
Confidence has a diminished impact on subjective career success with increased work experience, suggesting its greater relevance in early career stages. This intriguing finding suggests a need for future studies exploring distinct sets of career adaptability behaviors at different life stages. Notably, this aligns with age, where older and more experienced individuals prioritize connections at work over career adaptability behaviors in influencing subjective career success.
Respondents who initially enjoyed their jobs experienced a diminished impact of work connections on subjective career success. This suggests relationships may serve as a compensatory mechanism for those with lower initial job satisfaction. Another unexpected finding is that connections at work have a stronger effect on subjective career success for those without prior unemployment experience. Speculatively, unemployment trauma might diminish the belief in relationships, making positive work connections less influential in determining subjective career success.
Married individuals attribute greater importance to work relationships for subjective career success compared to unmarried counterparts. This could be tied to the younger age and limited work experience of unmarried individuals, making workplace connections less significant. Alternatively, unmarried individuals may face psychological barriers in forming connections, diminishing the overall importance of relationships, including at work.
Respondents with more children experience diminished effects of work connections and confidence behaviors on subjective career success. This suggests a shift in focus away from career priorities, potentially leading to reduced emphasis on behavioral or external factors. It also implies that family relationships may compete with work relationships for these individuals.
Respondents contributing more to the family budget and prioritizing work relationships experience increased subjective career success. For this group, concern behaviors play a crucial role, reflecting their heightened responsibility for dependents.
In larger companies, work relationships exert less influence on subjective career success compared to smaller counterparts. This is attributed to the dominance of a “formal organization” in larger organizations, where colleague relationships are less crucial. In contrast, smaller organizations, characterized by a more flexible “informal organization” (Child and Mansfield, 1972), place a higher importance on relationships between colleagues, aligning with theories of market and hierarchical coordination by Thompson (2003).
This study acknowledges limitations inherent in its methodology, particularly the reliance on an online survey that may introduce a self-selection bias toward digitally savvy white-collar workers. Geographical constraints also limit the generalizability of results beyond the specific location studied, despite efforts to ensure sample diversity. While the study boasts a satisfactory sample size, the acknowledgment of the potential benefits of larger samples highlights a trade-off in statistical power and precision. Despite these constraints, the study offers valuable insights for future research and practical applications in organizational settings, emphasizing the need for contextualized interpretation within the study’s parameters.
Conclusions
This study highlights that workplace relationships play a more significant role in predicting subjective career success than Career Adaptability Assessment System (CAAS) behaviors. Various personal and contextual factors influence the importance of these relationships, offering nuanced insights into career-development strategies.
The study introduces the concept of the “fifth C” – connections – as a crucial factor in CAAS, influencing subjective career success. Intriguingly, connections exhibit a negative effect on subjective career success for highly educated individuals, suggesting diverse career strategies based on life stages, family situations and sociodemographic factors.
Education negatively moderates the relationship between curiosity behaviors and subjective career success, possibly due to heightened expectations accompanying higher education. Job attractiveness at the start and unemployment experience also moderate the impact of connections on subjective career success.
These findings imply that tailored career strategies are essential, considering individual variations in family situation, age, experience and education. Organizations can create more opportunities for relationship-building, especially for less-educated and younger employees. Acknowledging that investments in workplace relationships can enhance or distract from career success, this study opens avenues for future research and provides valuable insights for individuals, career counselors and HR managers aiming for more meaningful and fulfilling careers.
This paper forms part of a special section “Innovation, Knowledge, and Digitalization: Building Trust to Face Today’s Challenges”, guest edited by Alba Yela Aránega and Juan Miguel Alcántara-Pilar.
