The effect of work pressure on audit quality within a Big-4
accounting firm
A view of young professionals
Name: Maurien Rikkert Student number: s4739264 Thesis supervisor: Dr. M. Visser Turn in date: 09-07-2020 Specialization: Accounting & Control Radboud University, Nijmegen, The Netherlands
i
Abstract
Recent incidents created a negative image of accountants and indicated a structural quality problem within the accountancy sector. Therefore, many rules and regulations are created to increase the quality again. This increased regulation can have unintended consequences, such as increased regulation pressure, especially for young professionals who are still in training to become a CPA.
The objective of this study is to investigate the effect of work pressure and regulation pressure on audit quality and the JD-CS model contains the theoretical foundation to analyze this relationship. Few studies have investigated this relationship in the accountancy sector. Also, these studies did not focus on the mediating role of decision latitude and social support. Audit quality is measured by using the questionnaire of Brown et al., (2016), which is based on the PCAOB framework. This study does not focus on the main effects, but instead focuses on the mediating role of decision latitude and social support in the described relationship. The corresponding hypotheses are tested by performing a multiple regression analysis with interaction effects. The results do not confirm the mediating role of decision latitude and social support, but only show the significant negative main effect of psychological demands on audit quality.
1
Content
1. Introduction ... 3 1.1 Problem definition ... 3 1.2 Research question ... 5 1.3 Relevance ... 61.4 Headlines of the research ... 6
1.5 Structure of the thesis ... 7
2. Literature review ... 8
2.1 Young professionals ... 8
2.2 Audit Quality ... 9
2.2.1 Definitions of audit quality in scientific research ... 9
2.2.2 Measuring audit quality ... 10
2.3 Work pressure... 13
2.3.1 Job Demand-Control-Support model... 14
2.4 Control variables ... 20
2.5 Conceptual models ... 21
3. Methodology ... 22
3.1 Field surveys... 22
3.2 Research design ... 22
3.2.1 Population, sample design and data collection ... 22
3.3 Operationalization of variables... 23
3.3.1 Dependent variable ... 23
3.3.2 Independent variable ... 25
3.4 Method for analysis ... 27
3.5 Reliability and validity of the study ... 29
2 4. Results ... 30 4.1 Missing data ... 30 4.2 Descriptive statistics ... 30 4.3 Cronbach’s alpha ... 33 4.4 Multiple regression ... 34 5. Conclusion ... 39 5.1 Conclusion ... 39 5.2 Theoretical implications ... 40 5.3 Practical implications ... 41
5.4 Limitations and future research ... 42
References ... 43
Appendices ... 50
Appendix 1: PCAOB Audit Quality Framework ... 50
Appendix 2: IAASB Audit Quality Framework ... 51
Appendix 3: Operationalization Audit Quality by Brown et al., (2016) ... 52
Appendix 4: Operationalization of audit quality ... 55
Appendix 5: Planning Master Thesis ... 56
Appendix 6: Survey Qualtrics ... 57
Appendix 7: SPSS output ... 68
3
1. Introduction
1.1 Problem definition
The public accounting sector has come under fire in recent years. Numerous scandals have taken place the last couple of years, such as the scandals of Enron (2001) and Ahold (2003). These occurring incidents create a negative image of accountants and cause a lot of criticism on the accountancy sector in recent years. According to Carnegie & Napier (2010) and Smith (2015) this negative image is created due to ‘stereotypes’ and the stereotype of Certified Public Accountant describes accountants as cold, aloof, impersonal, dull and uncreative individuals. This is conflicting with the fact that CPA accountants consider themselves as accountants with skilled inter-personal abilities necessary to maintain long-term relationships with clients. But the society is not only doubting the image of accountants, also their skills are being questioned. For example, one of the discussions that arose, is whether accountants are still professionals or whether they are more and more trying to meet the commercial interest of clients. Therefore, it can be said that the accounting profession is currently experiencing a period of crisis where the society often questions the professional role of the accountant and their existing revenue model (Dirsmith, Covaleski, & Samuel, 2015; De Vries, Blomme, & De Loo, 2018; Van Brenk & Renes, 2020).
This negative picture of accountants due to these critics has been reinforced in recent years by the shortage of quality that has been observed (Werkgroep Toekomst Accountantsberoep, 2014). In 2006 the Wet toezicht op accountangsorganisaties was brought to life. This law was meant for safeguarding the public function of the auditor’s report and fostering confidence in the financial market. The
responsibility for this law lays with the AFM (Autoriteit Financiële Markten) (Overheid.nl, 2006). The AFM started in 2009-2010 by publishing a report about the current quality within Big-4 accounting firms, which was meant to provide insight into the aspects that underlie the lack of quality within these firms (Autoriteit Financiële Markten, 2015). The next report was published in 2014 to see whether there were already observed changes in quality with regard to the measures taken (AFM, 2015). These reports eventually led to the creation of Werkgroep Toekomst Accountantsberoep within the NBA (Nederlandse Beroepsorganisatie Accountants), which is responsible for coming up with measures that restore the trust in the quality of the accountants (Werkgroep Toekomst Accountantsberoep, 2014). This workgroup published the report ‘In het publieke belang’ (2014) where they suggest measures which could recover the trust by assuring transparency of quality and independence. The report also mentioned that the concatenation of the quality incidents indicate a structural problem within the accountancy sector, where the society and politics do no longer trust the inner motivation of accountants to deliver high quality. One of the suggested measures was to create a monitoring committee to help restore the credibility and authority of the accountancy sector (Monitoring
4
Commissie Accountancy, sd). Now, 5 years later (2020), a new report is published by this monitoring committee (MCA). The report (‘spiegel voor de accountancy sector’) shows that in many cases the quality of auditing still does not meet the standard, despite the measures that have been suggested 5 years ago. The result of this finding is that there is a performance gap between what is expected from accountants and what is actually achieved by accountants (Monitoring Commissie Accountancy, 2020). This performance gap reflects the structural problem that was mentioned earlier and the underlying reason for this performance gap is mainly a lack of technical expertise. The main criticism of the report is that the accountancy sector felt pressure from the environment to change and therefore there was no internal motivation to change. Due to this pressure and the lack of self-will at some companies, there are still bottlenecks that need to be resolved before the desired level of quality can be achieved (Monitoring Commissie Accountancy, 2020). The report in 2014 also mentioned, among other things, a culture where there is not enough attention for aspects such as work-pressure, time-pressure and budget time-pressure. The latest report states that more attention is being paid to the high workload that was ascertained 5 years ago and that accounting firms themselves are taking several measures to reduce this workload.
Although measures have been taken to reduce the workload within the public accounting sector, there is still a very high workload among accountants.Several news articles have been published to show that there is still dissatisfaction among the accountants about the workload they experience (Piersma, 2018; Piersma, 2018; Isitman & Polman, 2019). Especially young accountants who are still in training to become a CPA, also called young professionals, experience difficulties combining work, study and social life (Piersma, 2018). But not only newspapers have been publishing more and more about work pressure, also scientific literature pays attention to this high workload. According to Ozkan &
Ozdevecioglu (2013) accountants face increased stress because accountancy mainly depends on intellectual activities and needs continuous attention. Also, the shortage on the labor market
contributes to the higher workload within the accounting sector. According to the NBA, the demand for accountants has increased with 26% in 2018 compared to 2017. Therefore, it will be extremely difficult for accounting firms to fill their vacancies in the future and this will ensure that there is not enough labor capacity for the accountancy sector to keep growing (Nederlandse Beroepsorganisatie voor Accountants, 2018; Maassen, 2018).
To be able to reduce this workload, it is important to get insight in the factors that underlie this problem. De Vries & Herrijgers (2018) have conducted a research into the factors that cause a high workload among young professionals. One of these factors is the structure of the audit firms in which young professionals work. The audit firms are characterized as hierarchical organizations in which they control on hard targets.These targets lead to a culture of ‘checking-the-boxes’ instead of actually
5
performing the job that has been studied for (Kornberger, Justesen, & Mouritsen, 201; Ladva & Andrew, 2014; De Vries & Herrijgers, 2018).
The checking-the-boxes’ culture has increased in recent years due to high regulation that is needed to improve the audit quality again (De Vries & Herrijgers, 2018; Knoop & Piersma, 2018; Piersma, 2018). The regulation covers all rules and regulation that apply to public accounting, which is
considerably larger than for accountants who do not exercise the public accounting profession(NBA, 2018). This increased regulation started in 2005 when International Financial Reporting Standards (IFRS) was implemented to create more corporate transparency, quality and better opportunities for comparability (Daske, Hail, Leuz, & Verdi, 2008). Besides this, the accountant needs to comply with a lot of rules and regulations that are included in the HRA. But in recent years, the pressure of AFM that is felt by accounting firms and the critique on the accountant increased this regulation pressure and enhanced the checklist culture. Accounting firms are creating more checklists and controls to assure good quality and at the same time they are also forced to apply the 53 measures proposed in the 2014 report from the NBA (Werkgroep Toekomst Accountantsberoep, 2014; Piersma, 2014). Therefore, a negative vicious circle seems to emerge. These measures are needed to tackle the low quality of audits, but at the same time they increase the workload and the regulation pressure. These measures can lead to more accountants leaving the public accounting sector which will lead to an even higher workload. Research conducted by the NBA (2016) has namely proven that almost 40% of the accountants that leave the public accounting sector do this because of the great emphasis on rules and regulations. This would mean that precisely this focus on quality causes a higher intention to leave the organization for accountants, which ultimately has a negative effect on the goal of increasing quality (Nederlandse Beroepsorganisatie van Accountants, 2016).
1.2 Research question
The negative vicious circle, described above, is alarming for the financial market and for accounting firms and therefore further research is needed. The aim of this research is obtaining knowledge and insight into the influence of work pressure on the quality of audit. Secondly, this research will focus on the measures that have already been taken to reduce the workload and whether these measures are perceived as effective. Therefore, the following research question has been formulated:
What is the effect of work pressure on the quality of audits within a Big 4 accounting firm? Within this research question, the focus is on young professionals. These young professionals feel work pressure because they are still in training to become a CPA. Also, the shortage in the accounting sector starts with students wanting to become an accountant. Young professionals are the lowest layer of the hierarchical audit forms and therefore it is important to focus on them (De Vries, Blomme, &
6
De Loo, 2018). Scientific literature also shows that young professionals show high turnover rates due to pressures related to the work they perform (Bao, Bao, & Vasarhelyi, 1986; Chong & Monroe, 2015).
1.3 Relevance
A lot of research has been done in recent years on the topic of work pressure and the quality of audit. However, according to De Vries, Blomme & De Loo (2018), there have been various calls in the literature to produce more research on the real perception and experiences of accountants, especially young professionals. First of all, this research contributes to the literature by giving a good insight into the difficult aspects that young professionals experience during their job and training. Secondly, not much research has been done into the relationship between work pressure and audit quality and especially not with the help of the Job Demand-Control-Support model (Yan & Xie, 2016; Del Pozo-Antúnez, Ariza-Montes, Fernández-Navarro, & Molina-Sánchez, 2018). Thirdly, regulation pressure is a subject that has not been widely studied in the literature and therefore this will also be included in the Job Demand-Control-Support model. Especially now, with the lack of consistently high audit quality, this subject is very relevant and timely and therefore it is necessary to study this phenomenon. Besides the scientific relevance, the practical relevance is also important. As already shown in the introduction of the research problem, there have been a lot of published news articles about the high workload among (young) accountants. For accounting firms to be able to recruit and retain
professional and well-trained accountants in the future, it is important to gain a good insight into the perceived workload. Only then accounting firms will be betterable to address the underlying problem of this high workload and ‘low’ quality. Failure to achieve increased quality can eventually lead to a downward spiral for the public accounting sector (Knechel R. , 2016).
1.4 Headlines of the research
This research focuses on accountants who are in training to become a CPA and at the same time work within a Big 4 accounting firm. The reason to focus on young professionals is because they experience a lot of work pressure according to the news and literature. This work pressure is mainly high in this stage of their career because they work, are still in training to become a CPA and try to maintain their social life as good as possible. It is also important to focus on young professionals because they are really important for organizations to recruit and retain.
An internship offers the (unique) opportunity to conduct this research within one of the Big 4
accounting firms. Therefore, it will be possible to conduct a survey among young professionals to find out what effect work pressure and regulation pressure have on audit quality.
7 1.5 Structure of the thesis
The remainder of this thesis is structured as follows. The next chapter provides a literature review on work pressure, regulation pressure and audit quality. Chapter 3 contains a description of the
methodology that is used and the organization that has been studied. Chapter 4 describes the results of the research based on a multiple regression analysis. Finally, chapter 5 will give a conclusion of the research and this chapter dives deeper into limitations and future research.
8
2. Literature review
This chapter gives an overview of audit quality and previous research that has been done on this topic. This chapter describes different definitions and audit quality frameworks. After that, work pressure will be defined and explained by using the Job Demand-Control-Support model. Finally, the control variables that need to be included in the study are explained.
2.1 Young professionals
As already mentioned in the introduction, this research focuses on young professionals. Young professionals are accountants who are still in training to become a CPA. Most of these young professionals choose to follow this trainee-trajectory right after completing their master. Therefore, they are called young professionals. It can be said that this group of employees occupies the lowest layer of the hierarchical audit forms. For this reason, it is important to focus on them, because they are the future of accounting (De Vries, Blomme, & De Loo, 2018). Accounting firms are characterized by an up-or-out culture (De Vries & Herrijgers, 2018). The up-or-out system consist of three or more layers. The lowest level contains the junior employees (young professionals). The middle layer consists of managers and senior managers. The highest layer consists of partners, which have the overall-responsibility for the audit. This hierarchical structure and the up-or-out system can be
compared with a pyramid (see figure 1). This comparison can be made because employees either make promotion or leave the organization in an up-or-out system (Kaiser & Ringlstetter, 2011). The
intention to leave the public accounting business is the highest among the young professionals (De Vries & Herrijgers, 2018). The NBA (2016) already identified that the main reason (40%) to leave the business is the emphasis on rules and regulation. With this high intention to leave, a lot of quality and expertise is lost already in the lowest layers of the pyramid. This may contribute to the lack of audit quality that has been identified.
9 2.2 Audit Quality
To be able to measure the influence of work pressure on audit quality, it is important to have a clear definition of audit quality. Audit quality has been researched a lot by academic researchers. The concept of audit quality is of great importance to standard setters and regulators, especially now that there is a lack of quality within the audits. But even after all these years and all these articles, there is, according to the literature, no universal agreed up-on definition of quality (Bedard, Johnstone, & Smith, 2010; Knechel et al., 2013). Even though there is much literature on audit quality, there is no consensus on how to define, let alone measure it (Financial Reporting Council, 2006; Knechel et al., 2013). In this paragraph, the different definitions and viewpoints are discussed. Also, different frameworks to measure audit quality are discussed and compared.
2.2.1 Definitions of audit quality in scientific research
A definition that is frequently used in quantitative literature is the definition of DeAngelo (1981). DeAngelo defines audit quality as follows:
“The quality of audit services is defined to be the market-assessed joint probability that a given auditor will both (a) discover a breach in the client's accounting system, and (b) report the breach. The probability that a given auditor will discover a breach depends on the auditor's technological capabilities, the audit procedures employed on a given audit, the extent of sampling, etc. The conditional probability of reporting a discovered breach is a measure of an auditor's independence from a given client”. Within this definition, DeAngelo (1981) makes a distinction between two elements. The first element is the expertise of the auditor which will eventually influence the
probability that an auditor will discover a breach in the accounting system. So, this element is related to the knowledge and competence of the auditor (Knechel R. , 2016). The second element is the likelihood that the auditor will actually report the breach that was found. This element relates to the degree of auditor independence (DeAngelo, 1981). Based on this definition, it becomes clear that audit quality depends on competence and auditor independence. These two elements are treated as separate elements which have a positive influence on audit quality. More independence leads to higher audit quality and more expertise leads to higher audit quality as well. But Knechel, R. (2016) questions the separability of these two constructs and he suggests that both constructs have an inverse relationship. Both elements are desirable and therefore there is an implicit tradeoff. However, Knechel (2009) also argues that it is impossible to measure audit quality according to the definition of DeAngelo (1981). This is due to the fact that it is impossible for users to estimate the competence and independence of auditors. Defond & Zhang (2014) also criticize the binary process of whether to detect a violation or not. They believe that the responsibility of an accountant goes beyond this.
10
Zhang and Defond (2014) define audit quality as “high quality auditors do not only consider whether the client’s accounting choices are in technical compliance with GAAP, but also how faithfully the financial statements reflect the firm’s underlying economics.” They also state that the supply of audit quality is determined by auditor independence. Auditor independence is, according to them,
determined by reputation, litigation and regulatory concerns. Francis (2011) argues that audit quality is not dichotomous, as stated by DeAngelo (1981), but that audit quality has different gradations across a continuum from low-quality to high-quality audits. Besides this, audit quality is affected by audit inputs (audit test, engagement team personnel), audit processes (implementation of audit test by engagement team personnel), accounting firms (engagement team work, hiring, training and
compensating auditors, providing audit guidance), audit industry & audit markets, institutions and the economic consequences of audit outcomes. Another frequently used definition is the definition of Palmrose (1998). She sees audit quality as the degree of certainty that the accountant has in the fact that the financial annual report is free from material errors. This definition differs from the definition of DeAngelo (1981) by means of focusing on the quality of the financial report instead of focusing on the accountant itself.
As already mentioned, both elements that are considered to reflect audit quality (competence and auditor independence) are very difficult to measure. For this reason, a lot of studies used proxies to be able to measure (a part of) audit quality. A lot of proxies have already been investigated. For example, audit firm size, partner tenure, provision, experience within industry of client, partner/manager’s attention to the audit, communication, knowledge of the audit partner, senior managers knowledge etc. (Association of Chartered Certified Accountants, 2014). This is quite a large variety of proxies and based on this it can be concluded that there is, indeed, no consensus on how to measure audit quality.
2.2.2 Measuring audit quality
Due to the lack of one unified definition of audit quality, it might be more helpful to look at different frameworks for measuring audit quality. This insight will give a broader understanding of what audit quality actually entails. But also here, there is no universal framework for audit quality (Knechel et al., 2013; Brown, Gissel, & Nelly, 2016). There are a number of academic papers providing theoretical frameworks for measuring audit quality (Bedard, Johnstone, & Smith, 2010; Francis, 2011; Knechel et al., 2013; DeFond and Zhang, 2014). In addition to these scientific papers, several regulatory bodies (such as PCAOB, KPMG and IAASB) proposed audit quality frameworks which are based on these scientific studies (Christensen, Glover, Omer, & Shelley, 2016). All these frameworks define audit quality in terms of a general framework of audit quality, namely audit inputs, audit processes, audit outputs & opinions and audit contexts (see figure 2). Each regulatory body creates their own audit
11
quality framework by adding components to this general framework consistent with scientific research.
Figure 2: general audit quality framework (Christensen et al., 2016)
PCAOB
The Public Company Accounting Oversight Board published an accounting quality framework in 2015. This regulatory body oversees the audit quality of public accounting companies to be able to protect the interest of the investors and shareholders. The PCAOB tries to identify quality indicators which define audit quality and with these indicators they try to come up with a portfolio of
quantitative measures that may provide new insights in the quality of the audit and how this can be improved in the future. Together with qualitative context, these indicators can enhance discussions about the audit process (PCAOB, 2015). The PCAOB defines quality as ‘meeting investors’ needs for independent and reliable audits and robust audit committee communications on financial statements, assurance about internal control and going concern warnings’ (2015). Also here, the quality
framework is based on previous scientific research and existing standards. In 2012, the PCAOB started with developing audit quality measures with a long-term goal of reporting collective measures over time. In 2015 they issued the concept release on Audit Quality Indicators. Without Audit Quality Indicators there would be a lack of transparency in the audit process and therefore it would be very challenging to asses audit quality (Brown et al., 2016). All the 28 quality indicators the PCAOB has developed fall within three categories. An overview of the framework can be seen in appendix 1. The first category is audit inputs (professionals), which includes indicators that focus on availability, competence and focus. The second category is audit process, which includes indicators that focus on leadership, incentives, monitoring, independence, tone at the top etc. The last category is audit results, which includes indicators that focus on the financial statement, going concern and communications between auditors and committees. This framework especially mentions that it is important to maintain talented people and that this indicator is an important aspect of audit quality within the audit inputs (PCAOB, 2013). The framework also states that it is of high importance to monitor these talented people well to maintain them. Besides this, the framework also cites the impact of workload on the quality of the audit, which is an important factor within this research (Persellin, Schmidt, & Wilkins, 2014). For example, the framework includes number of billable hours as an important quality indicator.
IAASB
12
quality standards for auditing. The IAASB sets standards under the supervision of the PCAOB. The IAASB “developed a framework for audit quality that describes input-, process- and output factors that contribute to audit quality engagement, audit firm and national levels, for financial statement audits” (IAASB, 2014). The IAASB sees audit quality as something that includes all key elements which maximize the chance that audit quality is higher on a consistent basis. According to the board, a quality audit is reached when: appropriate values, ethics and attitudes are in place, sufficient
knowledge, experience and regulation are in place, rigorous audit process and quality control procedures are in place that comply with the law and regulation, timely reports are provided and appropriate interaction with stakeholders is in place. The framework distinguishes inputs, process, outputs, key interactions with the Financial Reporting Supply Chain and contextual factors. An overview of the framework can be seen in appendix 2. Inputs include values, ethics, attitudes, culture, knowledge, skills and experience. Within the input factors, the framework makes a distinction between the audit engagement level, the level of an audit firm and the national level. Process factors focus on the rigorous audit process and quality control procedures that comply with laws and regulations. The output factors are the reports and documents that have been formally prepared and presented after the audit. This output is (partly) determined by the context. This context is depending on legislative requirements and regulation, but also on stakeholders. The key interaction with the financial reporting supply chain can have a major influence on the quality of the audit. These formal and informal activities allow a dynamic relationship between the inputs (the auditors) and the end product (output). Based on these interactions, different kind of skills or content can be brought into the process. Finally, the contextual factors are mainly environmental factors which can influence the quality (directly and indirectly). For example, a change in the law or an epidemic can influence the quality (COVID-19 for example) (IAASB, 2014).
NBA
As a result of the published report in 2014, the NBA published key general audit quality factors in 2016. These quality factors are important for audit firms to keep in mind and to pursue, because due to these factors audit firms are able to measure their audit performance. In 2017, the NBA published 10 drivers of audit quality and these drivers arose from root cause analysis of the Big 4 audit firms. After this publication, five other big accounting firms also created a root cause analysis. Therefore, the NBA published 7 themes that play a role within the audit quality process. These themes are thus based on a root cause analysis of 9 big accounting firms. The themes include planning & control, team
composition, process-management, quality of the organization as whole, learning organization, knowledge and competence and professional identity (Nederlandse Beroepsorganisatie van Accountants, 2019).
13
Knechel, Krishnan, Pevzner, Shefchick & Velury (2013) propose a quality framework after extensive analysis of different frameworks of regulatory bodies and other researchers. The framework is organized around a ‘balanced scorecard’ with 4 categories, namely inputs, process, outcomes and context. The inputs of the audit quality process are mainly related to the auditors self and their
individual characteristics. They look at incentives and motivation, professional skepticism, knowledge & expertise and within-firm pressures. The process factors are related to the characteristics of the audit process itself, such as judgement in the audit process, audit production, assessing risk, analytical procedures, obtaining & evaluating audit evidence, negotiations and review & quality control. All these inherent characteristics are different for every client. The outcome factors are dependent on both the input- and process factors. The outcome factors are observable characteristics, such as adverse outcomes, financial reporting quality, audit reports and regulatory reviews of audit firms. Finally, the context factors are related to the context of the audit and are dependent of different environmental institutions and situations. It also includes audit partner compensation, audit fees, auditor tenure and market perceptions of audit quality (Knechel et al., 2013).
The last framework that will be discussed is the framework of Francis (2011). Instead of seeing audit quality as a static object, Francis believes that audit quality is a continuum with different gradations. The framework states that audit quality is affected by each of the following units: audit inputs (audit tests, engagement team personnel), audit processes (implementation of audit tests by engagement team personnel), accounting firms (engagement teams, accounting firms that hire, train and compensate auditors and develop audit guidance, audit reports), audit industry and audit markets (accounting firms constitute an industry, industry structure affects markets and economic behavior), institutions
(institutions which affect auditing and incentives for quality, broader legal system), economic consequences of audit outcomes (audit outcomes that affect clients and users of audited accounting information). All these different levels of analysis show that audit quality is determined and created by every level, that every level can affect the other levels and that eventually every single level can affect and determine the overall audit quality (Francis, 2011; Knechel et al., 2013).
2.3 Work pressure
A possible and partial explanation for the reduced quality of audits in recent years is the increased work pressure that public accountants experience. According to Collins & Killough (1992) stress is unavoidable in the public accounting sector, but at the same time too much stress can produce negative consequences for the organization and for the accountants. One of these negative consequences is the decreased job performance and higher intention to leave the organization when stress is (too) high. With this high intention to leave comes also a loss of knowledge, expertise and skills. Therefore, organizations should pay more attention to the reduction of this stress to an acceptable level (because
14
stress will always be part of the job). This is because accounting firms will then be able to protect and retain their most valuable assets, namely the accountants (Collins & Killough, 1992).
There are a lot of definitions of stress in the literature and after all these years of research, there is still disagreement on the universal definition of stress (Cooper B. , 1987). According to McGrath (1976) stress contains an interaction between a person and the environment. In this interaction, there is a risk of stress when there is a requirement imposed from this environment, which cannot be met. This causes stress “under the conditions where the person expects a substantial differential in the rewards and costs from meeting the demand versus nog meeting it” (McGrath, 1976, p. 1352). But in this research, there is no particular interest in the concept of stress itself, but there is an interest in stress related to the work of the accountant (or the combination with the training to become a CPA). In this case, stress can best be defined as work stress or occupational stress. Cooper & Marshall (1976) define occupational stress as “negative environmental factors or stressors associated with a particular job” (p. 11). Karasek (1979) defines work pressure as the amount of work someone has in a particular period.
There has been extensive research into the concept of occupational stress in different fields, such as behavioral science and sociology. This research led to the creation of great achievements represented by Stimulation Theory, Social Exchange Theory, Response Theory and Interaction Theory (Yan & Xie, 2016). These studies found that stress can lead to affected physical and mental health, but also to a reduced quality of the performance. However, according to Yan & Xie (2016), the effect of work pressure on audit quality is not investigated a lot. Remarkably, in most of these previous studies in the field of accounting one of the most popular and widely used models related to occupational stress has hardly been used (Del Pozo-Antúnez et al., 2018). Therefore, by using the Job Demand-Control-Support model, this research will cover a gap in the accounting literature.
2.3.1 Job Demand-Control-Support model
Karasek developed the Job Demand-Control model (JD-C) in 1979. This model is mainly used to describe the physical health by measuring job demands and job decision latitude (Visser, Schouteten, & Dikkers, 2019). This model contains two constructs, namely job demands and decision latitude (job control). Job demands are the amount of requirements that the job entails. According to Bakker & Demerouti (2007) job demands are “those physical, psychological, social, or organizational aspects of the job that require sustained physical and/or psychological (cognitive and emotional) effort or skills and are therefore associated with certain physiological and/or psychological costs.” Decision latitude involves the ability of the employee to plan his or her own work and take own initiative (Karasek, 1979; Del Pozo-Antúnez et al., 2018). The fundamental assumption of the model is that high job demands cause strain when there is low decision latitude (job control) for the employee. The amount
15
of job demands reflect the occupational stress the employee has. This occupational stress can be reduced when the employee has enough decision latitude, which reflects the amount of control the employee has over his or her own work. Therefore, it can be said that there is occupational stress when someone experiences high job demands and low decision latitude (Karasek, 1979).
The job strain model (shown in figure 3) contains two predictions. The first prediction is that the strain increases when the job demands increase (diagonal A). This is because there also is a reduction in decision latitude and therefore this creates a ‘passive job’. A ‘passive’ job is related to dissatisfaction and therefore often related to lower productivity. The second prediction states that there are high job demands, but also high decision latitude (diagonal B). This will lead to the development of new behavior and therefore this can be called an ‘active’ level. Employees with an ‘active job’ will try to learn and therefore they will be able to live up to the high job demands. A ‘passive’ job, on the other hand, is related to dissatisfaction and therefore often related to lower productivity (Karasek, 1979). Even though this model has been widely used, it has also been criticized. For example, the model would present an incomplete model of the strain process and it would not incorporate equivalent psychosocial job characteristics (Johnson & Hall, 1988). To meet this shortcoming, the dimension ‘social support’ has been included.
In 1990, Karasek and Theorell extended the Job Demand-Control model into the Job Demand-Control-Support (JD-CS) model. They believed that the effect of high strain jobs could not only be lowered by greater decision latitude, but also by social support. As mentioned earlier, the authors state that high job demands and low decision latitude cause high strain jobs. Within the JD-CS model, the authors state that these high strain jobs are exaggerated if social support is also low. Instead of high strain jobs, these circumstances are called iso-strain jobs. Hereby, workers experience much higher strain because of social isolation (Karasek & Theorell, 1990; Fila, 2016). This extended model also takes into consideration social support, which is important when investigating the occupational stress of accountants. Accountants are mainly working in teams and therefore social support can be very important for accountants when carrying out their job (Karasek & Theorell, 1990).
16 Job demands
Prior research found out that work pressure can cause emotional exhaustion and reduced personal accomplishment (Persellin et al., 2014), also known as ‘burnout’ or work stress. Stress or a burnout arise when there are too many stressors and are therefore the consequences of stressors. A stressor is “something in the environment which presents a person with a demand or a constraint, or an
opportunity for behavior” (McGrath, 1976, p. 1352). According to Karasek, job demands are meant to measure the psychological stressors that are related to workload, unexpected tasks and job-related personal conflicts. Naturally, there are more job stressors that could have an influence on job demands, but the most cited sources of job pressure are the job demands that are directly related to
accomplishing the task. Therefore, the JD-CS model focuses on psychological job demands which includes work pacing/time pressure, exacting task requirements and overall workload demands (De Jonge & Dormann, 2006; Fila, 2016). Other studies searched for specific job demands, such as role conflict, role overload and role ambiguity (Del Pozo-Antúnez et al., 2018), but within this research the focus will be on the original JD-CS model which uses general psychological demands. Previous research found that a certain amount of job demands are necessary and beneficial to the psychological well-being, motivation and performance of the employee. However, when these job demands become too high, it will have the opposite effect (Fila, 2006). Previous studies also found that work demands of auditors consist mainly of time pressure. Especially in the busy season there is time limitation pressure because of hard deadlines. Auditors are sometimes caught in a dual pressure of prescribed time limit and cost control and therefore they need enough control to allocate their hours and arrange audit teams. This room for job control automatically needs to be greater when job demands are higher (Yan & Xie, 2016). Yan & Xie (2016) state that the quality of the audit is not affected by high job demands, provided that there is enough job control. It is also proven that excessive workload caused by high job demands and lack of control is positively related to emotional exhaustion (Chong & Monroe, 2015). This in turn can have a negative effect on the quality of the performance. Regulation pressure
Within this research, regulation pressure also plays a role and therefore regulation pressure will also be included as a main category of stressors. To be able to gain a good insight into the effect of regulation pressure on audit quality, it is decided to include this variable separately in the model and not to include it together with general psychological demands under the variable psychological demands. The main reason is that regulation pressure has been included in the JD-CS model in this research, but originally psychological demands consist of general psychological demands. The interest in regulation pressure comes from multiple reasons. The last couple of years there has been a major change in the regulation of the public accounting sector. In the beginning of this decade, accounting was more or less a self-regulated profession. This self-regulation was questioned after some accounting scandals
17
and this has led to a shift from self-regulation towards extensive supervision by government oversight and regulatory bodies (Canning & O’Dwyer, 2016; Knechel, 2016). Regulation pressure does not have an encompassing definition. This can be the case because it consists of two words: regulation and pressure. Regulation pressure encompasses all rules and regulations that apply to accounting firms. According to Knechel, R. (2016), there is proof that audit quality has improved as a result of
regulatory developments. But he also states that it is not self-evident that this regulatory pressure will continue to result in better audit quality. It is namely possible that new rules and regulation do not have the expected impact and instead have unintended consequences for the organization. But regulation pressure can have ‘unintended’ consequences consisting of lower satisfaction and higher workload and this can in turn influence audit quality in a negative way. When regulation pressure increases, this will have a negative (unintended) effect on the performance of the organization according to Vermeeren & Van Geest (2012). They conclude that regulation pressure is related to lower organizational performance and a higher intention to leave the organization. Too much
regulation can lead to work overload and lower satisfaction of employees and this, in turn, influences audit quality. Since regulation pressure is seen as part of job demands, the direct relation between regulation pressure and audit quality will be negative. The higher the regulation pressure is, the lower audit quality will be. Adler & Borys (1996) distinguish two different formalization systems, namely coercive and enabling. They state that both formalization systems can have different effects on stress, job satisfaction and intention to leave the organization. The coercive formalization system is positively related to absence, intention to leave the organization, stress and job satisfaction, since this type of formalization system sees work as disutility. This is contradictory to the enabling formalization system, which shows that employees are able to embrace rules and regulations when they see overlap between their own goals and the goals of the organization. Rules and regulations are, according to role stress theory, able to reduce role conflict and ambiguity and therefore able to increase work
satisfaction (Adler & Borys. 1996).
This research will focus on the buffering hypotheses of the JD-CS model. The strength of the JD-CS model lays in the interaction terms. These interaction terms state that high job demands are only problematic when there is not enough job control or social support. Therefore, job stress results from the interaction of job demands, job control and social support (Sargent & Terry, 2000). There are no consistent results regarding these interaction effects and therefore these buffering hypotheses need to be investigated more (Van der Doef & Maes, 1998; De Jonge, Dollard, Dormann, Le Blanc & Houtman, 2000).
Job control (decision latitude)
According to Karasek (1979), job control can moderate the negative effects of high job demands depending on the amount of job control the employee has. Therefore, this research takes the buffering
18
hypotheses of the model into account instead of the strain/main effect hypotheses. This hypotheses states, as already mentioned above, that high job demands are only problematic when there is no job control. This job control can manifest itself in two ways: skill discretion and decision authority (Karasek & Theorell, 1990). Skill discretion is the amount of space employees get to apply their own skills and knowledge. Decision authority is the amount of authority an employee has over his or her own decisions or decisions that affect the employee (Del Pozo-Antúnez et al., 2018). Previous studies provided evidence that job autonomy in decision-making moderates the amount of stress an employee experiences (Karasek, 1979; Del Pozo-Antúnez et al., 2018). More autonomy is also related to better work performance (Brink, Emerson, & Yang, 2016). In the accounting literature, on the one hand, it is found that more job autonomy can also lead to dysfunctional behavior. Employees are less controlled by other employees and can therefore make more mistakes without noticing this (Brink et al., 2016). This is because other controls within the control environment disappear, which leaves room for action and performance that deviates from the standard. On the other hand, many articles found a positive effect of job autonomy on job performance and job satisfaction. When an employee gets more
autonomy, the employee gets the feeling that the organization trusts him or her. This will lead to more intrinsic motivation to perform the job well and to deliver high quality (Gellatly & Irving, 2001; Langfred & Moye, 2004; Saragih, 2011). When accountants have more autonomy over their own work, they will also be more satisfied with more difficult tasks (more skill discretion) and they will feel more trust from the organization in their skills. Accountants are then able to express their abilities and control their own work, which will lead to less dysfunctional behavior and, at the same time, higher quality (Umar, Sitorus, Surya, Shauki, & Diyanti, 2017). Especially in the early career stages of accountants this higher job autonomy will lead to greater quality, because then decisions are not dominated by professional judgement but decisions are dominated by structure and minor professional judgement. Stressors that come with more difficult tasks will be moderated because of the ability to apply their knowledge and skills. Due to this situation, young professionals will gain experience in applying knowledge and skills and control their own work and this will, in the end, lead to a higher quality in the future (Del Pozo-Antúnez et al., 2018). This leads to the following hypotheses: H1:The negative relation between psychological demands and audit quality will be moderated by decision latitude
H2:The negative relation between regulation pressure and audit quality will be moderated by decision latitude
Social support
Like job control, this study also sees social support (in light of the buffering model hypotheses) as a moderate variable. These buffering hypotheses state that high job demands are only problematic when there is no social support. Social support is of great importance when studying the accounting sector,
19
because there mainly is teamwork. Social support can be defined as tangible and intangible support an employee receives from other people, such as supervisors or team members (Sargent & Terry, 2000). In the model, social support is divided into supervisory support (the frequency with which an
employee can ask the help of supervisors) and co-workers support (the frequency with which an employee can ask the help of colleagues). When this social support is high, the model states there is high work support. When this frequency is low, the model states there is low work support (Johnson & Hall, 1988; Theorell & Karasek, 1990). The support of superiors and team members is very important for employees, especially when tasks get harder. The superior creates the organizational culture and can therefore create team-spirit and dedication, but also serves the personal needs of the team
members and creates positive feedback. This will eventually lead to a better firm performance (Pratt & Jiambalvo, 1981). This positive attitude related to social support can, at the same time, neutralize the negative effects of job autonomy associated with dysfunctional behavior. When superiors are able to reduce stressors by, for example, rejecting stressful practices, the stress level of employees will be lower (Herda & Lavelle, 2016; Del Pozo-Antúnez et al., 2018). Lower stress levels will eventually have a positive effect on the performance of employees. Therefore, it is expected that social control will moderate the negative relationship between psychological demands/regulation pressure and audit quality. This leads to the following hypothesis:
H3:The negative relation between psychological demands and audit quality will be moderated by social support
H4: The negative relation between regulation pressure and audit quality will be moderated by social support
Finally, to find support for the JD-CS model, the three-way interaction between psychological
demands, decision latitude and social support, as well as the three-way interaction between regulation pressure, decision latitude and social support will have to be significant (De Jonge, Janssen, & Van Breukelen, 1996). Therefore, the following hypotheses are found:
H5:The negative relation between psychological demands and audit quality will be moderated by decision latitude and social support
H6: The negative relation between regulation pressure and audit quality will be moderated by decision latitude and social support.
20
Figure 4: The Job Demand-Control-Support Model (Johnson & Hall, 1988; Karasek & Theorell, 1990).
2.4 Control variables
Studies that have been examining the effect of work pressure on firm performance or audit quality using firm specific characteristics and individual characteristics as control variables. These studies state that personality traits are related to the ability of professionals to cope with work pressure (Iswari & Kusuma, 2013; Yan & Xie, 2016; Zadegan & Aqa’i, 2018). Karasek mentioned that it is not the goal of the Job Demand-Control model to ignore individual differences, because there are individual differences than can have an influence (age, gender, income, social class etc.) (1979). But there were no big differences when including these individual characteristics in the model. On the other hand, Karasek states that due to these different individual characteristics, the relationship between job demands and work pressure might not be found. Yin & Xie (2016) state that individual heterogeneities and firm heterogeneities can influence the perception of stress. Brown et al., (2016) suggest that audit quality can vary due to several factors, such as gender, experience level and firm size. Since the survey in this research will be conducted within one of the Big 4 accounting firms, there is no need to control for firm specific characteristics. Individual heterogeneities are for example auditor’s role, industry expertise, working period, educational level, gender and age (Karasek, 1979; Yin & Xie, 2016; Zadegan & Aqa’i, 2018). Educational level is in this case not a useful control variable, because all young professionals in this study are currently in training for their CPA. We can therefore assume that they all followed the same type of education. Therefore, in this study only gender, age, industry expertise, auditor’s role (where in the hierarchy) and working period are included as control variables.
21 2.5 Conceptual models
The overall conceptual model is the following:
22
3. Methodology
For this research a quantitative research method is chosen. To be able to answer the research question, a field survey will be conducted among the young professionals of a Big 4 accounting firm. This section provides an explanation for the chosen method and the operationalization of work pressure, regulation pressure and audit quality. First, the research method and research design will be discussed. After this, the variables will be operationalized and the reliability and validity of the study will be discussed.
3.1 Field surveys
According to O’Dwyer & Bernauer “quantitative research seeks to discover new knowledge by simplifying complexities in settings that tend to be more contrived” (2014). It is a scientific method which makes use of formulating, testing and refining hypothesis. Therefore, it relies on a deductive reasoning. The ontological idea of this deductive approach is that every phenomenon in our world can be measured (O’Dwyer & Bernauer, 2014). Survey research is a quantitative research method which provides an overview of a particular theme by surveying a large number of research units (Korzilius, 2000). Also, surveys are suitable for studying (causal) relationships (Muijs, 2004; Field, 2009). It is a non-experimental design and it is therefore not able to control the independent variables or add treatments. Instead, it measures the relationship between the dependent and independent variables by using statistical methods. This study will use a cross-sectional survey because the independent variables and the dependent variable are measured at the same point in time (Bhattacherjee, 2012). A field survey is suitable for gathering information from a large sample of young professionals (Groves et al., 2011). With this sample, provided that the sample is representative for the whole population, the results of this research can be generalized to the whole population of young professionals within the Big 4 accounting firm. Also, audit quality is a complex phenomenon to measure, because there is no universal, agreed up-on definition or framework. Due to this complexity, there is a lot of literature available on this topic and therefore it is possible to formulate hypotheses based on scientific literature. Most studies within this topic are quantitative and therefore quantitative research seems to be more suitable for measuring audit quality (O’Dwyer & Bernauer, 2014).
3.2 Research design
3.2.1 Population, sample design and data collection
Young professionals from all locations of the Big 4 accounting firm were selected. This was done by means of a single random sample. Due to an internship, there was easy access to all email contacts. For every location, the required email-groups could be selected. These groups were named NL – DL
23
AUDIT Location Trainees, NL – DL AUDIT Location Supervisors or NL – DL AUDIT Location Assistant Managers. The groups were very useful for selecting respondents, as this made it very unlikely that the survey would be completed by employees who are not in training to become a CPA. This was also mentioned in the e-mail in which the anonymous link was sent (see appendix 5). The survey was conducted in English, since there are multiple nationalities working for the Big 4 accounting firm.
The survey was online for almost 2.5 weeks and was sent to almost 1000 employees. Eventually a total response of 128 is collected, which comes down to a response rate of 12.8%. This is a quite low response rate, which may have been caused by the fact that many interns had to conduct surveys for their thesis. At the end, 20 respondents were deleted from the survey, as this group only partially completed the survey. The survey was presented to the employees by using an online setting. Qualtrics has been used to create the survey and the link for this survey was distributed via e-mail. The choice for Qualtrics was based on the fact that it offers the possibility to import the data with SPSS very easily.
3.3 Operationalization of variables
3.3.1 Dependent variable
Audit quality has many definitions in the scientific literature. Frequently used definitions are the definitions by DeAngelo (1981), Zhang and Defond (2014) and Palmrose (1998). Scientific literature has also shown that there is not one universal framework to measure audit quality. In this research, the PCAOB Audit Quality Framework has been chosen. In previous research, Brown et al., (2016) used a survey to measure audit quality. This survey is based on the PCAOB framework and therefore very useful for this study. According to Brown et al., (2016) junior-level auditors, which accountants in training largely are, prepare most of the audit procedures that define audit quality and at the same time these procedures function as the base for the rest of the tasks that determine audit quality. For this reason, the survey will be limited to audit inputs (professionals) and audit processes and will include only the items that are related to junior staff and senior levels (this is where most accountants in training will be in the hierarchy). Audit results and items related to this are more related to accountants higher up in the hierarchy (such as managers or partners). Previous literature, according to Brown et al., (2016), also suggests that upper-level auditor’s and investors’ definition of audit quality only relates to audit input and audit process (Christensen et al., 2016). The authors also state that the framework of the PCAOB does not include measures to find out if and how many times auditors work on multiple engagements at the same time. Scientific literature has proven that this relates to audit quality because it can cause auditors to be confused or forget things (Brown et al., 2016).
24
As already mentioned, the survey of Brown et al., (2016) will be used. The survey that the authors used for their research can be found in appendix 3. Since work pressure and audit quality both need to be measured in the survey for this research, it is important to pay attention to the length of the survey. Therefore, the operationalization of audit quality was examined and only the items that are important for this research (related to audit inputs and audit processes) are included. First of all, all the items are removed from the survey that are related to work pressure. This is due to the fact that work pressure is already measured by the Job Demand-Control-Support model in the survey. These are items such as “I feel overworked” and “I often feel time-pressured”. Due to this overlap with the JD-CS model, auditor mood or affect – audit professional has been removed from the survey. These items are already measured by the JD-CS model.
Some items are merged into one item, because this research is more interested in the overarching concept of multiple very specific questions. The survey asks a lot of questions on the confidence in the ability to apply or understand technical knowledge and procedures. Items that might not be clear for respondents (such as corroborative evidence or professional skepticism) are deleted from the survey. Questions that are about the knowledge or expertise of colleagues are removed, because not every CPA in training is able to answer this question. Also, the JD-CS model already contains questions about the competence of co-workers. There might not be enough insight into the expertise of colleagues in the short period they work for the firm. The questions about team effectiveness under Audit Team Activity – Audit Process are removed from the survey. These questions are added by Brown et al., (2016) based on previous literature and do not directly relate to the PCAOB framework. These items were included so that the difference between the view of staff and seniors on the quality of the team could be seen. Since choices needed to be made to shorten the survey, these items are removed. These items were mainly about the knowledge, expertise and skills of team members, which are not always clearly visible for accountants in training. Also, this overarching concept of the
competence of co-workers is already measured by the JD-CS model.
By shortening the survey where possible, it gives the opportunity to thoroughly measure the effect of work pressure on audit quality (input and process audit quality). Most of the items in the survey are measured using a 5 point Likert scale (totally disagree, disagree, neutral, agree, totally agree). Some questions are measured by a percentage, where a few options are given (0-20%, 20-40%, 40-60%, 60-80%, 80-100%)
25 Figure 6: Operationalization audit quality
3.3.2 Independent variable
There are four independent variables in this study, namely: psychological demands, regulation pressure, job control and social support. All these variables together will measure work pressure. This study makes use of the Job Content Questionnaire (JCQ) that is developed by Karasek (1985). This questionnaire was developed to measure social and psychological job characteristics, based on the JD-CS model. The questionnaire entails multiple subscales and is therefore very useful to measure all the different independent variables. The subscales are: decision latitude, which measures job control (9 items), psychological job demands, which measures job demands (5 items), and social support, which measures the support from co-workers and supervisors (8 items). In this research, one independent variable has been added to the model. This independent variable is regulation pressure. The subscale psychological job demands is divided into two dependent variables, namely
psychological demands and regulation pressure. Therefore, the questionnaire is supplemented with 4 items for regulation pressure. Regulation pressure has not been investigated a lot. Vermeeren & Van Geest (2012) created a theoretical model and questions to quantify regulation pressure. This model and the questions are based on the JD-R model and are therefore very suitable. This research took 4 items out of the 6 items for regulation pressure, which need to be answered on a 5 point Likert scale (totally disagree, disagree, neutral, agree, totally agree).
The variable work pressure (which consist of all the independent variables) consists of 28 items and these items are measured on a 5-point Likert scale (totally disagree, disagree, neutral, agree, totally agree). Research into the model confirms that the questionnaire is valid and reliable and therefore very useful for this research (Karasek, Choi, Östergen, Ferrario & De Smet, 2007; Landsbergis, 1988; Sale & Kerr, 2001).
26
Figure 7: Operationalization work pressure, decision latitude, psychological demands, regulation pressure and social support
Table 1: Measurement items
Constructs Items Adapted from
Dependent variable: Audit Quality
See appendix 4 (Brown et al., 2016) Independent
variable: Work pressure
25 items on a 5 point Likert Scale. Range from ‘totally disagree’ to ‘totally agree’
Decision latitude - Skill discretion
Learn new things (My job requires that I learn new things) Repetitive work (My job involves a lot of repetitive work) Requires creativity (My job requires to be creative) High skill level (My job requires a high level of skills) Variety (I get to do a variety of different things on my job) Develop own abilities (In my job, I can develop own
(special) abilities) - Decision authority
Allows own decision (My job allows to make a lot of
decisions on my own)
Little decision freedom (On my job, I have very little
freedom to decide how I do my work)
A lot to say (I have a lot to say about what happens in my
job)
Psychological demands/workload
- General Psychological demands
Work fast (My job requires working very fast) Work hard (My job requires working very hard) No excessive work (I am not asked to do an excessive
amount of work)
Enough time (I have enough time to get the job done) Conflicting demands (The demands that other people
make of me often conflict with my own demands)
Regulation pressure
It costs me a lot of time to comply with laws and
regulation at the organization I work
The rules and regulations are more important than my
experience or intuition at the organization I work
Some rules and regulation I’m dealing with conflict with
each other
(Karasek, 1985) (Karasek, et al., 1998)
(Vermeeren & Van Geest, 2012)
27
It is hard to do my job right because of the procedures and
regulations at the organization I work.
Social support
- Co-worker support
Co-workers competent (People I work with are competent
in doing their jobs)
Co-workers interested in me (People I work with take a
personal interest in me)
Friendly co-workers (People I work with are friendly) Co-workers helpful (People I work with are helpful in
getting the job done) - Supervisor support
Supervisor concerned (My supervisor is concerned about
the welfare of those under him/her)
Supervisor pays attention (My supervisor pays attention to
what I am saying)
Helpful supervisor (My supervisor is helpful in getting the
job done)
Supervisor good organizer (My supervisor is a good
organizer) Control variables: Gender Age Working period Current position Industry expertise
Dummy variable (0 = female, 1 = male) Dummy variable
< 20 years, 20 – 25 years, 26 – 30 years, 31 – 35 years, 36 – 40 years, > 40 years
Dummy variable
< 1 year, 1 – 2 years, 2 – 3 years, 3 – 4 years, 4 – 5 years, 5 – 6 years, > 6 years
Dummy variable
Trainee, Senior trainee, Supervisor, Assistant manager, Manager Dummy variable
Manufacturing, Finance or insurance, Health care or social assistance, Retail trade, Utilities, Wholesale trade, No specific industry experience
3.4 Method for analysis
This study is a social scientific study. Scientific studies focus on theoretical models and empirical results. The theoretical part is used to develop hypotheses. These hypotheses are then tested by use of empirical methods. The study uses a positivist method, aimed at testing theories and hypotheses. It can be stated that these empirical methods will show how well the theories reflect the observation of reality of the researcher. Social science is the science of individuals and groups of individuals and their behavior. In this case, the behavior of young professionals is studied by use of a field survey and based on this field survey conclusion can be drawn from the theoretical models and hypotheses
(Bhattacherjee, 2012). The data will be analyzed by using SPSS.
Originally, variables measured by a Likert scale are ordinal variables. This is because it is rank-ordered data (so no nominal data), but the difference between the different ranks cannot be assessed
28
(so no interval data). But there is a lot of discussion on whether Likert scale variables can be treated as interval variables. Even though, in the basis Likert scale variables are ordinal variables, they are often used as interval variables. An underlying reason for treating an ordinal variable as an interval variable might be that parametric tests (only able with interval or ratio variables) are more powerful than non-parametric tests. Another reason might be that the results from non-parametric tests are easier to interpret than the results from non-parametric tests (Allen & Seaman, 2007; Bhattacherjee, 2012). For this study, all the items in the survey are measured on a 5 point Likert scale. All the independent variables in the study are interpreted as subscales (because they are part of a larger variable, namely work pressure). These subscales have their own items in the survey. These items will eventually be merged into one complete subscale (for example general psychological demands which has 5 items). For this reason, these subscales can be interpreted as interval variables.
Since the independent variables as well as the dependent variables can be interpreted as interval variables, the next step is to determine the appropriate statistical test. For this research, multiple regression analysis will be used. This is because multiple regression analysis is generally easier to use than other statistical tests with interval variables. It could have been chosen to use Process v3.5 instead of linear regression. When using Process v3.5, variables are automatically centered and interaction terms are automatically created. This is more user-friendly (Field, 2009). But at the same time, Process v3.5 does not allow to use more than one independent variable for which it creates interaction terms. Therefore, in this case, it is better to use SPSS and center the variables manually and create interaction terms manually. This makes it possible to use more than one independent variable at the same time, which is needed for our regression model.
Before this test can be conducted, there are a number of other steps to take first. Firstly, Cronbach’s alpha needs to be conducted. This test is a way to determine if multiple items can form one scale together (so do the items measure the subscale) (Korzilius, 2008). After this, there are some assumptions that need to be tested before it is possible to conduct regression analysis. These assumptions are the measurement level of the variables, the independence of errors, the linearity between dependent and independent variables, homoscedasticity of residuals, multicollinearity, significant outliers or residuals, the normal distribution of residuals and testing of control variables. After all these steps, it is possible to perform multiple regression analysis. The regression equation will be as followed:
Audit quality = α + β₁ * psychological demands + β₂ * regulation pressure + β₃ * Decision latitude + β₄ * Social support + β₅ * psychological demands * decision latitude + β₆ * regulation pressure * decision latitude + β₇ * psychological demands * social support + β₈ * regulation pressure * social
