Moral Ethical Development
Chapter II: Literature Review
Researching the moral development, ethical decision making approaches, and the adoption of utilitarian ethics on the part of Taiwanese CPAs begins with a thorough literature review of these specific topics. The intent of this chapter is to evaluate the research instruments used for supporting and validating the hypotheses of this study as well. For purposes of clarity, this chapter is organization into three sections, starting with a review of moral development theories. The second section concentrates on the most commonly used instruments for completing measurement of ethical judgment and decision-making, with the last section of the chapter presenting an overview of empirical studies designed specifically to measure Taiwanese CPAs’ relationship to ethical evaluation and ethical intentions and overall moral development.
Introduction to Moral Development
Much of the foundational work completed in moral development is attributed to theorists who together have refined the key aspects of this field in the last twenty-five years. The evolution of moral development theories has concentrated on increasingly on its role in defining psychological aspects of moral development (Rest, 1979). Adding to the body of knowledge on moral development are the works of Kohlberg (1976) and Piaget (1932). In conjunction with Rest (1979), the works of these two theorists show specifically how the development of moral delineations and definitions are formative and not absolute, and have specific attributes associated with each.
As accountancy is a field that relies heavily on trust and the fulfillment of fiduciary responsibilities, the importance of ethical judgments in the field of study is clear. Accountants and CPAs need to be the sustainers of public trust in the accounting and auditing professions. Ironically however one of the world’s largest scandals, Enron, was architected by accountants and auditors, which lead the U.S. government to legislate into law the Sarbanes-Oxley (SOX) Act of 2002.
Later in this literature review the basis of the SOX Act will be discussed. As a result of the importance of ethical judgments in the accounting profession, several studies that are empirically based have had as their objective an assessment of the development of moral judgment, moral frameworks, and skills at assessing the ethical dilemmas that occur in the fulfillment of their professions.
In researching and defining the moral development of organizations, the work of Paiget (1932) serves as a foundation, based on his assessment of how children form ethical and moral values through the study of their respect for rules. Paiget (1932) wrote The Moral Judgment of the Child to add to the field of knowledge on moral development. Paiget writes that “all morality consists in a system of rules, and the essence of all morality is to be sought for in the respect which the individual acquires for these rules” (p.13).
Relying on observation as a research approach and completing interviews with a cross-section of boys to see how their comprehension of rules influencing the game was completed, Paiget was able to construct theories of moral development and the development of moral judgment. Paiget defined the concept of children developing morally through a progression from heteronomy, through stages of autonomy, finally to equity. Based on this research Paiget created a specific framework which describes the stages in conceptualization of moral decision-making and judgment, leading to the finding of rules presence and consciousness of rules as the foundation for developing moral judgment (Ponemon, 1990).
The development of stages of Cognitive Developmental Theory (Piaget, 1997) are defined as follows. The first phase is where sensory-motor intelligence (younger than two years old) is prevalent. The second stage is pre-operational thought (two to seven years old) that serves as the foundation for creating more concrete concepts in the future. The third stage is called concrete operations and occurs between seven to eleven years old, with the fourth stage being formal operations that occur from eleven to adulthood. Paiget observed that by age 7, children begin playing with and cooperating with one another. Secondly Paiget observed children in the 11 to 12 age range see rules as permeable, capable of being negotiable and mutually agreeable.
Theorizing that children are not born with the ability to understand and apply moral standards yet learn them over progressive stage, Piaget (1997) began creating a moral development theory that forms the foundation of Kohlberg’s creation of his own theories of moral development as well. Kohlberg refined and developed a six-stage sequence mode model and subsequently and empirically proved the assumptions of Piaget (1997), (Kohlberg 1976). His moral reasoning theory is a cognitive developmental theory that seeks to understand the reasoning behind how adults decide what course of action in morally right or wrong. Furthermore, Kohlberg’s theory develops sequentially through three levels of moral development,, with each level subdivided into two stages (Table 1), Kohlberg’s cognitive development theory is today regarded as a model and framework for ethics education according to Baxter and Rarick (1987) and Shenkir (1990). (Baxter & Rarick, 1987; Shenkir, 1990)
It’s clear that Piaget had a significant influence on the findings and conclusions of Kohlberg. This is evident for example in the second and third stages Kohlberg defines, which are exactly the same as Piaget’s. Despite this same approach to defining the specific level of moral development, Kohlberg differs from Piaget on the fourth through sixth stages of development. The factors that are driving the differences are the interpretations of internal moral convictions justice and that Kohlberg sees as overriding social expectations, norms, and values.
Kohlberg’s work on moral development provided a foundation for further analysis of moral development and the formation of moral judgment as well. Specifically focused the measurement of the changes by stage of morale development, Kohlberg had defined a hypothesis that over time and throughout maturity children, then adults, become more attuned to moral development.(Kohlberg, 1981). Throughout the remainder of this life Kohlberg concentrated his efforts on moral education and the propagation of his theories.
Despite the alignment of Kohlberg’s findings and analyses to Piaget (1997) and others in the area of normative moral development being aligned to the developmental phases of a child’s then adults’ life there are many critics of this view including the following academicians and theorists (Gilligan, 1993, , 1998; Gump, Baker, & Roll, 2000; Reimer;, Paolitto, & Hersh, 1990; Rest, Narvaez, Thoma, & Bebeau, 1999, , 2000; Shweder, 1982; Spohn, 2000; Sullivan, 1977). Specifically these theorists have centered on four specific questions that form the foundation of their critique of Kohlberg’s theories (et.al.).
First do the processes involved in moral reasoning foster and sustain moral behavior over time? Theorists remark that Kohlberg’s theory supports critical thinking regarding morality yet does not concentrating on what ought to be done versus what is actually accomplished (Sullivan, 1977). As part of this argument theorists contend that the lack of longitudinal empirical evidence to support the claim that the higher the level of moral judgment, the higher the level of moral behavior. During his lifetime, this was often discussed and presented as a critique of his work, an assessment he agreed with saying “I understand the theory of justice reasoning to be necessary but not sufficient for defining the full domain of what is meant by moral development” (Kohlberg, 1984). Whether or not knowing is equal to doing, knowing must typically come before doing, and so Kohlberg states that “the study of moral reasoning is valuable in its own right” (Kohlberg, 1984).
Second, the suppositions that the only aspect of moral reasoning is justice when in fact there are many other factors to consider. The theorists have said that Kohlberg’s theories tend to overemphasize the perception of justice as the foundation for making decisions. Theorists have long argued that in addition to justice, additional factors including compassion, caring, and other interpersonal feelings may play an important part in moral reasoning. (Rest, Narvaez, Thoma, & Bebeau, 1999; Spohn, 2000)
The third major question of Kohlberg’s analysis is the minimal amount of empirical evidences for post-conventional level thinking , a critical assumption that underscore much of what Kohlberg based his analysis on(Rest, Narvaez, Thoma, & Bebeau, 1999)The fourth question that emerges from the analysis is an assessment of Kohlberg’s theories and the over-reliance on the role and influence of Western culture. Critics and theorists contend that there is equally if not more emphasis specifically from Eastern cultures as well that have not been taken into account (J. C. Gibbs, 2003; Spohn, 2000).
As a result of these critiques specific to his work, in 1985 Kohlberg decided to eliminate the sixth stage of his work, citing a lack of empirical evidence and proven causality once thought to be present. In addition, there appears to be a dearth of evidence of support the fifth stage of his model, and in fact theorists pointed to little support from previous empirically derived research.
In addition to the above points, theorists contend that there is a dearth of evidence for the fifth level of scoring as defined by Kohlberg (Snarey, 1985). Gibbs (1979) is credited with being a co-developer of the scoring system and concentrates on moral judgment development throughout the fourth stage of the Kohlberg framework. In addition, Kohlberg defended the empirical data for the fifth and sixth stages stating that statistically significant findings had supported his framework, despite the critics and the lack of empirical studies that could be replicated by other theorists.
The fifth and sixth stages are clearly centered on post conventional thinking, which is a limitation of Kohlberg’s framework. The book Lawrence Kohlberg: Consensus and Controversy (Modgil & Modgil, 1986) defines the shortcomings of the fifth and sixth stages of the Kohlberg model in depth, with several of the theorists mentioned in this literature review being the primary critics of Kohlberg’s theories.
Analysis of Rest’s Theories
James Rest (1979) developed a revision of the developmental process of moral judgment by looking at the shortcomings of Kohlberg (et.al.) and Piaget (1997). As a result, Rest’s moral judgment model is significantly different than Kohlberg’s (Rest, 1979). According to theorists, “For the measurement of moral reasoning, Rest’s model assesses an individual’s tendency to use concepts of justice based on social cooperation in his or her moral thinking, while Kohlberg’s model assesses an individual’s use of justice concepts, focusing more on exchange and individual interests” (Elm & Weber, 1994), p. 346)
Rest (1986) consequently a Four Component Model that captures for types of psychological processes must take place for an individual to experience moral behavior. The Rest Four Component Model is summarized by the key points below:
Moral Judgment: the individual must make judgment on what should be the right thing to do. In other words, a person should be able to determine the appropriate action that is morally correct.
Moral Sensitivity: This is defined as the interpretation of individuals’ actions in terms of responding to their overall assessment of situation. The intent of this aspect of the model is to refer to how an individual’s conduct is analyzed in terms of how is going to influence themselves and those who are associates and friends.
Moral Motivation: the essence of defining and prioritizing moral decisions over competing contextual perceptions of morality. Moral values are, according to this model, above personal values. Moral values are always thought about first when making a decision.
Moral Action: The sense of this specific element of the model requires the development of competence in the development and use of strategies that develop a moral foundation. According to this specific aspect of the model, the individual needs to have the integrity and self-determination to stay in alignment with their need for behaving morally and ethically over time. The point of this specific aspect of the model is that self-determination is a critical aspect of the total model.
Assessing the Cognitive Moral Decision-Making Model
The ethical decision-making of accounting, auditing and financial professionals have increasingly come under fire due to the major ethical lapses of these professionals that led to some of the largest scandals in history. Enron, MCI, Tyco and many other scandals are directly responsible for SOX legislation in the U.S. and a marked rise in compliance globally, preceded by the savings & loan debacle of the 1980s, have together served as the catalyst of high levels of interest in the ethics of CPAs and finance professions (Cohen & Pant, 1993; Lampe & Finn, 1992; McNair & Milam, 1993; Ponemon & Gabhart, 1990; Shaub, 1994)
From this body of research with regard to compliance surrounding ethics, Lampe and Finn (1992) have defined three dominant types of ethical decision making models. These are defined with the context of audits, and include agency models, cognitive models, and professional code-implied models. The intent of the study is primarily focused on the cognitive model component first, as it is the most pervasively used in evaluating the decision-making process of auditors (Ponemon & Gabhart, 1993).
Accounting ethics research, in the majority of studies is based on a cognitive moral decision making model as first analyzed and published by Rest (1986). Modeling the individual moral decision processes that include the reasoning and action processes in completing and carrying out ethical decisions and actions is defined in the model as well. According to the theory the person with a strong sense of morals is that that evaluates an ethical dilemma and situation to ensure that actions are evaluated for their ethicity first. In defining this model Rest (1984) states that interpreting the situation including the decision to try and decide which choices are, and if and how a decision might affect others (i.e., they try to determine if an ethical issue exists).
The analysis next begins with an assessment that the use of moral judgment is also critical for the development of doing what ought to be done. The next step is to develop the moral intent and define just what exactly needs to be done. As part of this step the relative strengths and weaknesses of each decision is weighed by the emotions, perceptions, and socialization of the decision-maker. The final step is the selection of a given alternative that may or may not be taken based on the ethical judgment of the leader. (Rest, 1984).
Recognize Moral Issue
Make Moral Judgment
Establish Moral Intention
Act Moral Behavior
In summary, the defined model of cognitive moral decision making, which is often referred to as Rest’s four component model, is used pervasively through the majority of accounting ethics research, and is also used for the insights into causality it provides. As a result of its pervasive use, this model serves as the foundation of the analysis of survey results completed as part of this dissertation as well. The Cognitive Ethical Decision Making Model is specifically focused on the development of the research plan and methodology in addition to analyzing the key results and findings from this research effort.
Illustrating the Defining Issues Test (DIT)
One of the other major contributions Rest (1979) made in ascertaining the ethics levels of respondents throughout his research was the development of the Defining Issues Test (DIT), which is also used in the research completed on this dissertation as a means of measuring theory of moral reasoning. Rest contends that the DIT model continues to supplant and enhance knowledge in the key area of empirical ethics research (Rest, 1979).
How the DIT is constructed begins with a measurement of the individual’s responses to moral dilemmas often defined in the context of scenarios or short vignettes that explain the key concepts behind the concepts being tested. Through the use of these scenarios or vignettes, the respondents’ views of ethics and morality emerge (Rest 1979). The test is typically written yet has also been placed on websites to make it possible for respondents from many different locations geographically to take the test at the same time.
It is considered one of the more effective tests at standardizing moral and ethical dilemmas and issue questions. The test is constructed with six dilemmas (three in the short version) is accompanied by 12 stage specific questions (items). The respondent next reads the dilemma and chooses one of three decisions that align with their perception of the situation. Following this, the respondent rates the importance of each of the 12 items to their decision. Finally, the respondent is asked to rank the four most important items using multiple choice questions.
The Moral Judgment Interview (MJI) is a comparable test and allows the respondent to specifically respond with their statements about moral and ethical dilemmas presented. As is true with most research methods, the use of multiple-choice questions is more sound from a methodological standpoint and therefore improves the reliability factor of the DIT instrument over the MJI. The MJI requires the interviewer to assess the subject’s response and assign a value to that response, introducing bias into the analysis of results (Elm & Webber, 1994). This method of scoring used in MJI could potentially influence the reliability of the results since the analyst who is interpreting the responses of the subject rather than the subject choosing a response from alternatives (Elm & Webber, 1994).
A typical structure of the DIT is a multiple-choice questionnaire that contains multiple scenarios or vignettes, and has been designed to have up to 12 moral arguments in each dilemma relating to moral reasoning. Each of the questions on the questionnaire asks the subjects to select the most illustrative or definitive response for each situation given their perspective of it. The DIT then uses a point system on a four-point scale to measure the overall responses and create a score that indicates the respondents’ moral and ethical reaction to the points made.
A score of four points is given for the most important response, and one point given for the least important response. Both manually-based and machine-based approaches are used for tallying and analyzing the scores, with statistical analysis programs increasingly being used to manage this process. Often the scores are also measured as a series of indices as well. The Principled Index (P index or P score) is the most commonly used one in this field of research today (Rest, 1979).
The first score mentioned is often the “P” score (principled morality) which defines the level of a respondents’ ethical cognition. This is a metric that quantifies the level of respondents’ reaction to scenarios and vignettes that are identified as Stage 5 and Stage 6 in the Rest’s theory. In addition to the “P” score, the “D” score quantifies responses fro all stages rather than just those identified in Stages 5 and 6. Critics have contended that the “P” score is more reliable and therefore more trustworthy as a measure of ethical cognition, a point that Rest has agreed with from his analysis (Rest, 1990). In re-assessing the value of “P” as a measure of ethical cognition, Rest wrote “…represents the sum of the weighted ranks given to principled items and is interpreted as the relative importance given to principled moral consideration in making a moral judgment. (p.101).
Further supporting the reliability of the DIT test are the inclusion of control variables, or items that are purely included to provide a random check of consistency of responses (Rest, 1990). In more advanced research instruments, this approach to ensuring that the responses are consistent is commonly used. When incongruent or inconsistent responses are found they are given a code of “M” and tallied at the end of the survey. If this specific factor score is too high then the individual survey is considered unusable.
Since its initial development, the test has been used in more than 500 documented studies globally, as claimed by its author (Rest, 1990). In quantifying the value of the research instrument from a reliability standpoint, there have been a series of internal validity measures completed as well, with a test-retest methodology used to track internal validity. Scores on these tests using the statistical technical called Chronbach’s alpha index deliver a consistently high level of reliability, with a .70 on the score of this specific statistical metric (Rest, 1986). One of the findings that Rest (1990) has seen from the work completed cumulatively is that the DIT scores tend to have a high correlation level to education and a low correlation to gender, religious beliefs and affiliates, and gender.
The ability to empirically test moral development has been achieved with the DIT methodology, and by 1999 Rest and his colleagues created a new version, DIT-2 which is a revised updated stories and issue statements, was introduced. This new instrument is an improved version of the DIT, specifically including more measures of reliability and validity and the purging of questions that could lead to erroneous results despite the presence of control variables (Rest, Narvaez, Thoma, & Bebeau, 1999). The DIT was chosen as the moral development instrument has the DIT-2 version has not been thoroughly tested enough therefore limited comparative data is available. This is especially true for accountant specific studies undertaken since the introduction of this second rating methodology. Secondly, while the DIT-2 generally is considered to be just as reliable as DIT given the latters’approach to managing reliability (Rest & Narvaez, 1998). This specific dissertation will then focus on DIT as the measurement instrument for evaluating the ethics of Taiwanese CPAs.
Empirical Analysis of Accountants through the use of DIT
The concepts of moral development theory have provided the financial accounting profession with much opportunity for empirical research and investigation. While specific empirical studies have defined the current level of moral development of business majors including finance and accounting students, professional auditors, tenured and lecturing accounting educators, CPAs, and CMAs, there are many other empirically-derived studies that have focused on the relationship between moral development and various accounting specific behaviors.
The intent of this area of the literature review is to represent the specific aspect of moral development research that has been conducted with respondents who are accountants. The impact of the six stage theory Kohlberg has often discussed and analyzed in the works cited in this research are also analysed using Rest’s four component model as foundation for completing an analysis using DIT results as the foundation for these efforts to create a pragmatic approach to analyzing the research.
Validation of the DIT research instrument for use in conjunction with analyzing the ethical development of CPAs has been pioneered by Armstrong (1987) through the use of two samples of respondent group CPAs who were given the DIT. Using the mean DIT P scores of each respondent group and then comparing their relative means through the use of large samples of undergraduate college students, college graduate students, and adult respondents as reported by Rest (1979).
After completing the analysis and calculating the mean DIT P scores, the results showed scores between the two CPA respondent groups as 37.1 and 38.5 respectively. Calculated mean DIT P scores of the three respondent audiences were as follows: college students (42.3), college graduates (53.3), and the adults in general (40.0). Armstrong (1987) stated that with the CPAs averaging only 1.1 years of post graduate education there was a significantly lower DIT P score reported and achieved by the CPAs. What also was found from the Armstrong analysis was that significantly lower DIT P scores obtained from CPAs occurred compared to both college students and adults as a general respondent group. From this analysis it’s clear that there is a normal progression in moral development that generally takes place in college, yet did not take place with the CPA groups. Their moral development, as measured by the DIT, had not progressed beyond that achieved by the general adult population.
In addition to the research completed by Armstrong (1987) there has been significant work completed on the topic of moral development and theories of moral maturity relating to accounting students and professionals (St. Pierre, Nelson, & Gabbin, 1990). This second study defined a respondent base of 479 seniors divided into 10 groups by major who were then given the DIT instrument to measure their level of cognitive ethical development. What is unique about this specific study is the definition of specific disciplines in the study as well including math, psychology, and social work. What is noteworthy from this student is that the mean DIT P scores for the respondent base of accounting majors were significantly below those of psychology majors.
The researchers list the mean DIT P scores of 38.75 for the respondents who are male accountants and 45.85 for respondents are female accountants. This finding supported previous empirical studies relating to the development of ethical cognition in students, with female students receiving significantly higher scores, attributable to the previous results of empirical studies. Exacerbating these findings were also the DIT P score median value of 43.19 for college seniors. The use of the “M” value to ascertain the reliability of the study was also completed to better manage the sampling bias and potential errors. Based on these factors the research showed that there were no significant differences in DIT PI scores based on the exposure or not to an ethics course.
Clearly the DTP P scores were seen as delineators of cognitive ethical development independent of formalized processes; there is in fact indication that ethically-oriented people do tend to gravitate towards majors that have a relatively high level of accountability, as psychology does for example.
In an ancillary study, the work completed by Ponemon (1990) illustrated that through the study of 52 accountant respondents who ranged in position levels of staff, supervisor, manager, and partner were given the MJI to ascertain the relationship between moral stage development and hierarchical position. The results of the survey showed statistically significant results with the mean stage level increase from 3.4 at the staff level to 3.7 at the senior level and then to 4.1 at the supervisor level. While the MJI peaked at the supervisor level, mean stage levels decreased to 3.6 at the manager level and to a low of 2.9 at the partner level.
Clearly the decline in MJI scores in this specific study contradicts the core concepts of moral development theory. Citing both a decline due to socialization and self-selection, Ponemon (1990) has stated that the results of this specific project do not necessarily refute ethical moral development theories. Following this research an paradigm (independence level) and the DIT were next analyzed (Ponemon & Gabhart, 1990) in an empirical study of 119 respondents in a CPA firm group of audit managers and partners.
Results of this specific analysis indicate that managers and partners who achieved low DIT P scores had the propensity to be more cognitively focused on the potential penalties than those respondents who have higher DIT P scores from the research. Not surprisingly there is a key finding from this research, that shows a significant negative relationship between rank and DIT P scores (managers-35.7; partners=30.1) in addition to the age and experience of respondents relative to their age and experience scores.
Ponemon and Glazer (1990) also completed empirical research that attempted to add insights into the effects of a liberal college curriculum relative to the level of moral development. Respondents in this study are alumni practitioners in addition to two groups of students including freshmen, and accounting seniors. To ensure the study would be representative, respondents were randomly selected from two institutions that offer accounting degrees. The first institution included in the survey drew respondents from a private college where the accounting major curriculum is part of the liberal arts program.
The second institution is a state university where accounting majors were not required or encouraged to take liberal arts courses. There were 143 total respondents in the study that completed the DIT, and the analysis of the results highlighted statistically significant scores between respondent groups. Further, the analysis revealed mean DIT P testing of freshman from both learning institutions was not significantly different between each of the learning institutions. At both the senior and alumni levels, the mean DIT P scores increased.
Paradoxically the institution that had accounting as part of the liberal arts program had the higher DIT P scores than the institution that specifically focused on accounting as a core curriculum. In analyzing the results of the survey differences the researchers stated that to “…suggest that liberal learning in college may be an important factor in the development of the student’s and accounting practitioner’s moral reasoning” (p.204). The integration of accounting curriculum in liberal arts broader learning institutions has since become a foundational element in ethics research, and is relied on as a theoretical construct of the research completed in this dissertation.
The dynamics involved in the moral development and perception of “whistle blowing” was studied in an empirical study that had 106 internal auditors as respondents by Arnold and Ponemon (1991) These researchers relied on the DIT research instrument to specifically assess the moral development in the context of experimental scenarios and vignettes surrounding a fraud case and experimental treatments that elicited the perception of internal auditors of “whistle blowing”. The results of this research illustrated the fact that internal auditors who scored low on the DIT P scale perceived the act of “whistle blowing” from the standpoint of personal punishment as a risk of unethical behavior than those with high DIT P scores.
Clearly this supports the fact that the DIT P measurement instrument supports the contention that those internal auditors that have higher DIT P scores have a higher level of ethical cognition relative to those with lower DIT P scores. The internal auditors who scores lower on DIT P scores saw the act of “whistle blowing”, it can be inferred, as an external event that had negative consequences if a person was caught. This leads to the observation that internal auditors with low DIT P scores significantly externalize the ethical role of whistle blowers, while those with higher DIT P scores have already internalized ethical cognitions and see whistle blowing as incidental.
Paradoxically however the main affiliation factors including punishment of others had little if any influence on the perc