ETHICAL CONTROL ON REPAIRING A USER’S TRUST BASED ACCOUNTING
BELIEFS
By
Michelle C. Lau

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Business Administration – Doctor of Philosophy
2014

ABSTRACT
ETHICAL CONTROL ON REPAIRING A USER’S TRUST BASED ACCOUNTING
BELIEFS
By
Michelle C. Lau
Organizations rely on controls to minimize distortions of accounting information.
Nonetheless, users of accounting information often encounter preparer distortions from a
lack of preparer knowledge and/or objectivity. This study examines the process to repair a
user’s damaged beliefs following an encounter with preparer bias or preparer error. Users’
beliefs are examined in the context of trust to understand assessments of accounting
information as a subjective rather than objective judgment. The effects of an
organization’s control response to error (bias) discovery on repairing a user’s damaged
beliefs about accounting information are investigated. In general, the research question
examines whether a user’s trust based accounting beliefs are shaped only through one’s
direct experience with an information preparer, or indirectly by controls governing
preparer behavior. Based on experimental data from 159 participants, an organization’s
ethical control response repairs a user’s damaged beliefs following preparer bias less
effectively than a non ethical control response following preparer error. Controls
governing preparer behavior are found to shape a user’s beliefs about accounting
information in some scenarios, while in others only his or her direct experience with the
information preparer matters. Overall, findings from this study demonstrate economically
equivalent distortions of accounting information have psychologically different sequences

on a user’s beliefs about accounting information.

Copyright by
MICHELLE C. LAU
2014

ACKNOWLEDGMENTS

I would like to express my gratitude to both my committee co-chairs, Professor
Susan Haka and Professor Ranjani Krishan. I am truly inspired by their dedication to the
profession and passion for helping students succeed. Without their guidance and endless
support this dissertation would not have been possible.
I would also like to thank my committee members, Professor Harrison McKnight
and Professor Timothy Pleskac, whose expertise was a tremendous contribution. I am
grateful for the wonderful colleagues who have helped to shape and personalize my PhD
experience. Thank you Dara Marshall, Eric Marinich and all of the MSU doctoral students
for cheering me on. Financial support was generously provided by the Department of
Accounting and Information Systems at Michigan State University.
I am so fortunate and appreciative for finding such an incredible network of friends.
Thank you for a lifetime of joyful memories, laughter and tears with Chau Nyguen, Alda
Pires, Joyce Samuel, friends of Spartan Village and East Lansing; “How lucky I am to
have something that makes saying goodbye so hard” – A.A. Milne, Winnie the Pooh.
Lastly, I offer my sincerest and deepest gratitude to my family. Together you make
an irreplaceable base camp support team; without your help, no way, no how! Thank you
to SR for opening a world beyond my world. My dad told me once, when you set off to
climb a mountain sometimes you realize what you are searching for is what you left behind
– he was right.

v

TABLE OF CONTENTS

LIST OF TABLES

vii

LIST OF FIGURES

viii

INTRODUCTION

1

LITERATURE REVIEW & HYPOTHESES DEVELOPMENT
Ethical control on accounting beliefs
Assessment of Accounting Information as a Subjective Judgment
Subjective Judgment of Accounting Information as a Trust Based Judgment
Repairing Damaged Trust-Based Accounting Beliefs
Implicit Theory

4
4
6
7
9
11

METHOD
Experimental design
Participants
Procedures
Experimental treatments
Preparer Distortion
Ethical Control
Dependent Variable
Sample
Manipulation checks
Tests of Hypotheses
Subjective Beliefs
Hypotheses

16
16
16
17
19
19
20
20
22
22
23
23
24

APPENDICES
APPENDIX A
Excerpt from Experimental Materials

31
34
34

BIBLIOGRAPHY

45

vi

LIST OF TABLES

TABLE 1: Examples of Corporate Ethics Training and Development Programs

32

TABLE 2: Experimental Transfer Price Data

35

TABLE 3: Manipulation Checks

38

TABLE 4: Subjective Belief Scale Correlations (N = 72)

39

TABLE 5: Factor Analysis of Subjective Belief Dimensions (N = 72)

40

TABLE 6: How Preparer Distortions and Control Initiative Affect Subjective Beliefs
Test of H1

43

vii

LIST OF FIGURES

FIGURE 1: Hypothesized Relations Among Control, Preparer Distortion, Subjective
Beliefs and Transfer Price Decisions

36

FIGURE 2: Framework of Subjective Accounting Information Beliefs as a Subjective
Trust Judgment

37

FIGURE 3: Post Hoc Analysis

42

FIGURE 4: Tests of H2 and H3

44

viii

INTRODUCTION
Distortions of accounting information can violate expectations and damage a user’s
beliefs about accounting information. The purpose of this paper is to investigate the
effectiveness of ethical control on repairing a user’s damaged beliefs about accounting
information. Management accounting research grounded in agency theory has traditionally
focused on agency controls such as incentives, performance evaluation, monitoring, and
decision rights to minimize agent behaviors against the firm such as self-interested
preparer distortions (Christensen and Feltham 2005). Yet a growing body of research
acknowledges the boundaries of agency theory (e.g., Noreen 1988; Fehr and Schmidt
2006; Salterio and Webb 2006), and examines alternative mechanisms to control agent
behavior (Booth and Schulz 2004; Rowe 2004).
Organizations increasingly rely upon ethical climate1 in addition to or in lieu of
traditional agency controls to govern managerial actions when managerial actions are
perceived to be unethical and/or incongruent with the goals of the firm (Booth and Schulz
2004). Considering ethical climate (i.e., ethical control) on a user’s beliefs about
accounting information is important for at least two reasons. First, the effectiveness of
ethical control depends not only on its ability to govern agent behavior such as mitigating
preparer distortions and promoting a preparer’s communication honesty, but also on how it
influences a user’s beliefs about agent behavior and the characteristics of accounting
information. A user that distrusts accounting information will choose not to rely on the
information as a basis for decision making regardless of its inherent accuracy or objectivity
1

Ethical climate refers to the shared assumptions about normative and acceptable behavior, and how an
organization responds to ethical issues (Victor and Cullen 1988). Ethical climate can take many forms,
including the specific control mechanisms such as formal ethics training, codes of conduct, periodic reviews,
and reporting systems (Sims 1992).

1

(King 1996; Rowe, Shields, Birnberg 2012). Second, organizations are investing
increasing resources to communicate details of ethical climate (i.e., corporate social
responsibility reporting), therefore understanding how such programs affects a user’s
beliefs about accounting information is both relevant and important to organizations. This
study examines whether ethical control governing preparer behavior influences a user’s
beliefs about accounting information. The research question asks - can knowledge of a
preparer’s ethics training repair a user’s damaged beliefs about accounting information?
While a user’s beliefs about accounting information can be easily damaged by
encountering self-interested preparer distortions (King 1999), a user’s beliefs can also be
damaged by encountering a lack of preparer knowledge and/or experience limitations2
(Luft and Shields 2001; Gronewold, Gold, Salterio 2012). A natural question then arises –
does the nature of a preparer distortion affect controls governing preparer behavior on
repairing a user’s damaged beliefs about accounting information?
Drawing on implicit theory (Dweck, Chiu, and Hong 1995) and evidence from
psychology I predict the process to repair a user’s damaged beliefs about accounting
information depends not only on type of control governing preparer behavior, but also on
the nature of preparer distortion. Based on experimental evidence, I find users are more
responsive to controls governing preparer behavior following preparer distortions of error
relative to preparer distortions of bias. Specifically, I find a non-ethical control more
effectively repairs a user’s damaged beliefs following preparer error than an ethical control
following preparer bias. Results from the current study provide evidence that in some
scenarios a user’s beliefs about accounting information are influenced only by his or her

2

Although various heuristics and cognitive biases can give rise to preparer errors, in this paper preparer bias
refers to the objectivity (i.e., self-interest) of the information preparer.

2

direct experience with the preparer and/or information while in other scenarios a user’s
beliefs are shaped by the controls governing preparer behavior.
A 2x3x(4) mixed factorial design is used to test the hypotheses. I manipulate two
levels of preparer distortion preparer bias or preparer error. I also manipulate three levels
of control (none, ethical, and non-ethical). Participants’ beliefs about accounting
information are measured prior to and post manipulation. Participant choice behavior and
attitudes are also observed in a post distortion and control period.
The current study adds to a nascent but growing literature that examines belief
revision strategies following preparer distortions of accounting information. The current
study extends extant research on verbal and written strategies (e.g., Elliot, Hodge and
Sedor 2012) to investigate ethical control on repairing a user’s damaged beliefs about
accounting information. In this study, preparer errors are shown to have psychologically
different consequences than preparer biases on a user’s beliefs about accounting
information. Findings from this study demonstrate the consequences of those
psychological differences on the effectiveness of ethical control on repairing a user’s
damaged beliefs about accounting information. Overall, this study yields important
insights on how individuals process accounting information involving subjective
judgments. Moreover, this study offers an alternative perspective by examining the effects
of ethical control governing preparer behavior from the perspective of an information user.
The remainder of this paper is organized as follows. Section 2 reviews literature
and develops the hypotheses. Section 3 describes the research design. Section 4
summarizes results. Section 5 concludes the paper and offers a discussion of future
research.

3

LITERATURE REVIEW & HYPOTHESES DEVELOPMENT

Ethical control on accounting beliefs
Accounting research grounded in agency theory traditionally focuses on formal
controls such as incentives, performance evaluation, and monitoring to mitigate selfinterested behavior such as dishonest communication of private information. Yet
organizations also invest considerable resources to promote honesty, integrity, and trust to
mitigate agent behaviors against the firm (Dekker 2004; Emsely and Kidon 2007; Booth
and Schulz 2004; Webb and Salterio 2006). Ethical controls are often prescribed as an
alternative to or in lieu of traditional agency controls to govern managerial behavior. As an
example, The Sarbanes-Oxley Act of 2002 (SOX) requires public companies to disclose
whether they have (or not) adopted a code of ethics. The Federal Acquisition Regulations
(FAR) also requires companies doing business with the United States federal government
to provide periodic ethics training to all employees.
While prior research examines the influence of ethical control on ethical behavior
(e.g., Ponemon 1990; Booth and Schultz 2004), a nascent but growing area has begun to
consider how ethical controls influence beliefs about accounting information (e.g.,
Gronewold, Gold, Salterio 2013). In an experimental audit setting, Gronewold et al. (2013)
examine how ethical control alters perceptions about the willingness of auditors to selfreport preparer errors. Importantly, Gronewold et al. (2013) find auditors’ beliefs about
accounting information depend on how organizations respond to the discovery of preparer
errors. In their study, when organizations embrace errors as a cause for learning and
improvement (rather than punishment such as blame or sanctions), accounting information

4

is believed to be of better quality. Prior research finds ethical controls not only have a
significant effect on shaping the ethical behavior of agents, but also on beliefs about agent
behavior such as preparers’ willingness to communicate errors.
Gronewold et al. (2013) findings support the importance of how an organization
chooses to respond to error discovery. In practice, organizations often adopt formal
training programs to mitigate preparer distortions caused by both errors and biases.
Tangible skills training can be used to minimize preparer distortions caused by lack of
experience and/or knowledge. In contrast, many organizations rely on intangible skills
training such as formal ethics programs as a means to minimize deliberate preparer
distortions (Delaney and Sockell 1992). As an illustration, the formal ethics training
program of The Walt Disney Co. focuses not only on mitigating intentional preparer
distortions affecting external reporting but also internal accounting records such as such as
time cards, expense reports, invoices, performance evaluation and payroll (See Table 1 for
additional examples).
Extant studies on ethical training focus on the content, delivery and outcomes on
ethical behavior (Delaney and Sockell 1992). Delaney and Sockell (1992), for example,
find evidence to support a positive association between corporate ethics training and
ethical behavior by managers. Indeed, a primary objective of managerial accounting
systems is to facilitate learning and to improve decision making (Atkinson, Banker, Kaplan
1997; Sprinkle 2000). While management accounting systems (e.g., The Balanced
Scorecard) capture metrics such as employee participation rates, and dollars invested, the
conditions that make ethical training an effective control mechanism (or not) remains
relatively under explored. In particular, ethical training on beliefs about accounting

5

information is not well understood. A recent study of Fortune 500 companies reports 96%
of CEOs surveyed believe the impact of spending on programs such as ethics training is
important, however only 8% of CEOs report having an adequate understanding of this
investment (Phillips and Phillips 2009).

Assessment of Accounting Information as a Subjective Judgment
Accounting information is often assumed to be objective, precise, and reliable. An
important characteristic of accounting information is being reasonably free from preparer
error and preparer bias. When economic reality is routine and straightforward,
characteristics of accounting information can be easily verified (audited). However, many
types of accounting information lack an underlying basis for comparison are therefore hard
to verify3 (Rowe et al. 2012). Finding an underlying basis to assess management’s opinion
for example, can be costly or non-existent. When information is hard to verify, users often
form subjective beliefs about accounting information based on expectations about the
information and/or the information preparer (Hirst 1994; Kadous et al. 2012; Rowe et al.
2012). Specifically, users form beliefs based on expectations that information is
reasonably free of preparer errors and/or preparer biases4.
Importantly, decision makers often rely on their subjective beliefs about accounting
information rather than seeking methods to objectively assess the information (Rowe et al.
2012). Rowe et al. (2012) find when comparability of accounting information is low
(lacks standardization, uses technical language) managers trust the information has been

3

The IASB and FASB recognize that many types of information (e.g., management’s opinions) are
considered useful for decision making, but may not be directly or indirectly verifiable. As such, verifiability
is no longer included as an underlying characteristic of accounting information.
4
Although various heuristics and cognitive biases can also give rise to preparer errors, in this paper preparer
bias refers to the objectivity (i.e., self-interest) of the information preparer.

6

competently prepared by an expert with appropriate knowledge of accounting rather than
seeking methods to verify the contents for correctness. A manager who distrusts the
information will simply ignore the information as a basis for decision making (Rowe et al.
2012). That is, a user’s beliefs can have significant influence on whether information is
deemed useful, regardless of the inherent characteristics (e.g., accuracy, consistency,
verifiability) of the information.

Subjective Judgment of Accounting Information as a Trust Based Judgment
Relying on subjective beliefs about accounting information can expose the user to
unintentional and/or intentional distortions of accounting information by the information
preparer. Accounting based decisions involving subjective beliefs requires a willingness
of a user to be vulnerable to the information preparer based on positive expectations that
the information supplied is reasonably free from preparer errors and/or preparer biases.
This notion is consistent with trust - defined as the willingness of a party to be vulnerable
to the action of another party based on the expectation that the other party will perform a
particular action important to the trustor, irrespective of the ability to monitor or control
that other party (Mayer, Davis, and Schoorman 1995; Rousseau, Sitkin, Burt and Camerer
1998; Mayer and Davis 1999). As an illustration, a subsidiary manager enters into an
exchange relationship for goods and labor based on hard to verify cost estimates supplied
by another manager. To the extent the cost estimates are constructed based on asymmetric
information (e.g., customized product features, delivery scheduling, inside market
knowledge) the subsidiary manager risks being intentionally or unintentionally exploited
by the other manager’s information advantage. A manager choosing to continue in this
relationship notwithstanding the risk of being exploited reveals his or her expectations

7

about errors and/or bias embedded within the accounting information. The current study
draws upon the trust literature to better understand users’ beliefs about accounting
information as a subjective (trust-based) judgment.
Trust is a history dependent variable (Rotter 1980; Kramer 1999). Over time, as
one’s expectations are confirmed trust is strengthened (e.g., Coletti, Sedatole, Towry
2005). Individuals often demonstrate high levels of initial trust provided no reason to do
otherwise (Berg, Dickhaut, and McCabe 1995; Mcknight, Cummings, Chervany 1998).
This phenomenon is attributed to positive social expectations such as fairness, reciprocity
and honesty (Berg et al. 1995; McKnight et al. 1998; Colquitt, Scott, Lepine 2007).
Similarly, subjective beliefs about accounting information are also reinforced when
positive expectations of an information preparer and/or the information are confirmed.
Moreover, research finds information users demonstrate relatively positive beliefs about
accounting information even with limited or no prior experience with an information
preparer (Rowe et al. 2012).
Despite high levels of initial trust, trust can be easily damaged and is often hard to
re-establish once expectations are unmet or violated (Lewicki and Wiethoff 2000; Dirks
and Ferrin 2002; Kim, Ferrin, Cooper, and Dirks 2004; Emsley and Kidon 2007; Elliot,
Hodge, Sedor 2012). Similarly, users’ beliefs about accounting information are easily
damaged when expectations about the information and/or information preparer are violated
(King 1996). As an example, King (1996) finds buyers remain reluctant to rely upon a
seller’s accounting information after a single encounter with misreported costs even when
a seller is honest in subsequent interactions.

8

Accounting information beliefs are sensitive to users’ expectations about the
information preparer (Hirst 1994; King 1996; Maines and Wahlen 2006; Kadous et al.
2012). Kadous et al. (2012) find that investors view information with greater relevance
when it is received from a highly competent source rather than a less competent source.
Hirst (1994) finds auditors’ beliefs about an inventory valuation report depends on the
likelihood that the information preparer reports truthfully and accurately. As expectations
about the preparer’s objectivity to report truthfully and competence to measure inventory
accurately diminish, so too does the inferential value of accounting information. Hirst
(1994) also provides evidence that auditor judgments of audit information quality to be
unaffected by the degree of its verifiability. Findings from these studies suggest that users’
subjective beliefs about accounting information are sensitive to expectations of
competence and/or integrity of an information preparer. In addition, subjective judgments
about accounting information can differ from the actual characteristics inherent in the
information itself.

Repairing Damaged Trust-Based Accounting Beliefs
Accounting information users frequently encounter preparer errors and/or biases
that violate their expectations causing damage to beliefs as many factors such as economic
incentives (Healy and Wahlen 1999), knowledge limitations (Kadous et al. 2012),
mechanical or calculation error (Gronewold et al. 2013), and cognitive processes (Luft and
Shields 2001) cause preparer distortions of accounting information. For example, Luft and
Shields (2001) find cost-estimates are inherently less reliable (e.g., less accurate, more
inconsistent) when intangibles are expensed rather than capitalized because the choice of
accounting method affects how well managers are able to learn cost-profit relations. To

9

the extent preparer errors and biases distort accounting information from economic reality,
users’ expectations about information and/or the preparer can be violated. As such, users’
beliefs about accounting are easily damaged following encounters with preparer distortions
embedded within accounting information.
Maintaining positive beliefs about accounting information reduces transaction costs
(e.g., Dyer and Chu 2003), aids decision making (e.g., Zand 1972), and facilitates ongoing
relationships both within and across the firm (e.g., Kramer and Tyler 1996). Despite the
importance of a user’s positive beliefs about accounting information, beliefs can be easily
damaged. In particular, users’ beliefs can be easily damaged from encounters with preparer
distortions embedded within accounting information.
Devising strategies to repair a user’s damaged accounting information beliefs are
both relevant and important to organizations. However, repairing damaged beliefs can
prove more challenging than establishing initial positive beliefs about accounting
information. Repairing damaged beliefs differs from establishing initial beliefs as one must
not only (re-)establish positive expectations but also overcome any negative expectations
following an expectation violation (Kim et al. 2004). Assuming the process to repair a
user’s damaged beliefs requires a user to overcome negative expectations from preparer
distortions, the effectiveness of ethical control is predicted to depend on whether concerns
of subsequent preparer distortions can be alleviated.
A growing area of research investigates strategies to repair damaged trust. Extant
research investigates verbal responses on trust repair following an expectation violation
(Kim et al. 2004; Kim, Dirks, Cooper, Ferrin 2006; Elliot et al. 2012). Elliott et al. (2012)
examine a CEO’s verbal response to attribute the underlying cause for an unexpected

10

earnings misstatement. In their study, they find investors’ willingness to trust and
perceptions of management’s trustworthiness depends upon whether a CEO offers an
internal (i.e., personal) or external (i.e., situational) explanation for the underlying cause of
the earnings restatement. Although extant research focuses on verbal response as a belief
revision strategy, organizations also invest considerable resources on ethical controls to
manage expectations. The extent to which ethical control can be used to repair user’s
beliefs about accounting information following an expectation violation from preparer
distortion has yet to be fully examined.

Implicit Theory
Research on interpersonal evaluative and judgment processes in psychology offers
theory and evidence to predict the process to repair users’ beliefs about accounting
information depends not only on ethical controls, but also on the nature of preparer
distortion. Implicit theory describes a system of beliefs which guides how individuals
interpret and evaluate their social environment (Dweck et al. 1995). Hence, it is relevant
for understanding how preparer distortions and ethical control affect a user’s beliefs about
accounting information (Pedersen 1965; Dweck et al. 1995; Chiu, Hong, Dweck 1997).
Two assumptions about the personal attributes of others are central to the theory.
According to implicit theory individual attributes such as a person’s intelligence are either
assumed to be fixed and cannot be changed, or malleable - developed and changed with
effort and training over time (Dweck et al. 1995, Dweck 2008). Importantly, assumptions
about the malleability or fixedness of individual attributes are not mutually exclusive but
domain specific according to implicit theory. That is, a user may believe preparer

11

attributes (e.g., competence) to be malleable in one domain, while fixed in another (e.g.,
integrity) (Dweck et al. 1995).
When users assume the underlying cause of preparer distortions are caused by fixed
rather than malleable preparer attributes, then users will expect future preparer distortions
to re-occur with more certainty across time and context (Hong, Chiu, Dweck and Sacks
1997). Importantly, prior research suggests individuals tend to be more certain of future
preparer distortions related to integrity than competence (Reeder and Brewer 1979; Snyder
and Stukas 1999). As an illustration, King (1996) finds that once a user encounters a
single intentional preparer error he or she will anticipate subsequent dishonest
communication regardless of the level of honesty demonstrated by the preparer in
subsequent interaction periods. In contrast, preparer errors related to competence provide
less certainty about a manager’s future competence and the likelihood of subsequent errors
because even a very competent manager can perform poorly at times (Reeder and Brewer
1979; Snyder and Stuaks 1999).
Based on the foregoing, two conditions are necessary to alleviate a user’s concerns of
subsequent preparer distortions and repair a user’s damaged beliefs. First, a user must
believe preparer distortions are the result of a malleable (rather than fixed) preparer
attribute. To the extent a user believes the underlying cause of preparer distortions are the
result of fixed preparer attributes, ethical control will have little effect on repairing a user’s
beliefs. Users are predicted to be less responsive to ethical control following preparer
distortions caused by integrity based attributes (i.e., preparer bias) relative to competence
based attributes (i.e., preparer error).

12

Second, a user must believe the response used to govern preparer behavior will
adequately address the underlying cause of preparer distortion. As such, an ethical control
that does little to mitigate a user’s concerns of future preparer distortions is predicted to be
less effective on repairing a user’s beliefs than one that does. Given research that suggests
distortions of self-interest (i.e., integrity) provide more certainty of future distortions than
distortions of error (i.e., competence), an ethical control can exacerbate a user’s subjective
beliefs following preparer error by introducing concerns of agency problems than preparer
bias. In contrast, a non-ethical control that mitigates concerns of future preparer errors is
more likely to repair beliefs following preparer error than preparer bias. Taken together,
the formal hypothesis follows.
H: Ethical control on a user’s beliefs about accounting information depends on the
type of preparer distortion
Accounting based decisions involving subjective beliefs exposes the user to
potential distortions of information by the preparer, and thus reveals a user’s expectation
that the information has been prepared with appropriate objectivity and/or knowledge. As
previously stated, this notion is consistent with a common definition of trust5. Das and
Teng (2004) further clarify the definition of trust by differentiating the trust construct into
the following three components (1) subjective trust (2) trust antecedents, and (3)
behavioral trust. In this study, a user’s subjective beliefs about accounting information are
viewed in parallel with subjective trust. Subjective trust refers to a psychological state “a
5

Trust research in the accounting literature follows two streams. Some assume a substitutive relationship
between trust and control, and view trust as an informal control mechanism (e.g., Dekker 2004). Others
assume a complementary relationship between trust and control. Though the focus of this study is not to
reconcile these two perspectives, this study assumes trust and control to be related but distinct concepts.
Whereas control refers to explicit routines and procedures managers use to maintain or alter patterns in
organizational activities (Simons 1995), trust refers to expectations and interpersonal beliefs about others
(Mcknight et al. 1998). For the purpose of this study, trust is viewed as an important component of a
management control system, but is not in itself a control mechanism (Das and Teng 1998).

13

belief, attitude, or expectation concerning the likelihood that the actions or outcomes of
another individual, group, or organization will be acceptable or will serve the actor’s
interests” (Sitkin and Roth 1993, p368). Thus, a user’s subjective beliefs about accounting
information are considered as a psychological expectation that information is reasonably
free from preparer distortions. Behavioral trust refers to the actions or outcomes of
subjective trust such as the action to rely (or not) on one’s expectations of information as a
basis for decision making. In this study, a user’s decision to rely on information in a
subsequent period (behavioral trust) is predicted to be greatest following preparer error and
a non-ethical control relative to other combinations of ethical control and preparer
distortion. While it is possible to observe subjective trust in absence of behavioral trust and
vice versa, it is the simultaneous combination of subjective trust and behavioral trust that
constitutes trust (Kee and Knox, 1970; Das and Teng 2004). Analogously, it is the
simultaneous presence of a user’s positive beliefs about accounting information and a
willingness to base decisions upon those beliefs that indicates a user’s damaged beliefs
about accounting information have been repaired. Lastly, trust antecedents are the
situational and personal attributes which affect subjective trust.
To summarize, the current study examines how ethical control and preparer
distortions (trust antecedents) influence subjective beliefs about accounting information
(subjective trust), and the use of accounting information (or not) as a basis for decision
making (behavioral trust). Based on the foregoing, the formal hypotheses are stated as
follows:
H2: Following preparer error and a non-ethical control, users are more willing to
rely upon accounting information in a post-distortion period relative to all other
combinations of preparer distortion and control (behavioral trust).

14

H3: Subjective beliefs mediate the relation between controls, preparer distortion
and subjective accounting based decisions.

15

METHOD
Experimental design
A 2x3x(4) mixed factorial design is used to test the hypotheses. Participants are
randomly assigned to one of six experimental conditions. Two levels of preparer distortion
are manipulated (preparer bias or preparer error), and three levels of control initiatives are
manipulated (no control, ethical, and non-ethical). Participants complete four periods of
decision making. Participants’ beliefs about accounting information are measured via a
multi-item self-report Likert scale. Participants’ choice behavior is also captured in a post
distortion and control initiative period.

Participants
One hundred and fifty-nine students enrolled in a graduate level accounting course
from a large state university in the Mid-Western United States participated. Seventy
percent of the participants were male and the average age was 26. The average full-time,
professional work experience was more than two years. All experimental materials were
accessed online via laboratory computers. Eight participants ended the experimental
sessions prior to the final decision round, and were not included in the overall analysis.
Participants were informed that their compensation would be based on the outcome of their
decision for a randomly selected decision period. However, participants were not told how
many periods they might encounter. Participants were told that each decision period was
an independent from all other periods. Thus, the outcome of one decision period had no
effect on the outcome or decisions of any other decision period. At the end of the study,
one of the participant’s decision period was chosen at random by the computer to

16

determine his or her final compensation. Based on the outcome of the participant’s
randomly chosen decision period, participants could earn up to $10.

Procedures
At the beginning of each experimental session, participants are given a brief
introduction including an informed research ethics consent package, and guarantees for
anonymity. After signing the consent form, participants are provided further background
information about their role in the study. Participants are asked to assume the role of a
division manager for a large eye wear manufacturer. As division manager, participants are
tasked with the decision to manufacture a standard or custom order for sunglasses. The
custom order requires the purchase of a specialty lens produced by another division within
the company. Participants are told that the specialty lens is unique and has no outside
market price. As such, a transfer price for the lens must be agreed upon in order to
manufacture the custom order. While transferring the specialty lens at the highest possible
price is most profitable for the lens supplier, transferring lenses at the lowest possible price
is most profitable to the participant’s division. However, it is also described to participants
that finding a transfer price that distributes profits equitably between both divisions is
important because bonuses are based on divisional profits. Before proceeding, participants
respond to a series of questions to assess whether he or she understands the background
information presented.
At the beginning of each decision period, participants are provided with
information that details both the standard and custom order. The standard sunglass order is
described as an individual divisional project that consistently meets divisional profit targets
and guarantees a cash bonus of $5 at the end of each period. The custom project is

17

described as having the potential to generate a cash bonus up to $10 at the end of the
period if divisional profit targets are met (or no bonus if divisional profit targets are
unmet). The custom project requires the division manager to purchase custom lenses from
the internal lens division by accepting a proposed transfer price equal to variable cost per
unit of custom lens produced plus a 30% markup. Each period, participants receive the
Lens Division manager’s expected average variable costs of custom lens produced. The
expected average variable cost per unit of custom lens produced reported by the lens
division is described as being normally distributed with a standard deviation of $0.50/unit.
The participant also receives the proposed transfer price between divisions, and the
expected profits for each division. Based on the expected average variable cost per unit of
custom lens reported by the lens division, the proposed transfer price splits expected
profits equally between both divisions. The standard and custom orders differ in variance
but are equivalent in payoffs (see Appendix A for excerpt from experimental materials).
Participants are tasked with the decision to accept the proposed transfer price and
manufacture the custom order, or reject the proposed transfer price and manufacture the
standard order. At the end of each period, participants receive feedback on the outcomes
of their decision including the bonus earned (if any) based on his or her decision.
Overall, period 1 serves as a practice round for participants to familiarize
themselves with the study and task on hand. Since the focus of the current study is on the
repair of a user’s subjective beliefs about accounting information, period 1 also serves to
assess participant’s attitudes and expectations (e.g., fairness) on the proposed transfer
price. Period 2 serves to introduce the preparer distortion manipulation, the ethical control
initiative manipulation, and measure a user’s subjective beliefs about accounting

18

information. Period 3 and 4 provide a post preparer distortion period and enables the
measurement of behavioral trust.

Experimental treatments
Preparer Distortion
Following period 1, participants are informed of the start of period 2. In period 2,
participants encounter a preparer distortion when it is revealed that actual outcomes differ
from expected outcomes6. Specifically, the actual variable costs of custom lens produced
are greater than variable costs of custom lens produced initially reported by the Lens
Division manager. As a consequence, the actual transfer price is greater than the proposed
transfer price, and an inequitable split of profits arise between divisions. The preparer
distortion is either described to participants as the result of the Lens Division manager
being aware of additional material costs at the time of negotiations but deliberately
withholding this information for personal reasons (preparer bias), or the Lens Division
manager being unaware of additional material costs due to a lack of competence and
experience with estimating variable costs (preparer error). This manipulation is consistent
with the finding that a single encounter with a preparer distortion can be adequate to
damage a user’s subjective beliefs about accounting information (King 1999). Moreover,
prior research finds trust is easily damaged when social expectations such as fairness and
honesty are violated. Participants’ attitudes towards the consistency, expectations and
fairness of the expected and actual transfer price outcomes are also assessed via a Likert
self-report questions both prior and subsequent exposure to the preparer distortion
manipulation in period 2.

6

Participants who reject the proposed transfer price are not subject to a preparer distortion.

19

Ethical Control
At the beginning of period 3, participants are informed of one of two control
initiatives adopted by top management subsequent to the distortion discovery. The ethical
control condition is described to participants as a mandatory 6 hour in-classroom training
course on ethical conduct focused on keeping honest books and records. The non-ethical
control condition is described as a mandatory 6 hour in in-classroom training course on
cost estimation focused on keeping accurate books and records. Participants are also told
that the course has been added to the schedule of continuous learning and development
training to be received by all divisional managers in each subsequent year. A third
condition in which no control initiative is introduced provides a baseline for comparison.
Dependent Variable
In period 3, after responding to the decision of whether to accept or reject the
transfer price, but prior to receiving feedback on the outcome of the participant’s decision,
participants complete a multi item questionnaire to assess subjective beliefs about
accounting information. Participants were asked to report the extent to which they agreed
or disagreed with the following statement “I believe the information reported by the Lens
Division Manager to be reliable”. Participants were also asked to respond to statements
based on properties of accounting information such as accuracy (“I believe the variable
costs per unit of custom lens produced reported by the Lens Division Manager are
accurate”), truthfulness (“I believe the variable costs per unit of custom lens produced
reported by the Lens Division Manager to be true”), and faithful representation (“I am
confident the Lens Division Manager would not intentionally lie to me”). Participants also
responded to the following statements, “The Lens Division Manager lacks integrity” and
“The Lens Division Manager is trustworthy”. Participants responded to each statement

20

using a seven point Likert scale. After responding to the multi-item questionnaire,
participants receive outcome feedback based on their decision to accept or reject the
transfer price at the end of period 3. Next, participants complete period 4 by submitting
their decision to either accept or reject the transfer price. Participants respond to a final
questionnaire and receive feedback on their final round of decision making.

21

RESULTS

Sample
The following analyses are conducted based on the following sample. Of the 159
participants, 43 participants rejected the transfer price in period 2 and did not encounter an
expectation violation, and are therefore not included in the following analyses. The
remaining108 participants accepted the transfer price in period 2 and subsequently
encountered the preparer distortion manipulation (error or bias). Of the 108 participants, 36
participants were randomly assigned to the baseline condition, while 72 participants were
either assigned to the strong ethical control condition or weak ethical control condition.
Eight participants ended their experimental sessions prematurely and were excluded from
the final analyses.

Manipulation checks
The literature on trust finds violations of social expectations such as honesty and
fairness (i.e., equity) damages trust between parties (Das and Teng 1998). Similarly,
expectation violations arising from preparer distortions are expected to damage a user’s
subjective beliefs. To assess the effectiveness of the preparer distortion manipulation,
several measures were analyzed to establish whether 1) participants experienced an
expectation violation due to preparer distortion and 2) participants weighed the violation
equally across conditions. Results are reported in Table 3. While participants agreed the
initial transfer price proposed by the Lens manager to be equitable (Bias: µ = 2.20; Error: µ
= 2.51) across preparer distortion conditions, participants found the actual outcomes of the
transfer price to be highly inequitable in period 2 (Bias: µ = 6.00; Error: µ = 5.48) (Table 3,

22

Panel A). In addition, results reveal that actual variable cost reported by the Lens manager
was highly inconsistent with the expected variable cost of participants for both types of
preparer distortion (Bias: µ = 5.85; Error µ = 5.50). Taken together, these results indicate
participants experienced a similar expectation violation of equity and consistency
following both types preparer distortion. Moreover, when no control initiative was
introduced there was no significant differences found between subjective beliefs across
preparer distortion conditions (Bias: µ = 0.175; Error: µ = 0.022; F = .126, p =.73). The
preparer distortion manipulation appears to have been effective and consistent across both
preparer distortion conditions.
To assess the ethical control manipulation, participants were asked to identify the
type of training received by the Lens division manager (Table 3, Panel B). As expected,
participants in the ethical control condition correctly responded that the Lens Division
manager had received ethics training, but not cost training. In contrast, participants in the
non-ethical control condition responded correctly that the Lens Division manager had
received cost training, but not ethics training. Thus, it appears that both the control and
preparer distortion manipulations were successful.

Tests of Hypotheses
Subjective Beliefs
As discussed in Section III, subjective beliefs were measured based on six Likertscale questions. Descriptive statistics are reported on Table 3, Panel C. All six items
correlate significantly (p < 0.05, two-tailed) with all other items, with correlation
coefficients of at least .4 suggesting reasonable factorability (Table 4). The measurement
model with six measures provides a reasonably good fit with a minimum discrepancy χ2

23

=16.81(df.=9, p > 0.05), a Root Mean Square Error of Approximation (RMSEA) of 0.07,
Comparative Fit Index (CFI) of 0.99, and a Tucker-Lewis Index (TLI) of 0.987. Table 5
presents the factor analysis results. Panel A describes the set of six questions. Panel B
summarizes factor loadings and communalities. As shown in Panel C, the factor analysis
yields a single factor that accounts for 68 percent of the cumulative variation. The factor
loading is consistent with expectations, and represents subjective beliefs. Participants’
factor score measures subjective beliefs and is used to test the hypotheses.
Hypotheses
H1 predicts an interaction between ethical control and preparer distortion on
subjective beliefs. Figure 3 illustrates the pattern of results for the hypotheses. As shown,
there is a cross over interaction between controls and preparer distortion. Table 6 presents
the analysis of variance (ANOVA) model with subjective beliefs as the dependent variable.
Table 6, Panel A reports cell sizes, means and standard errors for subjective beliefs.
Results on Table 6, Panel B reveal a significant interaction between controls and preparer
distortion on subjective beliefs (F = 4.792, p<0.05, two-tailed).
A Post Hoc analysis was performed. As predicted, users’ responsiveness to ethical
control was found to be greater following preparer distortions from error than bias.
Furthermore, non-ethical control had a more positive effect on subjective beliefs following
preparer error than ethical control (F(1,68) = 5.61, p < 0.05). No significant difference was
found between either ethical or non-ethical control condition following a preparer
distortion of bias (F(1,68) = 0.58, p = .45). Additional analyses (untabulated) were
performed using the baseline condition (no control). In absence of an ethical control

7

As a rule of thumb, a model may be considered to have a reasonably good fit when RMSEA values are
close to 0.06, CFI and TLI are .95 or greater (Hu and Bentler 1999; Albright and Park, 2009).

24

initiative, no significant difference (F(1,34) = 1.26, p = .725) was found across preparer
distortion condition (error µ = 0.022 versus bias µ = 0.175) for subjective beliefs. These
findings provide support for the conclusion that the effectiveness of controls governing
preparer behavior on repairing a user’s subjective beliefs about accounting information
depends on the nature of preparer distortion.
H2 predicts repair of behavioral trust will be greatest given a non-ethical control
following preparer error relative to all other combinations of preparer distortion and
preparer control. As illustrated in Figure 4 panel A, the percentage of participants who
rejected the proposed transfer price in the post-distortion period did not differ across
control conditions following a distortion of preparer bias χ2 (2, N = 45) = 1.13, p = 0.570.
This sample included 11 participants in the no control manipulation, 15 in the ethical
training condition and 19 participants in the non-ethical training condition. The percentage
of participants that rejected the proposed transfer price in the post-distortion period
following distortion of preparer error differs significantly across control conditions χ2 (2, N
= 63) = 14.22, p > 0.01. As can be seen by the frequencies cross tabulated in Figure 4 panel
A, the percentage of participants who rejected the proposed transfer price was significantly
less following the non-ethical control relative to no control initiative condition. Following
a distortion of preparer error 48% of participants rejected the proposed transfer price in the
post distortion period when no control initiative was introduced compared with 13% of
participants rejecting the proposed transfer price following the introduction of the nonethical control initiative. Results from H2 provides evidence that the effectiveness of
control initiatives on repairing behavioral trust depends on the nature of preparer
distortion. While an ethical control initiative has little effect on repairing behavioral trust

25

following a distortion of preparer bias, a non ethical control repairs behavioral trust
following a distortion of preparer error. Overall, support for H2 is found.
H3 predicts a mediating effect of subjective beliefs on the relationship between
controls, preparer distortion and the decision to accept or reject a proposed transfer price in
a post-distortion period. The mediation analysis was conducted in AMOS. As a precondition for mediation the direct effect between control, preparer distortion and transfer
price decision was tested. A significant direct effect was found between preparer distortion
and transfer price decision (p < 0.05, two-tailed), but not controls (p = 0.23, two-tailed).
Thus, the mediation analysis was conducted on preparer distortion and transfer price
decision. As recommended for small samples, a nonparametric bootstrapping analysis was
used (Preacher and Hayes 2004). Results based on 1,000 bootstrapped samples indicate
that whilst controlling for variable cost knowledge of the Lens Division Manager, the total
effect of preparer distortion on transfer price decision is marginally significant (TE =.152,
SE = .119, p = 0.059), the direct effect is not (p = .152), and the indirect effect is
marginally significant (p = 0.08). Thus, subjective beliefs mediates the relationship
between preparer distortion and transfer price decision controlling for knowledge of
variable costs (IE lower 95% CI = -.004, upper 95% CI=.188). Based on results of the
mediation, given an ethical control a negative relationship was found between subjective
beliefs and the decision to accept the transfer price in a post-distortion period. Figure 4
summarizes the results.
Overall, findings in this study provide evidence that ethical control less effectively
repairs a user’s accounting beliefs following preparer bias than non-ethical control
following preparer error. While non-ethical control was found to repair both a user’s

26

subjective beliefs and behavioral trust relative to no control, ethical control had little effect
on repairing a user’s subjective trust or behavioral trust. As Rousseau, Sitkin, Burt and
Camerer (1988) argue “trust is not a behavior or a choice (e.g., taking a risk), but an
underlying psychological condition that can cause or result from such actions” (p 395). In
other words, it is the simultaneous presence of both subjective trust (e.g., positive
subjective beliefs about accounting information) and behavioral trust (e.g., decision to rely
on accounting information) that constitutes trust repair. Findings from this study reveal
ethical controls were less likely to repair damaged beliefs following distortion of preparer
bias compared to preparer error. Repairing a user’s damaged beliefs about accounting
information depends not only on re-establishing positive expectations about the objectivity
and accuracy of accounting information, but also on observing a user’s willingness to rely
on accounting information in a post distortion period. In this study, repairing a user’s
damaged beliefs about accounting information depends on both the nature of preparer
distortion, and also on the type of ethical control.

27

CONCLUSIONS

Organizations rely upon controls to mitigate distortions of accounting information
from economic reality. Distortions can lead to expectation violations and damage a user’s
beliefs about accounting information. The following study examines whether ethical
controls governing behavior of a preparer can influence a user’s beliefs about accounting
information, or whether a user’s beliefs are only affected by their direct experiences and
interactions with a preparer. Results from this study demonstrate the effectiveness of
controls on repairing a user’s damaged beliefs about accounting information depends on
the nature of control and type of preparer distortion. Following a distortion arising from
preparer error, a non-ethical control repairs a user’s beliefs more effectively than an ethical
control following preparer bias. Results support the conclusion that while in some
situations ethical control affects a user’s beliefs about accounting information, in other
situations only a user’s direct experience with accounting information and interactions with
a preparer affect those beliefs.
Findings in this study provide important insights for both practitioners and
scholars. First, I demonstrate distortions of information arising from preparer bias to have
psychologically different implications than distortions arising from preparer error. I find
the conditions conducive for non-ethical control are not parallel to the conditions
conducive for ethical control on repairing a user’s damaged beliefs about accounting
information. Implementing a non-ethical control governing preparer behavior was found to
be an effective strategy to repair damaged beliefs following distortions due to preparer
error. In contrast, an ethical control governing preparer behavior was found to be an
ineffective strategy to repair damaged beliefs following distortions due to preparer bias.

28

Practically speaking, results from this study provide important insights for understanding
the conditions of using ethical and non-ethical controls as a strategy to repair a user’s
damaged beliefs about accounting information.
Second, results from this study contribute to a deeper understanding of assessments
of accounting information as a subjective judgment. Prior research in accounting focuses
either on preparer incentives to distort accounting information, or controls used to mitigate
preparer distortions of accounting information. Results from my study add to extant
research by examining controls from the perspective of the information user. Moreover,
objective assessments of accounting information can often be difficult to ascertain in
practice. By examining a user’s subjective beliefs about accounting information in the
context of trust, I provide a framework to better understand assessments of accounting
information as a subjective judgment.
Findings in the current study also provide insights on corporate social responsibility
efforts. Prior research documents numerous benefits associated with ethics training. In
addition, an increasing trend by companies to report on their ethical initiatives exists. As
organizations seek to increase the visibility of their ethics programs, the current study
identifies conditions that exacerbate their effectiveness. Findings from this study provide
some insight on policy decisions aimed to improve corporate ethics (e.g., Sarbanes Oxley)
from the perspective of the accounting information user. Specifically, a user’s beliefs may
not reflect whether a control actually achieves desirable behavior from an information
preparer.
Several streams of future research exist. The current study examines specific forms
of controls, preparer distortions, and accounting based decisions. Future studies can

29

investigate how different variations of these variables can affect assessments of accounting
information. For example, under what conditions are ethical controls versus traditional
agency controls more or less effective on mitigating agency concerns? What are the
implications of implementing ethical controls following various internal or external audit
deficiencies? Are ethical controls more or less effective as a reactive versus preventative
control strategy?
While the current study focuses on beliefs from the perspective of the information
user, future research could also examine the social interactions between the information
user and preparer to examine accounting in a dyadic, or multi-party (rather than individual)
setting. This could bring together two streams of literature that focus on verbal
communication by the preparer (e.g., cheap talk) and formal controls on belief revision
strategies following an expectation violation. A comprehensive model that includes not
only a user’s beliefs about the information preparer, but also individual attributes of the
information user (e.g., risk preferences, trust propensity, optimism), and contextual
variables (e.g., gain/loss frame, interaction history, incentives) could be useful for
expanding the current understanding of factors affecting subjective beliefs involving
accounting information.

30

APPENDICES

31

TABLE 1
Examples of Corporate Ethics Training and Development Programs
Company
The Walt
Disney Co.

The Coca-Cola
Co.

Industry
Mass Media

Beverage

Excerpts from Corporate Ethics Training Programs (extracted from company’s website)
Compliance training, including training regarding the Company's Standards of Business Conduct
and ethics, is provided to employees and Cast Members worldwide through the Company's
learning management system known as Disney Development Connection. It is the Company's
intent, through its compliance training, to ensure that all of its employees and Cast Members have
the knowledge and training to act ethically and legally, in compliance with the Company's
Standards of Business Conduct.
Accurate and complete record keeping is essential to the successful operation of our company, as
well as our ability to meet our legal and regulatory obligations. You have the responsibility to be
accurate, complete, and honest in what you report, and record to meet regulatory requirements, as
well as in all Company documents including accounting records, time cards, expense reports,
invoices, payroll records, business records, performance evaluations etc…
To ensure an ongoing commitment to and understanding of our Code of Business Conduct, we
offer online training to all associates with Company-provided computers discussing topics related
to ethics and compliance. All newly hired associates receive the training upon hire and all others
receive the training at least once every three years. In 2010, approximately 22,000 employees
(management and non-management) certified their compliance with the Code of Business Conduct
and the Company's anti-bribery requirements. In addition to a number of optional training courses
on various topics, associates are requested to participate in ethics training on an annual basis,
resulting in an average of 60 minutes of ethics training per associate per year.

32

TABLE 1 (cont’d)
Examples of Corporate Ethics Training and Development Programs
Company
Boeing

Industry
Aerospace and
Defense

Excerpts from Corporate Ethics Training Programs (extracted from company’s website)
The Boeing Company is committed to fostering an environment where integrity is valued and
forms the foundation for every decision. Although maintaining ethical behavior is woven
throughout our daily communications and activities, the company conducts three mandatory and
educational activities annually as reminders about our commitment to ethics and business conduct
standards at Boeing. Annual activities include Recommitment to Ethics, Code of Conduct and the
Ethics Challenge. The Boeing Code of Conduct certification occurs annually. Employees certify
once a year that they will adhere to the Code of Conduct, which outlines the ethical business
conduct required of employees in the performance of their company responsibilities. Individuals
certify that they will not engage in conduct or activity that may raise questions as to the company's
honesty, impartiality or reputation or otherwise cause embarrassment to the company, among
other things. The annual Ethics Challenge training educates employees about situations they might
face in daily business using specific case scenarios. Participants answer questions about ethical
dilemmas; then learn which is the best answer and why. This training, which is administered by
managers, is typically completed in a group setting.

33

APPENDIX A

Excerpt from Experimental Materials
This period, your division can manufacture custom sunglasses that can be sold to retailers
for $26.00 per unit. The Lens Division manager proposes to transfer custom lenses to your
division at a price equal to: variable cost per unit of custom lens produced plus a 30%
markup. It is up to you whether you choose to accept or reject this transfer price. (Note:
Profits for the custom sunglasses will be shown to you at the end of each period regardless
of your decision.)
If you reject the proposed transfer price for custom lenses, then your division will
manufacture standard sunglasses, and you will earn a bonus of $5.00 at the end of this
period.
If you accept the proposed transfer price for custom lenses, then your division will
manufacture custom sunglasses. You will earn a bonus of $10.00 if your division’s profit
per unit is $3.00 or more at the end of this period. If, however your division’s profit per
unit is less than $3.00, then you will earn a bonus of $0.
The Lens Division manager reports an average variable cost per unit of custom lens
produced of $10.00 to you. The costs are normally distributed with a standard deviation of
$0.50/unit. Profits for the custom order are summarized below for both divisions:

34

TABLE 2
Experimental Transfer Price Data
Panel A: Expected outcome shown to participant
Sunglasses Division
Custom Order
Selling price of sunglasses to retailer

Expected
$ per unit

Lens Division
Custom Order

$18.00 Transfer price of lens to Sunglasses Division
(Variable cost per unit plus 30% markup)

Expected
$ per unit
$13.00

Minus:
Transfer price of lens from Lens Division
Other costs

Minus:
13.00 Variable cost per unit of custom lens produced
2.00

10.00

Profit Sunglasses Division

$3.00 Profit Lens Division

$3.00

Panel B: Actual outcome shown to participant accepting proposed transfer price
Sunglasses Division
Custom Order
Selling price of sunglasses to retailers

Actual
$ per unit

Lens Division
Custom Order

$18.00 Transfer price of lens to Sunglasses Division
(variable cost per unit plus 30% markup)

Actual
$ per unit
$15.60

Minus:
Transfer price of lens from Lens Division
Other costs

Minus:
15.60 Variable cost per unit of custom lens produced
2.00

12.00

Profit Sunglasses Division

$0.40 Profit Lens Division

$3.60

Your Bonus this period

$0 Lens Division Manager’s bonus this period

35

$10.00

FIGURE 1
Hypothesized Relations Among Control, Preparer Distortion, Subjective Beliefs and
Transfer Price Decisions
Panel A: Predicted relations of Preparer Distortion by Control Condition on Subjective
Beliefs

Subjective
Beliefs
Preparer Bias
Preparer Error

Non-Ethical
Ethical
Control Initiative

36

FIGURE 2
Framework of Subjective Accounting Information Beliefs as a Subjective Trust Judgment
Panel A: Subjective Beliefs as a subjective trust judgment

Trust Antecedents

Subjective Trust

Behavioral Trust

Preparer Distortion
Subjective Beliefs
about Accounting
Information

Transfer Price
Decision

Control Initiative

Panel B: Subjective accounting beliefs (subjective trust) are measured based on the
following six items:
1. I believe the variable costs per unit of custom lens produced reported by the Lens
Division Manager to be true.
2. I am confident the Lens Division Manager would not intentionally lie to me
3. I believe the variable costs per unit of custom lens produced reported by the Lens
Division Manager are accurate
4. I believe the information reported by the Lens Division Manager to be reliable.
5. The Lens Division Manager lacks integrity
6. The Lens Division Manager is trustworthy
Panel C: Transfer price decision (behavioral trust) is measured based on the following
item:
1. Accept proposed transfer price or Reject proposed transfer price

37

TABLE 3
Manipulation Checks

Mean
Mean
Bias
Error
n = 45
n = 63
Pre-preparer distortion based on initial costs reported by the Lens Division Manager
Q: Profits are equitably split between my division and
2.20
2.51
the Lens Division
(t = -.986, p = .33*)
Panel A: Preparer Distortion Manipulation Check

Post-preparer distortion based on actual costs reported by the Lens Division Manager
Q: Profits are equitably split between my division and
6.00
5.48
the Lens Division
(t = 1.68, p = 0.10*)
Q: Actual variable costs of custom lens produced are consistent
with initial costs reported by the Lens Division Manager
Panel B: Control Manipulation Check

5.85
5.50
(t = 1.01, p = .32*)
Ethical
n= 30

Non-Ethical
n= 42

Q: The Lens Division Manager received cost training

4.43
2.86
(F = 8.47, p < 0.05*)

Q: The Lens Division Manager received ethics training

2.07
4.07
(F = 16.06, p < 0.05*)

Panel C: Descriptive Statistics for Subjective Beliefs of participants who choose to accept
the transfer price in Period 2

Error
Preparer Distortion
Bias

Ethical Control
None
Ethical Non-Ethical
0.022
-.320
0.369
n=25
n=15
n=23
0.175
-0.052
-0.282
n=11
n=15
n=19

__________________
*two-tailed equivalent.

38

TABLE 4
Subjective Belief Scale Correlations (N = 72)

Item 1
Item 1
Item 2
0.597*
Item 3
0.689*
Item 4
0.667*
Item 5
-0.440*
Item 6
0.496*
See Figure 2 Panel B for item descriptions
____________________________________
*p <0.05, Two-tailed equivalent.

Item 2

Item 3

Item 4

Item 5

Item 6

0.724*
0.725*
-0.486*
0.627*

0.826*
-0.494*
0.603*

-0.537*
0.652*

-0.57*

-

39

TABLE 5
Factor Analysis of Subjective Belief Dimensions (N = 72)
Panel A: Descriptive Statistics (Mean and Standard Deviation)
Question item
I believe the variable costs per unit of custom lens produced reported
by the Lens Division Manager to be true
I am confident the Lens Manager would not intentionally lie to me
I believe the variable costs per unit of custom lens produced reported
by the Lens Division manager are accurate
I believe the information reported by the Lens Division Manager to
be reliable
The Lens Division manager lacks integrity.
The Lens Division manager is trustworthy
Panel B: Factor Loadings and Communalities
Question item
I believe the variable costs per unit of custom lens produced reported
by the Lens Division Manager to be true
I am confident the Lens Manager would not intentionally lie to me

40

Mean

Std. Deviation

4.69

1.47

4.49

1.56

4.60

1.33

4.48
3.92
4.16

1.42
1.50
1.43

Reliability Belief
Factor

Communality

0.62
0.72

0.79
0.85

TABLE 5 (cont’d)
Factor Analysis of Subjective Belief Dimensions (N = 72)
Reliability Belief
Factor

Communality

0.79

0.89

0.81
0.49
0.64

0.90
-0.70
0.80

Eigenvalue

Explained
Variance

Cumulative Explained
Variance (%)

4.07
0.66

67.82
11.07

67.82
78.89

0.46

7.63

86.52

0.34
0.29
0.17

5.75
4.91
2.82

92.27
97.18
100.00

Question item

I believe the variable costs per unit of custom lens produced reported
by the Lens Division manager are accurate
I believe the information reported by the Lens Division Manager to
be reliable
The Lens Division manager lacks integrity.
The Lens Division manager is trustworthy
Panel C: Principal Component Factors
Question item
I believe the variable costs per unit of custom lens produced reported
by the Lens Division Manager to be true
I am confident the Lens Manager would not intentionally lie to me
I believe the variable costs per unit of custom lens produced reported
by the Lens Division manager are accurate
I believe the information reported by the Lens Division Manager to
be reliable
The Lens Division manager lacks integrity.
The Lens Division manager is trustworthy

41

FIGURE 3
Post Hoc Analysis
a
Test of H1 (N = 72)

Preparer Distortion and Control Initiative on Subjective Beliefs

0.6
0.4

0.369

0.2
Preparer Bias
-0.052

0
-0.2
-0.282

Preparer Error
-0.32

Non-Ethical Control

Ethical Control

-0.4

_________________________
a
Variables described in Table 3

42

TABLE 6
How Preparer Distortions and Control Initiative Affect Subjective Beliefs
Test of H1
Panel A: Subjective Beliefs Factor, Mean [Standard Error], n = 72

Control Condition
Ethical Control

Non-ethical control

Preparer
Distortion
Condition
Bias

n
15

Error

15

Bias

19

Error

23

Panel B: ANOVA Model of Subjective Beliefs
Source of Variation
SS
df
Intercept
0.357
1
Controls
0.917
1
Preparer Distortion
0.639
1
Controls X Preparer Distortion
3.679
1
Error
52.21
68

Subjective Belief Factor
-0.052
[.980]
-0.320
[.807]
-0.282
[.985]
0.369
[.744]
MS F-statistic
0.357
0.465
0.917
1.194
0.639
0.832
3.679
4.792
0.768

p-value
0.498
0.278
0.365
p <.05*

* Two-tailed equivalent.
This table reports the results for tests of H1 using the factor score from a principal component analysis of
the six questions resented in Table 3 Panel A as the dependent variable.

43

FIGURE 4
Tests of H2 and H3
Panel A: Participants’ Decision to Accept vs. Reject Proposed Transfer price in
Post-Distortion Period (H2 Behavioral Trust) (N = 108)

Decision to Reject vs. Accept Transfer Price in Post
Distortion Period
# of Participants
25
20
1213
12
15
7
10
4
3
5
0
Bias x Error x Bias x

20
15

15
4

0

3

Error x Bias x

Reject
Accept

Error x

No
No
Ethical Ethical Non
Non
control control Control Control Ethical Ethical
Distortion x Control

Panel B: Observed relations of Preparer Distortion, Ethical Controls, Subjective
Beliefs on Transfer Price Decision in a Post Distortion Period (N = 72)
Variable cost
Knowledge
0.09
Ethical
Controls

-0.04

n.s
Subjective Beliefs

-0.31**

-0.18
.15 (0.06*)
Preparer
Distortion
_____________________
* One-tailed significance at less than 0.05

44

Transfer Price
Decision

BIBLIOGRAPHY

45

BIBLIOGRAPHY

Albright, J.J. and H.M. Park. 2009. Confirmatory factor analysis using Amos,
LISREL, Mplus, SAS/STAT/ CA LIS. IN: Indian University Information
Technology Service Publishing.
Atkinson, A. A., R.D. Banker, R.S. Kaplan, and S.M. Young. 1997. Management
Accounting. Upper Saddle River, NJ: Prentice Hall, Inc.
Berg, J, J. Dickhaut, and K. McCabe. 1995. Trust, reciprocity and social history.
Games and Economic Behavior 10: 122-142.
Chiu, C. Y., Y. Y. Hong, and C.S. Dweck. 1997. Lay dispositions and implicit
theories of personality. Journal of personality and social psychology 73: 19-30.
Coletti, A.L., K.L. Sedatole, and K.L. Towry. 2005. The effects of control
systems on trust and cooperation in collaborative environments. The Accounting
Review 80(2): 477-500.
Colquitt, J.A., B.A. Scott, and J.A. LePine. 2007. Trust, trustworthiness, and trust
propensity: A meta-analytic test of their unique relationships with risk taking and
job performance. Journal of Applied Psychology 92(4): 909-927.
Das, T.K., and B.S. Teng. 1998. Between trust and control: Developing
confidence in partner cooperation in alliances. Academy of Management Review
23(3):491-512.
Das, T.K. and B.S. Teng. 2004. The risk-based view of trust: A conceptual
framework. Journal of Business and Psychology 19(1): 85-116.
Dekker, H.C. 2004. Control of inter-organizational relationships: Evidence on
appropriation concerns and coordination requirements. Accounting, Organizations
and Society 29: 27-49.
Delaney, J.T., and D. Sockell. 1992. Do company ethics training programs make a
difference? An empirical analysis. Journal of Business Ethics 11(9): 719-728.
Dirks, K.T., and D.L. Ferrin. 2002. Trust in leadership: meta-analytic findings and
implications for organizational research. Journal of Applied Psychology 87(4):
491-512.
Dweck, C.S. 2008. Can personality be changed? The role of beliefs in personality
and change. Current Directions in Psychological Science 17(6): 391-394).

46

Dweck, C.S., Chiu, C., and Y. Hong. 1995. Implicit theories and their role in
judgments and reactions: A world from two perspectives. Psychological Inquiry
6(4), 267-285.
Dyer, J.H. and W. Chu. 2003. The role of trustworthiness in reducing transaction
costs and improving performance: empirical evidence from the United States,
Japan, and Korea. Organization Science 14(1): 57-68.
Elliot, W.B., F.D. Hodge, and L.M. Sedor. 2012. Using online video to announce
a restatement: Influences on investment decisions and the mediating role of trust.
The Accounting Review 87(2): 513-535.
Emsley, D. and F. Kidon. 2007. The Relationship between trust and control in
international joint ventures: Evidence from the airline industry. Contemporary
Accounting Research 24(3): 829-858.
Gronewold, U., A. Gold, S.E. Salterio. 2013. Reporting self-made errors: The
impact of organizational error-management climate and error type. Journal of
Business Ethics 117(1):189-208.
Healy P.M. and J.M.Wahlen. 1999. A review of the earnings management
literature and its implications for standard setting. Accounting Horizons
13(4):365-383.
Hirst, D.E. 1994. Auditor’s sensitivity to source reliability. Journal of Accounting
Research 32(1): 113-126.
Hong, Y., Chiu, C. Y., Dweck C. S. and R. Sacks. 1997. Implicit theories and
evaluative processes in person cognition. Journal of Experimental Social
Psychology 33: 296-323.
Hu, L. and P.M. Bentler. 1999. Cutoff criteria for fit indexes in covariance
structure analysis: conventional criteria versus new alternatives. Structural
Equation Modeling 6:1-55.
Kee, H.W., and R.E. Knox. 1970. Conceptual and methodological considerations
in the study of trust and suspicion. Journal of Conflict Resolution 14: 357-366.
Kim, P.H., D.L. Ferrin, C.D. Cooper, and K.T. Dirks. 2004. Removing the
shadow of suspicion: the effects of apology versus denial for repairing
competence –versus integrity-based trust violations. Journal of Applied
Psychology 89(1): 104-118.
Kim, P.H., K.T. Dirks, C.D. Cooper, and D.L. Ferrin. 2006. When more blame is
better than less: the implications of internal vs. external attributions for the repair

47

of trust after a competence – vs. integrity-based trust violation. Organizational
Behavior and Human Decision Processes 99: 49-65.
Kadous, K., K.L., Koonce, and J.M. Thayer. 2012. Do financial statement users
judge relevance based on properties of reliability? The Accounting Review 87(4):
1335-1356.
King, R.R. 1996. Reputation formation for reliable reporting: An experimental
investigation. The Accounting Review 71(3): 375-396.
Kramer, R.M. 1999. Trust and Distrust in Organizations: Emerging perspectives,
enduring questions. Annual Review of Psychology 50: 569-598.
Kramer, R.M. and T.R. Tyler. 1996. Trust in organizations: Frontiers of theory
and research. Thousand Oaks, CA: Sage Publications, Inc.
Lewicki, R. J., and C. Wiethoff. 2000. Trust, trust development, and trust repair.
The handbook of conflict resolution: Theory and practice 1(1): 86-107.
Luft, J.L. and M.D. Shields. 2001. Why does fixation persist? Experimental
evidence on the judgment performance effects of expensing intangibles. The
Accounting Review 76(4): 561-587.
Maines, L. and J. Wahlen. 2006. The nature of accounting information reliability:
Inferences from archival and experimental research. Accounting Horizons 24:
399-425.
Mayer, R.C., and J.H. Davis. 1999. The effect of the performance appraisal
system on trust for management: A field quasi-experiment. Journal of Applied
Psychology 84:123-136.
Mayer, R. C., Davis, J. H., and F.D Schoorman. 1995. An integrative model of
organizational trust. Academy of Management Review 20: 709–734.
McKnight, D.H., L.L. Cummings, and N.L. Chervany. 1998. Initial trust
formation in new organizational relationships. The Academy of Management
Review 23(3): 473-490.
Pedersen, D.M. 1965. The measurement of individual differences in perceived
personality-trait relationships and their relation to certain determinants. The
Journal of Social Psychology 65: 233-258.
Phillips, J.J. and P.P. Phillips. 2009. Measuring what matters: How CEOS view
the success of learning and development. Training and Development 63(8): 44.
Ponemon, L. 1990. Ethical judgments in accounting: A cognitive development

48

perspective. Critical Perspectives on Accounting. 1:191-215.
Preacher, K.J, and A.F. Hayes. 2004. SPSS and SAS procedures for estimating
indirect effects in simple mediation models. Behavior Research Methods,
Instruments, & Computers 36(4): 717-731.
Reeder G.D., and M.B Brewer. 1979. A schematic model of dispositional
attribution in interpersonal perception. Psychological Review 86:61-79.
Rotter, J.B. 1980. Interpersonal trust, trustworthiness, and gullibility. American
Psychologist 35:1-7.
Rousseau, D.M., S. Sitkin, R. Burt, and C. Camerer. 1998. Not so different after
all: A cross-discipline view of trust. Academy of Management Review 23: 393404.
Rowe, C. 2004. The effect of accounting report structure and team structure on
performance in cross-functional teams. The Accounting Review 79(4):1153-1180.
Rowe, C. M.D. Shields, and J.G. Birnberg. 2012. Hardening soft accounting
information: Games for planning organizational change. Accounting,
Organizations and Society 37(4): 260-279.
Salterio, S.E and A. Webb. 2006. Honesty in accounting and control: A
discussion of “The effect of information systems on honesty in managerial
reporting: A behavioral perspective.” Contemporary Accounting Research 4(23):
919-932.
Sitkin, S.B. and N.L. Roth. 1993. Explaining the limited effectiveness of legalistic
‘remedies’ for trust/distrust. Organization Science 4:367-392.
Sims, R. 1992. The challenge of ethical behavior in organizations. Journal of
Business Ethics 12:63-73.
Snyder, M. and A.A. Stukas, Jr. 1999. Interpersonal processes: The interplay of
cognitive, motivational, and behavioral activities in social interaction. Annual
Review of Psychology 50: 273-303.
Simons, R. 1995. Levers of Control: How managers use innovative control
systems to drive strategic renewal. Boston, MA: Harvard Business School Press.
Sprinkle, G.B. 2000. The effect of incentive contracts on learning and
performance. The Accounting Review 75(3): 299-326.
Victor, B. and J. Cullen. 1988. The organizational bases of ethical work climates.
Adminstrative science quarterly 33: 101-125.

49

Zand, D. 1972. Trust and managerial problem solving. Administrative Science
Quarterly 17: 229-239.

50