elm weber 94 measuring moral judgment

Measuring Moral Judgment: The MoralJudgment Interview or the Defining Issues Test?

fDawn R. Elmfames Weber

ABSTRACT. This paper compares and contrasts two distincttechniques for measuring moral judgment: The Moral Judg-ment Interview and the Defining Issues Test. The theoreticalfoundations, accompanying advantages and limitations, aswell as appropriate usage of these methodologies are dis-cussed. Adaptation and use of the instruments for husinessethics research is given special attention.

Attention paid to moral judgment and reasoning inthe business ethics hterature has dramatically in-creased over the past decade (Trevino, 1992). Ac-companying this emerging research focus is theincreased usage of two dominant moral reasoninginstruments: the Moral Judgment Interview (Colbyand Kohlberg, 1987) and the Defining Issues Test(Rest, 1979). Each of these measures significantlydiffer in their theoretical foundation, what they arecapable of measuring, how they are appropriatelyused, and their respective advantages and limitationsas a research instrument. Confusion regarding theseaspects of the instruments could lead researchers toimproperly select the wrong measurement given theresearch objective, inaccurately administer the in-

Daum Elm is currently an Assistant Professor of Management at the

University of St. Thomas in St. Paul, Minnesota. Dr. Elm has

written and published articles on ethical decision-making, moral

reasoning, doming and measuring honesty, and women's studies

and ethics. She also has research interests in socialization to work,

gender bias imagery in teaching business, and parental leave

policies.

James Weber is currently an Assistant Professor of Management at

Marquette University. He has published articles focusing on

managerial values and moral reasoning, teaching of ethics, and

methodological issues in business ethics research in Research in

Corporate Social Performance and Policy, Human Rela-tions, Business Ethics Quarterly, and Journal of BusinessEthics.

Strument, or incorrectly analyze the data collected.In an effort to guide future moral reasoning re-searchers, we discuss in this paper the underlyingtheories, objectives and capabilities, and advantagesand limitations of the Moral Judgment Interview(MJI) and the Defining Issues Test (DIT). We believethat a clearer understanding of the instruments willenhance future business etWcs research investigatingmoral reasoning.

Theoretical foundation

Lawrence Kohlberg

The moral development theory underlying both theMJI and the DIT was developed by Lawrence Kohl-berg (1969, 1973, 1976, 1981, 1984), although JamesRest expands beyond Kohlberg in developing theDIT. Kohlberg and his associates attempted tounderstand the moral development of individualsfrom childhood to adulthood by periodically assess-ing the highest stage of moral reasoning expressed bythe subjects. These observations led Kohlberg todevelop a six stage model of moral development.The critical perspective underlying Kohlberg's modelis the identification of the reasons given why certainactions are perceived as morally just or preferred.As the subjects express distinctly different moralrationales, these rationales are captured in the dif-ferent stages of moral development.

Kohlbergian research identifies three levels ofmoral development: preconventional, conventional,and postconventional (see Figure 1). Each level iscomprised of two stages of reasoning. The secondstage represents a more advanced and organizedform of the first stage within each level. The threelevels are briefly summarized below (for a more

Journal of Business £(/iia 13: 341 -355, 1994.© 1994 Kluwer Academic Publishers. Printed in the Netherlands.

342 Dawn R. Elm and James Weber

Level 1: PreconventionalStage 1: Punishment and Obedience Orientation

Stage 2: Instrumental Relativist Orientation

Level 2: ConventionalStage 3: "Good Boy - Nice Girl " Orientation

Stage 4: Law and Order Orientation

Level 3: PostconventionalStage 5: Social — Contract Legalistic OrientationStage 6: Universal Ethical Principle Orientation

Fig. 1. Kohlberg's stages of moral development.

complete description of Kohlberg's stages of moraldevelopment see Colby and Kohlberg, 1987, particu-larly pp. 18-19).

Preconventional level

At this primary level, the individual understandsnotions of "right" and "wrong" in terms of con-sequences of action (punishment, rewards, exchangeof favors) or the imposition of authoritarian power.At Stage 1 (Punishment and Obedience Orientation)the physical consequences of an action determine itsgoodness or badness. Avoidance of punishment andunquestioning deference to power are the key values.Right action is defined in Stage 2 (InstrumentalRelativist Orientation) as that which satisfies one'sown needs. Elements of fairness and equal sharingare interpreted in terms of the physical or pragmaticconsequences upon the decision maker.

Conventional level

As an individual progresses to the conventional level,the maintenance of expectations imposed upon thedecision maker by the family, group, or nation isperceived as valuable. Stage 3 ("Good Boy-Nice Girl"Orientation) emphasizes behavior that will please orhelp others, thus gaining approval from others forthe decision maker. At this stage there is a significantemphasis upon conformity to stereotypical images ofmajority or acceptable behavior. At Stage 4 (Law andOrder Orientation) the individual takes the per-spective of a generalized member of society. This

perspective emphasizes adherence to a consistent setof societal, legal, or religious procedures that areapphed impartially to all members of society.

Postconventional level

As the individual exhibits a postconventional level ofmoral maturity, there is a clear effort to define moralvalues and principles apart from various authori-tarian figures and bodies. Generally with utilitarianovertones. Stage 5 (Social-Contract Legalistic Orien-tation) defines right action in terms of generalindividual rights and socially accepted standards.Rather than rigidly maintaining laws (reflective ofStage 4, law and order orientation). Stage 5 empha-sizes the possibihty of changing law based uponrational considerations of social unity. At Stage 6(Universal Ethical Principle Orientation) right isdefined by the decision-maker's conscience in accordwith self-chosen ethical principles appealing tological comprehensiveness, universality, and consis-tency. (It should be noted that as the stages areapplied in a Kohlbergian empirical assessment ofindividuals. Stage 5 and Stage 6 are generally com-bined into a Stage 5 designation.)

Applying Kohlberg's theory

Kohlberg's stage theory has two basic applicationsfor business ethics research. First, it can be used toassess the level of moral development of individuals overtime. This reflects the original usage of the theory,basic to Kohlbergian research in moral education.Kohlberg and his associates found that moral rea-soning follows a clear pattern of development."Development is step by step, that is, the stages areinvariant" (Kohlberg, 1981: 136-137). Hence, re-search has found that a Stage 2 person does not leapinto Stage 4, but gradually moves from Stage 2through Stage 3 on the way to Stage 4. In addition,research has found that individuals are attracted toreasoning at one stage higher than their predomi-nant stage (Blatt and Kohlberg, 1975). Thus, adevelopmental assessment of moral reasoning wouldpredict a gradual upward progression through thestages contained in Kohlberg's model.

Since business ethics research often focuses on the

Measuring Moral Judgment 343

moral reasoning of mature business school studentsor managers, another apphcation of Kohlberg's stagetheory is the identification of moral principles being used.Rather than assessing the development of reasoningover time, research can explore the stage of reason-ing used at a point in time. Although developmentalanalysis has found that the individual progresses tohigher stages of reasoning, people are also capable ofexpressing their moral decision-making rationale atlower stages as well. A Stage 4 reasoner does not losethe capability of assessing a situation from a Stage 2or Stage 3 perspective. Research has focused on theinfluence exerted upon an individual's moral rea-soning by the organization (Elm and Nichols 1993;Weber, 1990), intensity of moral issue 0ones, 1991),personal value orientation (Weber, 1993a), and otherindividual and organizational influences.

Over the years numerous criticisms have beenlodged against Kohlberg's theory of moral devel-opment and scoring procedure. Kohlberg and hisassociates have responded to their major critics (seeKohlberg et al., 1983), addressing numerous theoreti-cal and methodological issues. In addition, challengesregarding the universality of Kohlberg's theory werecritically reviewed by Snarey (1985). He reportssubstantial support for the cross-cultural universalityof Kohlberg's theory. The criticisms voiced byGilligan (1982) and others, challenging Kohlberg'stheory with gender bias, have been explored bynumerous researchers (see Lifton, 1985; Nunner-Winkler, 1984; Walker, 1984). Each of these studiesprovide general, empirical support for the applica-tion of Kohlberg's theory to both sexes. Finally,Cortese (1984) suggested inadequacies in Kohlberg'sscoring methodology (e.g., standardization, researcherbias). The development of the Standard Issue Scoringmethod and manual (the third generation of a Kohl-bergian scoring instrument) provides the scorer withclearer distinctions between moral stages and pres-ents an abundance of additional examples of moralreasoning rationales representing each stage (Colbyand Kohlberg, 1987). These improvements haveminimized the frequency and probabihty of widevariations in scoring caused by researcher bias.

Additional criticisms have been leveled at moraljudgment research regarding the relationship be-tween moral reasoning and moral behavior (seeMischel and Mischel, 1976). Research reviewed byBlasi (1980), Higgins et al. (1984), and Rest et al.

(1986) suggests that moral reasoning J5 related tomoral behavior, but it is not the only causal elementin the process. Rest et al. (1986) discuss the relation-ship of moral reasoning to behavior as part of a fourcomponent model. Derry (1987), Higgins et al.(1984), and Trevino (1986) contend that moralbehavior is a function of moral reasoning level inconjunction with the socio-moral environment inwhich the individual is embedded.

James Rest

Expanding on Kohlberg's theory, James Rest devel-oped an adaptation of the developmental process ofmoral judgment. Rest's model of moral judgmentdiffers from Kohlberg's in several important ways(Rest, 1979, 1983; Schlaefli et ai, 1985; Thoma,1986).

Theoretically, Rest's model of moral reasoningdiffers from Kohlberg's in the core concept thatdefmes the different stages, in the conceptualizationof stage structures and in the means by which thecognitive structures are applied by an individual.Rest (1979) uses a slightly different conception of themorality of justice than Kohlberg. Although boththeories define stages using a concept of justice,Kohlberg's theory defines stages primarily in for-malistic terms (reversibility and universalizabilityimply that justice exists within the individual), whileRest's theory characterizes the concept of justice ateach stage based on how different concepts of socialcooperation can be organized (see Figure 2).

Conceptualization of justice

The theoretical foundation which underlies Rest'smodel of moral judgment begins with the idea of"social justice" (Rest, 1978: 18; Rawls, 1971). Indi-viduals are born into associations of people and mustbalance their own interests with those of others inthe association. Therefore, the problem of justicebecomes one of balancing interests in social coopera-tion and achieving equilibrium through that balance.Thus moral thinking is based on assignment ofrights and responsibilities in a social system toprovide cooperation and stability.

In particular. Rest conceptualizes moral reasoning


Stage Coordination of expectationsabout actions (how rules areknown and shared)

Schemes of balance (howequilibrium is achieved)

General concept fordetermining moral rightsand responsibilities

Stage 1 The caretaker makes knowncertain demands on the child'sbehavior.

Stage 2 Although each person isunderstood to have his owninterests, an exchange of favorsmight be mutually decided.

Stage 3 Through reciprocal role taking,individuals attain a mutualunderstanding about eachother and the on-goingpattern of their interactions.

Stage 4 All members of society knowwhat is expected of themthrough public institutional-ized law.

Stage 5 Formal procedures areinstitutionalized for makinglaws, which one anticipatesrational people would accept.

Stage 6 The logical requirements ofnon-arbitrary cooperationamong rational, equal, andimpartial people are taken asideal criteria for socialorganization which oneanticipates rational peoplewould accept.

The child does not share inmaking rules, but understandsthat obedience will bringfreedom from punishment.

If each party sees something togain in an exchange, then bothwant to reciprocate.

Friendship relationships establish astabilized and enduring scheme ofcooperation. Each party anticipatesthe feelings, needs and wants of theother and acts in the other's welfare.

Unless a society-wide system of co-operation is established and stabil-ized, no individual can really makeplans. Each person should followthe law and do his particular job,anticipating that other people willalso fulfill their responsibilities.

Law-making procedures are devisedso that they reflect the general "willof the people, at the same timeinsuring certain basic rights to all.With each person having a say inthe decision process, each will seethat his interests are maximizedwhile at the same time having abasis for making claims on otherpeople.

A scheme of cooperation thatnegates or neutralizes all arbitrarydistribution of rights and responsi-bilities is the most equilibrated, forsuch system is maximizing thesimultaneous benefit to each mem-ber so that any deviation from theserules would advantage some mem-bers at the expense of others.

The morality of obedience:"Do what you're told."

The morality of instru-mental egoism and simpleexchange: "Let's make adeal."

The morality of inter-personal concordance: "Beconsiderate, nice, and kind,and you'll get along withpeople."

The morality of law andduty to the social order:"Everyone in society isobligated and protected bythe law."

The morality of societalconsensus: "You areobligated by whateverarrangements are agreedto by due processprocedures."

The morality of non-arbitrary social coopera-tion: "How rational andimpartial people wouldorganize cooperation ismoral.

Fig. 2. From Rest, J.R. Development in Judging Moral Issues. Minneapolis, University of Minnesota Press, 1979.


as a function of two major factors. First is the set ofconcepts an individual holds of how people formmutual expectations ahout the coordination of theirbehavior. This begins with rudimentary concepts ofshared expectations (the norms and rules of a care-taker) and culminates in concepts of mutual expecta-tions based on the logic of requirements for an idealsystem of cooperation. The second factor is theindividual's perception of the distribution of benefitsand burdens; how various interests are balanced toachieve a just distribution. This begins with abalancing scheme of rules to be followed in stageone through a scheme based on maximizing thesimultaneous benefit to every member of the socialsystem in stage six (Rest, 1979: 19—20). Therefore,each of the stages in Rest's model has a distinctconcept of morality as justice, with justice as socialcooperation which underlies it. These two majorfactors which determine the central concept fordetermining moral rights and responsibilities foreach stage of development are outlined in Figure 2.

This theoretical foundation is somewhat differentfrom Kohlberg's in the conceptualization of moralityas justice described earlier in this paper. Moralthinking, according to Rest, is formed on the basis ofthe social justice achieved through balancing dif-ferent interests and assigning rights and responsi-bilities to provide cooperation. Kohlberg's concept ofjustice is similar, but has a slightly greater emphasison rights and responsibilities assigned to an indi-vidual by others and by himself/herself Thus, inKohlberg's theory, justice exists within the indi-vidual. In Rest's theory, it does not.

Stage structure

The above conceptual distinction, however, is notthe largest difference between the two models. Themajor difference is in the architecture of the stagestructure. Rest's model has the combination of thetwo elements (outlined in Columns 1 and 2 ofFigure 2) of shared expectations and balancingschemes underlying each stage. All responses givenby individuals in resolving moral dilemmas can becharacterized as a function of these two elementseven though the reasoning may be manifesteddifferently in different contexts. Kohlberg's modelhas a much more elaborate stage structure. His

model suggests that every moral judgment can beviewed as an interaction between three factors:addressing a certain kind of moral question, using aparticular type of justification, and emerging from aspecific social institution. His model creates a logicalgrid for every type of response rather than acceptingdifferent manifestations of a stage of reasoning. Thisis due to the conceptual difference in the relation-ship between the content (the values and philo-sophical principles) of the reasoning process and thestructure (cognitive organizational structures) of theprocess. Kohlberg considers stages of development asindependent from the philosophical distinctions. Rest(1979: 45) suggests that such fine distinctions are notparticularly psychologically meaningful. He arguesthat the most useful unit of analysis is the kind ofconsideration an individual brings up in resolving amoral problem. Such considerations may have both"content" and "structural" elements in them (accord-ing to Kohlberg). For Rest, an individual thinking ofsocial cooperation in terms of one-to-one relation-ships is at a different stage than a person thinking interms of a societal network of institutions. In Kohl-berg's model, these differences represent content(specific belief) differences, and since his stages arestructurally defined, would not represent differentstages of moral development.

To summarize, Kohlberg considers every responseto be distinctly, and separately, classified on the basisof the cognitive structures evoked; while Rest con-siders ranges in responses to represent differentmanifestations of the same types of reasoning. Stage2 reasoning is manifested in many different ways,but involves the same concepts and organizing struc-tures in Rest's model. In Kohlberg's model Stage 2Ais not only different, it is lower than a responsescored as Stage 2B, 2C, or Stage 3.

Application of cognitive structure

Rest's model of moral judgment employs a "soft-stage" concept in how cognitive structures are usedby an individual (Rest, 1979). This model posits thatan individual's level of moral reasoning is a com-posite of various types of thinking represented byseveral adjacent stages. Thus, an individual is neverin or out of a given stage. Kohlberg's model suggeststhat individuals can be located in a particular time.


and that the reasoning structures appropriate to thatstage will be consistent across situations. As such, inthe Kohlbergian model the stages are discrete("hard"), and no stage mixtures regarding a responseto a particular moral question are possible unless it isthe short "transition" phase as an individual passesonto the next stage. This is a different perspectivefrom Rest's regarding both the relationship betweencontent and structure (discussed above) and the wayreasoning structures are used.

For measurement of moral reasoning, then. Rest'smodel assesses an individual's propensity to useconcepts of justice based on social cooperation in hisor her moral thinking; while Kohlberg's assesses anindividual's use of justice concepts focused more onexchange and individual interests. The difference inthe conceptual foundation of the models, while notearth shattering, is sufficient to emphasize the factthat data and results obtained using these modelsand the instruments which correspond to each ismethod specific.

The larger imphcation for measurement of moralreasoning comes from the different conceptual rela-tionship between content and structure, and thus thestage structure itself, as discussed above. If an indi-vidual is concerned about law in the sense ofmaintaining social order, that represents Stage 4reasoning. If, however, an individual is concernedabout law in the sense of worrying about going tojail, that is Stage 2 reasoning. Rest's conceptualiza-tion of the stages of moral reasoning reflects thisdistinction without distinguishing between a multi-tude of kinds of concern for maintaining socialorder. Rest assumes the type of considerationbrought up 15 indicative of developmental level. ForKohlberg, the consideration represents content thatis independent of developmental level and structure. Inaddition, the use of discrete stage classifications inthe Kohlberg's model versus the composite of stagesused in Rest's model suggest considerably differentinterpretations of moral reasoning level.

Moral reasoning instruments: the MJI andthe DIT

As mentioned earlier, the MJI and the DIT representtwo primary methodologies for assessing the moralreasoning of individuals. Each is briefly described

below and a contrast of the MJI versus the DIT isshown in Figure 3.

Moral Judgment Interview (MJl)

In order to operationalize Kohlberg's theory andstage of moral development, the Moral JudgmentInterview was developed. The initial procedureinvolved interviewing a subject after being presentedwith a series of situations involving moral conflicts.For example, should Heinz steal a drug to save hisdying wife if the only druggist able to provide thedrug insists on a high price that Heinz cannot afFordto pay? The conflict is between the value of preserv-ing life and the value of upholding the law. Aftereach dilemma is presented, the subject is asked aseries of open-ended, probe questions designed toelicit information regarding the subject's moralreasoning in resolving the dilemma.

Specifically, the MJI is designed to "elicit asubject's (l) own construction of moral reasoning, (2)moral frame of reference or assumptions about rightand wrong, and (3) the way these beliefs andassumptions are used to make and justify moraldecisions" (Colby and Kohlberg, 1987: 61). Questionsare explicitly prescriptive so as to draw out norma-tive judgments about what one should do, ratherthan descriptive or predictive judgments about whatone would do.

Coding of the subject's responses has been prob-lematic for Kohlberg and his associates over theyears. The current scoring scheme evolved throughdistinct phases into the Standard Issue Scoringmethod (Colby et al., 1983; Colby and Kohlberg,1987). The current method is intended to overcomethe limitations present in prior instruments. Inherentin the development of the new scoring method isthe specification of clear and concrete stage criteriaand an improved definition of the developmentalsequences of the specific moral concepts within eachstage.

Anne Colby and Lawrence Kohlberg (1987) out-line and describe a 17-step process for coding thesubject's response into a stage score. The steps aredivided into three sections: (1) breaking down theinterview material into interview judgments (steps 1through 6), (2) matching the new interview judg-ments with previous (standardized) interview judg-

Conceptual foundation

Demand characteristics

Administration process

Data analysis

Data configuration

Reliability vahdity

Business application


Moral Judgment Interview Defining Issues Test

- Justice as fairness- Hard stage concept- Structural stage definition

- Formulation or production task

- Oral or written interview- Open-ended responses- Requires trained interviewer

- Content analysis coding- Stage score

- Discontinous variable

— Balance for social equilibrium/Justice— Soft stage concept

— Content/Structural stage definition

— Recognition task

— Written survey— Likert scale responses— Does not require trained administrator— Mathematical calculation (percentage)— P score

— Continous variable— Limited parametric statistical analysis - Full parametric statistical analysis- Correlated with external criterion

variables, including action- Moderate prior research data

- Reliable and valid

- Adopted to business context

- Correlated with external criterionvariables, including action

- Extensive prior research data

- Reliable and valid

- Not yet adapted to business context

Fig. 3. Comparison of the attributes of the MJI and the DIT.

ments found in the scoring manual (steps 7 through14), and (3) assigning stage scores (steps 15 through17) (Colby and Kohlberg, 1987: 159-160). A reviewof the correlation reliability data for the StandardIssue Scoring method indicates that the instrument iswell within the limits of acceptable reliability (Colbyet al., 1983: 25). A comparison with related measures(the Sentence Completion Test and the DefiningIssues Test) also supports the contention of anacceptable measure (see Colby et al., 1983 for athorough discussion).

James Weber (1991) found the elaborate scoringprocess to be somewhat cumbersome and developeda more streamlined, yet reliable, variation of theoriginal model. Central to Weber's adaptation is thedevelopment of an Abbreviated Scoring Guide toprovide a simpler, yet reliable, method to code moralreasoning responses into Kohlberg's moral develop-ment stages (focusing upon the original steps 7through 14 in the Standard Issue Scoring procedure).A comparison of Weber's results using the Abbre-viated Scoring Guide with the original lengthier

method revealed high correlations, well within thecommonly accepted limits (see Weber, 1991: 304).

To summarize, the MJI attempts to elicit the stageof moral reasoning predominantly formulated by theindividual in response to a series of open-ended,probe questions presented at the end of each moraldilemma. These responses enable the researchers toidentify a single or combination of stages of moralreasoning used by the individual to explain the rea-sons why a particular action should be taken in re-solving the moral dilemma. The open-ended formatembodied in the followup, probe questions possessesboth advantages and limitations for the researcherusing this measure, as discussed later in the paper.

Defming Issues Test (DIT)

Based on his adaptation of Kohlberg's model. Rest(1979) developed a non-interview measurementinstrument called The Defining Issues Test (DIT)to assess moral reasoning without relying on the


verbal skills of the individual. This test contains sixhypothetical dilemmas, three of them Kohlbergiandilemmas, that can be used to determine an individ-ual's moral reasoning skills. As noted, the dilemmascomprise a variety of social moral issues, rangingfrom stealing a drug to saving a life to discontinuinga school newspaper for its disturbing influence.None of the dilemmas in the DIT are directlyrelated to a business context. Subjects respond to thedilemmas by rating and ranking the importance of aseries of statements prototypical of the differentstages of moral reasoning. Requiring both the ratingand ranking tasks allows for a consistency check forindividuals who might check at random through theinstrument. Researchers can choose between theinitial version of the DIT containing six dilemmas oran abbreviated version containing only three.

Measurement of an individual's moral reasoninglevel is accomplished through the calculation of aweighted index of the percentage of stage five andsix reasoning used to resolve the dilemmas. Theresulting score is called a P score or P index. A Dscore or D index can also be obtained. (The D indexis an empirically weighted sum of responses based ondouble centered item ratings. See Davison, 1979 forfurther calculation information). The percentage ofStage 5 and 6 reasoning used is determined by theanalysis of the representative stage level of the fourtop ranked statements chosen by the individualregarding what to consider in resolving the dilemmaspresented. The actual score is calculated by summingthe weights for the top ranked statements anddividing by the appropriate total possible for theversion of the instrument being used (see Rest, 1979:100—102 for further explanation).

A potential difficulty using a non-interview assess-ment method is an individual's capability to inflatehis or her moral reasoning score by choosing state-ments which sound pretentious. (This is not aconcern with the MJI since subjects are required toformulate a response without a set of prototypicalstatements available.) Rest (1979) presented evidenceto show that individuals are unable to "fake upward"on the DIT. This is due to the inclusion of state-ments that sound lofty and philosophical, but actu-ally have no meaning. Subjects are informed thatsuch statements are included in the instrument inthe instructions. An individual who consistentlychooses these nonsense statements as important is

given an "M" score, signifying an attempt to artifi-cially inflate his or her moral reasoning level. Dis-regarding the responses of individuals with high Mscores insures that the reasoning skills used by thesubjects are accurately represented by scores on theDIT.

The reliability and validity of the DIT has beenwell established (Davison, 1979; Davison and Rob-bins, 1978). Numerous studies using the instrumenthave reported reliabilities in the 0.70 to 0.80 range,depending on the use of the abbreviated (threedilemma) or original (six dilemma) version (see Blasi,1980; Rest, 1979; Snarey, 1985).

Kay (1982) criticized cross-sectional and longi-tudinal studies using the DIT on conceptual andmethodological grounds. He argued that most of thestudies employed a quasi-experimental or correla-tional design rather than a true experimental design.As a result, Kay hypothesized that the DIT actuallymeasures educational achievement, direct moraltraining, intellectual skills, and social values ratherthan a distinct developmental process. This hypoth-esis is not supported by previous studies as describedin Rest (1979), Blasi (1980), and Snarey (1985). How-ever, quasi-experimental designs were used in amajority of studies in moral reasoning in order toassess certain cohort differences (e.g., the relationshipbetween age or educational level and moral judg-ment).

Comparing and contrasting the MJI and theDIT

The MJI, or an adaptation of the original procedure,has been used in business ethics research to assess themoral reasoning of business students (Brabeck, 1984;Stratton et al., 1981; Weber and Green, 1991) andbusiness managers (Derry, 1987; Weber, 1990).Similarly, the DIT has also been used by a variety ofbusiness ethics researchers to measure the moraljudgment of managers (see Elm, 1989; Elm andNichols, 1993; Poneman and Glazer, 1990). Thereare a number of advantages that have contributed,and continue to contribute, to the use of thesemethods in measuring moral reasoning. A com-parison and contrast between the two methodsfollows and is outUned in Figure 3.


Administration of the instrument

Although correlations between the MJI and the DITare consistently found to be highly reliable (McGrawand Bloomfield, 1987), the two approaches differmethodologically in a number of ways. One crucialdifference is that the MJI presents subjects with aproduction task. The subject must formulate a moralresponse without prompting from the researcher orthe instrument. Rather than presenting illustrationsof possible moral responses (as present in the DIT),the MJI allows the subject "free reign" in construct-ing the moral rationale to resolve the dilemma.While this also embodies some serious researchchallenges, the formulation of moral reasoning maybe a more fair (unbiased) assessment of the subject'smoral reasoning process.

Alternatively, the DIT presents subjects with arecognition task. Subjects using the DIT are presentedwith the hypothetical dilemma, as well as series ofstatements representing various stages of reasoning.They need only rate (and later rank) the statements interms of their importance in considering how toresolve the dilemma. Since the recognition task iseasier (and does not require any verbal capabilities bythe individual), it is likely that the DIT creditssubjects with more advanced reasoning than Kohl-berg's method does (Rest, 1979). This phenomenonhas been consistently demonstrated in the largenumber of studies which have used the instruments.Kohlbergian subjects do not tend to show reasoningcapability at stages five and six, while DIT subjectscan. In fact, as noted previously, the more recentscoring manuals for the Moral Judgment Interviewno longer include stage six judgments. Again, thismeans researchers must be cautious about compar-ing results obtained with the different methods. Ifsubjects reasoned at stage four using the MJI, theymight well have had a P score representing stage fivewith the DIT. No direct comparison is possible sincethe DIT does not generate a stage score as the MJIdoes. This refers to the "soft" stage concept discussedearlier. Rest has cautioned that ". . . all data shouldbe regarded as method specific unless proven other-wise" (1979: 68).

The second difference between the two methodsemphasizes the form in which the measure is admin-istered. As initially suggested by Colby and Kohlberg(1987), the MJI is to be administered by a trained

researcher in a face-to-face, oral interview with thesubject. This technique has often been avoided bymoral development researchers. The face-to-face,oral interview technique requires a significant timecommitment from both the researcher and thesubject, and/or may be difficult to administer dueto demanding interviewing skills required of theresearcher. However, a comparison of two similargroups of managers, one group was administered anoral interview procedure and the other group awritten interview (pen-and-paper) application, wasconducted by Weber (1991). He concludes that"there is no difference berween the two groups ofmanagers' stage of moral reasoning" (1991: 307).While additional research to confirm Weber's initialfindings should be undertaken, it does offer thepossibility that the MJI could be administeredthrough a pen-and-paper procedure, rather than thelengthier and more difficult oral interview structure.If it is found that the MJI can be administeredthrough a written survey, researchers may find thismoral reasoning instrument more attractive thanpreviously believed. This adaptation of the MJI moreclosely mirrors the administration of the DIT.

However, even if the pen-and-paper version ofthe MJI is administered, the demands place upon thesubjects are more severe than with the DIT. Thesubjects are asked to formulate responses to a seriesof open-ended questions which requires more effortand time than simply placing a series of marks on aLikert-scale grid. Also, use of the oral MJI mayconfound the measure of an individual's cognitiveskills with his or her verbal skills, since the indi-vidual must be capable of verbalizing his or herreasoning process.

Proponents of the DIT have always emphasizedits ease of administration. Since subjects are pres-ented with written scenarios and prototypical state-ments to rate and rank in resolving the dilemmas,researchers can confidently administer the instru-ment without significant training in interviewingtechniques or in reliably applying coding schemes.This creates a situation of efficient data obtainmentfor later analysis.

Data analysis

The coding of an open-ended response from the MJI


into a defined stage structure can be quite difficultand demanding upon the researcher. As mentionedearlier, Colby and Kohlberg (1987) outline a 17-stepprocedure to successfiilly accomplish this task. Weber(1991) attempts to address and minimize this cum-bersome operation by developing an AbbreviatedScoring Guide. Nonetheless, the commitment oftime and learning by the researcher to understandKohlberg's stage theory and code responses intoverifiable stage designations are serious challengeswhen using the MJI. In contrast to the DIT's mathe-matical calculations to arrive at a percentage ofprincipled moral reasoning (described earlier), thecontent analysis procedure of scoring the MJI datamay be less attractive to researchers.

Although less complex than the scoring process ofthe MJI, the DIT still requires that a subject consis-tently respond to all dilemmas presented to enablecalculation of the P score. If a subject misunder-stands the instructions or leaves one of the presenteddilemmas incomplete (e.g., does not rank order themost important statements in resolving the dilemma),a score cannot be calculated for that individual.This can present problems for researchers who donot have samples large enough to withstand someshrinkage in available data.

Another difference between the measures is theway moral reasoning level is indexed. This refers tothe content/structure relationship discussed previ-ously. The MJI uses complex procedures to assignsubjects to a moral reasoning stage. DIT research hasshown that the P score works best in theoreticallycorrelating vwth other psychological variables (Rest,1979, 1983; Schlaefli etai, 1985; Thoma, 1986). TheMJI produces a stage score for an individual, whilethe P score produces a percentage measure of laterreasoning stages used. The stage score allows fordirect assessment of the specific cognitive structuresbeing used to resolve the dilemmas. The P scoremeasures the individual's tendency toward usingprimarily postconventional reasoning.

The subject's responses to the dilemmas in theMJI are presented as discrete, ranked data; that is,particular stage scores associated with a normativehierarchy of stage categories. The problems ofstatistical analysis of such discontinuous data aremore significant than if the data were continuous.Researchers should be aware of the limited number

of statistical procedures available to assess this typeof data.

Since the DIT measures moral reasoning level as aweighted average index (P score) or an empiricallyweighted sum (D score), it provides researchers v̂ dtha continuous variable at the interval level of meas-urement. That means it can be used in analysis ofvariance, regression, or other parametric statisticalanalyses without violation of the assumption that thedependent variable is continuous (an assumptionvery commonly violated by many studies in businessethics, as reported by Randall and Gibson, 1990).

Data base comparisons

It should be reiterated that the results gleaned fromthe MJI (stage scores) are not directly comparable toresults from administering the DIT (P and D scores).Since the larger body of previous collected moralreasoning data is from utilizing the DIT, researchersusing the MJI have a more limited opportunity tocompare their results with others using the MJI.

An additional advantage of the DIT is its excellentcorrelation to numerous external criterion variables.Since 1979, the DIT has been used in thousands ofstudies in countries all over the world (see Rest et ai,1986; Snarey, 1985). The correlation of moral rea-soning as measured by the DIT to a large number ofexternal criterion variables has provided the advan-tage of theoretical confidence. Researchers knowthat the DIT measures the cognitive reasoning skillsan individual uses to resolve moral dilemmas. Thiscan be advantageous when studying the relationshipbetween moral reasoning and moral behavior. Whilethe exact form of this relationship is unclear, there isevidence to suggest moral reasoning is part of theprocess of behaving ethically. (For a more in-depthdiscussion of this relationship and the accompanyingarguments, see Blasi, 1980; Elm and Nichols, 1993;Jones, 1991; Rest, 1979; Rest etai, 1986). Confidencein the construct validity of the DIT can be a signifi-cant advantage when doing research in an area inwhich definitional and relationship issues are nottrivial.

Researchers utilizing the MJI measure to assessmoral reasoning will also find a substantial data baseof studies in the moral education literature (as


reviewed by Blasi, 1980, and noted in Colby et al.,1983), although not as extensive as with the DIT.Studies using the MJI have assessed a wide variety ofsubjects based on cultural (Snarey, 1985), gender(Walker, 1984) and/or age (Colby and Kohlberg,1987) variations. There is an evolving data base ofstudies using the MJI with managers of businessschool students as subjects, as referenced earlier inthis paper.

Reliability and validity

The reliability and validity of the MJI and the DIThave been very well established (see Colby andKohlberg, 1987; Rest, 1979). As discussed earher, theMJI's Standard Issue Scoring method has consistentlydemonstrated correlation reliability data well withinthe limits of commonly accepted levels (Colby et al.,1983: 25). Numerous studies have reported internalconsistency and test-retest reliabilities with variouspopulations of the DIT in very acceptable ranges.The construct (concurrent) validity of the instru-ment from related conceptual variables of moralreasoning has also been demonstrated (see Blasi,1980; Rest, 1979; Snarey, 1985).

Regarding the issue of face validity, the MJI hasbeen adapted for use in a business setting, providinga unique advantage for business ethics researchers.Weber (1991) developed two moral dilemmas in abusiness context to compliment the Heinz dilemmain assessing managers' moral reasoning. In addition,he incorporated key organizational values into thefollowup, probe questions asked of the subjects. Theface validity of Weber's dilemmas needs to beassessed, but it is an initial effort at developing abusiness-oriented MJI.

There is currently a lack of face validity regardingthe hypothetical dilemmas in the DIT when utilizedto assess managers' moral reasoning, since none ofthe dilemmas presented are related to a businesscontext. This lack of face validity of the dilemmascould cause subjects to abandon their managerialrole in favor of the role of a person in society atlarge. As a result, we might expect the dilemmasrelated to a business context to result in differentmoral reasoning levels than those that contain morebroad socio-moral dilemmas.

A related concern involves the possibility of"story pull," which suggests that individuals usedifferent levels of reasoning depending on theirfamiliarity and experience with a particular scenario(Freeman and Giebink, 1979; Magowan and Lee,1970; Weber, 1990). Both the DIT and the MJI sharethis weakness. As discussed more fully in the fol-lowing section of the paper, the influence of "storypull" could account for significant variations in thesubjects' responses to the moral dilemmas poseddepending upon their ability to associate with thecharacter in the dilemma or fully comprehend theconflict of the moral dilemmas posed, possibly dueto the occurrence (or lack) of a similar personalexperience.

Conclusions and implications

Theoretical implications

There are two major theoretical implications forusing the MJI or the DIT to measure moral reason-ing. The first concerns the inadequate recognition ofthe conceptual differences which underlie the twomethods. The second focuses on the use of a formu-lation or production task versus a recognition task toassess moral reasoning. We have attempted to pointout the differences between the theoretical founda-tions of the MJI and the DIT to provide a basis forconducting better research in business ethics usingthe concept of moral judgment. While the differ-ences between the two models are not phenomenal,they are sufficient to suggest that researchers need tobe aware of the framework in which the instrumentthey choose was developed. It is hkely that eachinstrument could be matched with specific researchobjectives for the most accurate interpretation of theresults. For example, the MJI may be more appro-priate for assessing the moral reasoning of publicspeeches or statements of managers or CEOs, whilethe DIT might be useful for assessing the reasoningof the listening audience. If the theoretical issues arenot fully understood, an accurate interpretation ofthe results is limited. In particular, as will be dis-cussed further, researchers need to consider theimplications for the conceptual differences regardingcontent and structure in stage definitions, as well as


those for utilizing a model and instrument based onhard or soft stage concepts.

The fact that the two instruments provide sub-jects with two different types of tasks has significantimplications for data interpretation and comparison.Researchers must be aware of the potential biases ofthe instrument used and consider the potentiallimitations when drawing conclusions. The recogni-tion task of the DIT will provide the researcher withresults that may be slightly skewed toward thehigher end of moral reasoning levels, while theproduction task of the MJI may confound themeasurement of cognitive skills with the individual'sverbal capabilities. In addition, researchers mustunderstand that the results provided by the twotechniques are not directly comparable. Stage scoresdo not compare directly with P scores. More care inreporting results appropriately for the method usedis desirable.

Research implications

To measure moral reasoning the DIT assesses anindividual's propensity to use concepts of justicebased on social cooperation in his or her moralthinking; while the MJI assesses an individual's use ofjustice concepts focused more on exchange and/or areliance on universal and irreversable ethical prin-ciples. This difference in the conceptual foundationis sufficient to emphasize the fact that data andresults obtained using these instruments is methodspecific. Rest (1979: 68) discusses this problem in thecontext of both the kinds of tasks to be performed bysubjects and the specific test stimuli, materials andcontent (different dilemmas). He suggests that boththe type of task (production, recognition, etc.) andthe test materials affect the cognitive structures thatare manifested. The concern with different organiz-ing structures for different dilemmas ("story pull")will be discussed later in this section.

A significant implication for measurement ofmoral reasoning comes from the different concep-tual relationship between the content and structure,and the stage structure itself, as discussed previously.If an individual is concerned about law for themaintenance of social order, that represents Stage 4reasoning. If, however, he or she is concerned aboutlaw because of a possible jail sentence, that is Stage 2

reasoning. The DIT statements reflect this distinc-tion through considerations raised rather than themore sophisticated, and numerous, types of concernfor adhering to ethical principles as assessed by theMJI. The DIT relies on considerations comprised ofboth philosophical values and beliefs and cognitiveorganizing structures (content and structure), whilethe MJI separates the two. Further, the MJI classifiesan individual's reasoning into a discrete stage, evi-denced by a predominant stage score; while the DITuses a composite of stages as suggested by thecalculation of the P score. This re-emphasizes theneed to match research objectives with the appro-priate model and instrument.

A critical advantage of both the MJI and the DITis one that is not shared with numerous method-ologies used in business ethics research, that is, wellestablished reliability and construct validity of theinstruments. Randall and Gibson, in their survey ofempirical studies on ethical beliefs and behavior,found that business ethics researchers have "littleconcern for the reliability and validity of theirinstruments" (1990: 462). In the past ten years,numerous studies have been conducted with instru-ments that have been developed by the researcherswithout pre-testing or regard for the reliability orvalidity of their instrument. Neither the MJI nor theDIT have this problem since the reliability andvalidity of both instruments to measure moralreasoning is very well established (see Colby andKohlberg, 1987; Davison, 1979; Rest, 1979).

In addition, both instruments positively compareto other measures assessing moral reasoning as wellas numerous external criterion variables. Weber's(1991) adaptation of the MJI also appears to be reli-able, although further validation of this method isneeded. However, the face validity of the two meas-ures for business differs. Business ethics researcherswho would prefer to minimize face validity concernsmight find the use of Weber's adaptation of the MJIto be the most fruitful approach. In additionalresearch, business contextual dilemmas need to bedeveloped for the DIT. This would allow businessethics researchers to take advantage of the large database of DIT studies at the Center for the Study ofEthical Development at the University of Minnesota,where instruments can be obtained, scored, recorded,and compared to existing data.

Using hypothetical scenarios represents a limita-


tion for doing research with both the MJI and theDIT. This is due to the fact that such situations canpotentially prime or cue a specific response. Differentstage scores for different dilemmas is a well-knownphenomenon in moral judgment research. Straughan(1985) suggested that hypothetical dilemmas lackimmediacy for subjects, and Randall and Gibson(1990) outlined several reasons why hypotheticalscenarios should be used with care. On the otherhand, Damon (1977) found no difference betweenchildren's behavior and their responses to hypo-thetical scenarios in his study.

Rest (1979) describes a variety of inconsistenciesin moral judgment scores due to test characteristicsincluding the work of Medinnus (1959) who foundchildren exhibited different moral reasoning levelswith different Piagetian stories and suggested thattheir familiarity with the story made a difference.This is supported by both Freeman and Giebink(1979) and Magowan and Lee (1970), who also foundthat higher levels of moral reasoning were associatedwith higher levels of familiarity with the story pres-ented. Lieberman (1971) also demonstrated "storypull" with various dilemmas in the Kohlbergianscheme such that certain stories pulled out or cuedcertain stages of reasoning.

In business ethics, several researchers have sug-gested that managers seem to use different reasoningwhen the problem is related to business than they dowhen it is a broader, societal issue. As mentionedpreviously, Weber has adapted dilemmas in the MJIfor business contexts. This has not been done withthe DIT. As a result, die face validity of the DITdilemmas to managers in their role as agents of theorganization is low. This could suggest a seriouslimitation to the use of the DIT, particularly in con-junction with the difference in story pull betweendilemmas of different content.

For example, Robert Jackail (1988) chronicled thisphenomenon in his description of corporate man-agers in United States businesses. He suggests thatwhat is morally acceptable at work is not acceptableat home or outside the corporation. He argues thatcorporate bureaucracies create their own sets ofinternal rules that supersede the moral rules ofsociety at large. Managers follows the bureaucraticrules when at the office, but not outside.

Elm and Nichols (1993) found that older, moreexperienced managers reasoned at lower levels on

the DIT than younger, less experienced managers intheir study. In addition, Weber (1990) found thatmanagers reasoned at different levels for two busi-ness related dilemmas of his own design (mean stagescores of 3.22 and 3.35) than on a broader moraldilemma (the Kohlbergian "Heinz" dilemma, meanstage score of 3.84). He suggested that the nature ofthe moral issue or organizational context factorscould have contributed to these differences in rea-soning. Interestingly, this research is not consistentwith the work of Magowan and Lee (1970), since themoral reasoning levels of these managers is lowerwhen resolving dilemmas that could be more fami-liar to them (business related) than the more un-familiar social dilemma of Heinz.

Jones (1991) took this further to present a theo-retical model that suggests that ethical decisionmaking (and moral reasoning) is issue dependent. Heargues that ethical decision making of managers inorganizations is a function of the moral intensity ofthe issue being considered. Moral intensity is de-scribed as a multidimensional construct that involvescharacteristics of the issue such as the magnitude ofthe consequences, the societal consensus regarding it,the probability of effect, and others. This model isthe first in business ethics research to attempt tocharacterize the dimensions that distinguish onemoral dilemma from another. For business ethicsresearchers an interesting question becomes "Dobusiness-oriented dilemmas pull out lower level(stage three and four) reasoning?" Jones' focus uponmoral intensity provides several avenues for furtherinvestigation of this question as a start for futureresearch (see Weber, 1993b for an initial empiricalexploration of this question).

Additional avenues for further research using thesemethodologies has begun in exploring the degree ofinfluence of a variety of variables on moral reasoning.For example, Trevino (1986) suggests individual andjob context factors, while Elm and Nichols (1993)examine the influence of organizational climate,self-monitoring, and organizational tenure on moralreasoning. Weber (1990) investigates the size of theorganization, while Barnett and Karson (1987) con-sider the type of situation presented. Dukerich et al.(1990) focus upon the impact of group interactionand leadership on moral reasoning in groups. A widevariety of studies contribute to our understanding ofmoral reasoning using both the MJI and the DIT.


Moreover, the use of these methodologies appealsto researchers interested in exploring the relation-ship between moral reasoning and moral action. AsBlasi (1980) discovered in his comparative assess-ment of moral judgment research, considerablesupport exists for a moderate statistical relationshipbetween moral reasoning and moral action. Thisrelationship is borne out by the research exploringstudents' tendency toward ethical whistleblowingand principled moral reasoning (Brabeck, 1984) andthe selection of the ethically correct decision (Weberand Green, 1991).

In conclusion, we have attempted to present anaccurate comparison of the two primary means ofassessing moral reasoning used today. In comparingand contrasting the methods we have pointed outadvantages and limitations to each, as well as dis-cussed some implications of the choice of methodfor research in business ethics. A great deal ofresearch in business ethics involves the concept ofmoral judgment, and care must be taken to appro-priately use and interpret results of our studies tofurther our understanding of ethical decision mak-ing and ethical behavior.

References

Barnett, J. H. and M. J. Karson: 1987, 'Personal Values andBusiness Decisions: An Exploratory Investigation', youma/of Business Ethics 6, 371-382.

Blasi, A.: 1980, 'Bridging Moral Cognition and Moral Action:A Critical Review of the Literature', Psychological Bulletin88, 1-45.

Blatt, M. M. and L. Kohlberg: 1975, The EfFects of Class-room Moral Education Upon Childrens' Moral Judg-ment'. Journal of Moral Education 4(2), 129—161.

Braheck, M.: 1984, 'Ethical Characteristics of Whistle Blow-ers', yoMr«a/o/iieje<jr<:/i in Personality 18, 41—53.

Colhy, A. and L. Kohlherg: 1987, The Measurement of MoralJudgment: Theoretical Foundations and Research Validations,vol. 1 (Cambridge University Press, Cambridge, MA).

Colby, A., L Kohlberg, J. Gibbs and M. Lieberman: 1983, 'ALongitudinal Study of Moral Development', Monographsof the Society for Research in Child Development, Series No.200,48(1 & 2), 1-107.

Cortese, A. J.: 1984, 'Standard Issue Scoring of Moral Rea-soning: A Critique', Merrill-Palmer Quarterly 30(3), 227—246.

Damon, W.: 1977, The Social World of a Child (Jossey-Bass,San Francisco).

Davison, M. L: 1979, 'The Internal Structure and thePsychometric Properties of the Defining Issues Test', in j .R. Rest (ed.). Development in Judging Moral Issues (Univer-sity of Minnesota Press, Minneapolis), pp. 223—245.

Davison, M. L. and S. Robbins: 1978, 'The Reliability andValidity of Objective Indices of Moral Development',Applied Psychological Measurement 2(3), 391—403.

Derry, R: 1987, 'Moral Reasoning in Work-Related Con-flicts', in W. C. Frederick (ed.). Research in Corporate SocialPerformance and Policy: Empirical Studies of Business Ethics andValues, vol. 9 (JAI Press, Greenwich, CT), pp. 25-50.

Dukerich, J. M., M. L. Nichols, D. R Elm and D. A. Vollrath:1990, 'Moral Reasoning in Groups: Leaders Make aDifference', Human Relations 43(5), 473-493.

Elm, D. R: 1989, 'Managers Reasoning About Ethical Issues:The Impact of Ethical Climate and Self-Monitoring',unpublished doctoral dissertation (University of Minne-sota, Minneapolis).

Elm. D. R and M. L. Nichols: 1993, 'An Investigation of theMoral Reasoning of Managers', Journal of Business Ethics,in press.

Freeman, S. J. M. and J. W. Giebink 1979, 'Moral Judgmentas a Function of Age, Sex and Stimulus', The Journal ofPsychology 102, 43-47.

Gilligan, C: 1982, In a Different Voice (Harvard UniversityPress, Cambridge, MA).

Higgins, A., C. Power and L. Kohlberg: 1984, 'The Relation-ships of Moral Atmosphere to Judgments of Responsi-bility', in W. Kurtines and J. Gewirtz (eds.). Morality,Moral Behavior, and Moral Development (Wiley, New York),pp. 74-106.

Jackail, R: 1988, Moral Mazes (Oxford University Press, NewYork).

Jones, T. M.: 1991, 'Ethical Decision Making by Individualsin Organizations: An Issue Contingent Model', Academy ofManagement Review 16(2), 366—395.

Kay, S. R: 1982, 'Kohlberg's Theory of Moral Development:Critical Analysis of Validation Studies with the DefmingIssues Test', Intemational Journal of Psychology 17, 27—42.

Kohlberg, L.: 1969, 'Stage and Sequence: The Cognitive-Developmental Approach to Socialization', in D. Goslin(ed.). Handbook of Socialization Theory and Research (RandMcNally, Chicago), pp. 347-480.

Kohlberg, L: 1973, Collected Papers on Moral Development andMoral Education (Laboratory of Human Development,Harvard University, Cambridge, MA).

Kohlberg, L.: 1976, 'Moral Stages and Moralization', in T.Lickona (ed.). Moral Development and Behavior (Holt,Rinehart & Winston, New York), pp. 31—53.

Kohlberg, L: 1981, Essays in Moral Development, Volume I: ThePhilosophy of Moral Development (Harper and Row, NewYork).

Kohlberg, L.: 1984, Essays in Moral Development, Volume II: The


Philosophy of Moral Development (Harper and Row, NewYork).

Kohlberg, L, C. Levine and A. Hewer 1983, Moral Stages: ACurrent Formulation and a Response to Critics (Karger, NewYork).

Lieberman, M.: 1971, 'Esrimarion of a Moral Judgment LevelUsing Items Whose Alternatives Form A Graded Scale',Unpublished Doctoral Dissertation, University of Chi-cago.

Lifton, P. D.: 1985, 'Individual Differences in Moral Devel-opment the Relation of Sex, Gender, and Personality toMoTsiity', Journal of Personality 53(2), 306-334.

Magowan, S. A. and T. Lee: 1970, 'Some Sources of Error inthe Use of the Projective Method for the Assessment ofMoral Judgment', British Journal of Psychology 61, 535—543.

McGraw, K. M. and J. Bloomfield: 1987, 'Social Influence onGroup Moral Decisions: The Interactive Effects of MoralReasoning and Sex Role Orientation', yowma/ of Personalityand Social Psychology 53(6), 1080-1087.

Medinnus, G. R.: 1959, 'Immanent Justice in Children: AReview of the Literature and Additional Data', Journal ofGenetic Psychology 94, 253-262.

Mischel, W. and H. N. Mischel: 1976, 'A Cognitive SocialLearning Approach to Morality and Self-Regulation', inT. Lickona (ed.). Moral Development and Behavior (Holt,Rinehart & Winston, New York), pp. 84-107.

Nunner-Winkler, G.: 1984, 'Two Moralities? A CriticalDiscussion of an Ethic of Care and Responsibility versusan Ethic of Rights and Justice', in W. M. Kurtines and J.L. Gevidrtz (eds.). Morality, Moral Behavior and MoralDevelopment (Wiley, New York), pp. 348-361.

Poneman, L. A. and A. Glazer: 1990, 'Accounting Educationand Ethical Developmenr The Influence of LiberalLearning on Students and Alumni in Accounting Prac-tice', Issues in Accounting Education, Fall, 195—208.

Randall, D. M. and A. M. Gibson: 1990, 'Methodology inBusiness Ethics Research: A Review and Critical Assess-ment', yoMroa/o/"Bi«mes;j£t/iics 9, 457—472.

Rawls, J.: 1971, /I Theory of Justice (Harvard University Press,Cambridge, MA).

Rest, J. R.: 1979, Development in Judging Moral Issues (Univer-sity of Minnesota Press, Minneapolis).

Rest, J. R.: 1983, 'Morality', in P. Musson (gen. ed.) and J.Flavell and E. Markman (eds.). Manual of Child Psychology,Vol. Ill: C^ognitive Development (Wiley, New York), pp.556-629.

Rest, J. R., R. Barnett, M. Bebeau, D. Deemer, I. Getz, Y.Moon, J. Spickelmier, S. Thoma and J. Volker 1986,Moral Development: Advances in Research and Theory (Praeger,New York).

Schlaefli, A., J. Rest and S. Thoma: 1985, 'Does Moral Educa-

tion Improve Moral Judgment: A Meta-Analysis of Inter-vention Studies Using the Defining Issues Test', Review ofEducational Research 55(3), 319-352.

Snarey, J. R.: 1985, 'Cross-Cultural University of Socio-Moral Development: A Critical Review of KohlbergianResearch', Psychological Bulletin 97, 202-232.

Stratton, W. E., W. R. Flynn and G. A Johnson: 1981, 'MoralDevelopment and Decision Making: A Study of Student'Exhics', Journal of Enterprise Management 3, 35—41.

Straughan, R.: 1985, 'Why Act on Kohlberg's Moral Judg-ments?', in S. Modgil and C. Modgil (eds.), Lawrence Kohl-berg: Consensus and Controversy (Fulmer Press, Philadel-phia), pp. 149-161.

Thoma, S. J.: 1986, 'Estimating Gender Differences in theComprehension and Preference of Moral Issues', Develop-mental Review 6, 165—180.

Trevino, L. K.: 1986, 'Ethical Decision Making in Organiza-tions: A Person-Situation Interactionist Model', Academyof Management Review 11 (3), 601 —617.

Trevino L. K.: 1992, 'Moral Reasoning and Business Ethics:Implications for Research, Education, and Management',Journal of Business Ethics 11, 445—459.

Walker, L.: 1984, 'Sex Differences in the Development ofMoral Reasoning: A Critical Review', Child Development55,677-691. .

Weber, J.: 1984, 'Managers' Mx r̂al Reasoning: Assessing theirResponse to Three Moral Dilemmas', Human Relations43 (7), 687-702.

Weber, J.: 1991, 'Adapting Kohlberg to Enhance the Assess-ment of Managers' Moral Reasoning', Business EthiaQuarterly 1 (3), 293-318.

Weber, J.: 1993a, 'Exploring the -Relationship Between Per-sonal Values and Moral Reasoning*, Human Relations, inpress.

Weber, J.: 1993b, 'Just the Facts, Ma'am: The Influence ofthe Moral Issue and Its Intensity', 1993 InternationalAssociation for Business and Society Proceedings, in press.

Weber, J. and S. Green: 1991, 'Principled Moral Reasoning: IsIt a Viable Approach to Promote Ethical Integrity?',Journal of Business Ethics 10, 325-333.

University of St. Thomas, ^Management Department,

St. Paul, MN 55105,U.SA.

Marquette University,Department of Management,

Milwaukee, WI53233,U.S.A.

elm weber 94 measuring moral judgment

Documents