Personality characteristics, preferences for rewards and the propensity to choose an auditing job

Abstract

Purpose

This study investigates theoretical relationships among personality characteristics, preferences for different types of rewards and the propensity to choose a job in auditing by management-related higher education students. Specifically, the authors consider motivation, locus of control (internal and external) and self-efficacy (SE) as personality characteristics and financial, extrinsic, support and intrinsic as types of rewards.

Design/methodology/approach

Data were collected through a questionnaire targeted at management-related higher education students in Portugal. Partial least squares structural equation modelling was used to analyse the data.

Findings

The full sample results show that different types of motivation, locus of control and SE are related to different reward preferences. The authors also find a positive association between a preference for extrinsic rewards and the propensity to choose a job in auditing. Moreover, when the authors consider the role of working experience in the model, the authors find that the reward preferences that drive the choice of an auditing job differ according to that experience.

Originality/value

This study enriches the literature by assessing preferences for different types of rewards, considering multiple personality characteristics and a comprehensive set of rewards. Furthermore, the authors identify the reward preferences that drive the choice of an auditing career. This knowledge empowers auditing firms to devise recruitment strategies that resonate with candidates’ preferences, which boosts the capacity of these companies to attract new auditors.

Citation

Alves, I., Gregório, B. and Lourenço, S.M. (2024), "Personality characteristics, preferences for rewards and the propensity to choose an auditing job", Journal of Accounting & Organizational Change, Vol. 20 No. 6, pp. 56-84. https://doi.org/10.1108/JAOC-04-2023-0080

1. Introduction

Management-related students in higher education are often interested in choosing an auditing career (Espinosa-Pike et al., 2021), usually working with clients from the private (instead of public) sector (Wen et al., 2018). Nevertheless, new entrants in the profession are not enough to replace auditors who leave (Khavis and Krishnan, 2021), reflecting the high rates of turnover in the auditing industry (Nouri and Parker, 2020). Additionally, worldwide fraud scandals have, to a certain extent, eroded the image and social perception of the auditing profession (Bhaskar et al., 2019; Herron and Gilbertson, 2004). The shortage in supply is even more acute because of an increasing demand for auditing services arising from environmental, social and governance reporting (Krasodomska et al., 2021; Pearson, 2014).

In addressing this phenomenon, it is crucial to explore possible ways to attract auditor candidates by analysing factors that may affect the supply of students, i.e. those entering the job market who are intent on pursuing a career in auditing. This topic is of high interest to auditing researchers and other stakeholders in the industry because the shortage of talent in auditing is usually identified as one of the major problems that auditing companies face [Association of Chartered Certified Accountants (ACCA), 2020; Association of Chartered Certified Accountants and Accounting and Corporate Regulatory Authority (ACCA and ACRA), 2012].

Prior research on the choice of an auditing job has focused on personality traits (Fisher, 2001; Holt et al., 2017) and on outcome expectations/career goals (Tetteh et al., 2022). We bridge these two streams of research and develop a theoretical model in which personality characteristics are considered as determinants of reward preferences, which ultimately have an effect on the choice of an auditing job. Rewards are important when considering possible future jobs (Victor and Hoole, 2021), notably in auditing, because they have the power to motivate, influence and determine the performance, satisfaction and turnover in a given job. Rewards also affect job attractiveness to the extent to which they match one’s preferences (Chiang and Birtch, 2007). Therefore, the growing interest in the drivers of self-reported preferences for rewards (Lourenço, 2020), such as employees’ personality characteristics, is not surprising.

In fact, personality characteristics stand out as substantial predictors of one’s preferences (Julian et al., 2021), as they are at the root of human behaviour (Becker et al., 2011). Personality has been increasingly incorporated in management accounting research, inasmuch as people do not make decisions only by applying cognitive elements (Abernethy and Wallis, 2019). That is, personality characteristics are known to have equally large effects on outcomes. Personality characteristics include, among others, motivation, locus of control and self-efficacy (SE). Studying these can contribute to a better understanding of reward systems’ efficacy (Vandenberghe et al., 2008). However, to the best of the authors’ knowledge, no study attempts to establish a relationship between multiple personality characteristics and preferences for rewards, nor a relationship between preferences for rewards and the propensity to choose an auditing job (PAJ).

To examine these issues, we collected data from Portugal. The accounting and auditing industry in Portugal is valued at €1.2bn and is ranked 16th in Europe in 2023 (out of 27 European Union countries). In 2023, the accounting and auditing industry in Portugal is the 77th largest industry (of the 294 total industries that IBISWorld tracks) [1]. The industry’s rank (77th) has remained stable since 2018 [2].

Although several accounting fraud scandals have affected the accounting and auditing professions all over the world, the professions in Portugal do not seem to have been significantly harmed (Leão and Gomes, 2022). The professional associations of accountants and auditors stand as the most prestigious and elitist corporatist organisations in Portugal (Rodrigues et al., 2003). The Big 4 auditing companies have largely contributed to this image with their international prestige, technical knowledge, human resources and contact with international clients (de Almeida, 2012). The Portuguese enforcement mechanisms, also play an important role in raising professional standards of ethical behaviour for professional auditors (de Almeida, 2014), which contributes to the consistent development of this profession in terms of status and performance (Rodrigues et al., 2003). According to Teixeira et al. (2015), auditing and accounting professions are common entry jobs for management-related students in Portugal. To attract new applicants to their firms, Big4 companies usually hold recruitment sessions in the large Portuguese universities with management-related programs, which encompass various courses in accounting, taxation and/or auditing. Therefore, in this context, management-related programs act as a pathway for future auditors.

We collect survey data from a sample of 652 management-related students from Portugal and use partial least squares structural equation modelling (PLS-SEM) to analyse them. Results from our full sample show that extrinsic motivation is positively associated with preferences for financial and extrinsic rewards and intrinsic motivation with support and intrinsic rewards. We also find that external locus of control (ELC) is positively associated with a preference for extrinsic rewards, and internal locus of control (ILC) relates positively to preferences for all types of rewards, except financial rewards. Additionally, SE has a positive association with preferences for financial, extrinsic and intrinsic rewards. Finally, we find a positive association between preferences for extrinsic rewards and the propensity to choose a job in auditing. Moreover, when we consider the role of working experience in our model, we find that the reward preferences that drive the choice of an auditing job differ according to that experience. Specifically, we find that for those with less or no working experience, a preference for extrinsic rewards is the only predictor of their choice of an auditing job, whereas for those with more working experience, a preference for financial rewards is the only significant predictor.

Our study contributes to recruitment and compensation research in auditing. Prior research highlights that the rewards used to direct and motivate audit staff (non-partners) are largely unexplored (Burrows and Black, 1998; Coram and Robinson, 2016; Fiolleau et al., 2023). The few studies in this regard show that the importance placed on various rewards is a function of auditors’ career stage (Lynn et al., 1996; Miao et al., 2009). Our study contributes to this stream of research by showing that rewards are also important to attract prospective job candidates and not only to motivate current auditors. Specifically, we identify the preferences for rewards that are related to students’ willingness to become auditors and how these preferences change with their working experience.

The next sections are organised as follows. Section 2 reviews the literature and develops the research hypotheses. Section 3 presents the methodology and Section 4 presents the results. Finally, Section 5 concludes.

2. Literature review and hypotheses development

2.1 Choice of an auditing job

Previous studies examining the choice of an auditing job have predominantly centred on individual personality traits (Fisher, 2001; Holt et al., 2017) and outcome expectations/career goals (Tetteh et al., 2022). We join these streams in our research by proposing a more holistic theoretical framework. In our framework, personality characteristics are posited to shape reward preferences, which in turn influence the decision to pursue an auditing job. The significance of rewards cannot be understated when evaluating prospective career paths (Victor and Hoole, 2021), especially in the auditing domain, because rewards play a pivotal role in driving motivation, shaping performance outcomes, enhancing job satisfaction and influencing turnover rates.

This is particularly important in the auditing industry, which has high rates of turnover (Nouri and Parker, 2020), especially in large firms (Khavis and Krishnan, 2021). Consequently, the imperative to attract new auditors has become even more pronounced. The literature provides several reasons for the high turnover rates, such as heavy workload and time pressure (Persellin et al., 2019), unsatisfying work–life balance (Khavis and Krishnan, 2021) and scandals that erode the image of the auditing career, entailing distrust (Holtzblatt et al., 2020). Rewards can serve as a pivotal mechanism to attract and retain talent, offsetting the departure of auditors from their careers.

2.2 Personality characteristics

Although several factors have been identified as determinants of career preference and selection, prior research highlights the role of personality in the choice of a given career (Barrick et al., 2001). Each career requires an individual to hold certain personal qualities to perform the tasks involved and find enjoyment in doing so (Raveenther, 2017; Roberts and Robins, 2000). In the context of the auditing profession, Fisher (2001) suggests a connection between personality, notably Type A (characterized by competitiveness, drive, ambition, anger, hostility, irritability, impatience and time urgency) and the role of an auditor. Holt et al. (2017) use the Big Five Personality Model (which encompasses extraversion, agreeableness, openness, conscientiousness and neuroticism) to link accounting students’ personalities to the choice of an auditing profession.

While the literature identifies personality types and traits that are related to the choice of auditing profession, the studies ignore other personality characteristics that shape the preferences for different incentives or rewards. Specifically, because personality characteristics will determine how a person adapts to their career (Holland, 1985), various job attributes such as working environment, monetary incentives, growth opportunities, social recognition and/or flexible working hours may be more or less appealing to persons with different characteristics. In this regard, motivation, locus of control and SE stand out as personality characteristics with the (theoretical) potential to influence reward preferences.

Motivation is a driver that leads human beings to do something, such as engaging in a certain action (Deci et al., 2017). Motivation can vary not only in its strength – high or low – but also in its source – extrinsic or intrinsic (Deci and Ryan, 2000). Deci and Ryan (2000) conceptualize extrinsic motivation (ExtMot) as a driver of behaviour when individuals expect a separable outcome from the task/action. Conversely, individuals with intrinsic motivation (IntMot) engage in a certain task due to the joy and pleasure they gain from performing the task itself, not because they are looking for ex post outcomes (Gagné and Deci, 2005). Hence, IntMot comes from internal factors, whereas ExtMot is driven by external causes. These two concepts are therefore different in nature and, as such, demand different approaches from organisations when designing a reward system (Victor and Hoole, 2021).

Locus of control can be defined as an individual’s beliefs about the influence s/he has over outcomes or what happens in his/her life (Rotter, 1966). The locus of control has two strands – external and internal. If someone is convinced that external causes play a dominant role in outcomes, leaving almost no space for personal influence, she/he has an ELC and she/he is known to be an external. Conversely, if an individual believes that his/her own actions are the main determinants of a given outcome, she/he is considered to have an ILC and she/he is called an internal.

Finally, SE refers to the extent to which an individual thinks that she/he is able to undertake a task successfully (Bandura, 1978). Therefore, SE refers to the individual’s perception of being competent (Lei et al., 2021).

2.3 Preferences for rewards

The aim of a reward system designed and used by an organisation is to attract, motivate and retain high-quality employees (Gomez-Mejia et al., 2004). Reward systems are designed according to what is considered to be employees’ preferences (Lawler, 2000; Rynes et al., 2004). Preferences for rewards and their effects on job outcomes and job choices (Acheampong, 2021) are the subjects of ongoing debate in academia. For instance, contemporary research suggests that matching a stated preference for a reward with the actual receipt of that reward can either lead to a better job outcome (e.g. performance) or have no effect (Lourenço, 2020). A possible explanation for the no-effect outcome is that self-reported preferences may come with some pitfalls, namely, social desirability bias (Rynes et al., 2004) and a lay rationalism effect (Hsee et al., 2003). Social desirability bias occurs when individuals express their preferences according to social norms instead of their own genuine will. Lay rationalism refers to human beings’ tendency to overrate “hard” characteristics, such as money, due to its tangible nature and fungibility, compared to “soft” characteristics, such as non-financial rewards that are difficult to quantify. These biases can also affect the relationships between preferences for rewards and job choices. Therefore, it is even more essential to consider various preferences for different rewards rather than framing the issue as a singular preference for rewards.

Aligned with the idea that there are multiple preferences for rewards, Bussin et al. (2019) suggest that reward preferences can differ based on demographic cohorts and describe rank preferences for types of rewards across the three generations currently active in the labour market. They find generation Y (those who are currently joining the workforce or have recently done so) to be the most demanding one, as they prioritize growth opportunities and learning experiences when choosing a job, placing these considerations immediately above compensation [3], [4]. Other studies suggest that generation Z individuals prioritize non-financial motivating rewards, such as career advancement opportunities for personal growth and development (Iorgulescu, 2016), interesting and meaningful work (Schroth, 2019) and personal fulfilment (Grow and Yang, 2018). However, they also place significant value on tangible rewards such as monetary compensation (Goh and Lee, 2018), benefits at work such as health plans (Maloni et al., 2019), family leave (Grow and Yang, 2018) and job security (Goh and Lee, 2018).

The multiplicity of rewards is also increasingly evident in the auditing industry. A recent study by Fiolleau et al. (2023) shows that beyond monetary rewards for non-partners, audit firms leverage non-monetary rewards such as gift card-based and points-based schemes that are redeemable for merchandise, etc. These are implemented for various purposes, from performance incentives to employee recognition and business development. Notably, while the audit industry has historically favoured penalty-based schemes, there has been growing advocacy for shifting towards more reward-centric approaches to inspire high-quality audits (Peecher et al., 2013).

In response to this evolving landscape, our study draws upon Chiang and Birtch’s (2007) reward typology. This typology ensures a comprehensive exploration of the diverse reward structures that auditing firms are now adopting. These authors classify rewards into four categories: one financial and three non-financial (extrinsic, support and intrinsic). The strength of using this classification resides in its ability to capture the vast diversity of non-financial incentives, which are conceptually distinct and serve different purposes. This framework enables us to state our hypotheses with greater precision, considering the unique nature of each reward type and corresponding personality characteristics.

Financial rewards (FinRew) are cash-flow based rewards such as individual performance incentives, basic salary, overall benefits, team performance incentives, job security and an organisation incentive plan (Chiang and Birtch, 2007). Prior research shows that FinRew can increase the acceptance of difficult performance goals, leading to an enhancement of productivity (Jenkins et al., 1998). Moreover, FinRew attracts above-average workers because only those who are likely to have good performance have a reason to self-select themselves into variable compensation agreements (Gerhart and Fang, 2014).

Extrinsic rewards (ExtRew) are non-financial (not cash-flow-based) and are related to causes external to the task in itself. Following Chiang and Birtch’s (2007) classification, ExtRew includes relationships with co-workers, management style, authority/power, job pressure, job title/status, equity and team spirit. Past research has been especially focused on management style, power and status (Gkorezis and Petridou, 2012).

Support rewards (SuppRew) are non-financial (not cash-flow-based) rewards related to the working conditions that aim to provide joy from performing the task. Thus, they can be considered as more “internal” than the previous category. SuppRew may encompass organisational support, job location, alternative work arrangements, flexible benefits, work–life balance and working environment (Chiang and Birtch, 2007). Berenyi (2022) highlights the importance of non-financial rewards such as organisational support, work–life balance and the working environment for employees to fully perform their duties. Organisational support refers to someone feeling that his/her firm values employees’ contributions and cares about their emotional well-being (Akingbola and van den Berg, 2019). In parallel, work–life balance offers more flexibility to workers to manage their lives, professionally and personally, which enhances job satisfaction (Carleton and Kelly, 2019). Finally, working environment refers to interpersonal relationships, that is, the extent to which ties between colleagues are guided by mutual help and fellowship, which can affect job satisfaction (Paramitha and Indarti, 2014).

Finally, intrinsic rewards (IntRew) are non-financial rewards related to the task in itself, i.e. internal factors. Examples of IntRew, as illustrated by Chiang and Birtch (2007), are: challenge, job variety, accomplishment, autonomy, responsibility, nature of work, opportunity to use skills/ability, learning opportunities and job satisfaction. Organisations that provide their employees with non-financial rewards such as autonomy, challenging tasks, opportunities to use skills and learning moments, are intrinsically rewarding their workforce (see Clay et al., 2022).

2.4 Personality characteristics and preferences for rewards

Theoretically, employees prefer certain rewards over others, depending on their personality characteristics. Regarding motivation, we argue that different types of motivation will likely lead to different types of preferences for rewards. For example, FinRew and ExtRew are based on the instrumental value of doing something (Derfler-Rozin and Pitesa, 2021), namely, the possibility of being eligible to receive money or social recognition (Stajkovic and Luthans, 2001). Hence, ExtMot should be a driver of these preferences, as this type of motivation refers to external motives to engage in a certain task. FinRew and ExtRew encourage people to undertake a given task by providing them reasons to do so beyond the task itself. Hence, we argue that ExtMot should be positively related to a preference for financial rewards (PreFinRew) and a preference for extrinsic rewards (PreExtRew) because these rewards have an external nature. Therefore, our hypotheses are the following:

H1a.

ExtMot is positively associated with PreFinRew.

H1b.

ExtMot is positively associated with PreExtRew.

Conversely, IntMot is about the joy and pleasure of doing a task just for the sake of doing so. This joy may arise because the person finds the job to be interesting or because it represents what she/he stands for (Gagné and Deci, 2005). Both SuppRew and IntRew aim to provide conditions that foster the joy of undertaking a given job. SuppRew aims to provide good working conditions so employees can perform their work at their maximum, thereby establishing an emotional tie between the employee and the employer. Hence, SuppRew target internal factors, enhancing a sense of relatedness with the company (Akingbola and van den Berg, 2019). IntRew are means to design more interesting jobs that favour the inclusion of, for example, learning opportunities and challenging tasks (Victor and Hoole, 2021). Hence, we argue that intrinsic motivation should be positively related to a preference for support rewards (PreSuppRew) and a preference for intrinsic rewards (PreIntRew) as they have an internal nature. Thus, our hypotheses are the following:

H1c.

IntMot is positively associated with PreSuppRew.

H1d.

IntMot is positively associated with PreIntRew.

Locus of control (LC) is also a predictor of preferences for rewards (Heywood et al., 2017) because it refers to the perception that individuals have over outcomes. Specifically, externals usually present low levels of self-esteem (Ng et al., 2006) and high levels of neuroticism (or lack of emotional stability) (Bono and Judge, 2003), which leads to a lack of confidence. Therefore, externals push themselves away from situations that, in spite of being positive in nature, entail a payoff. This is because externals feel they will not be able to meet others’ expectations, i.e. they shield themselves from the possibility of being, in their minds, a potential disappointment (Majerczyk et al., 2020) by rejecting the situation from the outset to avoid feeling anxious (Spector, 1982). Rewards in general (financial, extrinsic, support and intrinsic) are assigned to employees to motivate them, so rewards aim to reach a positive outcome. However, rewards also convey an idea of employees’ obligation to deliver results (Baker et al., 1988). This idea of delivering better results – especially for those who are more self-doubting, such as externals – triggers anxiety. Therefore, we expect that externals shield themselves from these feelings, leading them to discard being rewarded. Thus, our hypothesis is the following:

H2.

ELC is negatively associated with preferences for rewards.

In opposition to externals, internals are usually more optimistic individuals (Ratnawati et al., 2021). This optimism may arise from the belief they have in being in charge of their lives (i.e. in control of what they get), which leads them to think they will achieve desirable outcomes. Moreover, internals also seek help to attain the goals they have set for themselves (Singh et al., 2020), which increases their probability of achievement. According to Ng et al. (2006), in a work setting, internals search for jobs/tasks that have high motivating characteristics, such as rewards. In fact, this idea is consistent with Malik et al.’s (2015) argument that internals do not see incentives (“external interventions”) as a pressure condition but rather as an aid to reach the goals of both themselves and their organisation. Therefore, even if rewards, as stated above, may be seen as a means that organisations have to demand a higher performance (that ultimately is difficult to attain), we argue that internals will see this as an opportunity to defy themselves instead of being afraid not to match others’ expectations (unlike externals). Hence, our hypothesis is the following:

H3.

ILC is positively associated with preferences for rewards.

Because SE refers to the perceptions that individuals have about their probability of success when executing a task, SE is also an important determinant of preferences for rewards. Individuals with more confidence in their personal ability (high SE) perceive greater expectations of their actions and are likely to be more attracted to compensation systems that link their individual behaviour to rewards (Cable and Judge, 1994). In fact, SE is a significant predictor of performance (Kader, 2022) both at work and in academic settings. Individuals with high SE seek positive reinforcements, others’ approval and the feeling of thriving at work (Ashraf et al., 2014). Thus, we argue that individuals with high SE – who like to demonstrate their potential – perceive all rewards as an opportunity to show their abilities and also be acknowledged for their performance. For instance, the prospect of being rewarded with money (financial reward), power (extrinsic reward), a healthy working environment (support reward), or learning opportunities (intrinsic reward) is perceived by individuals with high SE as an opportunity to use their technical, leadership, interpersonal and cognitive skills, respectively. Therefore, we hypothesize the following:

H4.

SE is positively associated with preferences for rewards.

2.5 Preferences for rewards and propensity to choose an auditing job

Preferences for rewards can have a strong influence on job preferences and career decisions (Cable and Judge, 1994). Reward systems in companies may act as signalling devices for potential candidates, affecting job and organizational attractiveness by providing information about less visible organizational attributes (Gerhart and Milkovich, 1992; Rynes and Miller, 1983). Prospective job candidates make search and application decisions based on their perception of the alignment between their own preferences and organizational characteristics (Turban and Keon, 1993). Central to this argument is the nature of the organization’s human resource system (Bretz and Judge, 1994), with a particular emphasis on the rewards available.

Therefore, in our theoretical model, we conceptualize how preferences for rewards drive the choice of an auditing job. Specifically, we consider the PAJ as the extent to which an individual is willing to apply for a job in auditing based on its attractiveness to him/her according to the reward preferences she/he has. In this regard, we expect that PAJ should be aligned with how well auditing companies offer rewards (financial, extrinsic, support and intrinsic) that match the preferences of the potential candidates. For example, prior research suggests that a factor that could affect students’ propensity to choose accounting-related jobs is the salary scale (Jackling and Keneley, 2009; Awadallah and Elgharbawy, 2021). However, other studies (see Frecka et al., 2022) show that auditing firms do not provide above-average starting salaries and that they even underperform similar job options (such as tax or corporate accounting). Moreover, this gap persists over the years (Hoopes et al., 2018). An auditing job is also known for its long hours of work, which makes the financial package even less competitive in relative terms. This means that if students place a great emphasis on financial rewards, i.e. have a strong PreFinRew, they will be less interested in an auditing-related career. Thus, we hypothesize the following:

H5a.

PreFinRew is negatively associated with PAJ.

Similarly, other types of extrinsic preferences, such as power, status or a certain type of leadership style, provided by auditing companies are expected to influence PAJ. Specifically, auditing companies are known by their fast promotion ladder (Pruijssers et al., 2020) and also by their open and communicative leadership style (Duh et al., 2020; Nekhili et al., 2021). Moreover, auditors are viewed as professionals with a positive image regarding their reliability, timeliness, courtesy and respect (Tetteh et al., 2022). As such, students who value these extrinsic factors, i.e. have strong PreExtRew, will be more interested in an audit-related career choice. Hence, we hypothesize the following:

H5b.

PreExtRew is positively associated with PAJ.

Intrinsic factors are also important in the choice of a specific job (Awadallah and Elgharbawy, 2021). In this regard, SuppRew, which is related to perceived organisational support and a healthy working environment, is a special challenge in the auditing profession. Work overload is a structural feature in auditing, as documented by the literature over many years (Hegazy et al., 2023; Persellin et al., 2019). The lack of work–life balance is also often reported by those who have chosen to become auditors (Khavis and Krishnan, 2021), because it is a consequence of work overload. Work overload and work–life (dis)balance are also a concern for auditing regulatory bodies, which see them as threats to the attractiveness of an auditing job [Association of Chartered Certified Accountants and International Federation of Accountants (ACCA and IFAC), 2021]. Furthermore, the lack of auditors’ determination to whistleblow regarding unethical behaviours (Erkmen et al., 2014; Wainberg and Perreault, 2016) suggests a lack of organisational support from auditing firms. Therefore, students who place a great emphasis on organisational support and a healthy working environment, i.e. have a strong PreSuppRew, will be less likely to pursue an auditing job. Thus, we hypothesize the following:

H5c.

PreSuppRew is negatively associated with PAJ.

Finally, intrinsic rewards such as challenging and variable tasks, autonomy, increased responsibility and learning opportunities, are also important for the choice of an auditing profession. Wen et al. (2018) argue that students who believe that accounting-related professions provide a wider variety of professional experiences are more positive about an accounting-related career than others. In fact, individuals who work at auditing firms report constant apprenticeship moments (Kusaila, 2019), classifying auditing as a challenging profession not only because of its content but also because of the variety of industries requiring audits (Low, 2004). Similarly, Jackson et al. (2023) suggest that the dynamic nature of the accounting-related profession is one of the most attractive factors for students. Moreover, professional independence in the workplace positively influences students’ intention to choose an accounting-related job (Wen et al., 2015). Concurrently, those who choose to become auditors expect to have great responsibility as well, because investors and, more broadly, stakeholders rely on auditors’ opinions to make decisions (see Chen et al., 2021). Therefore, students who place a great emphasis on intrinsic rewards, i.e. have a strong PreIntRew, will be more likely to pursue an auditing job. Thus, we hypothesize the following:

H5d.

PreIntRew is positively associated with PAJ.

Figure 1 illustrates our conceptual model.

3. Research method

3.1 Data gathering technique and survey design

Due to the nature of the variables in our model, which are not available via secondary sources, we use survey-collected primary data. The survey, hosted on Qualtrics, was distributed to undergraduate and graduate students in management-related areas at a large university and a large polytechnic institute in Portugal via a link shared with potential respondents [5]. Additionally, other contacts with students and professors in other schools in Portugal were made to increase the number of responses. Linkedin was also used to reach more potential respondents (only within the network of students enrolled in management-related courses) due to its network visibility. To minimize social desirability bias, which is a common issue in behavioural studies, we clarified that neither wrong nor right answers exist, we ensured respondents of the confidentiality of their answers, and we provided the possibility of choosing the option “do not know/do not answer” in all questions.

To increase the response rate, we indicated that it took only 10 min to complete the questionnaire, and we used both reminders and financial incentives (participants who completed the questionnaire were eligible to win one of three vouchers of €50 each). Additionally, we offered participants the possibility of receiving the conclusions of our study as an additional (non-financial) incentive.

We collected 1,077 responses over a two-month period between May and July 2022, but only 652 were usable, i.e. had less than 15% missing values (Hair et al., 2017) [6]. We are not able to compute a response rate because participants could share the survey link with their peers. Additionally, we cannot test for potential (non-)response bias, as we are not able to differentiate early from late respondents as participants were invited to participate in the survey on different days. Nevertheless, we address the concern of common method bias (CMB). CMB can arise because we use the same measurement instrument to collect data for both our endogenous and exogenous variables (Speklé and Widener, 2018). Therefore, we perform a full collinearity assessment, following Kock’s (2015) procedure. Because all values are below 3.3, CMB is not a matter of concern in our model (Kock, 2015). Moreover, we also perform Harman’s one factor test to assess CMB (Podsakoff et al., 2003). The results suggest, once again, that CMB is not a concern because the average variance extracted (AVE) of a single latent variable was 20.8%, which is lower than 50% (Kock, 2021).

The analyses of our sample show that female participants account for 67% of the respondents. The age of our respondents varies from 18 to 65 years, with a mean of 23 and a mode of 19. 52% of the participants report having working experience, but about half of them were in the labour market for less than one year.

3.2 Variables measurement

The questions and items were adapted from previously validated scales, and we pre-tested our survey to ensure its clarity and syntax (Speklé and Widener, 2018). Items’ loadings are presented in the Appendix. Table 1 presents descriptive statistics for all constructs.

Extrinsic motivation (ExtMot) and intrinsic motivation (IntMot) were measured using scales conceived by Gagné et al. (2010). The IntMot scale has three indicators. For ExtMot, we added two items from van der Kolk et al. (2019), thereby making a scale comprising five items. Responses were made on a seven-point Likert scale, for which 1 = Not at all and 7 = Very strongly.

ELC and ILC were measured with Lumpkin’s (1985) scale, which has six items (three for each construct). Respondents were asked about their level of agreement with each indicator, responding on a five-point Likert scale ranging from 1 = Strongly disagree to 5 = Strongly agree.

SE was measured with a scale initially proposed by Chen et al. (2001) and modified by Imperial College London (2019). This scale has six items responded to on a five-point Likert scale (1 = Not at all confident, 5 = Extremely confident).

Preference for financial rewards (PreFinRew), Preference for extrinsic rewards (PreExtRew), Preference for support rewards (PreSuppRew) and Preference for intrinsic rewards (PreIntRew) were assessed with Chiang and Birtch’s (2007) scales. We asked respondents to express their level of appreciation for each reward (item) in a future job on a five-point Likert scale ranging from 1 (I do not give importance) to 5 (I give great importance). This scale has 28 items divided into four types of preferences.

PAJ was measured with questions and items adapted from Bartlett et al. (2017). Respondents were asked about their level of agreement with each item and responded on a seven-point Likert scale (1 = Strongly disagree to 7 = Strongly agree).

3.3 Planned analyses

For the PAJ scale, we start by performing an exploratory factor analysis (EFA) using principal component method. We identify a single latent variable and use it in further analyses.

Next, we test our theoretical model using PLS-SEM with SmartPLS 4.0 (Ringle et al., 2022). We use PLS-SEM rather than a covariance-based (CB-SEM) technique for model estimation due to the nature of our data (Hair et al., 2017). Indeed, CB-SEM requires the distribution of variables to follow a bell-shaped curve (Hair et al., 2017). That condition is seldomly observable, especially in behavioural research that uses psychometric variables (Goodhue et al., 2012). Conversely, PLS-SEM is a distribution-free data analysis technique that is more suitable for our data.

PLS-SEM model assessment is a two-step procedure. In the first stage, we compute a measurement model to infer constructs’ reliability and validity (convergent and discriminant). In addition, all constructs in our measurement model are reflective because latent variables give rise to manifest variables and there are no second-order constructs (Bedford and Speklé, 2018). In the second stage, we use a structural model to test the research hypotheses, after checking that collinearity was not an issue (all variance inflation factors were below 5).

4. Results and discussion

4.1 Results of exploratory factor analysis

The Kaiser–Meyer–Olkin (KMO = 0.872) and the Bartlett’s test of sphericity (p < 0.01) support the EFA for the variable PAJ, and the Kaiser’s criterion allowed extracting one factor with an eigenvalue of 3.64 (the factor explained 73% of total variance and Cronbach’s alpha = 0.9).

4.2 Reflective measurement model evaluation

To examine the measurement model, we start by assessing indicators’ reliability using outer loadings. Loadings above 0.708 should be kept, and those below 0.4 must be excluded (Hair et al., 2017). For those loadings that range between these limits, we delete them (each one at a time) only if, by doing so, it would improve the composite reliability (CR) and/or AVE beyond threshold values – which are 0.7 and 0.5, respectively (Hair et al., 2017).

Furthermore, regarding internal consistency, we focused on analysing Cronbach’s alpha (hereinafter, alpha) and CR, because true reliability of measures is lower bounded by the former and upper bounded by the latter (Hair et al., 2017). Alpha’s acceptable threshold value is 0.5–0.6 (Nunnally, 1978). Analysing the alphas and CR of each construct, we conclude that all latent variables pass the alpha criterion, with the exception of two – ELC and ILC – even though both of them meet CR’s standard. Because the alpha measure is truly conservative in its computation formula, it could provide underestimated reliability (Hair et al., 2017). We opted to keep these constructs in our model so that all hypotheses could be tested, but in a robustness test, we dropped them to infer the quality of our inferences without these constructs.

To examine convergent validity, AVE was assessed. Each latent variable attained the 0.5 threshold, which means that every construct explains at least half of its items’ variance (Hair et al., 2017). In addition, all indicators are statistically significant (all of them at the 1% significance level) with a bootstrapping procedure using 5,000 subsamples. In conclusion, there is no concern about convergent validity in our model.

To analyse discriminant validity, we used Fornell and Larcker’s (1981) yardstick, which demands that each construct share more variance with its items than with other latent variables’ indicators. Table 2 presents the results of these procedures with the square root of AVE in italic diagonal. All inter-construct correlations are lower than items’ square root variance explained by their construct for any variable, thus indicating that each latent variable is unique and, therefore, discriminant validity exists.

4.3 Structural model evaluation

Table 3 shows the empirical results of our conceptual model (base model).

The base model column in Table 3 shows that ExtMot is positively and significantly related to both PreFinRew and PreExtRew, thus supporting H1a and H1b. These findings are consistent with Stajkovic and Luthans’ (2001) arguments, conveying the idea that individuals whose motivation is based on the instrumental value of doing something prefer rewards that target external factors, such as FinRew and ExtRew.

We also find support for H1c and H1d, because IntMot is positively and significantly associated with both PreSuppRew and PreIntRew. These findings are also aligned with previous research (Gagné and Deci, 2005), suggesting that individuals who engage in tasks either because these tasks are interesting by themselves or because they represent what the individual stands for will look for rewards that enhance this feeling of inner motivation, namely, IntRew and SuppRew.

Interestingly, H2, predicting a negative relationship between ELC and preferences for rewards, is not supported. Specifically, ELC is not significantly related to PreFinRew, PreSuppRew and PreIntRew. However, ELC is positively and significantly related to PreExtRew, which is contrary to our hypothesis of a negative relationship. A possible explanation for the positive relationship is that externals tend to feel more unattended and socially excluded (Ye and Lin, 2015), and when ExtRew are provided, they may release a sense of social recognition given by others (Stajkovic and Luthans, 2001), thereby reducing feelings of social exclusion or unattendance. This may lead externals to state a preference for ExtRew. Alternatively, because externals usually experience powerless feelings (Desai et al., 2018), they may attempt to offset these feelings by seeking power in the workplace. In fact, ExtRew provides power, among other things.

H3, referring to a positive relationship between ILC and preferences for all rewards, is partially validated. Specifically, ILC is positively associated with PreExtRew, PreSuppRew and PreIntRew. These findings are consistent with prior literature (Malik et al., 2015; Ng et al., 2006), suggesting that internals see rewards as a way to reach their professional desires. Nonetheless, ILC is not significantly related to PreFinRew, which means that the previous argument does not hold for financial incentives. A possible explanation for this could be social desirability bias (Rynes et al., 2004). Internals are individuals who like to project a good image of themselves (Valentine et al., 2019). In western societies, stating an open preference for money may lead to an association with greed (Zeelenberg et al., 2020), which has a negative connotation. Thus, internals may refrain from openly stating a preference for FinRew, leading to a non-significant coefficient between ILC and PreFinRew.

H4, referring to a positive relationship between SE and preferences for all rewards, is also partially validated. Specifically, SE is positively related to PreFinRew, PreExtRew and PreIntRew. These results are aligned with our argument that high-SE individuals like to display their potential and perceive rewards as an opportunity to exhibit their abilities, which leads to a high preference for rewards. Our results show, however, that this reasoning does not apply to SuppRew, as the relationship is not significant. A possible explanation for this is that individuals who perceive themselves as being effective are usually high achievers. According to Rayburn and Rayburn (1996), high achievers have a Type A personality, which means they are chronically competitive. Their competitiveness, in turn, leads them to be emotionally detached from others (Jia et al., 2022). Because SuppRew aims to build an emotional tie between the employee and the employer, high-SE individuals’ emotional detachment may explain the absence of a preference for SuppRew.

Finally, we find a positive association only between PreExtRew and PAJ, thus validating H5b, but we fail to find statistically significant relationships for other preferences. Therefore, our data do not support H5a, H5c and H5d. The positive and significant relationship between PreExtRew and PAJ advances prior qualitative studies suggesting that social recognition is the main driver to becoming an auditor amongst management-related students (Bekoe et al., 2018; Tetteh et al., 2022), and ExtRew provides, among others, social recognition. Figure 2 shows the final structural model with significant relationships.

4.4 Robustness tests

To analyse the strength of our findings, we perform several robustness tests. In Model 1, Table 3, we drop both ELC and ILC because both constructs were below the lower bound reliability range threshold (alpha < 0.5). Model 1 shows that our results are similar to the base model: all statistical inferences are unaffected as all path coefficients’ significance levels remain similar.

In Model 2, Table 3, we include risk-aversion (hereinafter, RA) as a control variable related to the propensity to choose a job in auditing [7]. RA can be defined as the extent to which a person likes to mitigate risks. In fact, an auditing job is a work of great responsibility (Chen et al., 2021), so planning plays a key role and aims to mitigate risks. Therefore, it is likely that those who are risk-averse tend to like auditing, thus establishing a positive association between RA and PAJ. Indeed, we obtain a positive and statistically significant association (coeff. = 0.091, p < 0.05). More importantly, all results remain qualitatively unchanged (with similar significance levels) when compared to those of the base model.

We perform a third robustness test (untabulated) by running our model on a subsample of only management-related students. Even though our questionnaire was distributed mainly to management-related students, we have a few observations that come from other areas. Because students from non-management-related courses (e.g. law, data management and quantitative methods) can have different motivations and preferences regarding their career choice, we run our main model without this group of students (50 observations). We find that all statistical inferences are qualitatively unchanged, as all path coefficients’ significance levels are similar to those in the main model.

4.5 Additional analyses

Because about half of our sample includes students with working experience, which can influence students’ career decisions to become auditors (Navallas et al., 2022) and their reward preferences, we perform some additional analyses.

First, we include a moderator variable for working experience (measured as the number of months of working experience) in the main model (linking it to each type of reward preferences–PAJ relationship) and find that all statistical inferences remain as all path coefficients’ significance levels and their signals are similar to those of the main model. We also find that working experience positively and significantly moderates the relationship between PreFinRew and PAJ (coeff. = 0.145, p < 0.05).

To further investigate the effect of working experience, we split our sample into two groups – one without working experience and the other with working experience – and ran our main model in these subsamples (applying a bootstrapping procedure with 5,000 subsamples). Table 4 shows the results for these subsamples.

We find that for the sample of students without working experience, the relationship between PreExtRew and PAJ is still the only statistically significant one (coeff. = 0.156, p < 0.05), and all the other preferences for rewards are not related to PAJ. However, for the sample of students with working experience, we find that the relationship between PreFinRew and PAJ becomes significant (coeff. = 0.158, p < 0.1), while the PreExtRew–PAJ relationship becomes nonsignificant (p = 0.267) [8]. However, the test for the difference in coefficients between the two groups is not statistically significant for these relationships. The remainder of the coefficients for these subsamples is in general similar to those reported for the base model (Table 3), except for the locus of control [9].

Nevertheless, business practice indicates that after three years of professional experience, professionals are positioned to pursue roles with greater responsibility, with the corresponding financial incentives. Following this rationale, we split our sample into two groups – one without or with low working experience (i.e. fewer than three years) and the other with high working experience (i.e. 3 or more years) – and ran our main model in these subsamples (applying a bootstrapping procedure with 5,000 subsamples). The last three columns of Table 4 show the results for these subsamples. We find that for the sample of students with less or no working experience, the relationship between PreExtRew and PAJ is the only statistically significant (coeff. = 0.164, p < 0.01). For the sample of students with high working experience, we find that the relationship between PreFinRew and PAJ becomes significant (coeff. = 0.340, p < 0.05), while the PreExtRew–PAJ relationship becomes nonsignificant. Moreover, the test for the difference in coefficients between the two groups is statistically significant. In this analysis, the remainder of the coefficients for these subsamples are in general similar to those reported for the base model, except for the locus of control.

These findings regarding differences in the relationships between preferences for rewards and PAJ based on working experience provide important insights. They show that prospective candidates may be looking for different things in the auditing profession. Those without or with low working experience look for extrinsic rewards such as power, status and leadership style. Conversely, those with high working experience may enter the auditing profession as seniors or as managers. For them, extrinsic rewards may not be as important because they may already be established players in the job market. For them, financial rewards are valuable as they capitalize on their prior working experience. Additionally, because participants with working experience are also older (correlation coeff. = 0.77, p < 0.01), they may already have financial responsibilities and therefore place greater value on stability and pay in an auditing career. In fact, untabulated analyses performed on two subsamples (using a median split based on the age of the participants) yielded results similar to those derived from the subsamples based on working experience, especially concerning the relationships between preference for different types of rewards and PAJ [10].

These results advance prior research showing that the importance that individuals place on various rewards differs according to their career stage (Lynn et al., 1996; Miao et al., 2009) or demographic cohort (Bussin et al., 2019). Specifically, employees in the exploration stage (usually between the ages of 20 and 30, during which an individual is primarily concerned with finding an occupation in which she/he can succeed and grow as an individual) prefer support and peer acceptance; in the establishment stage (usually between the ages of 30 and 45, during which a conscious commitment is made to a particular occupational field, and effort is expended to stabilize oneself) individuals are motivated by promotion and challenge; employees in the maintenance stage (starting generally around the late 30s to mid-40s, a time of holding one’s own and maintaining what has already been achieved) are less competitive and value security (as well as salary, bonus and fringe benefits); while those in the disengagement stage (the final critical adjustment for employees is the transition from working to retirement) are moving into retirement and are likely to focus on issues related to their life after retiring (Lynn et al., 1996; Miao et al., 2009). Our results, using a recent cohort of applicants, unlike those used in prior studies, suggest that for those without or with low working experience (prior to or at the beginning of their exploration stage), extrinsic rewards, such as power, status and leadership style, are important in the decision to apply to an auditing job. These participants have not yet started (or just started) their careers, and, therefore, they may consider an auditing job as a means to gain recognition in the job market, which is key if they aim to move to another job in the future. Our study shows that those with high working experience (in the exploration stage or more advanced) stress more financial rewards, such as basic salary, performance incentives, job security and overall financial benefits, in their decision to seek an auditing job. This may reflect their more demanding financial responsibilities and the more competitive compensation packages offered by auditing companies for those who already have some working experience.

Regarding demographic cohorts, various studies explore the reward preferences of different generations. For example, prior studies found that individuals from generation Z are motivated by rewards of a non-financial nature, such as career advancement, opportunities for learning, growth and development (Iorgulescu, 2016), interesting and meaningful work (Schroth, 2019), personal fulfilment (Grow and Yang, 2018) and good working relationships with their co-workers (Goh and Lee, 2018; Grow and Yang, 2018). Nevertheless, in some studies, pay (Goh and Lee, 2018), benefits at work such as health plans (Maloni et al., 2019) and family leave (Grow and Yang, 2018), as well as job security (Goh and Lee, 2018), also rank high in generation Z reward preferences. Conversely, millennials place a strong emphasis on financial performance-based rewards and job security (Magni and Manzoni, 2020), among other factors. Money could serve as a motivational tool, satisfying millennials’ need for achievement or a sense of entitlement (Ng et al., 2010). Another factor influencing millennials’ preference for financial-based rewards is their life stage, which often involves significant financial commitments such as purchasing homes and starting families (Barhate and Dirani, 2022). Our untabulated results support these arguments. Specifically, using Seemiller and Grace’s (2016) classification, we split our sample into two groups: generation Z (i.e. up to 28 years old) and millennials (i.e. 28 years old or more) and ran our main model in these subsamples. The results suggest that, in their decision to seek an auditing job, millennials value financial rewards more than generation Z (the difference in coefficients between the two groups is statistically significant, coeff. = −0.374; p < 0.10). Therefore, our findings substantiate the claim that for older individuals (millennials), financial rewards are important in their decision to apply for an auditing job. This is consistent with prior research showing that millennials have a pronounced preference for financial rewards, which may reflect their more demanding financial responsibilities. Our study adds to this prior research by showing that this preference for financial rewards is a driver of their choice to pursue an auditing career.

5. Conclusions

Auditing companies struggle to find new auditors to replace those who leave (Khavis and Krishnan, 2021), with high turnover rates being a structural feature in the auditing industry (Nouri and Parker, 2020). Our study addresses this issue by providing insights regarding the choice of an auditing job by management-related students, considering different personality characteristics and preferences for different types of rewards.

We find a positive association between ExtMot and both PreFinRew and PreExtRew. We also find a positive association between IntMot and PreSuppRew, as well as between IntMot and PreIntRew. Such findings reinforce previous research positing that people whose energy is propelled by external drivers prefer money and extrinsic incentives (Stajkovic and Luthans, 2001), whereas individuals whose motivation is driven by internal sources call for rewards that target internal factors (Gagné & Deci, 2005). Regarding LC, ELC is positively associated with PreExtRew, suggesting that externals try to offset their common feelings of social exclusion/unattendance (Ye and Lin, 2015) with social recognition provided by ExtRew assignment. ILC is positively associated with preferences for all rewards, with the exception of PreFinRew. This nonsignificant relationship might be due to social desirability bias, because previous research documents a highly significant association between social desirability and ILC (Valentine et al., 2019). We also find that SE is positively associated with preferences for all rewards, with the exception of PreSuppRew. A possible explanation for this is that high-SE individuals are high achievers with Type A personalities (Rayburn and Rayburn, 1996), which means that they are overcompetitive and tend to be emotionally detached (Jia et al., 2022). Because SuppRew aims to build an emotional tie between the employee and the employer (Chiang and Birtch, 2007), high-SE individuals’ emotional detachment leads to a nonsignificant relationship between SE and PreSuppRew. Finally, we find that only PreExtRew is positively associated with students’ propensity to pursue a job in auditing.

Moreover, when we consider the role of working experience (age) in our model, we find that the reward preferences that drive the choice of an auditing job differ according to that experience (age). Specifically, we find that for those without or with less working experience (younger), PreExtRew is the only predictor of the choice of an auditing job, whereas for those with more working experience (older), PreFinRew is the significant predictor. These are important findings, as they show that prospective candidates may be looking for different things in the auditing profession. Those without working experience (younger) look for extrinsic rewards such as power, status and leadership style. Conversely, those with working experience (older) may enter the auditing profession as seniors or as managers. For them, power, status and leadership style may not be as important because they may already have an established name in the job market. For them, financial rewards are important as they capitalize on their prior working experience, and as they grow older, they likely also have greater financial responsibilities (Barhate and Dirani, 2022), such as purchasing homes and starting families.

Our findings provide a more detailed picture of the drivers of the choice of a career in auditing. Specifically, students not only expect certain outcomes, such as money and social prestige (Tetteh et al., 2022), but the outcomes they expect differ according to their working experience (age). Additionally, these results advance the auditing literature by suggesting that the importance placed on various rewards is not only a function of auditors’ career stage (Lynn et al., 1996; Miao et al., 2009) or age but also influences the decision of prospective candidates (i.e. management-related students) to choose an auditing job.

Our study has practical implications as our findings provide valuable insights for the recruitment strategies of auditing firms, which struggle with high turnover rates (Nouri and Parker, 2020). For example, by highlighting the rewards most important to the choice of an auditing job – non-financial extrinsic and financial – our study can help auditing companies to develop effective recruitment strategies to attract new applicants (Espinosa-Pike et al., 2021), considering their previous working experience. However, past research shows that auditors with high intrinsic motivation are more likely to prioritize the profession itself, are less inclined to engage in unethical behaviour, and are more prone to whistleblowing (Erkmen et al., 2014; Wainberg and Perreault, 2016). Therefore, audit firms should also consider the risks of an excessive focus on financial and extrinsic rewards, as this may discourage prospective candidates with strong intrinsic motivation.

As with any research, this study is not exempt from limitations, and its findings should therefore be interpreted with caution. First, our analysis relies on cross-sectional data, so it is not possible to empirically establish causal relationships (Lourenço, 2019). We can only argue causal relationships at a theoretical level. Second, our study uses a self-reported measure of the PAJ and not an actual choice of applying to such a job (or actually entering the profession). Even though this measure shows high reliability and validity, it ultimately may not be translated into an actual choice.

Our study provides several avenues for future research. Although our setting is in the auditing profession, we believe that our findings can be generalizable to other settings. The relationships between personal characteristics and reward preferences reflect innate traits and therefore could be applicable to a variety of different settings. The relationships between reward preferences and the choice of an auditing job could be generalizable to professions with work environments comparable to the auditing sector, emphasizing extrinsic and financial rewards. Future research could explore these theoretical relationships in different industries and investigate how the specific context affects the relationships between preferences for rewards and career choices. This line of inquiry could provide further insights for the recruitment and compensation literature. Longitudinal research can help document causal relationships from an empirical perspective, which in our research are only posited at a theoretical level. Furthermore, other moderating effects could also provide a more in-depth understanding of how preferences for rewards affect the choice of an auditing job. For example, future research might investigate the role of academic background in the relationships we document. It would also be interesting to investigate whether our findings are specific to auditing within for-profit organizations. Auditing in the public sector can have different drivers, as auditors play a role in the democratic accountability of these organizations (Hay and Cordery, 2021). Similarly, auditing charities or non-governmental organizations may be seen as a different job, as auditors are gatekeepers for the different stakeholders of these organizations. Country comparative analysis would also be an interesting extension of our work because cultural differences may influence the perception of an auditing job by management-related students and the drivers of that choice. Finally, subsequent studies could delve deeper into an individual’s intrinsic motivation as opposed to the extrinsic rewards commonly offered by auditing firms, as these can influence the profiles of potential hires and impact their propensity to speak up and share information within an auditing firm (Kadous and Zhou, 2019).

Figures

Figure 1.

Proposed conceptual model

Figure 2.

Final structural model with significant relationships

Table 1.

Descriptive statistics for the measurement variables

Variable	N	P25	Mean	Median	P75	St dev
ExtMot	645	4.40	5.05	5.20	5.80	1.09
IntMot	639	4.33	4.96	5.00	5.67	1.18
ELC	639	2.00	2.63	2.67	3.00	0.68
ILC	648	3.33	3.57	3.67	4.00	0.60
SE	648	2.50	2.97	3.00	3.50	0.85
PreFinRew	641	3.67	4.00	4.00	4.33	0.56
PreExtRew	637	3.29	3.71	3.71	4.07	0.55
PreSuppRew	648	3.83	4.15	4.17	4.50	0.55
PreIntRew	643	3.78	4.13	4.11	4.55	0.52
PAJ	597	3.40	4.42	4.60	5.60	1.40

Note:

The number of observations (N) per variable is lower than the total sample (N = 652) due to missing values

Source: Authors’ own creation

Table 2.

Convergent and discriminant validity evidence

	1	2	3	4	5	6	7	8	9	10	Alpha	CR	AVE
1. ExtMot	0.762										0.831	0.873	0.580
2. IntMot	0.249	0.842									0.796	0.880	0.710
3. ELC	−0.009	−0.157	0.807								0.466	0.789	0.652
4. ILC	0.143	0.190	−0.279	0.806							0.482	0.785	0.649
5. SE	0.081	0.259	−0.251	0.187	0.868						0.935	0.948	0.754
4. PreFinRew	0.268	0.046	−0.027	0.106	0.130	0.747					0.733	0.834	0.558
7. PreExtRew	0.401	0.134	0.035	0.158	0.221	0.510	0.805				0.712	0.843	0.648
8. PreSuppRew	0.184	0.215	−0.076	0.117	0.053	0.499	0.373	0.746			0.631	0.784	0.557
9. PreIntRew	0.120	0.381	−0.160	0.225	0.365	0.301	0.319	0.470	0.709		0.836	0.876	0.503
10. PAJ	0.248	0.202	−0.096	0.141	0.097	0.129	0.165	0.093	0.091	0.839	0.889	0.921	0.704

Source: Authors’ own creation

Table 3.

Base model and robustness checks

Expected sign	Independent variable	Dependent variable	Base model	Model 1	Model 2
			Coeff.	Coeff.	Coeff.
+	ExtMot	PreFinRew	0.252***	0.260***	0.252***
+		PreExtRew	0.371***	0.386***	0.371***
+	IntMot	PreSuppRew	0.202***	0.217***	0.202***
+		PreIntRew	0.288***	0.307***	0.288***
−	ELC	PreFinRew	0.016		0.016
−		PreExtRew	0.116***		0.116***
−		PreSuppRew	−0.029		−0.029
−		PreIntRew	−0.016		−0.016
+	ILC	PreFinRew	0.056		0.056
+		PreExtRew	0.100**		0.100**
+		PreSuppRew	0.075*		0.075*
+		PreIntRew	0.116***		0.116***
+	SE	PreFinRew	0.103**	0.107**	0.103**
+		PreExtRew	0.202***	0.188***	0.202***
+		PreSuppRew	−0.020	−0.000	−0.020
+		PreIntRew	0.265***	0.285***	0.265***
−	PreFinRew	PAJ	0.052	0.050	0.055
+	PreExtRew	PAJ	0.126***	0.127**	0.130***
−	PreSuppRew	PAJ	0.004	0.007	−0.014
+	PreIntRew	PAJ	0.033	0.033	0.041
	RA	PAJ			0.091**
Model fit		N	652	652	652
		SRMR	0.068	0.070	0.066
		rms Theta	0.124	0.123	0.124

Note:

*, ** and *** indicate significant path coefficient at p-value level of 10, 5 and 1%, respectively (two-tailed)

Source: Authors’ own creation

Table 4.

Main model per subsamples of working experience

Independent variable	Dependent variable	Sample without working experience	Sample with working experience	Difference in	Sample with low working experience	Sample with high working experience	Difference in
		Coeff.	Coeff.	coefficients	Coeff.	Coeff.	coefficients
ExtMot	PreFinRew	0.239***	0.244***	−0.005	0.202***	0.338***	−0.136
	PreExtRew	0.446***	0.316***	0.130**	0.367***	0.346***	0.021
IntMot	PreSuppRew	0.187***	0.199***	−0.013	0.157***	0.297**	−0.139
	PreIntRew	0.347***	0.231***	0.116	0.288***	0.229**	0.059
ELC	PreFinRew	−0.084	0.087	−0.170*	0.017	−0.057	0.074
	PreExtRew	0.052	0.177***	−0.125	0.119***	0.057	0.063
	PreSuppRew	−0.016	−0.028	0.013	0.008	−0.188	0.196
	PreIntRew	−0.054	−0.013	−0.041	−0.027	−0.026	−0.001
ILC	PreFinRew	0.024	0.118**	−0.093	0.077	0.118	−0.041
	PreExtRew	0.018	0.161***	−0.143*	0.116***	0.146	−0.030
	PreSuppRew	0.085	0.113	−0.028	0.139***	0.104	0.035
	PreIntRew	0.091	0.125**	−0.034	0.105**	0.283	−0.178
SE	PreFinRew	0.114	0.077	0.037	0.118**	0.068	0.051
	PreExtRew	0.207***	0.182***	0.025	0.217***	0.116	0.100
	PreSuppRew	−0.014	−0.017	0.002	0.015	−0.097	0.112
	PreIntRew	0.236***	0.288***	−0.053	0.282***	0.194*	0.088
PreFinRew	PAJ	0.043	0.158*	−0.115	0.018	0.340**	−0.322*
PreExtRew	PAJ	0.156**	0.087	0.069	0.164***	−0.116	0.280*
PreSuppRew	PAJ	−0.012	−0.056	0.044	−0.010	−0.042	0.031
PreIntRew	PAJ	0.064	0.017	0.047	0.023	0.110	−0.087
	N	312	340		543	109

Note:

*, ** and *** indicate significant path coefficient at p-value level of 10, 5 and 1%, respectively (two-tailed)

Source: Authors’ own creation

Appendix.

Variables and indicators’ outer loadings

Variables	Abbreviation	Indicators	Loading
Extrinsic motivation	ExtMot	Because this job has the potential to afford me a certain standard of living	0.658c
		Because it allows me to make a lot of money in the future	0.852
		I do this job because it will provide me a good wage in the future	0.727
		Because it provides me with economic certainty	0.796
		For the status it provides me	0.761
Intrinsic motivation	IntMot	Because I enjoy this work very much	0.842
		Because I have fun doing my job	0.837
		For the moments of pleasure that this job brings me	0.848
External locus of control	ELC	Many of the unhappy things in people’s lives are partly due to bad luck	0.797
		Getting a good job depends mainly on being in the right place at the right time	0.101a
		Many times I feel that I have little influence over the things that happen to me	0.817
Internal locus of control	ILC	When I make plans, I am almost certain that I can make them work	0.697b
		Getting people to do the right things depend upon ability; luck has nothing to do with it	0.698c
		What happens to me is my own doing	0.901
Self-efficacy	SE	How confident are you that you will be able to achieve most of the goals you have set for yourself?	0.870
		How confident are you that you will accomplish difficult tasks?	0.871
		In general, how confident are you that you can achieve outcomes that are important to you?	0.886
		How confident are you that you will succeed at most any endeavour you set your mind to?	0.874
		How confident are you that you will successfully overcome many challenges?	0.881
		When things are tough, how confident are you that you can still perform well?	0.824
Preference for financial rewards	PreFinRew	Individual performance incentives	0.828
		Basic salary	0.634c
		Overall benefits	0.753
		Team performance incentives	0.759
		Job security	0.487b
		Organization incentive plan	0.622b
Preference for extrinsic rewards	PreExtRew	Relationships with co-workers	0.443b
		Management style	0.591c
		Authority/power	0.889
		Workload/job pressure	0.475b
		Job title/status	0.887
		Fair treatment/equity	0.250a
		Team spirit/morale	0.433b
Preference for support rewards	PreSuppRew	Organizational support	0.896
		Job location	0.347a
		Alternative work arrangements	0.555c
		Flexible benefits	0.653b
		Balance work–personal life	0.577b
		Working environment	0.748
Preference for intrinsic rewards	PreIntRew	Challenge	0.763
		Job variety	0.708
		Accomplishment	0.689c
		Autonomy	0.626b
		Responsibility	0.641c
		Nature of work	0.497b
		Use skill/ability opportunities	0.772
		Learning opportunities	0.693c
		Job satisfaction	0.686c
Propensity to choose an auditing job	PAJ	In a near future, I see myself applying for a job in auditing	0.885
		If I was invited to take a job offer in auditing, I’d accept it	0.852
		I’m interested in a job in auditing	0.919
		I see myself pursuing a career in auditing	0.902
		Finding a job in auditing has a positive connotation	0.596c

Notes: ^aItems removed because their value was below 0.4; ^bitems removed because their value was above 0.4, but below 0.7, and removing them led to an increase of CR or AVE above threshold values; ^citems preserved because, in spite of their value being above 0.4 and below 0.7, removing them would lead to a decrease of CR or AVE below threshold values. All items preserved in the measurement model are statistically significant at 1% level

Source: Authors’ own creation

Notes

1. IBISWorld is a global industry research firm that provides detailed reports and analysis on various industries used by banks and financial firms, academic institutions, government organizations, consulting firms, and corporations.

2. Data retrieved from www.ibisworld.com/portugal/industry-statistics/accounting-auditing/3880/ on 21 August 2023.

3. Generation Y, also known as millennials, typically includes individuals born between 1981 and 1995, whereas Generation Z, or post-millennials, covers those born after 1995 (e.g. Seemiller and Grace, 2016). However, Bussin et al. (2019) classify generation Y as those born between 1981 and 2000, thereby merging what many recent studies treat as two distinct groups: generation Y and Z.

4. Generation Z closely resembles the demographic of the students we surveyed in this study, as we discuss in Section 3.

5. We targeted management-related students because they are the most likely to become auditors (Espinosa-Pike et al., 2021).

6. To avoid duplicate responses, we limited responses to one questionnaire per IP address and we requested an email address.

7. RA is a four-item variable that was adapted from Payan et al. (2012), whose responses were given on a five-point Likert scale (1 = Strongly disagree, 5 = Strongly agree). After dropping two indicators due to low outer loadings, we managed to reach all thresholds: alpha = 0.71; CR = 0.87; AVE = 0.77. We also checked for discriminant validity and collinearity, and no problems were detected.

8. We conduct similar analyses excluding respondents aged 24 years and older (who are above percentile 75) and observe that the results are inferentially identical. Therefore, we find no evidence that outliers in terms of age are driving our results.

9. The non-results for the locus of control relationships may be the product of the smaller sample size (the number of observations in the main model is roughly twice the number of observations in each subsample), the weaker construct validity of the locus of control variables, and/or it may suggest that locus of control is not an important determinant of reward preferences for participants without working experience. Because these participants may not have experienced the different rewards, they may not associate them with internal or external causes.

10. Specifically, for younger participants, the relationship between PreExtRew and PAJ is the only statistically significant one; conversely, for older participants the relationship between PreFinRew and PAJ is the only statistically significant one.