Trust and professional scepticism : can auditors have both? : what is the relation between trust and professional scepticism?

(1)

Trust and professional scepticism:

Can auditors have both?

What is the relation between trust and professional scepticism?

Anouska Wage 6071643 Date: 18-6-2015 Word count: 15.212 MSc Accountancy and Control, variant Accountancy Faculty of Economics and Business, University of Amsterdam Supervisor: Dhr. A.Th.A. Koet Academic year: 2014-‐2015

(2)

Statement of originality

Statement of Originality

This document is written by student Anouska Wage who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

2

Abstract

The purpose of this study is to examine the influence of an auditor’s trust in general and in his client and professional scepticism. Trust can be defined as the willingness to be vulnerable, having positive expectations regarding the other person’s actions and take the risk that the other party might not meet those expectations. Professional scepticism is defined as the doubt an auditor needs to have about the reliability of the information provided by the client and the requirement to gather sufficient evidence to be able to decide whether or not the received information from the client is free of errors. The relation between trust and professional scepticism is examined with a questionnaire. The questionnaire consists of two parts, part one is about trust and part two is about professional scepticism. The questionnaires are analysed with a multiple linear regression. Trust is the independent variable of the regression and professional scepticism is the dependent variable. The analysis of the results of the questionnaire shows there is no significant relation between trust and professional scepticism. The trust an auditor has in people in general and in the client and his professional

scepticism are separate aspects of the audit. This means an auditor as a person can have high trust in people in general and in his client and have professional scepticism at the same time. The hypotheses that an auditor is a person who does not trust

anybody due to the requirement to be professionally sceptical is not true: the results of this research show that the auditors who were part of the research have a

“professional” distrust. The limitations of this study were a relatively small sample and the questionnaire could have been more reliable or more extensive by covering external factors of trust and scepticism too or by being more client specific. The opportunities for future research are to do the same study with a bigger or different sample and to use a different measure for trust and professional scepticism.

(4)

1. Introduction and motivation

1.1 Background

Every day we interact with other people, we help others, they help us or we just talk with them. An important aspect of those social interactions is trust (Johnson-George and Swap 1982). Trust is needed between parents and their children, friends, shops and their customers but also between the auditor and his client. Trust is needed between the auditor and the client because they interact in a way that the auditor controls the financial statements of the client. The auditor needs to have some trust in the client because the client provides information the auditor uses during the audit of the financial statements and the client has more specific knowledge about his

organisation. If the auditor does not have any trust in the client and the information he gives, he needs to investigate every account and all the given information which makes the audit very time-consuming and inefficient (Rennie et al., 2010).

The need for trust to make the audit more efficient and less costly can be explained with the theory of Covey (2006) which is explained in his book Speed of trust. Covey (2006) describes trust as a function of speed and costs, the more trust one has in the other party, the faster he can make decisions and the less costly those decisions are. This phenomenon also happens in the audit, the more trust the auditor has in his client, the less evidence he mostly collects (Rose 2007). If the auditor decides he needs to collect less evidence before he gives his opinion or judgement, the audit takes less time and is less costly.

Trusting another person is never without risk, there is always a chance the client did not tell the truth and misled the auditor for his own self-interest (Shaub 1996). To reduce the risk of misstatements made by the client, intentional or

unintentional, the auditor needs to stay professional sceptical towards the information provided by the client (Quadackers et al., 2014). The importance of professional scepticism is also emphasized in different auditing standards (e.g. PCAOB 2008), because the lack of professional scepticism has been a major reason that auditors did not detect fraud (e.g. Benston and Hartgraves, 2002). According to SAS No. 1, a professionally sceptical auditor is not satisfied with less than persuasive evidence. This shows that professional scepticism requires the auditor to collect more evidence, which makes the audit more time consuming and expensive.

(6)

Even though trust and scepticism seem to be in conflict with each other, they are both important aspects of the audit (Rennie et al., 2010). The conflict for the auditor is that on one hand the audit needs to become less costly, so the auditor needs to have more trust in the client, but on the other hand the auditor needs to be more sceptical to increase the quality of the audit and reduce the risk of fraud (Ettedge et al., 2014). The relation between trust and professional scepticism has not been studied before, but since the auditor needs to have more trust in the client and also be more sceptical, it is important and interesting to know whether trust and scepticism

complement each other or are two different aspects of the audit. Shaub (1996) argues that trust and scepticism complement each other for instance, but Hurtt (2010) states that trust and scepticism are two different concepts. In case trust and professional scepticism are two different concepts, it can still be possible they influence each other, but no empirical research has been conducted yet to examine the possibility of this relation.

1.2 Development of the research question

It is interesting to research whether trust and scepticism are two separate constructs or compliment each other. If trust and scepticism compliment each other it might be impossible for the auditor to increase trust and scepticism at the same time. It is also possible that trust and scepticism are two separate constructs and this can be

explained as follows. An auditor who has good past experiences in working with a particular client might have a lot of trust in that client (Rennie et al., 2010). If an auditor knows this client has motives, pressureand opportunities to misstate amounts in the annual report and/or commit fraud, it could be possible that the auditor is more sceptical towards the evidence received from the client, but still trust the client. He might still trust the client, because the client never committed fraud or had a

misstatement in his disclosures despite his motives. The following research question is formulated to examine what the relation between trust and professional scepticism is and whether they complement each other or are two separate aspects of the audit:

(7)

6

1.3 Motivation

The aim of the study is to examine what the relation is between the trust an auditor has in general and in his clients and his professional scepticism. The trust an auditor has in people in general and in his clients and his professional scepticism towards the client have both been studied separately, but the relationship between the auditor’s trust in general or in his client and professional scepticism has not been studied before. The contribution to prior research of this study is that it looks at the relation between trust and professional scepticism instead of studying trust and professional scepticism separately. The possible outcome of the study might make it clearer for society how the relation between the auditor and the client works.

1.4 Structure

The remainder of this paper is as follows. In the next chapter prior literature is studied. The literature will not only be economic literature, but since trust and

scepticism are psychology topics, literature with regard to psychology about from this is studied too. The literature is used to develop the hypotheses and to formulate the expected outcome of the study. The third chapter gives a description of the sample and an explanation of the questionnaire used for this study. It also defines the

variables and how the results are going to be analysed. The results are analysed in the fourth chapter. The fifth chapter discusses the results and gives a conclusion to answer the research question of this paper. Finally the sixth chapter gives the limitations of this study and the possibilities for future research.

2. Literature review and hypothesis development

2.1 Literature review

2.1.1 Trust

2.1.1.1 What is trust

There are multiple different definitions of trust in the literature. Rotter (1967) defines trust as the expectancy held by an individual or group that the word, promise, verbal or written statement of another individual or group can be relied upon. According to Rousseau et al. (1998) and Mayer et al. (1995) trust is the acceptance of vulnerability based on positive expectations. Schoorman et al. (2007) define trust as the willingness

(8)

to take risk. Most definitions share the idea that trust is the willingness to be

vulnerable, having positive expectations regarding the other person’s actions and take the risk that the other party might not meet those expectations.

Kee and Knox (1970) argue that the level of trust someone has results from a subjective evaluation of the other person’s trustworthiness. Deutsch (1958: 268) describes a trustworthy person as someone who is aware of being trusted and that he is somehow bound by the trust invested in him. According to Mayer et al. (1995) trustworthiness is related to the characteristics of the person trusted, a so called trustee. The first characteristic they mention is ability. Ability is the knowledge and skills needed to perform a certain action (Gabarro, 1978). The personality of a trustee is split in benevolence and integrity (Mayer et al., 1995). Benevolence relates to the extent to which the trustee is believed to “do good” for the one trusting him, called a truster, besides any personal motives (Mayer et al., 1995: 178). Integrity is defined as the extent to which trustee is believed to follow moral and ethical principles when serving the needs of the truster (Colquitt et al., 2007: 910). Mayer et al. (1995) state that a person is trustworthy when he has the abilities to complete a certain action and is benevolent and integer.

Trustworthiness is related to the context in which a person is trusted. One might believe someone is trustworthy in a certain situation, but not trustworthy at all in another situation (Lewis and Weigert, 1985). This phenomenon is further explained in the next paragraph “trust and distrust”. Trust is also related to trust propensity. Trust propensity is defined as a personal factor or characteristic that affects the likelihood someone trusts the other party, it is the general willingness to trust others (Mayer et al., 1995: 715). It is influenced by someone’s cultural background or prior experiences. (Mayer et al., 1995). Trust propensity is linked to dispositional factors, which are a person’s personality and personal motivations (Colquitt et al., 2007). A person’s personal motivation (e.g. does he have a reason to trust), personality factors (e.g. does he have trust in humanity in general or is he always suspicious about everything) and attitude (e.g. does he has an open mind and attitude towards others or is he very close minded) are dispositional factors that influence trust (Kee and Knox, 1970). Trustworthiness is not only influenced by the circumstances of a particular situation, but based on the above literature it can be said that someone’s personal characteristics, so his general willingness to trust others, also influences the evaluation of the other party’s trustworthiness.

(9)

8

Rousseau et al. (1998) describe different types of trust, either between persons or organisations, in their paper. The first type of trust is deterrence-based trust. Based on different literature they have studied (e.g. Sitkin and Roth, 1993), they argue that deterrence-based trust is not actual trust, because it is based on the belief someone is trustworthy because of his costs of opportunistic behaviour when he abuses this trust. The second type of trust is calculus-based trust. This type of trust is not only based on the costs of opportunistic behaviour, but also on the information one has about the other party regarding his intentions and reputation. Thirdly Rousseau et al. (1998) mention relational trust. Relational trust is the trust people have because of positive previous interactions with the other party. The last type of trust is institution-based trust. This is a combination between calculus-based trust and relational trust. Institution-based trust is based on the information one party has about the other party’s intentions and reputation and his positive past experiences with that party. Based on the literature they have studied, McKnight et al. (1998) divide institution-based trust in situational normality and structural assurance. Situational normality is described as the trust people have because the situation is normal, there is no reason to think something might go wrong (Garfinkel, 1963). Structural assurance is based on trust safeguarded with externalities such as guarantees and regulations (Shapiro, 1987b). This shows institution-based trust is mostly related to past experiences, to decide whether a situation is (ab)normal, and the possibility to safeguard oneself.

It has been argued by Lewis and Weigert (1985) that trust is based on a cognitive process, making a rational choice based on the circumstances of a certain situation, and an affective process, an emotional bonding between two parties. Cognitive trust is trust based on rational reasoning. Rational reasoning is a choice based on the circumstances and evidence of the trustworthiness of that specific person or situation (McAllister, 1995). This choice is influenced by the competences or integrity of the other party for instance. Other considerations for the cognitive process are past interactions and the extent of social similarity between the two parties

(Zucker, 1986). Affective trust, also called emotional trust, is trust based on the emotional bond between the two parties (Lewis and Weigert, 1985). Affective trust can be influenced by the care and concern one has in the welfare of the other party and trying to make sure his welfare does not decrease (McAllister, 1995). Trust based on only cognitive processes, so only based on choices concerning the specific

(10)

on an emotional bond. Since both, rational reasoning and an emotional bond, are important for establishing trust, people base their decisions on the cognitive and affective processes concerning trust (Lewis and Weigert, 1986).

As can be seen from this paragraph and prior literature, trust is a complex phenomenon to understand with different researchers having a different definition and view about it. The literature shows that trust can be based on rational reasoning about the circumstances of the situation, like the evaluation of someone’s abilities, past experiences with the trustee or the trustee’s reputation for example. Besides trust based on situation specific reasoning it can also be based on the emotional bonding between the two parties. The next section further elaborates on the factors that influence the trust between two parties.

2.1.1.2 Factors influencing trust between two parties

Initial trust must be present in order to build a new relationship. Trust cannot be based on experiences or knowledge about the other person or organisation in the beginning of a relationship (McKnight et al., 1998). According to McKnight et al. (1998) a truster’s personality is an important factor in the decision to have initial trust in the other. His personal characteristics influence how much faith he has in humanity in general and also his trusting stance, which is the willingness to depend on another. Besides the influence of his personality, a truster’s initial trust is also influenced by situational normality and structural assurances. This means that the personality of truster does not only influence his choice to have initial trust, but he also considers the fact whether the circumstances are (ab)normal, and whether regulations,

guarantees, etcetera, are present.

Rousseau et al. (1998) state that without the presence of interdependency and risk there is no necessity for trust. Once the decision is made to trust, or form a relationship where trust is necessary, different factors influence the amount of trust between the two parties. Communication and openness is one of the most important aspects for trust (e.g. McKnight and Chevany, 1998; Kee and Knox 1970; Butler, 1991). The competence of the other party is also important for the truster. Trust decreases if the trustee does not have the right competences or expertise. (e.g. Mayer et al., 1995; Butler, 1991). Finally, most literature mentions the integrity of the other party. If the trustee is not integer the truster would have less trust in him (e.g. Mayer et al., 1995; Butler, 1991). The trustee’s personal motivation, characteristics and

(11)

10

attitude are also important factors the truster takes into consideration in the decision to trust (Kee and Knox, 1970). The role all those factors play for auditors in their

decision to trust is further elaborated at the end of this section in paragraph 2.1.1.4 Trust and the auditor.

2.1.1.3 Trust and distrust

According to Lewicki et al. (1998) trust and distrust are two different elements of the relationship between two different parties. Trust and distrust can happen at the same time, because they are two different aspects of a relationship (McKnight and

Chevany, 2001). The context of a situation is an important aspect in order to trust or distrust the other party, one might trust someone to do one job, but distrust him to perform another job. (Schoorman et al., 2007). One might trust a dentist to take care of his teeth for example, but not trust him to fix the roof of his house.

Trust is defined by Lewicki et al. (1998) as the positive expectations someone has about the other party’s actions. Those positive expectations are influenced by hope, faith, confidence, assurance and initiative. Lewicki et al. (1998) argue that having high trust means that one hopes, and is convinced that the other party fulfils the positive expectations, so he has faith and confidence in the other party. High trust can also be based on the assurance and willingness to take initiative, because of past experiences, that the other party will fulfil the expectations. Distrust on the other hand, is defined as the negative expectations about another party’s actions. Lewicki et al. (1998) argue that distrust is driven by fear, scepticism, cynicism, wariness and watchfulness, and alertness. They also state that having high distrust is the fear the other party will fulfil negative expectations. Someone might be sceptical, wary and cynical towards the promises of the other party. They can also have an increased level of watchfulness and be more alert to negative signs in their relationship.

McKnight and Chevany (2001: 43) state that when someone has the intention to distrust he is not willing to depend on the other party. The factors that contribute to the decision to trust, like past experience, quality of communication and one’s

personal motives, also play a role in the decision to distrust (Kee and Knox, 1970). If a party has no trust in humanity in general and has a distrusting stance, generally assuming that people are not reliable or trustworthy, he will be more likely to distrust others (McKnight and Chevany, 2001). They also describe distrust behaviour as the lack of cooperation, increased controls, not granting much autonomy and no business

(12)

transactions. Trust and distrust are both partially based on the same type of factors, but having trust or distrust is also dependent on the situation, which is shown with the example of the dentist. This shows that even though trust and distrust are connected, they are two separate constructs and it is possible they both exist at the same time in the same relation (Lewicki et al., 1998).

2.1.1.4 Trust and the auditor

Mayer et al. (1995: 710) argue that working together involves interdependencies, people have to depend on others to accomplish their goals. This means that from the moment the auditor is hired by the client it is necessary the client and the auditor have initial trust in each other. Without initial trust it is not possible to have a good

relationship. If the auditor does not trust his client from the start of the relationship, he should decline the job, or when the trust is lost during the audit he should disengage from that client (Rennie et al., 2010).

An auditor receives pieces of information from the client in order to perform an audit. The auditor needs to trust the client to the extend that the information he received is reliable. Without having trust in the client the auditor needs to verify every piece of information received by the client and every assumption made by the client, which will make the audit very time consuming and inefficient (Rennie et al., 2010). The auditor must consider carefully which information needs further verification and which information does not need proper evidence. For each client the auditor needs to decide how much audit risk he is willing to take. The lower the audit risk, the more evidence the auditor collects to ensure the financial statements do not contain any material misstatements (Arens et al., 2012: 231).

The level of trust an auditor has in his client is influenced by different factors. Shaub (1996) mentions two factors that play a role in an auditors decision about the level of trust he has about the received information and evidence from the client. First he describes the significance of the information provided for his opinion. When the information is less significant for the overall opinion the auditor will put less effort in verifying the information compared to important and significant information. The other factor is the trustworthiness of the personnel of the company being audited. The more trustworthy the personnel is, the more trustworthy the overall client seems. This means that before the auditor gives his judgement about the trustworthiness of the company as a whole, he will evaluate its personnel. As mentioned before,

(13)

12

trustworthiness is based on the ability, benevolence and integrity of the trustee. This means that the auditor evaluates the abilities of the personnel of the client, whether he believes the personnel will act in a moral and ethical manner (Mayer et al. 1995). This can be evaluated based on the competences, openness and reputation of the personnel.

The level of trust someone has results from a subjective evaluation of that other person’s trustworthiness (Kee and Knox 1970). The subjective evaluation of the client’s trustworthiness is influenced by different factors. In the model of Kee and Knox (1970) trust may vary amongst others due to past experience, the quality of communication, the client’s incentives, his power, his personal characteristics and his attitude. Considering those different factors, auditors will have expectations of the client’s reasoning to tell the truth or making errors and misstatements (Shaub, 1996). Since the level of trust is dependent on past experiences, an auditor will be able to better evaluate the proper level of trust for clients they have worked with in the past compared to new clients. For this reason Rennie et al. (2010) state that auditors will most likely have less trust in new clients and that it takes time to establish a high level of trust between the auditor and the client.

Shaub (1996) and Rennie et al. (2010) both mention the importance of good communication between the client and the auditor. The better the quality of their communication, the higher the level of trust will be. The communication between the auditor and his client needs to be open and honest in order to trust each other. Rennie et al. (2010) also mention the client’s demonstration of concern at the moment the auditor has found an error. If the client does not show any concern once a

misstatement is found the auditor will have less trust in the client and spend more time on the audit to look for other misstatements (Rennie et al., 2010).

Other factors influencing the trust of an auditor in his client, mentioned in the paper of Shaub (1996), are whether or not the client has been involved in unsuccessful business decisions and the importance of the client. A more important client is a client whose payments to the audit company are a relatively higher percentage of the total revenue of the audit company compared to less important clients. The more important the client, the higher the level of trust in that particular client (Shaub, 1996). The auditor also evaluates the external environment of the client. With regard to this, one of the most important factors the auditor considers is the industry in which the client operates. Some industries are seen as more risky than others and Shaub (1996) found

(14)

in his study that once a client operates in a higher risk industry the auditor has less trust in the client.

2.1.2 Professional scepticism

2.1.2.1 Professional scepticism in the auditing standards

According to SAS No. 1 (AICPA, 1997), an auditor needs to stay professionally sceptical during the audit. Professional scepticism in SAS No. 1 is described as the attitude that includes a questioning mind and a critical assessment of audit evidence. The auditor must use his knowledge, skills and ability to perform, in good faith and with integrity, the collection and evaluation of the evidence and the auditor should not be satisfied with less than persuasive evidence (AU 230.07-09). Professional

scepticism is also mentioned in the international standards. ISA 200.16 describes professional scepticism as the critical assessment, with a questioning mind, of the validity of audit evidence obtained. He also needs to question the reliability of documents, responses to inquiries and other information obtained from management.

SAS No. 99 is closely related to SAS No.1. SAS No. 99 provides guidelines for auditors in the risk assessment of the fraud risk of their clients. It states that the auditor should evaluate the evidence with an attitude that a material misstatement due to fraud could be present, regardless of any past experiences and beliefs of the

honesty and integrity of the management of the entity. The auditor also needs to consider the human aspect of an audit. The consideration of the human aspect of the audit means the auditor must be aware of the fact that people can make

(un)intentional errors and can have incentives to commit fraud (Hurtt, 2010). SAS No. 99 also emphasizes the importance of identifying the incentives and objectives of a client to present misleading evidence or commit fraud and to recognize the possibility for untruthful communication (AU 316.87.A2-A3). Professional scepticism requires an on-going questioning when there is a risk of fraud (AU 316.13).

SAS No. 57 includes the importance of professional scepticism when evaluating the subjective and objective inputs into clients’ accounting estimates (AICPA 1998). SAS No. 67 mentions the need for professional scepticism in confirmation procedures, when auditors design, perform and evaluate those procedures (AICPA 1992). SAS No. 82 describes the increased importance of professional scepticism in relation with an increased fraud risk assessment. ISA

(15)

14

204.63 also mentions the importance of increased professional scepticism in engagements with a higher fraud risk. As this paragraph shows, professional

scepticism is emphasized in different accounting standards, but none of the standards give an exact definition of professional scepticism. It is only described with vague terms like “a questioning mind”. The vagueness about the definition of professional scepticism makes it harder for accountants to perform their jobs properly and in accordance with the standards (Hurtt et al. 2003).

2.1.2.2 Professional scepticism in academic literature

Bunge (1991: 131) argues sceptics want to have unusual events checked before they admit those events are real. According to Shaub and Lawrence (1996: 126),

professional scepticism is the choice to prevent or reduce the harmful consequences of another person’s behaviour by the willingness to doubt, question or disagree with the information provided by the client. Kadous (2000: 335) states that professional scepticism refers to the independence from the client and the auditor’s duty to the public is more important than the duty to the client. Nelson (2009: 4) defines

professional scepticism as the judgements of the auditor and his decisions that reflect the risk of an incorrect assertion, depending on the information that is available to the auditor. The overlap between all definitions is that professional scepticism requires the auditor to doubt the reliability of the information and evidence provided by the client and the requirement to gather sufficient evidence to be able to decide whether or not the information is free of errors.

Professional scepticism can be divided into two perspectives, neutrality and presumptive doubt (Quadackers et al., 2014). Neutrality is defined as the fact that auditors should be unbiased before they give their judgement about the information received from the client. According to Nelson (2009), neutrality is the focus of the current accounting standards. Neutrality means the auditor assumes the received information is reliable and based on the truth, but he needs to verify this in order to give his judgement. Presumptive doubt means an auditor assumes some level of dishonesty of the management unless the information and assumptions are verified (Bell et al., 2005). The auditor who has presumptive doubt focuses more on evidence of misstatements and fraud (Quadacker et al., 2014).

An auditor’s scepticism is influenced by different factors. The auditor’s knowledge and his experience have a positive influence on professional scepticism

(16)

(Nelson 2009), but Nelson (2009) and Shaub and Lawrence (1996) also conclude that experienced auditors were more likely to assume a possible explanation for a

misstatement is correct. This means the auditor’s experience can have a positive influence on his professional scepticism, because he has more knowledge about how to conduct an audit and staying sceptical, but his experience with clients might also decrease his scepticism since the experienced auditor tends to believe the client’s explanations easier because of positive past experiences with that client.

The client’s characteristics and actions also influence the auditor’s

professional scepticism. Two important characteristics are the client’s integrity and motives of the client (Nelson, 2009; Shaub and Lawrence 1996). When a client has motives to commit fraud and is not integer an auditor needs to be more sceptical compared to a client with less motives to commit fraud and more integrity. Other factors that influence the amount of professional scepticism are a problem solving ability or self- confidence (Nelson, 2009). According to Shaub and Lawrence (1996) a client’s actions that can influence the auditor’s professional scepticism are related party transactions, transactions between the mother and daughter company for instance, the quality of communication between the auditor and the client, and (in)accurate assertions prior year. The client himself influences the amount of professional scepticism too, if he has no motive to commit fraud and open communication for example, the auditor will be less sceptical.

This paragraph shows that professional scepticism does not have one single definition. Most researchers have different interpretations of professional scepticism, but in general they agree to the fact that an auditor needs to have a questioning mind and question the reliability of the information provided and assumptions made by the client. There are different factors influencing the amount of professional scepticism. Those factors can be based on the auditor’s personal traits, but also on the client’s actions and characteristics. This study only focuses on the characteristics of the auditor and not the characteristics of the client.

2.2 Summary of the theory and hypotheses development

In this section the hypotheses are developed. A short summary of the theory is given to make it easier to understand what the theoretical background is of every

(17)

16

There is no general definition of trust in the literature, but most definitions have in common that trust is having positive expectations about the other party’s actions, taking the risk this party might not meet those expectations and the willingness to be vulnerable due to this risk. Professional scepticism does not have a general definition, but most literature argued that an auditor needs to have doubt about the reliability of the information provided and assumptions made by the client. Trust is the willingness to be vulnerable for the risk that the client might not meet the expectations of the auditor, which can mean that he commits fraud for example, and by staying

professional sceptical the auditor needs to decrease this risk as much as possible. This leads to the following hypothesis:

H1: If the general trust of an auditor increases, his professional scepticism

decreases.

Trust is measured by combining the trustworthiness scale and the general client trust scale. The trustworthiness scale is more focused on the belief the

respondent has that people in general are trustworthy and the general client trust scale focuses on the trust auditors have in their clients. The following two hypotheses can be formulated with this information:

H2a: The auditor’s belief a person is rustworthy does not have any significant

effect on professional scepticism.

H2b: Professional scepticism is negatively influenced by the trust an auditor

has in clients in general.

Professional scepticism consists of six sub-elements. Those elements are a questioning mind, suspension of judgement, search for knowledge, interpersonal understanding, autonomy and self-esteem. The six sub-elements are defined in the paper of Hurtt (2010) and a short definition of each of the sub-elements is given before the hypotheses are developed. This study also tests what effect the auditor’s trust in people in general, his belief people are trustworthy and the trust an auditor has in clients, have on each separate element of professional scepticism.

A questioning mind is defined as the suspicion of the auditor towards the client even though he might believe the client is honest and integer. The following hypothesis is formulated regarding a questioning mind:

H3: Trust in people in general, the belief people are trustworthy and trust in

(18)

Suspension of judgement means that the auditor needs to collect a particular amount of information and evidence and objectively evaluate this before he gives his opinion. Even though the evaluation of the information needs to be objective, the auditor decides for himself when he believes he has enough evidence to give his opinion. This leads to the following hypothesis:

H4: Suspension of judgement is negatively influenced by the trust the auditor

has in his clients, but is not influenced by the trust the auditor has in people in general and his belief people are trustworthy.

Search for knowledge is defined as the personal curiosity of the auditor to find more information and evidence about the client or a particular account. Since search for knowledge is a person’s own curiosity, it is more a personal trait. Therefore the following hypothesis is formulated:

H5: Search for knowledge, or personal curiosity, is not significantly

influenced by the trust an auditor has in people in general, his belief people are trustworthy and trust in the client.

Interpersonal understanding is understanding the motives and integrity of the person providing the information and evidence to the auditor. Trust is also based on the understanding of the motives and integrity of the other party. A better

understanding of the other party’s motives and integrity leads to more trust. This leads to the following hypothesis:

H6: Interpersonal understanding is positively influenced by the auditor’s trust

in people in general, his belief people are trustworthy and his trust in the client.

Autonomy is defined as the personal evaluation of the auditor about the amount of information and evidence he needs to collect before he is convinced that the client is speaking the truth. Since it is based on the auditor’s personal evaluation, trust might play a role in his judgement of the information. The more trust an auditor has in the client, or in people in general, the less evidence he will need in order to give his judgement about the reliability of the information provided by the client. The following hypothesis is formulated about the relation between autonomy and trust:

H7: The auditor’s trust in people in general, his belief people are trustworthy

and his trust in the client have a negative influence on the autonomy of the auditor.

(19)

18

The last sub-element of professional scepticism is self-esteem. Self-esteem is a personal characteristic of the auditor. An auditor with more self-esteem is harder to persuade and dares to rely on his own judgements. Self-esteem is not influenced by the amount of trust an auditor has, because it is a personal trait. This leads to the last hypothesis:

H8: The amount of self-esteem is not significantly influenced by the trust the

auditor has in people in general, his belief people are trustworthy or an auditors trust in his client.

3. Research method

3.1 Sample

The participants for this study are 57 auditors from one of the Big 4 companies. 33 respondents were male and 24 were female. The average age, in years, of the respondents is 26.25. The function of the respondents varied from intern to partner and on average they had 3.47 years of experience. The risk of having a bias towards auditors with a lot of experience, or with no experience at all, is decreased by having a variety of function and years of experience. All demographic information about the respondents is shown in table 1.

TABLE 1

Demographics on respondents

Male Female Total

n = 33 n = 24 n = 57 Age (years) Mean 26.39 26.04 26.25 Standard deviation 6.64 5.38 6.09

Years of experience (years)

Mean 3.36 3.60 3.47 Standard deviation 6.67 5.25 6.06 Function Intern 14 11 25 Trainee 9 5 14 Senior Trainee 1 1 2 Supervisor 4 1 5 Assistant Manager 2 4 6 Manager 1 - 1 Senior Manager 1 - 1 Partner 1 2 3

(20)

3.2 Instrument

All participants were asked to answer a 4-page questionnaire. The complete

questionnaire can be found in appendix A, but in the next two paragraphs are the trust part and the professional scepticism part of the questionnaire shown separately. The first page of the questionnaire starts with a few demographic questions. Those

questions give information about the gender, age, years of experience and function of the participant. The questionnaire itself consists of two parts, the first part is about trust in people in general but also trust in clients, and the second part is about the auditor’s professional scepticism. The scales to measure trust and professional scepticism are developed by other researchers (Wrightsman 1974; Shaub 1996 and Hurtt 2010) and used for this study. The next two paragraphs discuss the theory and development of the scales. It also defines which measurements are used for each scale and how to read and analyse the answers given by the respondents. To test the

reliability of the scales the Chronbach’s alpha is calculated. Due to the Chronbach’s alpha some items had to be excluded from the statistical analysis, but a reasonable explanation is given for all excluded items. The exclusion of some items from the statistical analysis does not mean a different scale is developed, but that some items were not usable for this study due to the setting or sample of this study. Besides the exclusion of some items, the scales are unchanged.

3.2.1 Trust

The first part of the questionnaire is about trust. This part consists of three scales, the trustworthiness scale, the independence scale and the general client trust scale. The same scales are used in the research of Shaub (1996) about the effects of situational and dispositional factors on the trust an auditor has in his clients. The three scales of the questionnaire can be found below.

Trustworthiness scale Statement Strongly disagree Neutra l Strongl y agree

1 Most students will tell the instructor when he has made a mistake in adding up their score, even if he has given them more points than they deserved

(21)

20

2 If you give the average person a job to do and leave him to do it, he will finish it successfully

-3 -2 -1 0 1 2 3

3 People usually tell the truth, even when they know they would be better off by lying

-3 -2 -1 0 1 2 3

4 Most students do not cheat when taking an exam

-3 -2 -1 0 1 2 3

5 Most people are basically honest -3 -2 -1 0 1 2 3

6 If you act in good faith with people, almost all of them will reciprocate with fairness toward you

-3 -2 -1 0 1 2 3

7 Most people lead clean, decent lives -3 -2 -1 0 1 2 3

8 People claim they have ethical standards, but few people stick with them when chips are down

-3 -2 -1 0 1 2 3

9 If you want people to do a job right, you should explain things to them in great detail and supervise them closely

-3 -2 -1 0 1 2 3

10 If most people could get into a movie

without paying and be sure they were not seen, they would do it

-3 -2 -1 0 1 2 3

11 Most people are not really honest for a

desirable reason; they are afraid of getting caught

-3 -2 -1 0 1 2 3

12 Most people would tell a lie if they could

gain by it

-3 -2 -1 0 1 2 3

13 Most people would cheat on their income

tax, if they had a chance

-3 -2 -1 0 1 2 3

14 Nowadays people commit a lot of crimes

and sins that no one else ever hears about

-3 -2 -1 0 1 2 3 Source: Wrightsman (1974: 227) Independence scale Statement Strongly disagree Neutral Strongly agree

1 Most people will change the opinion they express as a result of an onslaught of criticism, even though they really do not change the way they feel

-3 -2 -1 0 1 2 3

2 Nowadays many people will not make a move until they find out what other people think

(22)

3 The important thing in being successful nowadays is not how hard you work but how well you fit in with the crowd

-3 -2 -1 0 1 2 3

4 The typical student will cheat on a test when everybody else does, even though he has a set of ethical standards

-3 -2 -1 0 1 2 3

5 It is a rare person who will go against the crowd

-3 -2 -1 0 1 2 3

6 Most people have to rely upon some else to make their important decisions for them

-3 -2 -1 0 1 2 3

7 The average person will rarely express his opinion in a group when he sees that the others disagree with him

-3 -2 -1 0 1 2 3

8 Most people have the courage of their convictions

-3 -2 -1 0 1 2 3

9 Most people can make their own

decisions, uninfluenced by public opinion

-3 -2 -1 0 1 2 3

10 It is achievement, rather than popularity with others, that gets you ahead

nowadays

-3 -2 -1 0 1 2 3

11 The average person will stick to his opinion if he thinks he is right, even if others disagree

-3 -2 -1 0 1 2 3

12 If a student does not believe in cheating, he will avoid it even if he sees many others doing it

-3 -2 -1 0 1 2 3

13 The person with novel ideas is respected in our society

-3 -2 -1 0 1 2 3

14 Most people will speak out for what they believe in

-3 -2 -1 0 1 2 3

Source: Wrightsman (1974: 229-230)

General Client Trust scale

Statement Strongly

disagree

Neutral Strongly agree

1 I believe it is highly unlikely that any of my clients would attempt to deceive me

-3 -2 -1 0 1 2 3

2 It is in the client’s best interest to always tell the auditor the truth

-3 -2 -1 0 1 2 3

3 I have learned that clients are willing to be vague about the facts on issues that they

(23)

22

think might influence the audit opinion

4 Trustworthy clients are the norm -3 -2 -1 0 1 2 3

Source: Shaub (1996: 172)

The trustworthiness scale is developed by Wrightsman (1974) and reflects the belief people are trustworthy, moral and responsible (Wrightsman, 1974: 53). The independence scale is also developed by Wrightsman (1974) and indicates whether people are able to stand up for their own beliefs despite group pressure to do otherwise (Shaub, 1996: 162). The last scale is the general client trust scale. This scale is especially developed for the research of Shaub (1996) and measures, like the title already reveals, the general trust auditors have in clients.

All three scales are measured with a Likert-scale and have a range from “-3”, strongly disagree, till “3”, strongly agree. The respondent can choose “0” when he is neutral about the statement. Some of the items are negatively worded and need to be reversed for the statistical analysis. For the trustworthiness scale statement 8 till 14 needs to be reversed, for the independence scale statement 1 till 7 and for the general client trust scale statement 3.

The Chronbach’s alpha is calculated to test the reliability of each scale. After calculating the Chronbach’s alpha, some statements needed to be excluded from the statistical analysis to make the scale more reliable. Table 2 shows the Chronbach’s alpha of each scale and which statements are excluded from the statistical analysis. The variable “trust” is constructed from the trustworthiness scale and the general client trust scale. The answers of the trustworthiness scale and the general client trust scale are added up and divided by eighteen to get the average value of trust for each respondent.

A Chronback’s alpha of 0.7 is applied in most literature to indicate whether the scale is reliable. The alpha of trust and the trustworthiness scale are above 0.7, which indicates the scales are reliable measures. The independence scale has a Chronbach’s alpha of 0.691 which means the scale is less reliable, but since this is still very close to 0.7 this might not be a problem for the statistical analysis and results. To get a Chronbachs’s alpha of 0.691, statement 4, “the typical student will cheat on a test when everybody else does, even though he has a set of ethical

(24)

it even if he sees many others doing it”, of the independence scale need to be excluded from the statistical analysis. Statement 4 and 12 were the only two statements of the independence scale related to situations of students. Almost half of the respondents were interns which means they are still students. Their answer might have been different compared to the answers of the respondents that work fulltime. The situations, described in the excluded statements, are situations students might have experienced recently, but for the other respondents it might have been a year or maybe years back. It is more difficult for non-students to envision the situations or times might have changed and people respond different in the described situations nowadays. For this reason the independence scale is more reliable when those items are excluded.

The Chronbach’s alpha of the general client trust scale is 0.695 which is a little bit below 0.7. The general client trust scale is also a less reliable, but since 0.695 is close to 0.7 it is probably not a problem too. Statement 3, “I have learned that clients are willing to be vague about the facts on issues that they think might influence the audit opinion”, and 4, “trustworthy clients are the norm”, needs to be excluded from the statistical analysis to get a Chronbach’s alpha of 0.695. The mean of statement 3 is -0.88 and the mean of statement 4 is 0.98. For both statements the respondent has to give his opinion about the honesty of the client, but statement 3 is in negative wording. Statement 3 and 4 do not need to be included in the statistical analysis since they balance each other’s effect. For this reason only statement 1 and 2 are used for the statistical analysis of the general client trust scale.

3.2.2 Professional scepticism

Hurtt (2010), an associate professor from the Baylor University, Wisconsin,

developed a measurement scale for professional scepticism. The scale is divided in six

TABLE 2

Chronbach's alpha questionnaire part I: Trust

Alpha

Excluded from the statistical analysis

Trust 0.704 -

Trustworthiness scale 0.714 -

Independence scale 0.691 4, 12

(25)

24

different aspects of professional scepticism. The six aspects, or sub-elements, in Hurtt’s model are a questioning mind, suspension of judgement, search for

knowledge, interpersonal understanding, autonomy and self-esteem. These aspects explain the characteristics a sceptical auditor needs to have and how he collects and evaluates the evidence. Hurtt studied literature of different researchers, for example Nelson (2009) and Rose (2007), about professional scepticism, and she studied the accounting standards, like SAS and ISA. She used this literature study to define the sub-elements of professional scepticism.

Hurtt (2010) defines a questioning mind as the doubt an auditor needs to have about the information/assumptions he received from the client. An auditor cannot assume all information is correct without proper evidence. Nelson (2009) argues that a questioning mind means the auditor needs to have some suspicion and disbelieve towards the client. Arends et al. (2012: 361) define a questioning mind as the

consideration of the auditor how likely the client might commit fraud, even though he believes the client is honest and integer.

Suspension of judgement is the second sub-element of professional judgement. Suspension of judgement is the requirement that the auditor must collect and evaluate enough evidence about all the elements of the financial statements of the client before he gives his opinion or judgement.

Search for knowledge is more related to personal curiosity or interest to find more r evidence (Hurtt, 2010). This aspect is a personal characteristic of the auditor. It is in someone’s character to have personal interest in gathering more knowledge and staying curious. A questioning mind, suspension of judgement and the search for knowledge are all related to the way an auditor collects and evaluates evidence. The first three sub-elements of professional scepticism indicate that a sceptical auditor has some disbelieve in the information received from the client and will gather more knowledge and evidence until he is convinced enough to give his judgement.

The next sub-element of professional scepticism in the model of Hurtt (2010) is interpersonal understanding. She defines interpersonal understanding as the

auditor’s understanding of the motives of a person to give misleading information and the understanding of the integrity of that person. An auditor gathers some knowledge about the person who provides the information in order to understand his motives and judge his integrity.

(26)

The last two sub-elements of professional scepticism are autonomy and self-esteem. Autonomy and self-esteem are more related to the personal characteristics of the auditor. Autonomy is defined as the ability of an auditor to rely on his own judgement about the amount and quality of evidence he needs before he believes the information/the assumptions received is based on the truth (Hurtt, 2010: 154). It is the auditor’s own evaluation and personal judgement on the amount of information he needs for a particular part of the audit before he gives an audit opinion about the client. Hurtt (2010: 155) defines self-esteem as the ability to resist persuasion and challenging assumptions and conclusions made by someone else. Boush et al. (1994) argue that a person without self-esteem does not dare to rely on his own judgements, faster relies on the facts given to him and is easy to persuade.

Hurtt (2010) formulated 30 different statements, each of them part of one of the six subjects, or sub-elements, of professional scepticism. Appendix A shows the questionnaire developed by Hurtt (2010) used for this research and the categorisation of all statements to one of the sub-elements is shown below.

Professional scepticism sub-elements Questioning mind

1 My friends tell me that I often question things that I see or hear 2 I frequently question things that I see or hear

3 I often reject statements unless I have proof that they are true 4 I usually notice inconsistencies in explanations

5 I enjoy trying to determine if what I read or hear is true Suspension of judgement

1 I take my time when making decisions

2 I do not like to decide until I have looked at all of the readily available information 3 I dislike having to make decisions quickly

4 I like to ensure that I have considered most available information before making a

decision

5 I wait to decide on issues until I can get more information Search for knowledge

1 I think that learning is exciting 2 I relish learning

(27)

26

The professional scepticism scale uses a six-point Likert-scale which starts at “1”, strongly disagree, till “6”, strongly agree. Statement 1, 10, 16, 17, 19, 25 and 26 from the professional scepticism questionnaire in appendix A are written in negative wording and need to be revised for the statistical analysis.

Table 3 shows the Chronbach’s alpha for the professional scepticism scale and for each of the sub-elements of professional scepticism. The Chronbach’s alpha is calculated for this scale to check the reliability of the questionnaire and the different sub-elements. The table also shows which statements are excluded from the statistical analysis to make the results more reliable.

4 I like searching for knowledge 5 The prospect of learning excites me

6 I enjoy trying to determine if what I read or hear is true Interpersonal understanding

1 I like to understand the reason for other people's behaviour

2 I am interested in what causes people to behave the way that they do 3 The actions people take and the reasons for those actions are fascinating 4 I seldom consider why people behave in a certain way

5 Other people's behaviour does not interest me Autonomy

1 I tend to immediately accept what other people tell me 2 I usually accept thing I see, read, or hear at face value

3 I often accept other people's explanations without further thought 4 It is easy to convince me

5 Most often I agree with what others in my group think 6 I usually notice inconsistencies in explanations

Self-esteem

1 I have confidence in myself 2 I do not feel sure of myself 3 I am self-assured

4 I am confident of my abilities 5 I feel good about myself

(28)

The professional scepticism scale has an alpha of 0.772 which indicates it is a reliable measure. Two of the sub-elements of professional scepticism, a questioning mind and search for knowledge, have a Chronbach’s alpha below 0.7. The

Chronbach’s alpha of a questioning mind is 0.674, which is slightly lower than 0.7.

The scale to measure a questioning mind is not very reliable, but it is close to 0.7. An explanation could be that a questioning mind is hard to measure. It is a personal trait to have a questioning mind which is affected by many other external factors. The Chronbach’s alpha of search for knowledge, 0.696 is also below 0.7, but this is only a small difference. The scale might be a little less reliable, but this will probably not lead to problems in the results.

To make the scale for a questioning mind more reliable statement 3, “I often reject statements unless I have proof that they are true”, needs to be excluded from the statistical analysis. Statement 3 can be excluded because all items of a questioning mind are about having doubt and questioning the information and assumptions the respondent receives, but not about immediately rejecting something. Immediately rejecting something without having enough evidence is not having a questioning mind, but more immediate disbelieve. This means someone already assumes something is not true before he has this confirmed by gathering extra evidence. For this reason the statement can be excluded from the statistical analysis because it is not a proper measure of a questioning mind.

Statement 6 of the sub-element autonomy, “I usually notice inconsistencies in explanations”, needs to be excluded from the statistical analysis to get a higher Chronbach’s alpha and a more reliable result. As can be seen in the theory described earlier in this paragraph, autonomy is a personal trait of the auditor. Being able to find inconsistencies is not a personal trait, but a skill. Autonomy is about what the

TABLE 3

Chronbach's alpha questionnaire part II: Professional scepticism

Alpha Deleted item from the scale

Professional scepticism 0.772 -

Questioning mind 0.674 7

Suspension of judgement 0.755 -

Search for knowledge 0.696 -

Interpersonal understanding 0.776 -

Autonomy 0.706 18

(29)

28

respondent would do after he has find the inconsistencies and how much extra

information he wants to gather once he has found the inconsistencies and autonomy is not just being able to find the inconsistencies. Statement 6 is therefore not a good measure for autonomy and can be excluded from the statistical analysis.

3.3 Data analysis

The collected data is analysed with a multiple linear regression. The data, collected with the questionnaire, is divided in dependent variables, independent variables and control variables. The influence of trust on the auditor’s professional scepticism is studied, therefore professional scepticism is the dependent variable and trust the independent variable. The statements of the professional scepticism scale are

categorised among the six elements of professional scepticism to get a more detailed result. This division can be seen in Appendix B or the previous section instrument, professional scepticism. Trust is constructed of the trustworthiness scale and the general client trust scale. The sub-elements of professional scepticism are also put in separate regressions as dependent variables and the two scales of trust as separate independent variables to get a more detailed result. The control variables in this study are the independence scale and experience. The independence scale measures the extent to which someone stays with his own believes despite what others say. The independence scale is used as one of the control variables, an increase in professional scepticism might be caused by an increase the independence of the auditor. As described in the theory of chapter 2, a sceptical auditor dares to stand up for his own beliefs and does not form an opinion before having the right amount of evidence. Professional scepticism is also referred to in the theory chapter as the independence of the auditor: the public’s interest is more important than the client’s interest.

According to Nelson (2009) and Shaub and Lawrence (1996), experience can have a negative influence on professional scepticism. They state that even though the auditor is more experienced and has more knowledge about being sceptical, the chance that he believes a possible explanation from the client about a misstatement has also increased.

Fourteen regressions are used for this research. A 90%-confidence interval is used for all regressions. The first regression is to examine the relation between trust and professional scepticism and to test hypothesis 1. The next regression tests

(30)

hypotheses 2a and 2b, and shows whether the auditor’s belief people are trustworthy or his trust in the client influences professional scepticism more when taken

separately. The following twelve regressions are between trust and the six sub-elements of professional scepticism and between the trustworthiness scale and the general client trust scale as separate independent variables and the six sub-elements of professional scepticism. Those twelve regressions test hypothesis 3 till 8. The results of the regressions show whether trust has a positive or negative influence on one of the sub-elements of professional scepticism and it also shows if the trustworthiness scale or the general client trust scale has a bigger influence on one of the sub-elements of professional scepticism. The control variables of all fourteen regressions are the independence scale and experience. The fourteen regressions together are used to describe the relation between trust and professional scepticism, to examine whether the trustworthiness scale or the general client trust scale has the bigger influence and which of the sub-elements of professional scepticism is most influenced by the auditor’s trust in people in general, his belief people are trustworthy and his trust in the client.

4. Results

The results of the tests are analysed in this chapter and the analysis is used to answer the research question. The first paragraph gives the descriptive statistics of all variables used in this study. The next paragraph discusses the correlation matrix. At last the results from the fourteen regressions are analysed.

4.1 Descriptive statistics

Table 4 gives a summary of the descriptive statistics of all variables used for this study. The mean and standard deviation are calculated for each variable, as well as the range of answers given by the respondents.

The first section of table 4 are the descriptive variables of the dependent variables. The mean of professional scepticism is 4.30, with a standard deviation of 0.50, which indicates that overall the respondents were sceptical. Professional scepticism is calculated by taking the sum of the answers of all statements of the professional scepticism scale of each respondent and divide this answer by 30. The

(31)

30

mean of all sub-elements also scored in the upper part of the scale, above 3.50, which implies the

respondents overall possessed the characteristics of a sceptical person. None of the respondents answered a statement of the professional scepticism scale with “1” and only the questions with regard to a questioning mind and suspension of judgement have a few respondents who answered the statements with “2”. The range of the answers, mostly between “3” and “6”, also indicates the respondents are

professionally sceptical, because the higher the score on the scale, the more sceptical the respondent.

The next section of table 4 gives the descriptive statistics of the independent variables, trust, the trustworthiness scale and the general client trust scale. Trust is calculated by taking the sum of the answers of the trustworthiness scale and the general client trust scale and divide the outcome by eighteen. The mean of trust, the trustworthiness scale and the general client trust scale are very close to 0.

TABLE 4 Descriptive statistics

Dependent, independent and control variables (n = 57)

Mean

Standard

deviation Min Max

Dependent variable

Professional scepticism (anchors: 1 = strongly disagree; 6 = strongly

agree) 4.30 0.50 4 6 Theoretical sub-elements: A questioning mind 4.32 0.76 2 6 Suspension of judgement 3.88 0.73 2 6

Search for knowledge 4.42 0.65 3 6

Interpersonal understanding 4.58 0.76 3 6 Autonomy 4.23 0.60 3 6 Self-esteem 4.60 0.59 3 6 Independent variables

Trust (anchors: -3 = strongly

disagree; 3 = strongly agree) 0.21 0.62 -1 2

Subscales:

Trustworthiness scale 0.14 0.69 -1 2

General client trust scale 0.23 1.64 -3 3

Control variables

Independence scale -0.02 0.719 -1 2

(32)

This indicates the respondents overall were quite neutral about most statements. The respondents do not have high trust or distrust in people in general and the client.

The last section of table 4 covers the descriptive statistics of the control variables. The independence scale had a mean of -0.02 and a standard deviation of 0.719. The overall answers of the respondents on the independence scale were

therefore also neutral and the answers of the respondents revealed they were not really dependent nor independent. The mean of experience was 3.47 years with a standard deviation of 6.06. As can be seen in table 1 most respondents are interns and trainees and for this reason the mean of experience is not very high.

4.2 Correlations

Table 5 presents the correlation matrix of the dependent, independent and control variables.

First of all the correlations between trust and the trustworthiness scale and between trust and the general client trust scale are positive and significant. A positive correlation of trust with the trustworthiness scale and the general client trust scale is expected since trust is composed of the trustworthiness scale and the general client trust scale. A positive correlation means that if the score of trustworthiness and/or general client trust increases, the overall trust increases too. The correlation matrix also shows a positive and significant correlation between professional scepticism and the six sub-elements of professional scepticism. Since professional scepticism is constructed of the six sub-elements, it was expected they would have a positive and significant correlation. The control variables, the independence scale and experience, have a positive significant correlation with trust and trustworthiness. The correlations between trust and the control variables and between trustworthiness or general client trust and the control variables do not exceed 0.8. Therefore the results do not have a problem with the multicollinearity of the regressions (Gujarati, 2003).

Trust does not have a significant correlation with any of the dependent

variables. There is no significant relation between trust and professional scepticism or with the sub-elements of professional scepticism that can be seen from the correlation

(33)

32 TABLE 5 Correlation matrix 1 2 3 4 5 6 7 8 9 10 11 12 1 Professional scepticism - 2 Questioning mind 0.454*** - 0.000 3 Suspension of judgement 0.346*** _0.008 0.391*** _0.003 - 4 Search for knowledge 0.539*** 0.000 0.483*** 0.000 0.147 0.275 - 5 Interpersonal understanding 0.387*** _0.003 0.672*** _0.000 0.131 _0.332 0.366*** _0.005 - 6 Autonomy 0.546*** 0.153 0.431*** 0.070 0.177 - 0.000 0.256 0.001 0.606 0.188 7 Self-esteem 0.414*** 0.090 0.007 0.078 0.132 0.515*** - 0.001 0.507 0.958 0.566 0.327 0.000 8 Trust 0.082 0.198 0.215 0.174 0.155 0.013 0.090 - 0.543 0.140 0.108 0.195 0.250 0.925 0.508 9 Trustworthiness scale 0.083 0.537 0.186 0.167 0.210 0.117 0.183 0.174 0.218 0.104 0.008 0.955 -0.034 0.804 0.762*** 0.000 - 10 General Client trust scale -0.238* 0.075 0.257** 0.054 -0.051 0.708 0.009 0.948 0.094 0.489 -0.309** 0.019 -0.051 0.708 0.322** 0.015 0.472 0.097 - Control variables 11 Independence scale 0.015 0.913 0.295 0.141 0.199 0.138 0.130 0.335 -0.014 0.919 -0.032 0.813 0.264 0.150 0.449*** 0.000 0.327** 0.013 0.205 0.170 - 12 Experience -0.126 -0.164 0.069 -0.003 -0.035 -0.013 -0.081 0.278** 0.356*** -0.103 -0.023 - 0.349 0.222 0.609 0.983 0.799 0.926 0.549 0.036 0.007 0.447 0.867

***: Significant at the 0.01 level (2-tailed) **: Significant at the 0.05 level (2 tailed) *: Significant at the 0.1 level (2-tailed)

Trust and professional scepticism : can auditors have both? : what is the relation between trust and professional scepticism?