Introduction

Ethical sensitivity (Shawver and Sennetti 2009) and ethical reasoning (Dellaportas 2006) are the common objectives for accounting ethics instruction, but perhaps the most important objective is that, as professionals, accountants develop the confidence to make ethical choices when confronted with ethical dilemmas (Arce and Gentile 2015; Kish-Gephart et al. 2010). Robbins (2012, p. 143) defines ethical confidence as “(t)he courage to exhibit leadership in ethically confusing environments by confronting and engaging with ethical issues publicly and openly.” However, there is a great deal of uncertainty regarding how to effectively teach ethics (McWilliams and Nahavandi 2006) so that future accountants will acquire the necessary skills, possess confidence, and exhibit leadership when faced with ethical issues. After reviewing numerous accounting ethics studies, Bampton and Cowton (2013, p. 558) suggest this uncertainty regarding the effectiveness of accounting ethics pedagogy stems from “little research into actual behavior rather than attitudes” (Robertson 1993) and “the absence of a theoretical framework or explicit hypotheses” (Randall and Gibson 1990; Weber 1992; Robertson 1993). Edwards and Kirkham (2014) posit that it is essential researchers and ethicists evaluate new theories and pedagogical approaches to business ethics.

Despite the uncertainty surrounding how best to teach ethics to ensure accounting professionals develop ethical confidence and competence, the Association to Advance Collegiate Schools of Business (AACSB) Accounting Accreditation Standards require accounting degree programs to establish expectations for ethical behavior and to provide effective learning experiences in ethical reasoning (AACSB International 2013). In addition, several state boards of accountancy mandate that CPA candidates complete an ethics curriculum in the form of identifiable accounting ethics coursework (e.g., California, New York) and/or a Board-approved stand-alone ethics course (e.g., Maryland, Texas). The Pathways Commission suggests that education should “build skills in ethical decision-making and responsible judgment” (American Institute of Certified Public Accountants and American Accounting Association 2012, p. 133). Given these requirements, it is even more important that theories and pedagogical approaches to business ethics instruction are tested.

The purpose of this study is to evaluate the effectiveness of two accounting ethics pedagogies: one based on a more traditional accounting ethics approach and one based on Giving Voice to Values (GVV) (Cote et al. 2011; Gentile 2010). Two cohorts of similar students participated in each type of ethics instruction, and then we observed their behavior when they encountered ethical challenges.

The traditional ethics approach is based on the ethics literature of Abdolmohammadi (2005), Armstrong et al. (2003), Bernardi et al. (2002), Dellaportas (2006), and Granitz and Loewy (2007), as well as the theoretical underpinnings of principled reasoning and universality. Universality refers to accepting a consensus of ethical values, such as trustworthiness, respect, responsibility, fairness, caring, and citizenship (Josephson 2002). The traditional approach includes student exposure to philosophical theories of ethical decision making and ethical decision-making frameworks,Footnote 1 studying the accounting standards of professional conduct, preparing cases in which accounting professionals engaged in unethical behavior (e.g., Enron, WorldCom), and discussing instances in which individuals were whistle-blowers and spoke out against fraud.

The instruction approach, Giving Voice to Values (GVV), is an “action oriented pedagogical approach for developing the skills, knowledge and commitment required to implement values-based leadership” (The Giving Voice to Values Curriculum 2016). With theoretical foundations in the value identification and performative ethics literature (Edwards and Kirkham 2014; Kidder 2005; Sheehan and Schmidt 2015), GVV specifically focuses on skill-building to develop confidence in the face of ethical challenges by identifying values, both one’s own and those others commonly hold, as well as identifying common reasons, rationalizations, and justifications for unethical behaviors. GVV further encourages identifying the best approaches and actions to take when values are challenged. These actions are based on one’s own strengths as well as scripting responses to value challenges.

We investigate the association between the levels of observed unethical behaviors, cheating and being untruthful about cheating, among two different cohorts of accounting students. Our results indicate consistent unethical behaviors within the cohort that participated in the traditional pedagogy, but not in the cohort that participated in the GVV pedagogy. In addition, a debriefing of the students from both cohorts indicates the GVV cohort had confidence in their ethical competence and that of their peers; however, this was not the case for the traditional cohort.

The remainder of the article proceeds as follows: In the next section, we describe accounting ethics education research and the theoretical framework and background for both the traditional and the GVV interventions. We also present our hypotheses and then follow with a discussion of the methodology. Finally, we present our findings and discuss our results, the study’s limitations, and suggestions for future research.

Background and Hypotheses Development

Accounting Ethics Education and Instruction

We focus our study on ethics education in an accounting academic environment. Two recent comprehensive reviews provide background information on the variety of educational methods and pedagogies employed in accounting ethics education scholarship (Bampton and Cowton 2013; Christensen et al. 2016). Loeb (2007) also provides information regarding accounting ethics education approaches in his 18-year retrospective study. He identifies common techniques that support ethics instruction, including pedagogical frameworks, case discussions, experiential methods such as role-playing, readings, lectures, self-reflections, and recognition of moral heroes, though some of these techniques are yet to be investigated empirically (Loeb 2007). More recently, Loeb (2015) posits that active learning strategies are more effective than passive learning strategies.

Christensen et al. (2016) conducted a meta-analysis of 47 studies spanning a 25-year period, which employed accounting students as subjects and used Rest’s (1986) Defining Issues Test (DIT/DIT-2, hereafter DIT) to evaluate the impacts of various study variables, including ethics instruction, on students’ moral reasoning.Footnote 2 Bampton and Cowton (2013) conducted a literature review of accounting ethics scholarship over a 20-year period, including studies that employed accounting students as well as those that utilized other subjects, such as professional accountants. These two reviews focus on research that explored ethics education and instruction with accounting students, which is a subset of both articles. Although the Christensen et al. (2016) meta-analysis focuses on studies that use the DIT only, Bampton and Cowton (2013) note that a significant percentage of studies on accounting students’ ethics use Rest’s DIT (1986).

Bampton and Cowton (2013, p. 556) posit accounting ethics education includes two primary areas of research: “whether accounting students have a higher or lower moral standard than other students” and “whether the teaching of ethics should be an integral part of accountancy courses and, if so, whether it has any long-term effect on moral reasoning.” Within this second area, the focus of this study, Dellaportas (2006) suggests that ethics interventions in accounting curricula encompass either the “discrete” approach (a separate course) or the “pervasive” approach (infused in one or more accounting courses). Dellaportas (2006) found increased ethical reasoning resulted from the discrete approach, which is a separate ethics course. In contrast, Christensen et al.’s (2016) meta-analysis of 16 ethics instruction studies shows that the pervasive approach, i.e., embedding or infusing the instruction, rather than the discrete (stand-alone) course has a significant positive effect on moral reasoning as measured by the DIT regardless of the type of intervention. The accounting student instruction studies in their meta-analysis include Abdolmohammadi (2005), Armstrong (1993), Bancroft (2002), Bernardi et al. (2002), Bonawitz (2002), Buell (2009), Delaney (2005), Dellaportas (2006), Earley and Kelly (2004), Hickman (2008), Klimek and Wenell (2011), Ponemon (1993), Shaub (1994), St. Pierre et al. (1990), Venezia (2005), and Wilhelm (2010). The accounting student instruction studies in the Bampton and Cowton (2013) review include Armstrong (1993), Fulmer and Cargile (1987), Green and Weber (1997), Hiltebeitel and Jones (1992), Jeffrey (1993), Lampe (1996), Loeb (1991), McCarthy (1997), Ponemon (1993), Shaub (1994), and St. Pierre et al. (1990). This large group of studies predominantly explored the effectiveness of introducing ethical frameworks, case discussions and problems, and exposure to the AICPA Code of Professional Conduct.

Taken as a whole, research investigating the efficacy of this wide variety of interventions is mixed (Bampton and Cowton 2013). As examples, Bampton and Cowton (2013) point to Lampe (1996), Ponemon (1993), and St. Pierre et al. (1990), who found no increase in moral reasoning as the result of ethics education, whereas Armstrong (1993) and Hiltebeitel and Jones (1992) did report positive results. Using the case approach with ethical dilemmas, Abdolmohammadi (2005) showed an improvement in graduate accounting students’ ethical reasoning. Martinov-Bennie and Mladenovic (2015) explored the impact of providing a pedagogical framework alone or as a component of an integrated ethics education. In one unexpected finding, increases in students’ ethical judgments were greater if they were not exposed to the framework prior to participating in the integrated ethics education. Green and Weber (1997) reported that students possessed higher levels of moral reasoning after exposure to the AICPA’s Code of Professional Conduct; however, McCarthy (1997) did not find that students’ ethical development was enhanced through this exposure. In sum, Bampton and Cowton (2013, p. 556) conclude that due to the method and ethical content of the interventions differing among the studies and the studies primarily emphasizing capturing attitudes and intentions as opposed to behavior, and “perhaps because of inadequacies of research design,” it is difficult to determine which interventions are effective.

Hassan et al. (2014) contrasted attitude, intention, and behavior in their review of ethics consumption research. They noted that this research has “identified gaps between what consumers think, what they intend and what they actually do” (2014, p. 220). Within the literature, two theories, the theory of reasoned action (Fishbein and Ajzen 1975) and the theory of planned behavior (Ajzen 1985, 1991), dominate the discussion of attitudes (what an individual thinks), intention (what an individual intends), and behavior (what an individual actually does). The theories suggest that a person’s behavior stems from his or her intention to act (behave in a certain manner) and that the intention is influenced by his or her attitude toward the behavior. However, as noted, Hassan et al. (2014) highlight a gap identified in ethics consumption research, which suggests that what an individual intends to do, as influenced by his or her attitude, is a weak predictor of behavior. Similar to Bampton and Cowton (2013), the authors highlight a lack of research that measures what individuals actually do, or their actual behavior (Hassan et al. 2014).

Sims and Felton (2006, p. 297) suggest that “no agreement exists among business schools and their faculties regarding what should be the context and pedagogy of a course in ethics.” This sentiment, in addition to the theoretical and methodological concerns of Bampton and Cowton (2013), supports the need for continued research incorporating Bampton and Cowton’s suggestions as well as Edwards and Kirkham’s (2014) call for testing ethical theories and pedagogies.

The Traditional Curriculum

Velasquez et al. (1987) discuss Socrates’ belief that ethics is knowing “the right thing to do” when confronting a situation in which ethical issues are at play, and most importantly, that this knowledge can be taught. Kohlberg (1984) also posits that an individual’s ability to deal with moral issues is not formed right away but develops over time, and as such, it is influenced by educational efforts. With this in mind, Hasnas (2013) describes two approaches that are traditionally used to teach business ethics and notes that most curricula incorporate aspects of both. The first employs a philosophical perspective that “derives guiding normative principles from abstract theories of philosophical ethics” (Hasnas 2013, p. 275). The other approach incorporates an atheoretical perspective in which students analyze cases that involve ethical dilemmas and recommend solutions. A combination of both of these approaches often includes an ethical decision-making framework, which integrates philosophical perspectives as guides to assist students in analyzing cases (see, for an example, “A Framework for Ethical Decision Making” at the Markkulla Center for Applied Ethics 2016). As another example specific to accounting contexts, the first two chapters of Mintz and Morris’s (2014) text on ethical decision making in accounting describe modern moral philosophies and an ethical decision-making framework that integrates Rest’s Model. These chapters initiate a discussion of the AICPA Code of Professional Conduct and present an extensive list of cases centered on accounting topics, such as earnings management. Further, accounting ethics research indicates that faculty teaching ethics predominantly employ ethical decision-making frameworks, case discussions and problems, and instruction related to the AICPA Code of Professional Conduct (cf. Armstrong et al. 2003; Granitz and Loewy 2007; Green and Weber 1997).

The traditional curriculum at our university involves a mixture of theoretical and atheoretical approaches in which students are introduced to ethical theories, such as the utilitarianism and deontological perspectives, and then discuss paradigmatic cases (including both “villains” and “exemplars”), ethical decision-making frameworks, and the AICPA Code of Professional Conduct. Our faculty members employ resources that focus on ethical theories and ethical decision-making frameworks, which include those available at the Markkula Center for Applied Ethics and the Josephson Institute of Ethics in particular (Josephson 2002). Josephson (2002) emphasizes embracing universality, or examining issues from a universal point of view, in which universality refers to a consensus of ethical values, such as trustworthiness, respect, responsibility, fairness, caring, and citizenship.

In contrast to the traditional approach, Treviño et al. (2014) advocate research investigating action-oriented ethics instruction, such as the GVV curriculum with its emphasis on the development of individual voice efficacy and confidence. The GVV initiative, created by Mary Gentile, prepares students and business professionals to respond effectively to ethical challenges faced in the workplace (Gentile 2010). A series of interviews of MBAs conducted by the Aspen Institute revealed that while the individuals could recognize ethical challenges, could relate ethical theories, and could even describe step-by-step approaches to resolving ethical conflicts, they were unable to determine the appropriate course of action to respond effectively and believed that their business education did not prepare them to do so (Gentile 2010). Thus, the GVV curriculum addresses this shortcoming. Approximately, 600 education and business settings, including Lockheed Martin, Walmart, General Mills, Prudential, Northrop–Grumman, General Dynamics, Mayo Clinic, Kaiser Permanente, and The World Bank, employ the GVV curriculum (The Giving Voice to Values Curriculum 2016).

The GVV Curriculum

The GVV curriculum adopted by our university replaces the traditional curriculum with exercises, short cases, teaching strategies, and annotated bibliographies that are available to all faculties, with flexibility for instructors to choose among the many resources to structure what works best for their classrooms. Three primary themes are present in the program: (1) Understanding Yourself, (2) Understanding Others, and (3) Preparing to Respond Effectively (Cote and Latham 2016). The curriculum includes the identification of values, both one’s own and those commonly held by others, and the identification of which communication style and approach works best for each individual. Action-oriented learning exercises provide guidance in strategizing approaches to take action in a values challenge based on one’s own strengths, which helps develop individual voice efficacy and confidence. The process of identifying common reasons and rationalizations used to support unethical behavior helps individual’s script responses to values challenges. A key characteristic of GVV is that it is post-decision making; that is, it assumes an individual has already identified a challenge to one’s values exists and wishes to develop the optimal personalized strategy to respond. A second key underpinning of GVV is the interconnectedness of the instruction with one’s peers. A significant portion of the exercises involve communication with others about one’s own personal values, perspectives, and potential strategies as well as soliciting feedback. Gonzalez-Padron et al. (2012, p. 255) refer to this GVV component as reciprocity, which is the “notion of learning from others as much as from the facilitator of the training.” Over time, it evolves into a community of understanding among peers. A third component in the instruction is the understanding that ethical challenges should be normalized. They are common occurrences, and repeated practice in safe settings will better prepare individuals to address them when they occur. As Arce and Gentile (2015, p. 535) suggest, “(t)he internalization of GVV’s constructs allows faculty to comfortably raise and endorse ethics as part of the natural order of business decision-making because the stakes of doing so have been normalized.”

Theoretical Support and Empirical Evidence

The GVV curriculum draws its theoretical foundation from the values identification or clarification and the performative ethics literature (Edwards and Kirkham 2014; Kidder 2005; Sheehan and Schmidt 2015). Sheehan and Schmidt (2015, p. 186) note empirical evidence that shows personal values are fairly constant beliefs influencing one’s behavior and there exists “a positive link between an individual’s values and his or her ethical behavior.” The authors conducted a retrospective review beginning in the 1960s on the values clarification literature whose roots reside in education and state that though “interest in values clarification in education is currently off its peak, there is still interest in using value clarification tools” (Sheehan and Schmidt 2015, p. 187). They further note scholars who suggest that increasing students’ awareness of their values is a first critical phase in ethics instruction. Sheehan and Schmidt (2015) provide positive empirical evidence from a values clarification exercise involving senior-level undergraduate accounting students to support their contention. Values identification represents a key introductory component in the GVV curriculum, which is supported by values clarification theory (Kirschenbaum 1976).

Edwards and Kirkham (2014) explore the belief that action-oriented ethics instruction operates within a performative space. They define performative ethics as “the ability for discourse and conversation to produce new ethical realities and, in turn, be guided by that productive process” (Edwards and Kirkham 2014, p. 64). Maclagan (2003) advances the idea that performative ethics help individuals respond to ethical challenges effectively through developing the communication abilities and confidence needed when encountering conflicts. The theory proposes that holding a dialogue to draw out beliefs and align expectations increases the potential of future ethical action. Tudway and Pascal (2006, pp. 107–108) suggest that this discourse aligns “what we say with what we do.” Edwards and Kirkham (2014) postulate that the GVV curriculum resides in performative ethics with the interconnectedness of the instruction with one’s peers, the sharing of past experiences, the development of understanding one’s own and others’ values through discourse, and the emphasis on repeated voice practice in preparation for future action. In describing GVV, they state it is “…one of a number of business ethics theories that focus on the expression of values within the context of organizational culture, social communication and the analysis of moral disengagement and rationalis(z)ations…” (Edwards and Kirkham 2014, p. 493). Salvador and Folger (2009) suggest that ethical decision making is not just another decision-making process as emotion plays a critical role in some cases. Reynolds et al. (2010) propose a two-cycle process in which a pattern match is attempted first when encountering an ethical dilemma. In effect, the GVV approach creates the pattern match through normalization and practice.

Given the relative newness of the curriculum, there are a limited number of empirical studies that measure the impact of GVV. Cote and Latham (2016) investigated the impact of a peer-to-peer approach by employing upper-division Beta Alpha Psi accounting students and teaching them how to use the GVV curriculum to instruct and prepare introductory accounting students to manage ethical challenges they will face in the workplace. They used a pre- and post-GVV instruction assessment of students’ perceptions of their skills and abilities to respond to ethics challenges. Five constructs tied to the GVV themes emerged from the assessment: Ability to Influence Others, Empathy and Appreciation of Others’ Differences, Knowledge of Self, Knowledge of Others, and Speaking-Up Abilities. The mean comparisons across the constructs pre- and post-instruction indicated statistically significant higher post-scores, which showed that students perceived they possessed ethical confidence after the instruction. Shaw (2013) also identified a positive statistical difference in MBA students’ perceptions of their ability to address ethical challenges as a result of their instruction in a course incorporating the GVV curriculum.

Sheehan and Schmidt (2015) employed a values identification exercise, which they suggested serves as a successful introduction to the remaining GVV framework. Almost all of the accounting students subjects (97%) found the exercise served as a meaningful reflective way to discover their own values and advocated faculty at other universities adopt it. Ingols (2011) presented positive findings related to a Simmons College experiment. After implementing GVV across the university curriculum in response to student performance on ethics essays related to their Assessment of Learning (AOL), they found 39% more students were able to link ethics and values to recommendations for action (Ingols 2011).

With its specific exercises that help guide the identification of one’s own and others’ values, create a shared space, develop an understanding of common reasons and rationalizations used in ethical challenges, determine one’s own optimal communication style, and script responses to challenges, GVV develops students’ voice efficacy and confidence regarding ethical challenges. Therefore, the GVV curriculum develops students’ personalized toolkit via repeated practice, thus increasing the propensity for individuals to effectively respond to ethical dilemmas.

McCabe et al. (2012) speculate that students enter college believing that cheating and/or dishonesty is not as prevalent as it might have been in their high school experiences, and thus they do not plan to engage in the behavior; however, once they learn it occurs, they are more likely to participate in unethical behavior, rationalizing it as a means to survive in the culture. We posit that the shared understanding of values as well as the stronger voice efficacy, which results from the GVV curriculum, contributes to students refraining from unethical behavior. As such, we investigate the association between levels of observed unethical behavior and types of ethical instruction using the following hypothesis:

H1

Levels of observed unethical behavior are lower for accounting students who participate in GVV instruction than that of accounting students who participate in traditional ethics instruction.

Other Variables in Accounting Ethics Research

In addition to ethics instruction, accounting student ethics research explores several individual characteristics to determine whether they are related to ethical (unethical) behavior: cognitive moral development, age, gender, political ideology, major (i.e., whether students majoring in accounting possess more or less moral reasoning than other students), and university class level (Bampton and Cowton 2013; Bernardi and Bean 2008; Christensen et al. 2016). The subjects in this study are all senior-level accounting majors; hence, we limit our discussion to the first variables highlighted. Thus, cognitive moral development, age, gender, or political ideology may explain decisions to engage in ethical or unethical behavior rather than or in addition to type of ethics instruction.

Cognitive Moral Development

Kohlberg’s (1969) theory of cognitive moral development (hereafter, CMD) describes the progression and changes in how individuals think about behavior in the face of ethical dilemmas as they age. Kish-Gephart et al. (2010) note the theory emphasizes individuals’ moral reasoning processes and justifications in ethical dilemmas rather than the final decision, and they define five stages of development. At the highest stage, which is five, an individual utilizes sophisticated reasoning incorporating ethical principles and considers the societal impact, whereas in lower stages, an individual employs a self-interested or punishment avoidance orientation in the face of ethical dilemmas. The authors note that most adults operate at the conventional level three in which ethical decisions are influenced by expectations of peers or level four in which ethical decisions are predominantly rule-based.

The Defining Issues Test, or DIT, developed by (Rest 1979, 1986) is arguably the most widely used measure of CMD, or moral judgment (Bailey 2011; Kish-Gephart et al. 2010). The test provides vignettes incorporating ethical dilemmas and asks the test-taker to evaluate what one should consider in rank order when deciding how to behave. Rest (1979, 1986) presents a four-component model to describe the totality of moral function: moral sensitivity (stage 1), moral judgment (stage 2), moral motivation (stage 3), and moral character (stage 4) with CMD, as measured by the DIT’s P score, focusing on the second stage. Cote et al. (2013, p. 119) note that the “(l)iterature generally supports a positive relationship between CMD and ethical judgment, yet less clear evidence exists concerning the relationship between CMD and ethical action.” They did not find a statistically significant relationship between CMD and an ethical reporting choice. However, Kish-Gephart et al.’s (2010) meta-analysis, which incorporated 19 studies employing the DIT P score, showed a positive (negative) relationship between CMD and ethical (unethical) intention or behavior. As such, we propose the second hypothesis:

H2

Levels of observed unethical behavior by accounting students are lower for individuals with higher moral reasoning as evidenced by higher P scores.

Age

Prior literature shows that advancement in ethical decision-making abilities occurs as individual’s age and acquires instructive experiences in life (Kohlberg 1984; Rest 1986). As such, age is a common variable of interest in ethics intervention research, yet the results are inconsistent (see for example, Ford and Richardson 1994 and Treviño 1992). Kish-Gephart et al.’s (2010) meta-analysis, which investigated antecedents of unethical choices in behavioral ethics research studies, indicated no age effect. Christensen et al.’s (2016) meta-analysis included 14 accounting ethics research studies involving students’ age as a variable (Abdolmohammadi 2005; Abdolmohammadi et al. 2009; Armstrong 1984, 1993; Bancroft 2002; Buell 2009; Cote et al. 2013; Dellaportas 2006; Hickman 2008; Klimek and Wenell 2011; Shaub 1994; Thorne 1999, 2001; Venezia 2005). The authors suggest “we should witness advancement in moral reasoning not within a semester or possibly even an academic year, but over a longer span of time” (Christensen et al. 2016, p. 147). In addition, they did not find that age was a significant factor; however, they noted a significant difference in moral reasoning scores between lower division students and seniors. Christensen et al. (2016) classified university students as traditional and non-traditional ages, which is commonly defined as students older than 25 (Pelletier 2010), and called for future research investigating the impact of education on moral reasoning for both traditional and non-traditional students. Pelletier (2010) provides the statistic that 47 percent of students enrolled in United States colleges and universities are older than 25. Based on the majority of evidence to date, we do not anticipate age in our study of senior-level accounting students to be associated with unethical behavior; however, we explore whether traditional versus non-traditional age groups vary. The theory of cognitive moral development suggests the older, non-traditional age group would be less likely to engage in unethical behavior (Kish-Gephart et al. 2010). Thus, our third hypothesis is:

H3

Levels of observed unethical behavior by accounting students are lower for non-traditional students than traditional students.

Gender

Bampton and Cowton (2013) identified a gender effect, citing some empirical studies indicate higher moral reasoning from female accounting students (e.g., Bernardi and Bean 2008; Shaub 1994), while others find no statistical difference (e.g., Rogers and Smith 2001; Stanga and Turpin 1991). Christensen et al. (2016) included 20 studies in a meta-analysis that compared the moral reasoning scores of females versus males (Abdolmohammadi 2005; Abdolmohammadi and Baker 2008; Abdolmohammadi et al. 2009; Armstrong 1984, 1993; Bernardi 1995; Bernardi et al. 2002; Brown-Liburd and Porco 2011; Buell 2009; Christensen et al. 2010; Cote et al. 2013; Douglas and Schwartz 1998; Eynon et al. 1996; Klimek and Wenell 2011; Ponemon 1993; Shaub 1994; St. Pierre et al. 1990; Thorne 1999, 2001; Venezia 2005). Although the correlation effect size was statistically different with females scoring higher than males, the authors noted the heterogeneity between studies is significant, indicating the potential of other factors impacting the results.

Borowski and Ugras’ (1998) meta-analysis of 48 studies indicated that gender (female) is significantly related to ethical behavior. Kish-Gephart et al.’s (2010) meta-analysis also showed a greater frequency of unethical choices for males than for females, but the correlation was weak. Further, their findings indicated these gender differences disappeared when individual psychological traits were controlled. There are also studies that show no effect for gender in relation to ethical behavior (Gupta et al. 2011). Based on these collective findings, we anticipate that there may be a marginal gender effect. Thus, our fourth hypothesis is:

H4

Levels of observed unethical behavior by accounting students are lower for female students than male students.

Political Ideology

Christensen et al. (2016) analyzed five accounting student ethics research studies that explored the relationship between political ideology and CMD as measured by the DIT (Eynon et al. 1996; Fisher and Sweeney 1998, 2002; Klimek and Wenell 2011; Sweeney and Fisher 1998). The results suggest that accounting students who classify themselves as politically liberal have significantly higher DIT scores than those who identify as politically conservative. However, they noted that there is a debate associated with scaling issues related to the measurement of political attitudes, and as such, the findings should be interpreted cautiously (Bernardi et al. 2004).

In a study of 93 government budget officials, Reck (2000) found a statistically significant relationship between DIT scores and political ideology but no significant relationship between budget allocations, a moral judgment issue, and DIT scores. She did find a marginally significant (p = .0794) relationship between budget allocations and political ideology. In a study of 286 Flemish-speaking consumers in Belgium, Van Kenhove et al. (2001) found a significant link between political ideology and ethical beliefs regarding benefiting actively or passively from illegal activities for some political party preferences but not others. Kjærnes (2012) argues that ethical choices related to purchasing food closely align with political ideology. These studies suggest a potential link between political ideology and ethical choices. Thus, our fifth hypothesis is:

H5

Levels of observed unethical behavior by accounting students are lower for students who self-identify as politically liberal, than students who self-identify as politically conservative.

Research Methods

We investigate the impact of traditional and GVV ethics instruction in an academic setting using a between-subject design and by collecting student responses to an ethical challenge for the traditional and GVV ethics instruction cohorts. The semester-based university is an urban, commuter campus where the average student age is 26, and most students are employed while attending school. We collected data from a senior-level auditing course for two student cohorts following either the implementation of traditional ethics instruction or GVV instruction. There was no participant overlap between the two cohorts. The same faculty member, hereafter referred to as the ethics instructor, taught the two cohorts and implemented the ethics instruction methods in four courses, including the senior-level auditing course. The other three courses were Intermediate Accounting I and II, and Accounting Information Systems and Internal Control.

The study university’s College of Business faculty members employ several approaches supported by the literature (see, for example, Granitz and Loewy 2007) to encourage academic integrity. These approaches remained constant between the two cohorts. Upon admission to the college, students participate in an orientation and affirm their commitment to the academic integrity code of conduct. This code, as well as the zero-tolerance consequences of violating it, is clearly articulated in all syllabi. The ethics instructor utilizes an honor statement on all exams and assignments. Granitz and Loewy (2007) suggest faculty members should act as role models and promote good professional behavior. The ethics instructor served on the university’s Academic Integrity Task Force, which develops the code of conduct and academic integrity processes, as well as on the Academic Integrity Hearing Board throughout the entire study. She also conducted ethics workshops for graduate teaching assistants as well as for CPAs. She described these roles to students and emphasized the belief that academic integrity and behaving ethically are the cornerstones of education and professional life. The students were required to subscribe to the Wall Street Journal, take weekly quizzes, and participate in discussions of current events with an emphasis on the accounting profession, including real-world ethical violations with serious consequences.

We describe below the details of the two forms of ethics instruction, traditional and GVV, leading up to the time of the experiment. The instruction supports the learning objective in the courses, which states, “identify and evaluate an ethical issue and support a recommendation concerning the issue.” The first class session in which ethics was presented for both cohorts required a 50-min time commitment with student preparation in advance. Subsequent case discussions and exercises occurred within the context of the technical topics covered during class.

Ethics Instruction: Traditional Cohort

As described previously for the traditional cohort and summarized in Appendix 1, we implemented traditional ethics education in the accounting program in four core courses at the junior and senior levels. The traditional ethics education approach consisted of several practices supported in the literature, such as the presentation of ethical theories and paradigmatic cases and the use of ethical decision-making frameworks (see, for example, Armstrong et al. 2003).

In Intermediate Accounting I, during the second class session covering the standards setting process and public interest, the ethics instructor presented ethical philosophies, such as utilitarianism, consequentialism, deontology, relativism, and virtue ethics, and followed with a practice case discussion emphasizing steps one might take in ethical decision making. The case appears in the Intermediate Accounting text and focuses on rule-making when an individual confronts opposition to early adoption of a standard that would adversely affect the net income for the year (Kieso et al. 2011, 2005). In the remainder of the semester and in Intermediate Accounting II, students completed four additional ethics case analyses (two each semester). The vignettes were taken from the Intermediate Accounting text and presented ethical dilemmas involving chapter topics, e.g., inventory reporting or compensation plans (Kieso et al. 2011, 2005).

The Accounting Information Systems and Internal Control class incorporated control environment failures, ethical violations, and whistleblowing through class discussions on Enron, WorldCom, and Phar-Mor. Students completed an ethics essay in an examination, which required them to detail what they would do in the face of an ethical dilemma, such as one faced by an individual in a situation like Phar-Mor.

In the Auditing class, in the first three weeks of the semester leading up to the study of professional ethics, students completed the DIT as well as readings from the Josephson Institute of Ethics’ Six Pillars of Character (Josephson 2002). The DIT results served as a vehicle to discuss moral reasoning during discussions on the AICPA Code of Professional Conduct and professional ethics in the fourth week of class. The students used the six-step approach to resolve ethical dilemmas in the Auditing textbook and employed it to address ethical dilemmas as homework (Arens et al. 2014, 2006). The students also completed an ethics case. The homework and ethics case analysis had no points specifically assigned to them; however, student participation was rewarded.

Ethics Instruction: GVV Cohort

As noted earlier for the GVV cohort and summarized in Appendix 2, we implemented the GVV ethics education approach in the accounting program in four core courses at the junior and senior levels, with Intermediate I again as the starting point. The ethics instructor followed the guidance provided by Cote et al. (2012) for integrating GVV into the curriculum. She structured the instruction into three segmentsFootnote 3:

  • Segment 1: Setting the Stage. Students are assigned the GVV reading, due in the second class week of the first Intermediate semester, “Ways of Thinking about Our Values in the Workplace,” which appears on the GVV website in the Foundation Readings and Exercises Module. Cote et al. (2012, p. 14) noted that “(t)he reading establishes the curriculum, provides research supporting its effectiveness and key pillars.” During the class session, the instructor used the two-page “An Action Framework for Giving Voice to Values—To-Do List” to review the key concepts from the reading, including developing a list of commonly held values. The instructor then utilized the “A Tale of Two Stories” foundational exercise in which student groups with two to four members were asked individually to first recall a situation in which their values were challenged, and they answered the challenge by speaking up, and then to recall a second circumstance in which they did not speak up. As a class, the students developed a list of enablers, including organizational, situational, and individual factors that encouraged them to speak up as well as a list of dis-enablers that made it difficult to do so. The instructor expanded on this exercise with a short practice case discussion, “Profit Maximization and Layoffs,” which introduced the GVV discussion questions. Students used these questions in all subsequent assigned cases in the text, current event discussions, and examinations.

    1. 1.

      What are the main arguments you are trying to counter? Specifically, what are the reasons and rationalizations that you need to address?

    2. 2.

      What is at stake for the key parties, including those with whom you disagree?

    3. 3.

      What levers can you use to influence those with whom you disagree?

    4. 4.

      What is your most powerful and persuasive response to the reasons and rationalizations that you need to address?Footnote 4

  • Segment 2: Enhancing Self-Knowledge. Students were assigned the GVV “Personal–Professional Profile” from the Self-Knowledge and Self-Assessment Module due in week 7. This exercise required students to provide a personal assessment of commonly held values as well as which values are most important to them.

  • Segment 3: Enhancing Ability to Script. Students completed “Reasons and Rationalizations: An Exercise” and “Guidelines for Peer Coaching” from the Scripts and Skills Module with a partner prior to class and individually practiced scripting their responses to the GVV case “The Part-Time Job with a Full-Time Challenge,” both due in week 10. As the final case assignment of the semester, students scripted their responses to an ethics case in Kieso et al. (2011, 2005), which was due in week 12.Footnote 5

Key to GVV instruction, as assignments were turned in, the class discussed their approaches to develop shared understanding, communication styles and successful practices for responding. The ethics instructor also continued the same approach in Intermediate II, Accounting Information Systems and Internal Control, and Auditing as part of the curriculum described previously, including the use of the GVV questions on the examination. For example, in Intermediate Accounting II, students employed scripting as they completed two additional ethics case analyses from the Intermediate Accounting text as homework assignments. As with the traditional cohort, the Accounting Information Systems and Internal Control class incorporated control environment failures, ethical violations, and whistleblowing through class discussions on Enron, WorldCom, and Phar-Mor. Students completed an ethics essay during an examination that required them to script what they would do in the face of an ethical dilemma, such as the one faced by an individual in a situation like Phar-Mor. By the third week of the semester in Auditing, the cohort had practiced scripting responses to ethical dilemmas eight times and shared those responses with their peers.

Restricted Access Trial

We collected data from the traditional and the ethics cohorts at two points during the semester in which the Auditing class was taught. In the first class session, the first data collection point, subjects were assigned a code and completed the Defining Issues Test. In the third week, students accessed the course management website to obtain a brief description of moral reasoning to assist them in interpreting their own DIT scores, which they received in the next class session, along with a discussion of the AICPA Code of Professional Conduct and professional ethics. Students then completed a case analysis of Leigh Ann Walker, Staff Accountant (Knapp 2013, 2006) prior to the next session.

The case involves the true story of a young woman (pseudonym “Leigh Ann Walker”) who recently joined a public accounting firm after graduation. While on an audit, her supervisor asks her if she has taken the CPA exam yet and offers to share study materials if she has not. Although she has taken the exam, Leigh Ann lies to her supervisor, fearing she did not pass and hoping to avoid future embarrassment. Leigh Ann does pass the exam and informs the supervisor. The supervisor reports Leigh Ann’s lack of integrity, her lie, to firm partners. Leigh Ann is let go from the firm due to a concern over her trustworthiness. The ethics instructor indicated that this case generated lively discussions in class, with student sentiments ranging from shock that someone would be fired over a “white lie” to support for the action taken by the firm.

The students face a decision at this juncture as well as in the class session, which was the second data collection point. On the course management website, the solutions to Leigh Ann Walker, Staff Accountant were placed in the same section as the moral reasoning document. The ethics instructor posted an explicit statement, in bold and italics, that students should not access the solutions prior to the class discussion of the case. The instructor utilized a statistical tracking feature in the course management software on all posted materials and turned it off at the beginning of assignment deadlines. Although students received instructions not to access the solutions, the tracking feature was turned on for the Leigh Ann Walker, Staff Accountant case solutions prior to the class session and was turned off directly before the class session actually began. During the class session, identifying themselves by their code only, students answered a series of questions on the assignment, including whether or not they followed the restricted access instructions. Subsequent to the class session, an authorized staff person downloaded the tracking information and replaced student names with their codes so that the ethics instructor had no knowledge of student identity in the trial. In each cohort, students were debriefed in a subsequent class session by the same faculty member, who was not the ethics instructor.

It should be noted that the ethics instructor’s intent in the posting was not to encourage students to behave unethically, but to simulate a warning students may face in their professional lives in which accountants often have access to confidential information. She assumed, given the substantial prior ethics instruction, the students would follow the posted instructions, but if they did not, they would be truthful when asked anonymously and would not offer a “white lie,” similar to Leigh Ann, the very case they were assigned. The ethics instructor’s intention was that the exercise would provide evidence of professional growth by the students in the face of a values choice. The results from the traditional cohort prompted the ethics instructor and other faculty members to explore the effect of adopting the GVV action-oriented ethics instruction and to conduct a between-subjects’ comparison of the two instruction methods. In the first study, 89% of students who participated in traditional ethics instruction accessed the unauthorized material, and 96% were not truthful about their behavior (Christensen et al. 2010).

Dependent Variable for Testing Hypotheses 1–5

In both the traditional and GVV cohorts, we measured (un)ethical behavior by tracking whether each student followed the instructions given (whether he/she accessed the unauthorized or restricted material) and whether or not he/she was truthful when asked about his/her behavior (lied/did not lie). For the GVV cohort, none of the non-traditional and none of the females accessed the case materials; thus, we did not use accessing the materials as a dependent variable. Hence, we only used whether or not the individual told the truth about accessing the unauthorized material as the dependent variable.

Independent Variables for Testing Hypotheses 1–5

We coded instruction type as traditional = 0 and GVV = 1 (H1). We used subjects’ responses on the Defining Issues Test to capture moral reasoning as measured by DIT P scores (H2), age (H3), gender (H4), and self-identification of political ideology (SIPI) (H5). Traditional age (under 26) was coded as 0 and non-traditional age as 1. Gender was coded as female = 0 and male = 1. The SIPI coding is a 1–5 scale in the DIT, where 1 = very liberal and 5 = very conservative. We perform a logistic regression analysis to test our five hypotheses.

Results

The traditional cohort and GVV cohort consisted of 28 accounting students (57% females, 43% male) and 34 accounting students (56% female, 44% male), respectively. The average age of the participants was 33.00 for the traditional instruction cohort and 29.71 for the GVV cohort, which reflects a slight decrease in non-traditional students for the latter cohort. The two cohorts were similar in terms of SIPI (political ideology) and average P score (moral reasoning). Table 1 provides the descriptive statistics, and Table 2 provides the correlation matrix. Not surprisingly, there was a significant positive correlation between non-traditional students and age. We found significant negative correlations between GVV instruction and accessing the restricted case, not being truthful (lying) about access, and between the SIPI and P scores. The significant negative relationship with the DIT P score translates to lower levels of SIPI (more liberal) being associated with higher P scores (higher moral reasoning).

Table 1 Descriptive statistics
Full size table
Table 2 Correlation matrix
Full size table

Table 3 provides the frequency counts of the restricted access trial by cohort and further classification based on demographic variables, gender, and traditional/non-traditional age. Only one student in the GVV cohort accessed the case and was untruthful about that access. Three students in the GVV cohort said they accessed the case when they had not, a result potentially stemming from not understanding the question when the faculty member posed it (i.e., did you access the Leigh Ann Walker, Staff Accountant case solutions?) or not paying attention. Table 4 provides our logistic regression results with not being truthful about access (lying) as the dependent variable.

Table 3 Restricted access trial frequency counts
Full size table
Table 4 Logistic regression results
Full size table

H1 states that the levels of observed unethical behavior by accounting students are lower for students participating in GVV instruction than that of students participating in traditional ethics instruction. Our results support H1; that is, GVV instruction shows a positive relationship with ethical behavior (B = 4.797, p = .000, Exp(B) = 121.159). H2 predicts that higher levels of moral reasoning as evidenced by higher P scores are associated with lower levels of observed unethical behavior. Although the direction of the association is as anticipated, the results are not statistically significant and do not support H2 (B = −.031, p = .379, Exp(B) = .969), indicating a lack of relationship between moral reasoning as measured by the DIT P score and the (un)ethical action taken by the students. H3 postulates that levels of observed unethical behavior by accounting students are lower for non-traditional students than traditional students. Again, the direction of the association is as predicted; however, the results are not statistically significant, and H3 is not supported (B = −376, p = .700, Exp(B) = .687).Footnote 6 H4 posits that levels of observed unethical behavior by accounting students are lower for female students than male students, and it is not supported (B = −.823, p = .369, Exp(B) = .439). Finally, H5 states that levels of observed unethical behavior by accounting students are lower for students who self-identify as politically liberal than for students who self-identify as politically conservative. H5 is supported (B = −1.279, p = .015, Exp(B) = .278).

Discussion and Implications

Prior research contends that academic dishonesty correlates with future unethical workplace behavior (Curtis and Williams 2014; Shawver and Shawver 2013). Thus, assessing approaches to reduce unethical student behavior is critical to the profession. Our primary finding is that the type of ethical instruction matters; that is, students who participate in GVV ethics instruction respond to an ethics “pop quiz,” a term referred to in Christensen et al. (2010), in a dramatically different manner than students who participate in traditional ethics instruction.

Comparing the tone and content when each cohort was debriefed provides insights into the perspectives each brought to the situation. In both cohorts, the discussion was engaging, and students were not hostile or aggressive. The faculty member who conducted the debriefing was not the course instructor. Using a third-person perspective, students were asked, “Why do you think a student would look at the file when expressly instructed not to?” In the traditional cohort, the tone was somewhat defensive. Commonly provided responses were as follows:

  • “Perhaps they accidently clicked on the link.”

  • “Student did not see the instructions.”

  • “They looked because they wanted to check answers so that they would not be embarrassed in class discussion.”

  • “Student did not want to disappoint the professor with incorrect answers.”

  • “They assumed everyone would look, so they needed to have the same access as other students.”

In the traditional cohort, there was concern that the professor would now perceive the class differently. They were reassured that this was not the case, and care was taken to discuss this as a mistake in judgment rather than a flaw in ones’ character. However, the tone of their comments implied a gap in their knowledge of peers’ thoughts or potential behavior.

The tone and content of the debriefing discussion differed with the GVV cohort. Students used GVV terminology to describe their thoughts and feelings. The comments below characterize the predominant views expressed during the debriefing when asked the same question:

  • “Trust among our peers and group members is important.”

  • “How my actions might impact others and their views of me are important.”

  • “The respect of the professor is important to us.”

This debriefing session provides insight into the impact curriculum-embedded GVV instruction has on students. Their comments indicate an evolution in reasoning that emphasizes the importance of a community of understanding among peers as key to decision making. They expect ethical behavior from their peers. The GVV cohort expressed confidence, not only in their own ability to act ethically but also in the ability of their peers. The emphasis within the GVV curriculum on the importance of shared values and making an effort to understand the views of others results in a culture in which academic honesty is perceived as a norm.

Weaving the GVV framework throughout these four accounting courses has demonstrated the efficacy of universality (Josephson 2002) and reciprocity (Gonzalez-Padron et al. 2012) as critical elements in the evolution of students’ ability to address ethical challenges. With active learning where peer-to-peer sharing occurs, students recognized the commonality among their values. Through repetition using a variety of exercises and small group discussions, a community evolved where the culture of common values and expectations resulted. McCabe et al. (2012) noted that when students witness cheating behavior in college, they conclude it must be widespread and therefore necessary for them to survive in the academic environment. An academic intervention, such as the GVV framework, might have the capacity to normalize positive ethical values if introduced early in the academic career.

The lack of significance between the DIT P score (moral reasoning) and behavior supports the argument that the P score captures an earlier stage in the Rest model (1979, 1986) rather than an ethical action. Bailey et al. (2010) suggest that the mixed findings related to the DIT P score may be tied to researchers employing it to capture stages other than moral reasoning in the ethical decision process. Our study results are consistent with Chan and Leung (2006), Cote et al. (2013), and Patterson (2001). We find no difference between the two cohorts on moral reasoning as measured by P scores, regardless of the instruction type, suggesting that embedded GVV instruction is not associated with levels of moral reasoning in a different way than traditional ethics instruction. This finding is consistent with GVV being a post-decision-making tool. Students first identify the values challenge (i.e., a component of moral reasoning) and subsequently take action. However, the behavior that students exhibited in that subsequent action was significantly different between the two cohorts.

Our results also suggest that age (traditional versus non-traditional) and gender are not related to the level of observed unethical behavior. An additional sensitivity analysis conducted by Kish-Gephart et al. (2010) indicated that demographic variables, such as age and gender, became insignificant and were dominated by psychological variables in their study. Christensen et al. (2016, p. 157) suggest “that gender may be interacting with other factors to produce their observed results.” Unlike gender and age, there was a significant negative (positive) relationship between self-identified liberals (conservatives) and the level of observed unethical behavior. We note relatively the same percentage of liberals versus conservatives between the two cohorts, yet almost all of the cohort that underwent the GVV instruction behaved ethically, providing additional evidence for the positive impact of the approach regardless of political views.

Research exploring ethics instruction in an accounting curriculum is sometimes criticized for assessing attitudes and perceptions rather than observing behaviors, though the challenges of the latter are acknowledged (Bampton and Cowton 2013). Our study contributes to the literature by addressing this criticism through an exploration of the impact of incorporating two different ethics instruction approaches in an accounting curriculum. For similar students in two cohorts, we find dramatically different and encouraging results in developing ethical confidence related to embedding an action-oriented ethics framework.

Our study has limitations, and in calling attention to them, we provide suggestions for future research. The generalizability of the findings is limited by sample size. However, the statistically significant result with a robust effect size related to the type of instruction suggests that assessing GVV’s impact using a similar comparison of traditional and GVV instruction methodology in other educational settings is a promising avenue for future research. In addition to academic environments, the education settings could include ethics instruction conducted in public accounting firms. Furthermore, we do not capture measures of behavior taken over longer periods of time, e.g., subsequent years, following the ethics instruction. Longitudinal studies that capture how people use the GVV framework over time could also have considerable potential. Cote and Latham (2016) suggest that students are unaware of the many gray areas they will need to navigate in their professional lives. A note from a recently graduated accounting student who participated in GVV instruction illustrates this point and provides support for the curriculum’s continuing impact. Permission was obtained from the former student to include this excerpt from the note:

…I can’t begin to tell you all the ethical issues I have faced in the last six months working for my company. Honestly, when you’re taking the classes, you don’t think you will really face any of those issues. At least I didn’t. If I hadn’t had the training from the classes, I would have been completely overwhelmed and felt bullied. It has given me the strength to stand up for myself on certain issues. Having gone through the GVV curriculum, I felt confident in my ability to stand up and express my concerns because I had practice how to approach situations. Having that training has not only given me strength, but it has also made me feel that I was not alone…

Additional research in these areas will add to the collective knowledge of “what works and what does not” in ethics instruction. Faculty in search of an innovative, practical approach to teaching ethics and developing ethical confidence are encouraged to consider adding GVV to their instructional toolkit.