Disentangling Related Domains of Moral, Cognitive, and Ego Development

Research andEvaluation CommitteeLucy CromwellMary DiezAustin DohertyZohreh EmamiMarcia MentkowskiGeorgine LoackerKathleen O'BrienWilliam RickardsGlen RogersTimothy RiordanStephen Sharkey

Educational Researchand EvaluationMarcia MentkowskiWilliam RickardsJudy Reisetter HartGlen RogersKathleen Schwan MinikDenise SandersLynn Chabot-Long

Paper presented at the annualmeeting of the AmericanEducational ResearchAssociation, Seattle, WA.April, 2001

Disentangling RelatedDomains of Moral,Cognitive, and EgoDevelopment

Glen Rogers Marcia Mentkowski

Judith Reisetter Hart Kathleen Schwan Minik

Educational Research and EvaluationALVERNO COLLEGE

cite as:Rogers, G., Mentkowski, M., Reisetter Hart, J., & Schwan Minik, K. (2001, April). Disentangling related

domains of moral, cognitive, and ego development. Paper presented at the annual meeting of theAmerican Educational Research Association, Seattle, WA.

Printed 2003This publication is available from:Alverno College Institute3400 South 43rd StreetPO Box 343922Milwaukee, WI 53234-3922Phone: 414-382-6000www.alverno.edu

Graphic Design: Lynn Chabot-Long, Project Specialist, Educational Research and Evaluation

Copyright 2001. Alverno College Institute, Milwaukee, Wisconsin. All rights reserved under US, International andUniversal Copyright Conventions. Reproduction in part or whole by any method is prohibited by law.

Table of ContentsIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

Relationships Among Ego, Moral, and Cognitive Development . . . . . . . . . . . . . . . . . 1Development of Self and Ethical Commitment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5Empirical Limitations to Disentangling Related Developmental Domains . . . . . .10

Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

Sentence Completion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11Measure of Intellectual Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11The Defining Issues Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12Critical Thinking Appraisal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12Test of Cognitive Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12

Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12Confirmatory Factor Analysis Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17Critical Thinking and Integration of Self in Context . . . . . . . . . . . . . . . . . . . . . . . . . 17Ethical Aspect of Integration of Self . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18Moral Reasoning and the Consolidation of Moral Development . . . . . . . . . . . . . . .21Moral Self as Developmental Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23Moral Self, Moral Ideals, and Moral Identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25Development, Self-Reflection, and Identity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27An Educational Theory of the Person . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28

Epilogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41

List of AppendicesAppendix A: Factor Loadings for Human Potential Measures for Assessments at Times

1 and 2 Previously Reported in Mentkowski and Associates: Exploratory FactorAnalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35

Appendix B: Factor Loadings for Human Potential Measures Including the MoralJudgment Interview: Exploratory Factor Analysis Ignoring Time of Assessment 37

Appendix C: Fitted Coefficients for Base Confirmatory Factor Analysis . . . . . . . . . . . .39

List of Figures and Tables

Figure 1: Confirmatory Factor Analysis Base Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13Table 1. Coefficients for Indicators and t-values in the Confirmatory Factor Analysis .14Table 2. Factor Loadings for Human Potential Measures for Replication Sample:

Exploratory Factor Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15Table 3. Correlations Among Three Latent Developmental Constructs with DIT as

Indicator of Moral Reasoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

INTRODUCTION

Theories of human development of adultcapacities that are rooted in develop-mental psychology have an array ofstances that overlap and differ in a vari-ety of ways (Alexander & Langer, 1990).At a very broad level, a common overlapis their inclusion of a constructiviststance, whereby they explain andobserve structural development in termsof the individual's active meaning-mak-ing. Differences arise in how domainssuch as cognitive development, episte-mological development of self, moraldevelopment, and ego development aredefined and how they are thought to berelated to one another (Westenberg,Blasi, & Cohn, 1998). For example,Kegan (1982, 1994) argues for a singledeep overarching developmental processunderlying cognitive, interpersonal, andintrapersonal growth, while Loevinger(1976, 1986) and Kohlberg (1984a, 1986)have held that cognitive and ego devel-opment are related but distinct process-es. Articulating the relationship betweenego and moral development has been aparticular source of contention betweenLoevinger’s and Kohlberg’s theories (Lee& Snarey, 1988; Loevinger, 1986;Kohlberg, 1984a, 1986).

Relationships Among Ego, Moral,and Cognitive Development

Loevinger (1976, 1986) has consistentlymaintained that moral development is aninseparable facet of a single coherentprocess of ego development, the overalltransformation of the organizing struc-ture of meaning-making. The course ofego development entangles change inimpulse control/character development,interpersonal relations, conscious preoc-cupations, and cognitive complexity. In

contrast, Kohlberg (1976, 1986) consis-tently maintained that moral judgment isa distinct domain with its own structurallogic and transformation. Kohlberg’sdescription of the relationship betweenego and moral development evolved overtime, and his position came to resembleLoevinger’s stance that the ego subsumesor is more general than the moral (cf. Lee& Snarey, 1988). 1

The strong stage by stage parallelbetween Loevinger’s description of egodevelopment and Kohlberg’s descriptionof moral development (Loevinger, 1976;Kohlberg, 1984a; Lee and Snarey, 1988;Snarey, Kohlberg, & Noam, 1983)reflects not only common moral content,but also some common structuralassumptions about “(a) an invariantsequence of (b) hierarchical transforma-tions, which are (c) structured wholes”(Noam, Kohlberg, & Snarey, 1983, p.111). Even so, Kohlberg (1986) leanedtoward Kegan’s theory of ego develop-ment rather than Loevinger’s. In collab-oration with Gil Noam and John Snarey,he noted that “in contrast to Loevinger,who has not spelled out the theoreticalinner structural logic of each stage or thelogic of the sequence from one stage tothe next, Piaget, Kegan, Kohlberg, andother structuralists have defined stagessolely in terms of cognitive structures orways of thinking” (Noam, Kohlberg, &Snarey, 1983, p. 112). 2

Loevinger, Kohlberg, and Kegan dis-agreed with each other about therelationship of cognitive and egodevelopment. Kegan’s (1982, 1994)subject/object theory of developmenthas held that Piagetian-like cognitivedevelopment and the development ofthe meaning-making capacities of theself are aspects of a single unified

Disentangling Related Domains of Moral, Cognitive, and Ego Development. 1

developmental process (also seeSouvaine, Lahey, & Kegan, 1990).Loevinger (1976) held that although egodevelopment overlaps with some aspectsof cognitive development, the overallprocess of ego development is distinctfrom intelligence. At the same time, shespeculated that some particular kinds ofcognitive development might berequired for progression to particularstages of ego development (pp. 175-181).Kohlberg’s (1976, 1984a, 1986) descrip-tion of the relation between cognitiveand ego development continued toevolve, and was strongly conditioned byhis concerns about the relation betweencognitive and moral development.

Kohlberg’s rational reconstruction of thecourse of the “logic” of the individual’smoral development in terms of “justiceoperations” gives his theory a distinctivecognitive-developmental stance. Theidea of justice operations parallelsPiaget’s structural description of thecourse of cognitive development interms of increasingly complex and equi-libriated logical operations. Such opera-tions are conceptualized as generalizedand interiorized forms of actions and aredistinguished from second order reflec-tion where the actions themselves arethe focus on conscious attention andawareness. Reversiblity, which is a keyjustice operation, reaches full maturityin an impartial perspective taking actionof “moral musical chairs,” which calls forequal respect for all persons.3 Kohlberg(1981) also specifically located the phi-losophy of his moral theory in the socialcontract theory of John Rawls (1971)where an individual who presumes anoriginal position of not knowing his ownposition in society would have rationalreasons for preferring distribution ofgoods according to principles of justice. 4

Kohlberg rejected the idea that moraldevelopment was the application of cog-nitive development to the moral domain.He held that “the higher moral stages bydefinition require a high level of logicaland social-cognitive sophistication, butadvanced logical thinking and social cog-nition can be expressed within the moraldomain without constituting thinkingthat is morally advanced” (Colby &Kohlberg, 1987, p. 12). This is how therelation between cognition and moraldomains is expressed when it is stipulat-ed to be analytically true, which meansthat it would not be subject to empiricaltest. However, Kohlberg and others havealso commonly held what they feel is anempirically testable conceptualization ofa relationship between distinct moraland cognitive domains. In particularKohlberg (1976, 1986) has consistentlyargued that particular stages of cognitivedevelopment are necessary but not suffi-cient conditions for more specificallydefined (partially isomorphic) stages ofmoral development. The consistency ofthis particular argument for cognitivedevelopment as a precursor for moraldevelopment suggests its importance toKohlberg, and this is all the more sogiven his relatively less stable descrip-tion of the structural relations amongmoral, cognitive, and ego constructs(Kohlberg, 1976, 1986; Loevinger, 1986).

Although Kohlberg (1986) saw it as morespeculative and outside of his main con-cern of articulating the course of moraldevelopment, he collaborated with JohnSnarey and Gil Noam to articulate a dis-tinct theoretical expression of the struc-tural relation of moral development withother distinct strands of development. Inparticular, Snarey, Kohlberg, and Noam(1983) postulated three separate structur-al subdomains within a unified ego that

2 Disentangling Related Domains of Moral, Cognitive, and Ego Development

stand in cognitive relation to each other.The three subdomains are epistemologi-cal (e.g., Piaget, 1972), ethical (e.g.,Kohlberg, 1986), and metaphysical (e.g.,Fowler, 1981). Each is considered a sub-domain of “reasoning” in a broad mean-ing-making sense that “reunites what aperson thinks, feels and does” (p. 309). Inturn, each implies a particular referencepoint: “the natural environment,” “thesocial environment,” and “the ultimateenvironment.” At the same time, a subdo-main might in some way operate on anenvironment that is not its implied refer-ence point. For example, the epistemolog-ical meaning-making stance of WilliamPerry (1970) was seen as focusing on thesocial and ultimate environments ratherthan the natural environment, eventhough the latter is the implied referencepoint of the epistemological. Loevinger(1986) characterized Kohlberg’s matrix ofstructural subdomains and environmen-tal reference points as creating a theorywhere “not only is moral developmentgoverned by rigid structure, all of egodevelopment is rigidly, though differentlystructured” (p. 185).

Loevinger held that the logical opera-tions of cognitive development were ill-suited to defining the structural develop-ment of the ego (Blasi, 1976). In particu-lar, the synthetic process of meaning-making that was the structural core ofego development called for a recognitionof content in particular meanings notcaptured by the generalized structuralforms of Piagetian cognitive develop-ment. Individual purposes and valuesarising in the context of culturally devel-oped forms of life cannot be reduced toforms of reasoning, except by losingtheir concrete meaning, which play acentral role in the structure of ego devel-opment.5 Loevinger (1976) included

structural transformations of the indi-vidual’s motives as reflected, for exam-ple, in the stage names “impulsive” and“self-protective.” Kohlberg (1986, p.498) in contrast disavowed this stance:

Many readers of my earlier writings(Kohlberg, 1969, 1984, Ch. 1) were leftwith the belief that my moral stages wereeventually expressions of a developmen-tally hierarchy of motives. In this inter-pretation, preconventional judgmentswere motivated by, or expressed motivesof, concern for extrinsic punishment andreward, conventional reasoning by affilia-tion, approval and respect-seeking, andpostconventional judgments by motiva-tions of conscience or moral self-judg-ment. In fact, however, I have avoidedthis emotivist view….I have claimed thatin some sense there is a primary motiva-tion ‘to do the right thing’ in thesociomoral world as Piaget assumed aprimary adaptation a ‘truth’ motivationfor the infant and child’s actions towardthe physical world.

The transformation of motive structurethat Kohlberg disavows, Loevinger seesas central (cf. Kohlberg, 1984c, p.243).For Kohlberg, such transformation couldnot be formalized or universalized, andthis was essential to his cognitivist phi-losophy of moral development and allthat it entailed, which included a philo-sophical characterization of distinctlynormative demands and his rationalreconstruction of the development ofmoral judgment (cf. Kohlberg 1984c, p.246). Among Kohlberg’s criticisms ofLoevinger’s theory was its lack of philo-sophically grounded normative claimsfor higher stages being better.

Much of the power of the ideas ofKohlberg and Loevinger comes from thedevelopment of strong measures of theirconstructs and the resulting program-


matic research they have sustained.Following Kohlberg, James Rest devel-oped another measure of the develop-ment of moral reasoning that he linkedto an influential revision of Kohlberg’stheory. Rest (1979a) significantly depart-ed from Kohlberg’s stage model by argu-ing that individuals might use moral rea-soning that reflected several stages ofmoral judgments. In this view, an earlierand less complex form of justice reason-ing might be elicited by a situation andstill be a useful part of the individual’smoral judgment repertoire. Thus, Restsignificantly relaxed the structuralholism requirements of Kohlberg’s theo-ry. He made other concomitant relax-ations. Notably, he argued that the con-tent of concerns that individuals focuson (e.g., friendships, law) are entangledwith the structure of their moral reason-ing (pp. 41–47) and that measurement ofreasoning about justice was betterviewed as probabilistic. Here, he tookmetatheoretical positions that resem-bled Loevinger more than Kohlberg.

Rest’s most distinctive contribution wasto develop a measure of moral reasoningthat used an innovative selection/ratingformat rather than a production one.The instrument measures the partici-pant’s preferences for statements thatreflect particular types of moral ratio-nales for an action choice rather thanspontaneous demonstration of thecapacity and proclivity toward producinga moral rationale. Rest, Narvaez, Bebeau& Thoma (1999) have offered a vigorousargument for their instrumentation.Crucially, they have summarized evi-dence that individuals who prefer state-ments linked to lower stages of moraljudgment are unable to comprehend thearguments for higher stages, while thosewho prefer statements linked to higher

stages comprehend the arguments forboth lower and higher stages.

Rest et al. (1999) clarified a divergencefrom Kohlberg in terms of their substitu-tion of the concept of moral schemas formoral stages. These moral schemas arecontent representations that structuretop-down processing of information asopposed to Kohlberg’s process-orientedjustice operations. Despite these andother differences, Rest et al. (1999) havegood reasons to submit their theory asneo-Kohlbergian. Their rational recon-struction of moral schemas stronglyresembles Kohlberg’s (1986) rationalreconstruction of moral stages. Bothbroadly track adolescent and adultdevelopment from a maintaining societyway of thinking (a social systems form ofconventional moral reasoning) to a cre-ating society way of thinking (postcon-ventional moral judgments of humanrights or general ethical principles).Although Rest distanced himself fromthe concept of justice operations, hebroadly retained Kohlberg’s ordereddescription of socio-moral perspectives,which Kohlberg considered key to struc-turally defining moral development.Like Kohlberg, he argued that cognitivedevelopment was a key component tothe development of moral judgment, butalso that moral judgment was distinct.As evidence, Rest and his colleagueshave cited patterns of correlations,unique prediction of variance in regres-sion analyses, and discriminant validityin at least one intervention study (Rest,1979a; Rest et al., 1999). UnlikeKohlberg, Rest did not develop a theo-retical or empirical stance toward thedevelopment of a unifying or syntheticego. But, like Kohlberg, he focused ondeveloping an account of how moraljudgment processes related to moral


behavior. Rest (1986a) and Rest et al.(1999) articulated a four componentmodel of moral processes for predictingmoral behavior: (1) moral sensitivity tothe presence of moral issue, (2) moraljudgment about what action is most jus-tified, (3) moral motivation to commit tomoral action and place moral valuesahead of other values (4) and moralcharacter, which includes havingcourage and overcoming fatigue ortemptation. Rest (1986a, p. 5) describedthese as processes in the production of amoral act and did not conceptualizethem as general traits of people.Moreover, they were not held to be anoverall unity of processes.

Development of Self and EthicalCommitment

Perry’s (1970, 1981) theory of intellectu-al and ethical development focuses onthe epistemological development of boththe individual’s way of understandingknowledge and making personal com-mitments in a pluralistic and contextual-ly relativized world. Perry preferred toassess growth through the PerryInterview (Perry, 1970), but alsoacknowledged the usefulness of theMeasure of Intellectual Development,which used a short essay format(Knefelkamp, 1974; Knefelkamp &Slepitza, 1976; Widick, 1975). Perrycharacterized the individual’s growth asmoving from belief in absolutism (intruth and morality) through positionsthat first recognized and then increas-ingly grappled with the epistemologicaland personal implications of multiplicityunbounded by absolutes. Perry recog-nized each position as a coherent wholeand then observed an ordered progres-sion in them related to internal conflictsthey had left unresolved. Each position

“both includes and transcends the earli-er ones, as the earlier ones cannot dowith the later” (Perry, 1981, p. 78). Inearlier positions (2 to 3) the existence ofpluralism is accommodated by seeing itas either illegitimate or a limited andtemporary area of uncertainty. Thelinchpin position of relativism (position5) heralds metathought or the coordinat-ed capacity to think about thinking.Relativistic thought is better equilibriat-ed, Perry suggests. Epistemologicalunderstanding of the capacity formetathought yields not only legitimacyto pluralistic positions, but also the legit-imacy of their mutual relevance andtransformative interchange. The impli-cations of this new capacity formetathought across perspectives andsystems then yield. in the higher posi-tions (6 to 9), transformation in the sub-jective meaning of commitment andidentity. Commitment is now under-stood as a contextualized capacity andresponsibility to think for oneself, tointellectually engage others, and toreflect on the choices one makes in allareas of life. Such a commitmentembraces contextual relativism byacknowledging that it will not be possi-ble to rely on any form of absoluteknowledge and values. Not only areauthorities relativized, but so are one’sown committed stances and now inter-nalized standards for self-evaluation.The relativistically committed individualis open to the capacity of external posi-tions to critically inform and even (whenpersonally thought through) to poten-tially transform his or her commitments.Thus, taking full responsibility for one-self ironically means relativizing oneselfand one’s chosen standards. Positivedevelopment embraces this flux withexistential commitment, rather thanescaping into cynicism or nihilism.


Loevinger (1976) saw Perry’s theory asbroadly congruent with her own, describ-ing potential stage by stage comparisonsbetween the two. She appreciativelyobserves that “of all the theories of egodevelopment, the one with greatestpoignancy for most college students isthat of Perry (1970)” (p. 126). Both theo-ries include a focus on the structuraltransformation of motives and, morebroadly, of the self as a meaning-makingprocess. This meaning-making self devel-ops toward increasing complexity, inte-gration, and autonomy, which is markedby individuation, but also by increasingrespect for and awareness of interde-pendence and contextual relativism.

The kind of ethical development Perryarticulated in his theory is different fromthat articulated in Kohlberg’s stage theo-ry of moral judgment. A clear differencelies in the locus of concern toward socialcooperation (Kohlberg) versus individ-ual development (Perry). A related dif-ference is Kohlberg’s deontologicalemphasis on justice or respect for per-sons and Perry’s existential emphasis onself-definition of the good life.6 WhereasKohlberg described increasingly equilib-riated perspectives on moral norms,obligations to others, and the generalwelfare of all, Perry’s theory describesthe ethical dimensions of choosing toaccept the difficult personal conse-quences that are inherent in personalgrowth. Perry describes some of thecompeting personal interests in terms ofan “urge to conserve” that can over-whelm a call to growth. One’s emergingunderstanding of new perspectives thatoffer potential for growth might be evad-ed because they endanger “the wish tomaintain community in family andhometown values and ways of thinking,the reluctance to admit one has been in

error, the doubt of one’s competence totake on new uncertainties and responsi-bilities, and, most importantly, the wishto maintain a self one has felt oneself tobe” (Perry 1970, p. 52). Perry (1970,1981) very eloquently empathized withthe validity of these concerns. But, hisexistentialist stance grounded hisdescription of growth and his commit-ment to describing the higher positionsas better.

Kohlberg (1984a) saw Perry’s theory asmetaethical (e.g., “what is morality?”)rather than normative (e.g., “what is theright thing to do?”). Kohlberg saw this aspart and parcel with Perry’s existentialstance. Like Gibbs (1979), he saw all the-ories built on developmental growth inexistential meaning-making or secondorder reflection as necessarily involving“soft stages.” Kohlberg & Ryncarz (1990)specify five criteria that distinguish hardstages from soft stages. Briefly these are(1) universality of stages and sequence,(2) equilibriated interiorized forms ofaction, (3) strict separation of contentfrom structure, (4) amenability torational reconstruction, and (5) “theabsence of the ego or a self in the con-struction of the stages” (p. 205). Perry’s(1970) description of existential develop-ment acknowledges potential regres-sions in development (failing the firstcriterion for hard stages), embraced con-tent cues in describing development(failing the third criterion), and centeredon the development of the self as a cen-ter of totality in meaning making (failingthe fifth criterion).

Kohlberg and Ryncarz (1990) speculatedon a seventh stage of moral developmentthat broaches the existential question of“why be moral?” but held that it was nec-essarily a soft stage, like all of Perry’s


positions of development. Kohlberg(1984b) acknowledged some particularconnections between his theory andPerry’s. Notably, he held that the indi-vidual’s confrontation with subjectivismand relativism (what Perry’s called mul-tiplicity) provides the essential service ofdisequilbriating conventional moral rea-soning.7 Kohlberg saw this phenomenonas being so powerful that it temporarilyconfused the transparency of develop-mental growth on his scheme. For awhile Kohlberg characterized extremerelativism as Stage 4½, but Kohlbergabandoned this position when heobserved that a relativistic stanceoccurred at other stages as well(Kohlberg, 1984b). Kohlberg (1984d, p.364) acknowledged findings that post-adolescent growth pushed into Perry’scontextual relativism and that thisgrowth “may occur after the attainmentof principled stages of justice reasoning,”but disputed Gilligan and Murphy’s con-clusion that this also entailed the erosionof principled moral judgment in favor ofa concern for responsibility and care (cf.Habermas, 1996; Gilligan & Murphy,1979; Murphy & Gilligan, 1980). Instead,he argued that such growth after post-conventional justice reasoning repre-sented increased awareness of factualambiguities and thereby appropriatelycontextualized the search for the appli-cation of principled morality.

Kohlberg also augmented his structuralstage theory of moral development witha theory of ideal types conceptualized interms of both structure and content.Following Piaget, Baldwin, and Kant, hedistinguished moral autonomy (Type A)from heteronomy (Type B). His descrip-tion of the relation between his moralstage theory and the development ofmoral autonomy evolved considerably

over time (Tappan et al., 1987). He aban-doned the position that Type A and TypeB moral judgments represented struc-tural substages within each stage ofmoral development, because thesequence was only predominantly fromType A to Type B. Instead, he settled onthe position that autonomy was definedby both structure and content and spec-ulated that the developmental path intoautonomous moral judgment wouldtend to occur as a one-time shift, a shiftthat might happen at any of severalstages of moral development.

There are some potential convergencesof Kohlberg’s conception of autonomousmoral judgment and Perry’s theory. Forexample, Kohlberg says autonomousmoral judgments “reflect more freedomfrom fixed norms, external authority fig-ures, and pragmatic considerations”than heteronomous judgments “evenfrom within the same sociomoral per-spective” (Tappan et al., 1987, p. 378).Perry’s theory is perhaps broadly con-gruent with three criteria that Tappan,Kohlberg and their colleagues say distin-guish autonomous moral judgment fromheteronomous moral judgment (free-dom from external parameters,reversibility via full consideration ofothers perspectives, constructivism asunderstanding that rules are developedby people). Kohlberg speculates on amacro one-time shift to autonomousjudgment, while Perry articulates adevelopmental logic for a sequence ofpositions that reflect a gradual epistemo-logical emergence from a dualisticreliance on external authority and rules.There are other divergences between thetheories.

Tappan et al. (1987) distinguished arange of other criteria for autonomous


moral judgment, and these can be con-trasted to how Perry took up the ethicalself. These criteria extend a Kantian per-spective on autonomous morality andjustice. They include: hierarchy as plac-ing moral values over nonmoral ones,intrinsicalness as valuing persons asends, prescriptivity as experiencingmoral obligations as a moral necessity ofinner compulsion regardless of the indi-vidual’s inclinations to do otherwise, anduniversality as a willingness to general-ize moral judgments so that they applyto everyone. Of these, Perry’s claim toinclude the ethical domain most clearlyrelies on the criterion of compellinglyfelt obligation or prescriptiveness.

Perry (1970) observed that those indi-viduals who felt they could not follow thecall to increased integrity that they heard“revealed in their accounts the specialkinds of stress, and of distress, associat-ed with internal denial and disassocia-tion.” They “seemed to be actively deny-ing or fighting off within themselvesawareness of their urge to progress.”When they made these denials explicit, itwas “always with acknowledgment ofsome dis-ease or even shame” (p. 53).Shame as an emotion is definitively amoral one and theorists besides Perryhave also tended to relate it to one’ssense of self.8 For example, Tugendhat(1993) has argued that the experience ofshame is linked to moral identity and thedesire to be a member of the moral com-munity.

The role of personal identity in explain-ing moral commitment has led a numberof theorists to articulate the concept ofmoral self (see Noam & Wren, 1993).Personal ideals, which may includemoral ones, define the experienced selfin important ways. Frankfurt (1993, p.

25) holds that what the person experi-ences as unthinkable—the ideals that arenot allowed to be forsaken—constitutesthe kind of identity required by genuineintegrity:

If someone has no ideals, there is nothinghe cannot bring himself to do. Moreover,since nothing is necessary to him, there isnothing that he can be said to essentiallybe. To be sure, he may have a number ofpersistent psychological dispositions ortraits; he may exhibit various consistentpatterns of inclination and choice. Butany stable volitional characteristics hemay have are products of impersonalinfluences. They are not consequences ofhis wanting to be a person of certain sortor to devote himself to a certain kind oflife; they are not fixed by his will itself butby contingencies external to it.

This kind of commitment to self-defin-ing ideals echoes Perry’s theory of intel-lectual and ethical development. But, asFrankfurt observes, commitment toideals need not be to moral ideals (wheremoral is now limited to concern for oth-ers, such as justice or care). Perry (1981)likewise observes that a student who hasembraced contextual relativism mightcommit to a set of values, a particularcareer, or a relationship with another.Thus, it is important to distinguish twodifferent senses of the moral in relationto Perry’s theory. On the one hand, thetheory includes a focus on the individ-ual’s sense of moral obligation to theirown growth. On the other hand, the the-ory does not privilege, or even focus on,moral ideals in the interpersonal realm(e.g., justice or care) relative to personalideals. This contrasts with Kohlberg’sgeneral conception of the moral domainand his conception of autonomousmorality. These differences in metaethi-cal assumptions extend into instrumen-


tation. For example, Kohlberg’s MoralJudgment Interview elicits verbal reflec-tions on moral dilemmas while theMeasure of Intellectual Developmentelicits short essays on a best class, arecent decision, or career choice.Although the Perry essays do not gener-ally elicit interpersonal moral concerns,personal growth in the epistemologicalunderstanding of commitment hasmetaethical implications for the capacityand proclivity to reflectively commit tomoral ideals.

One question, then, that arises fromPerry’s description of sequenced growthin the epistemological basis of commit-ment is “how do the range of positionsrelate to moral development?” Thiscould be asked in a stage by position wayor more stochatistically across a range ofpositions and stages. Theoretically, thereseems to be some agreement that at theupper reaches of moral development thedistinction between personal concernsand interpersonal moral concerns maycollapse. This may be expressed as theself becoming a legitimate moral con-cern alongside moral concern for others(Gilligan, 1982), as a perceived unitybetween self and others (Staub, 1993), asa perceived unity between the self andmorality (Colby & Damon, 1993), or as aKantian autonomous willing of themoral in principled moral judgment thatunites moral judgment and personalinterest (Armon, 1989; Kohlberg,1984a).9 Wren (1993) notes that theimpersonal demands of justice are para-doxically deeply part of the self and itsneed to step outside itself. Perry’sdescription of existential commitment incontextual relativism would likewise col-lapse the distinction between personaland interpersonal moral concerns bothby emphasizing that the individual is

responsible for creating values (cf.Abelson & Nielson, 1967, p. 107) and byemphasizing the capacity for meta-thought to reach across value systems toengage others in authentic and dynamicinterchanges (cf. Armon, 1989).

As our above description of the contestedrelationship between ego and moraldevelopment attests, this loose agreementon the integration of morality and self-interest (where the moral is now various-ly understood to include, subsume, or beenlightened self-interest) does not extendto less advanced stages of moral and egodevelopment. Even at the upper reachesof theories of development, the looseagreement on integration of personal andmoral interest is largely hypothetical dueto the dearth of empirical observations ofthis level of development in the popula-tion. At lower stages of development,Nisan (1990, 1993, 2000) describes evi-dence for a kind of “limited morality” or“moral satisficing” whereby individualsallow personal interests to override whatthey accept as the priority of a moral obli-gations as long as such an indiscretiondoes not upset their sense of moral bal-ance, which requires they not allow theirperceived morality go below a certainlevel.10 Colby and Damon (1993) reportthat the integration of moral and selfinterests begins toward the end of child-hood and that the degree of integrationvaries on into adulthood. Evidence that itis not predominantly a post-conventionalphenomena includes Colby and Damon’s(1992) finding that 23 exemplars ofmorally committed lives had MJI scoreswell distributed across stages 3 to 5,though also, only two were scored stage 3.

In general, it seems that most investiga-tors, including Kohlberg, believe that thedevelopment of moral motivation to do


what is right is at least partly linked tothe development of the self or ego.Kohlberg (1984a) articulated this moti-vation as a function of the developmentof autonomy and responsibility.11 At thesame time, there is general disagreementon how to conceptualize the relationshipbetween ego, moral, and cognitive devel-opment. Kohlberg has strongly arguedfor the distinct developmental structureof moral judgment, as a separate compo-nent and strand of development within aunifying ego. But, as we have seen,developmental psychologists hold arange of positions on how to identify anddefine domains of development andtheir relationships with one another.

Empirical Limitations toDisentangling Related

Developmental Domains

One barrier to making sustained theoret-ical progress in understanding domainsof development is insufficient empiricalstudy of how measures derived from dif-ferent theories relate to one another.Empirical studies have generally relateddiscrete pairs of developmental meas-ures (e.g., Lee & Snarey, 1988;Loevinger, 1979; Rest, 1979a) or usedsmall samples (e.g., Commons et al.,1989; King, Kitchener, Wood, & Davison1989; King & Kitchener, 1994). 12

However, Mentkowski and Associates(2000) recently reported findings from abattery of longitudinally administeredhuman potential measures for a largesample. They found two broad factorsthat replicated across the four occasionsof measurement from college entrance tofive years after graduation.13 One factorwas Critical Thinking and the other wasIntegration of Self in Context, Althoughexternal critics of developmental theoryhave suggested that all developmental

differences might be reduced to generalintelligence (e.g., Sanders, Lubinski, &Persson Benbow, 1995), these findingssupport the distinction between cogni-tive development and development ofmeaning-making or self-processes.

The Critical Thinking factor included theTest of Cognitive Development (Renneret al., 1976) and three subscales fromWatson and Glaser (1964) CriticalThinking Appraisal (inference, recogni-tion of assumptions, and deduction).The Test of Cognitive Development(TCD) is based on Piaget’s theory of for-mal operations. The Integration of Selfin Context factor included the Measureof Intellectual Development andLoevinger's Sentence Completion Test.The Measure of IntellectualDevelopment (MID) was constructed inthe context of Perry’s (1970) theory ofintellectual and ethical development,and the Sentence Completion Test (SCT)was constructed in the context ofLoevinger’s theory of ego development.

Appendix A shows the factor loadings fortwo of the four times of assessmentreported in Mentkowski and Associates(2000). 14 Moreover, Mentkowski andAssociates found that Kohlberg’s MoralJudgment Interview (Colby et al., 1987)15

also loaded on the Integration of Self inContext factor, reinforcing the interpre-tation that this factor reflected develop-ment of the deep moral self. Appendix Bdisplays the factor loadings for theanalysis that included the MJI.16

These findings are important to develop-mental theory because they help definetwo distinct domains of development inrelation to a range of measurements withdiffering origins and formats. In particu-lar, the empirical overlap of measures


derived from diverse developmental the-ories of moral judgment (Kohlberg), syn-thetic ego functioning (Loevinger), andepistemological schemes of commitmentin a relativistic world (Perry) supportsarguing for a core commonality amongthe theories that empirically rides abovethe differing theoretical stances andexpressions. In part, this may more nar-rowly reflect general agreement onbroad hierarchical distinctions in adultdevelopmental for the levels in whichthey are typically achieved in society, asthe highest stages of these developmen-tal schemes were not scored. At the sametime, the replication of two factors in thecontext of such a restriction of range ofscores even more strongly supports thedistinction between structural develop-ment of the self and structural develop-ment of cognition, a distinction that hasbeen held more firmly by some develop-mental psychologists (e.g., Loevinger)than others (e.g., Kegan).

A further finding reported byMentkowski and Associates (2000) sug-gests that the widely used Defining IssuesTest (Rest, 1979a) may be “somewhereintermediate between critical thinkingand moral development” (p. 115), as theP% index from this six story measureconsistently loaded ambiguously on thetwo factors across time. This findingillustrates the need to use multiple meas-ures of latent constructs in the process ofclarifying developmental theory.

Although the factor analyses replicatedthe same two-factor solution acrossoccasions of assessment, it is possible toquestion the robustness of the ability togeneralize from these replications. Thefour occasions of assessment spannedten years, but each of the four factoranalyses substantially overlapped

because the individuals were part of thesame longitudinal sample. The presentpaper empirically focuses on the replica-tion of the two factor solution in a dis-tinct sample of Alverno College students

METHOD

Instruments

Sentence Completion Test

The SCT is based on Loevinger’s (1976)theory of ego development. This theorycomprehensively subsumes moral devel-opment, cognitive development, inter-personal relations, character develop-ment, and personality developmentunder the construct of ego development,specifying nine stages. The SentenceCompletion Test consists of 36 sentencestems, such as “My mother and I...” or“Education,...” which participants areasked to complete (Hy & Loevinger,1996; Loevinger & Wessler, 1970;Loevinger, Wessler, & Redmore, 1970).Response to each stem is independentlyrated as manifesting one of the nine lev-els. Estimates of inter-rater reliability(pearson r) summarized across the threetimes of assessment in the Alverno longi-tudinal study was .94 (n=75). Internalconsistency estimates averaged .89.

Measure of IntellectualDevelopment

The MID is intended to assess epistemo-logical development as described byPerry (1970). Participants write threeshort essays describing (a) a best class,(b) a recent decision, and (c) their careerchoice (Knefelkamp, 1974; Knefelkamp& Slepitza, 1976; Widick, 1975). The rat-ing procedures and criteria, initiallydeveloped by Knefelkamp (1978), wereextensively modified during use at

Disentangling Related Domains of Moral, Cognitive, and Ego Development 11

Alverno College (Mentkowski, Moeser, &Strait, 1983; Reisetter Hart, Rickards, &Mentkowski, 1995). Most refinementshave focused on the nature of knowledgeand learning (Mines, 1982; Moore,1983), but we have found that the role ofthe self and the decision-making processhave also emerged as important(Reisetter Hart, et al., 1995). Inter-raterreliability in the Alverno LongitudinalStudy was .82 for essay A, .78 for essayB, and .75 for essay C.

The Defining Issues Test

The DIT is intended to be a measure ofmoral reasoning (Rest, 1979a, 1979b,1986a); it is based on Kohlberg’s (1969)theory of moral development. Restdeparts from Kohlberg’s “hard stage”model, arguing for a more complexmodel that presumes “stage mixture.”Participants are presented with sixmoral dilemmas; for each, they are askedto choose among alternative considera-tions for making a moral judgment. Theresulting “P percent” score, which weused in our analyses, denotes the relativeimportance a person gives to principledmoral considerations. Rest’s criteria forconsistency and meaningfulness werenot used to exclude data from the repli-cation sample. Rest (1979a; 1979b) esti-mates both inter-item consistency (.77)and test-retest reliability (.82).

Critical Thinking Appraisal

The CTA (Watson & Glaser, 1964) is ameasure of critical thinking abilities thathas been widely used to measure collegeoutcomes. With a recognition format,the instrument tests for several compo-nents of critical thinking skills through aseries of multiple choice exercises. Weadministered three of five subscales

Inference, Recognition of Assumptions,and Deduction. They report low alphareliability estimates for the subscales(.55, .54, .41) respectively.

Test of Cognitive Development

The TCD developed by Renner et al.(1976) focuses on Piaget’s most complexstage of cognitive development, formaloperations (Inhelder & Piaget, 1958).Participants completed five paper-and-pencil tasks two tasks in proportionality,two in conservation, and one on separa-tion of variables—which elicit writtenjustifications for answers. The scoringkey was provided by McBer andCompany (Klemp, circa 1977), and wasrevised for rescoring at Alverno College(Schwan Minik, Rogers, & Ben-Ur,1994). In the Alverno LongitudinalStudy, estimates of alpha reliability forthe TCD ranged between .62 and .68.Inter-rater reliability for the total scorewas estimated at .84.

Sample

The present report draws from data col-lected in the Alverno College LongitudinalStudy (Mentkowski & Doherty, 1983), butnot reported on by Mentkowski &Associates (2000). The included partici-pants in the present analyses were not eli-gible for longitudinal analyses eitherbecause they only completed the batteryof assessments at entrance to the college(n=114, listwise deletion) or else becausethey completed the battery as part of across-sectional comparison group ofgraduates (n=54, listwise deletion).Because the structure of the factors hadnot varied across time in the prior analy-ses, these two samples were combined forthe purposes of maximizing the samplefor confirmatory factor analyses. PRELIS


FIGURE 1: CONFIRMATORY FACTOR ANALYSIS BASE MODEL

CriticalThinking

Integrationof Self inContext

DIT

CTAInference

CTAAssumptions

CTADeduction

TCD

MIDBest Class

MIDDecision

MIDCareer

SCT

Disentangling Related Domains of Moral, Cognitive, and Ego Development 13

inputted raw data into LISREL, withmissing data treated pairwise (n rangingfrom 174 to 378). Approximately 25% ofthe sample was missing the CTA and 50%the MID, which accounts for the differ-ences between the listwise and pairwisesample sizes. The mean age at time ofassessment was 23 with a standard devia-tion of 7.7. All were women who enteredor graduated from Alverno College in thelate 1970s.

Confirmatory Factor AnalysisModel

The results of four exploratory factoranalyses reported by Mentkowski andAssociates (2000) were used to specify aconfirmatory factor analysis. Theexploratory factor analyses they reportedconsistently supported a two factor solu-

tion (both scree plots and eigen valuesgreater than 1). Figure 1 displays the con-firmatory factor analysis model tested.Specified indicators of Critical Thinkingwere the TCD and three CTA subscales.Indicators of the Integration of Self inContext latent factor were the SCT andthree MID essays. In addition, the basemodel included the Defining Issues Test(DIT) as an indicator of both factors,which directly tests whether it is an indi-cator of either or both correlated latentconstructs.17 As Figure 1 shows, LISRELalso estimated error for the multiple indi-cators including the DIT.18

RESULTS

Age at the time of assessment tended tohave a small correlation with the human


Latent Factor Indicators

Indicators

CriticalThinking Path-

Coefficients t-values

Integration ofSelf in Context

Path-Coefficients t-values

CTA-Inference .77 7.15

CTA-Assumptions .43 4.72

CTA-Deduction .57 5.93

TCD .63 Referent

MID-best class .53 4.17

MID-decision .66 4.55

MID-career .54 4.22

SCT .44 Referent

DIT-P% .38 2.04 .26 1.33ns

Table 1. Coefficients for Indicators and t-values in the Confirmatory FactorAnalysis

Note. The analysis reported in this replication sample uses data that did not meet criteriafor inclusion in sample frames used by Mentkowski and Associates (2000).The replicationsample includes those who only completed assessments at Time 1 in the Alverno CollegeLongitudinal Study (who took the inventories as freshman) and those from a cross-sectionalcomparison sample (who took the inventories as graduating seniors). After listwise deletion,the sample size is 168. The MJI was not included in the factor analysis due to limited sam-ple sizes for the MJI.

potential measures. Pearson correlationsof age with the indicators of CriticalThinking—CTA Inference (.23), CTARecognition of Assumptions (.19), CTADeduction (.20) and TCD (–.02)—weresimilar to those for the indicators ofIntegration of Self in Context—MID-A(.19), MID-B (.23), MID-C (.14), and SCT(.22). The DIT had a similar low associa-tion with age (.20).

The overall fit of the base model wasgood: GFI=.97 and Chi-Square(25df)=27.6, p=.33. Table 1 displays thecompletely standardized path coeffi-cients from the latent construct to each ofthe indicators and the test of their statis-

tical significance. In this base model, theDIT is a weak indicator for both latentconstructs, but only a statistically signifi-cant indicator of Critical Thinking.Appendix C presents the completely stan-dardized solution for all components ofthe specified model. To better under-stand this finding, it is useful to compareit to alternative models that do not asdirectly test the hypothesis, but that limitoverinterpretation, and suggest howresearchers might construct future meas-urement models. First, if DIT is specifiedas an indicator of Integration of Self inContext (but not Critical Thinking), thenthe overall model fits about the same(GFI=.96, Chi-Square 26df=30.8, p=.24)


Factor

Human Potential Measures Critical ThinkingIntegration of Self

in ContextCritical Thinking Appraisal,Inference .75 .33

Critical Thinking Appraisal,Recognition .44 .10

Critical Thinking Appraisal,Deduction .56 .26

Test of Cognitive Development .52 .43Defining Issues Testa .45 .45Sentence Completion Testb .19 .36Measure of IntellectualDevelopment, Essay A .30 .36

Measure of IntellectualDevelopment, Essay B .16 .70

Measure of IntellectualDevelopment, Essay C .17 .46

Table 2. Factor Loadings for Human Potential Measures for ReplicationSample: Exploratory Factor Analysis

Note. The analysis reported in this replication sample uses data that did notmeet criteria for inclusion in sample frames used by Mentkowski andAssociates (2000). After listwise deletion, the sample size is 168. Orthogonalfactors were extracted using principal axis factoring and rotated through vari-max. The two-factor solution explains 36 percent of the variance in the bat-tery of measures. The MJI was not included in the factor analysis due to lim-ited sample sizes for the MJI. a P percent score.b Total protocol rating.

and the path for the DIT as its indicator is.62 rather than the meager .26 shown inTable 1. However, similar results alsooccur if DIT is specified conversely as anindicator of Critical Thinking (but notIntegration of Self in Context). The over-all model fits about the same (GFI=.96)and now the path from Critical Thinkingto the DIT indicator is .60. This is consis-tent with the findings from the direct testof whether the DIT is an indicator of bothlatent factors, and affirms the prior con-clusion of an ambiguous relationship.

Table 2 presents a follow-up exploratoryfactor analysis on this replication sam-ple. This analysis used principal axis fac-toring and Varimax rotation. The follow-up exploratory factor analysis is notindependent of the confirmatory analy-sis and is presented in order to explorethe stability of the solution and to facili-tate a visual comparison with the factoranalyses Mentkowski and Associatespresented for each of the four times ofassessment in the longitudinal study(Appendix A shows two of these four fac-tor analyses). The factor analysis inTable 2 generally replicates the fourprior exploratory factor analyses, withthe two factor solution again supported(eigen’s values 3.45, 1.03, .83, .64, .41).Although the Test of CognitiveDevelopment was an effective indictor ofCritical Thinking in the confirmatoryanalysis, it did not load as cleanly on thisfactor in this exploratory analysis (whichin this case was post-confirmatory) as ithad previously done for Mentkowski andAssociates. The Best Class essay (EssayA) from the MID also did not as cleanlyload on Integration of Self in Context.Nonetheless the overall factor structurewas similar to that found by Mentkowskiand Associates. 19 The DIT again loadedambiguously on the two factors, .45 on

each factor, Thus, both the confirmatoryand exploratory factor analyses contin-ued to indicate that the DIT is statistical-ly adding little clarity to the measure-ment of these two latent developmentalconstructs.

Two measurement models now presentthemselves as ways to fit the data, and asa guide to future research. One modelwould pursue parsimony and eliminatethe DIT as a measure of either construct.Such a model fits the data well, Chi-Square, 19df= 21.1, p= .33). It is prefer-able to the prior models both in terms ofparsimony and its lower Chi-Squarevalue, even though it is not directly com-parable (see Hayduk, 1987, p. 172). But,another model that represents the DITas an indicator of a third intermediatelatent construct might be warranted asan interim solution to representing therange of empirical data. Here, moralreasoning is represented as distinct notonly from critical thinking but also froma more deeply structural moral develop-ment inherent in the construct ofIntegration of Self in Context. This is thestrategy used by Mentkowski andAssociates (2000), and it fits the presentdata just as well as the other models thatincluded the DIT. Whether the parsimo-nious two-factor or speculative three-factor model is used, there is a strongcorrelation among the constructs.

If the parsimonious two factor model isused, then correlation between CriticalThinking and Integration of Self inContext is estimated to be .80. Unlikeordinary correlations between observedvariables (in any linear combination),which underestimate true correlations,this is an “estimated correlation betweentwo latent unobservable variables”(Joreskog & Sorbom, 1989, p. 90). If the


still unsupported three factor model isused and DIT is estimated to have a reli-ability of approximately .80, this yieldsan almost identical correlation amongthe two clearly identified factors, and asomewhat smaller estimate for theirrelationship to moral reasoning (seeTable 3).

DISCUSSION

The present paper crossvalidates a twofactor model of Critical Thinking andIntegration of Self in Context, andexplores alternative measurement mod-els, acknowledging that the data can beaccounted for by more than one theory.Mentkowski and Associates (2000) find-ing of a distinct developmental factormarked by ego, epistemological/ethicalcommitment received crucial support.The present study’s estimation of the cor-relation between the two latent con-structs (.79) is even higher than thatreported by Mentkowski & Associates(.37 and .42), who also found some evi-dence for a divergence among thedomains after college. By modeling theseconstructs with multiple indicators it waspossible to statistically disentangle theirdistinctiveness and relatedness. Thevalue of disentangling these constructs isthat one is not reduced to the other. Thishas great educational implications, asdifferent factors appear to support thedevelopment of Critical Thinking and

Integration of Self in Context (seeMentkowski & Associates, 2000). Forexample, Mentkowski and Associates(2000) observed that in PerspectivesInterviews participants attributed theirpersonal growth to engaging multipleapproaches to learning and the perspec-tives of others. Likewise statistical causalmodeling indicated that breadth ofpreparation in college was an indirectcause of after college growth forIntegration of Self in Context.

Critical Thinking and Integrationof Self in Context

The confirmatory factor analyses repli-cated the two developmental factorsidentified by Mentkowski andAssociates (2000). The exploratory fac-tor analysis of the replication samplealso recovered the two factor structure.The exploratory factor analysis of thereplication sample provided onlyambiguous support for the stability ofthe TCD as an indicator of CriticalThinking. One interpretation mightfocus on the theoretical distinctionsbetween formal operations and a broaddomain of critical thinking tasks, thatmight, for example, be solved usingbroad knowledge content in the con-crete operational domain. This interpre-tation might link the more ambiguousloading to the role of formal operationsas a presumptive precursor to the devel-


Table 3. Correlations Among Three Latent Developmental Constructs withDIT as Indicator of Moral Reasoning

Critical ThinkingIntegration of Self

in Context Moral Reasoning

Critical Thinking 1.00

Integration of Self in Context .79 1.00

Moral Reasoning .64 .61 1.00

opment of Integration of Self in Context.However, the TCD consistently loadedon the Critical Thinking factor in thelongitudinal sample reported byMentkowski and Associates (2000) andit is a strong indicator of the CriticalThinking in the confirmatory factoranalyses. Thus, the lack of stability inthe exploratory analyses (which invitenonsensical factorial solutions) maysimply reflect unreliability of measure-ment and the high correlation betweenthe two latent constructs. In the longitu-dinal study, estimates alpha reliabilityfor the TCD ranged between .62 and .68.Even though the inter-rater reliabilityfor the total score was estimated at .84,the replicability of an individual’s scoreswith a broader sample of formal opera-tions tasks would not be high. Theexploratory analyses are noteworthy asthey again establish the ambiguousloading of the DIT, showing that thefinding is not dependent on our specifi-cation of the confirmatory model.

The confirmatory factor analyses also sup-ported the latent construct of Integrationof Self in Context identified byMentkowski and Associates (2000). Theybroadly interpreted this factor as a deepstructural and developmental construct.The replication of the loadings for thePerry measure and measure of ego devel-opment was interpreted in light of thebroad overlap in the theories. Theyobserve that “both theories shed light ondifferent aspects of the same largerprocess of increasing intellectual complex-ity and self-differentiation, which corre-sponds to the individual’s self-definingand integrative relationship with the widerworld” (p. 114). They also used the loadingof the MJI on this factor as evidence thatthis “deep structure of the self entails aparticular kind of moral self” (p. 115).

Ethical Aspect of Integration of Self

The present replication of the Integrationof Self in Context factor did not testwhether the MJI again loaded on this fac-tor. Thus, it is useful to see whether wecan infer an ethical dimension of person-al growth from the present findings. Thismeans articulating some kind strand ofnormative structure in the measuredconstructs. On the one hand, the ethicaldimension of the Integration of Self inContext is suggested by both Loevinger’sand Perry’s theories. Loevinger’s meas-ure, in particular, includes prompts thatelicit moral concerns scores for a dimen-sion of impulse control and characterdevelopment as an aspect of overall per-sonal growth. Most of the women in thereplication sample were scored at I-3/4(self-aware) or I-4 (conscientiousness),39 percent and 35 percent , respectively.The self-aware level is the adult normtransition from a much less frequentlyscored (5 percent) position of I-3 (con-formist), which implies orienting oneselfin relation to conformity to external rulessanctioned by social groups. The self-aware individual recognizes a greatermultiplicity and differentiation of stan-dards, while the conscientious individualhas internalized or self-chosen standardsthat focus on consequences and long-term goals and ideals (Loevinger, 1986).A substantial minority (11 percent) werescored for I-4/5 (individualistic), whichincludes a greater respect for individuali-ty. This path of development into self-awareness (and beyond) seems to imply adevelopmental change in orientation tonorms, a change that leads toward theemergence of ideals that have motiva-tional force in the person.

On the other hand, this moral dimensionof growth in relation to the orientation to


group norms is not separated out from abroader development of the self.Moreover, the measure derived fromPerry’s theory does not generally elicitorientations to ethical/norm content orexplicitly score for moral development,in particular, at the ranges of adultdevelopment generally scored in thisstudy. In general,, the essays were scoredas already reflecting some degree oftransition from position 2 (46 percent)or already in position 3 (36 percent).This suggests that they were either in theprocess of abandoning the idea that amultiplicity implied error or hadembraced the legitimacy of multiplicityof viewpoints in some areas as the bestthat can be done for now. Such growthimplies a kind of moral courage inrestructuring ones meaning-makingstance, but does not particularly focus onorientations to group norms. This con-trasts with the kind and level of mean-ing-making process implied byLoevinger’s measure. The characteriza-tion of development from the Loevingermeasure includes changes in normativeorientations, as well as achievement ofpostconventional levels not generallyobserved in these Perry essays.

In this context, the replication of a par-tial overlap of Integration of Self inContext with Rest’s measure of post-con-ventional moral judgment can be inter-preted as evidence for a moral aspect tothe deep structure of the development ofthe self. By itself, however, this replicat-ed finding leaves ambiguous the con-struct of moral self as an aspect of a uni-tary process of ego development. First,as a recognition measure. post-conven-tional moral reasoning on the DIT doesnot imply the same kind of commitmentto moral principles as normative ideals.Second, the partial loading of Rest’s

measure on the Critical Thinking factormight be interpreted as suggesting thatmoral reasoning, as a developmentaldomain, falls intermediate betweenIntegration of Self in Context andCritical Thinking.

If the DIT is seen as intermediatebetween ego and cognitive development,this could suggest that the developmentof moral judgment is separate from andpartially isomorphic with cognitivedevelopment and ego development,which would be in proximity toKohlberg’s position. This interpretation,however, runs against Mentkowski andAssociates (2000) finding that theKohlbergian MJI loaded cleanly onIntegration of Self in Context rather thanwith the DIT. If the development ofmoral judgment entailed a separate anddistinct structure (as Kohlberg suggest-ed), then the DIT and MJI would pre-sumably load together, even if they werepartially isomorphic with other domains.

We interpret the full set of indicators(MJI, SCT, MID) reported byMentkowski and Associates (2000) asdisconfirming Kohlberg’s insistence onthe “absence of an ego or self in the con-struction of the stages” (Kohlberg &Ryncarz, 1990, p. 205). AlthoughKohlberg conceptually abstracts the“epistemic subject” from the “function-ing ego of the self,” the findings castdoubt on the presumption that the “logi-cal” progression of moral judgment isunentangled with existing motive struc-tures and other content. The subjectiveexperience of the search for commit-ments, meaning, and moral stanceappear to be empirically intertwined, asLoevinger suggests. Indeed, the structur-al transformation of motives in relationto normative constructions that


Kohlberg disavowed may be key toreconstructing his theory (cf. Noam,1990) and to accommodating these find-ings. The loading of the MJI onIntegration of Self in Context providesessential support to a developmentalconstruct of a moral self. As Blasi (1976,p. 44) noted:

Kohlberg recognizes that one can ask:Why should I be moral? But he considersfacing and answering this question asanother stage…. This question, however,needs to be answered in some way at eachstage to translate to moral action. The cri-teria that an individual resorts to in hissearch for an answer (not necessarilyexplicitly and consciously) cannot bestrictly formal but must refer to the con-tent of his personality.

In such a reinterpretation, Kohlberg’sdistinction between hard stages and softstages and concomitantly between struc-ture and content would not fully hold.Kohlberg’s epistemic subject behind thecognitive construction of moral struc-tures does not remain distinct from thefunctional ego and its subjective motivesand commitments (cf. Noam, 1990).Instead of Kohlberg’s distinct moraloperational structures, the unity ofmeaning-making as a developmentalstructure seems to empirically entanglemoral and personal development inadulthood. This apparent unity of thesestructural aspects of moral and egodevelopment generally occurs withoutthe benefit of having generally achievedthe upper reaches of development on anyof the measures.

Granting that the evidence supportsLoevinger’s position that moral develop-ment is an inseparable strand of thedevelopment of the ego, does not, how-ever, mean that Kohlberg’s position is

without merit. For example, Kohlberg’sdistinction between structure and con-tent may remain useful as a normativeideal in idiographical assessment ofmoral competence (Lind, 1985). Ournomothetic analyses do not challengesuch a position. Lind (1985) has madethe related argument that the normativeideal for moral structure exists as a theo-retical requirement rather than beingsubject to an empirical conclusion.There are also compelling philosophicalreasons for keeping a normative mean-ing for “ought” where it necessarilyremains as an opening to the possibilityof critiquing any existing state of affairs(e.g., Boyd, 1986).

With that said, the nomothetic data gen-erally supports Loevinger’s empiricalclaims: namely, the distinction of a cog-nitive developmental domain from abroad and synthetic meaning-makingdomain of ego functioning that includesmoral development as an inseparablestrand (cf. Mentkowski & Associates,2000). The cognitive developmentaldomain is distinct from this broad devel-opmental domain of meaning-making,but also highly related as indicated bythe even higher correlation between thelatent constructs in the present replica-tion. This may account for how the Textof Cognitive Development in the oneanalysis did not load as cleanly on theCritical Thinking factor. This perhapsanomalous result and the high correla-tion between the latent factors supportthe idea of a partial isomorphism oroverlap of this broad ego domain withthe cognitive domain. This interpreta-tion would not go as far as asserting asingle underlying developmental pro-cess, but would be more compatible withKegan’s position.


Moral Reasoning and theConsolidation of Moral

Development

The task of disentangling such highlyrelated developmental domains is great-ly facilitated by a large sample, multi-trait, multimethod approach. In thiscontext, the findings for the MJI report-ed by Mentkowski and Associates(2000) are both enlightening and some-what ironic. They found that key featuresof structural moral development appearto be an aspect of the broader structuraldomain of the self, which is distinct frombut highly related to a domain of cogni-tive competence. This means thatKohlberg’s rigorous and sustained dis-tinction between structural operationsand content in cognitively definingmoral structure appears to have createda measure that is a useful indicator of alatent construct of the self that inter-twines structure and content. But, theirony is even greater. Rest (1979, 1986b)abandoned a strong distinction betweenmoral structure and content, but thestructurally “purer” MJI loaded moreclearly on a developmental factorladened with content scoring cues.

It is conceivable that if additional instru-ments were included, the DIT might loadon a third intermediate factor, separatefrom the MJI. Rest (1979a) maintainedthat the DIT and MJI are measuring dif-ferent constructs, but what is the differ-ence? The structure versus content dis-tinction confronts the empirical ironynoted above. Rest et al. (1999) haverecently argued that the DIT has somepotential to measure a broader frame ofpostconventional thinking than the MJI,but how does this account for the DIT notloading with MJI on an even broaderconstruct of self? Rest (1979a) has also

concluded that the “the DIT can beregarded as tapping the earlier and moretacit moral understanding of a subjectwhile the Kohlberg test taps the moreconsolidated understanding and thatwhich the subject can put into words” (p.158). Is it possible that such a consolidat-ed understanding in the domain of moraljustice is why the MJI appears to meas-ure the development of a self system?

Before we explore the idea of consolidat-ed understanding as a developmentalindicator of growth of a moral self, wemust ask, does the DIT systematicallyconfound the identified latent constructsof Critical Thinking and Integration ofSelf In Context? In other words, does theDIT measure a mixture of the two fac-tors, rather than something distinct?There is some evidence in the AlvernoLongitudinal Study that the DIT mightbe distinct from the cognitive and egodomains. Growth on the DIT leveled offafter graduation (or more positively wasmaintained), while the indicators ofIntegration of Self in Context (SCT, MJI,and MID) and the indicators of CriticalThinking (TCD and CTA) showedupward growth five years after college(see Mentkowski and Associates, 2000).Still, the argument that the DIT meas-ures something distinct needs to explainwhy the MJI loaded on the Integration ofSelf in Context construct, while the DIThas consistently only partially done so.

Mentkowski and Associates (2000) sug-gested that the MJI taps into the “moralaspect” of the “deep structuring of theself, an individual’s way of morally con-structing and being in the world. Thismoral development construct contrastswith a more narrow moral reasoningconstruct, which we locate in the DIT, assomewhere intermediate between criti-


cal thinking and moral development” (p.115). A further theoretical articulation ofsuch an intermediate moral reasoningconstruct is needed. In particular, Rest’spost-conventional moral schemas some-how need to be related to but distinctfrom a structural dimension of personaland moral development. Although Restdid not articulate such a set of theoreti-cal constructs, one of the distinctions hemade between the MJI and the DIT maybe useful here. He suggested that theMJI might measure the developmentalstructure of moral judgment through theconsolidated understandings of verbalconstructions, while the DIT mightmeasure structure through the compre-hension of and appreciative preferencefor post-conventional moral judgmentarguments (cf. Rest, 1979; Rest et al.,1999). If DIT measured schemas are notbroadly consolidated understandings,but only sketchy, albeit readily evoked,thought processes, they may have a loos-er relationship to the motivational struc-tures of ego development. Otherwiseput, Rest’s moral schemas do not seemto fully capture the construct of the indi-vidual’s emerging self-direction.20

Instead moral reasoning on the DITmight be a more specific (if still macromorality) developmental construct of“social knowledge structures” refractedthrough post-conventional justiceschemas. Such moral reasoning schemaswould be a more specific (if still macro)construct relative to the moral develop-ment of the self. Depending on one’smetapsychology, these moral schemasmight be seen as purer cognitiveprocesses and capacities or, perhaps, anarrowing of the structural domain ofmeaning-making relative to a constructof a deep moral self. Either articulationof an intermediate position for moralreasoning, may have as yet unarticulated

implications for the four componentmodel offered by Rest and his col-leagues. Even though the componentmodel has the more specific purpose ofaccounting for the relation betweenmoral judgment and behavior, its con-ception of moral motivation may need toalso account for the changing structureof personal and moral commitment.

Whether the DIT measures a more spe-cific or confounded construct, or some-thing else, it has proven itself useful.Even if it confounds cognitive develop-ment with moral dimensions of egodevelopment, it may do so usefully. Thefactorially identified cognitive and egodomains are, after all, related. The DIThas proven useful as a particularly sensi-tive and easily administered measure ofcollege outcomes (Mentkowski andAssociates, 2000; Rest, 1986a; Rest &Narvaez, 1994). The present findingssupport the sustained argument by Restand his colleagues that the DIT cannotbe reduced to intelligence or cognitivefunctioning. At the same time, the find-ings may support the call for supple-menting DIT studies of moral develop-ment more often with just as easilyadministered cognitive measures. Suchmeasures would enable investigators tocontinue to investigate and potentiallydisentangle the relationship betweenmoral reasoning and cognitive develop-ment. Such a disentangled relationshipmight be a way of indirectly and readilyassessing the developmental structure ofa moral self with greater precision.

Moral Self as DevelopmentalStructure

As we more carefully disentangle thedomains of development and the dimen-sions of the person, our language must


more sensitively track nuances of mean-ing. Mentkowski and Associates (2000)used the construct of the moral self topoint to the developmental structure ofthe self apparently reflected in the MJI.Such an empirically defined moral selfreflects an underlying congruence ofstructure that needs to be articulated.Aside from the stage by stage congru-ence, what seems to be the structuralcenter of such a moral self?

At the broadest level the factorial con-vergence of the SCT, MID, and MJI scor-ing systems may imply a broadly com-mon articulation of the individual’sprogress toward creating self-directedframeworks of meaning-making. Giventhe widely diverging instrumentationand the collection of constructedresponses, the convergence must be seenas more than just broad agreement oncriteria, however. Across the differentformats, content domains, and diversescoring cues, the individual’s develop-mental construction of meaning-makingalso converged.

One structurally binding feature of self-direction that may unite the moral andego domain is the developmentalrestructuring of commitment as articu-lated by Perry. The existing ego structurebecomes transformed through its con-frontation with multiplicity and the needto take a stance. Such a need to take astance when living in pluralistic worldconfronts the demands of rationality andit limitations. Given disagreementamong respected authorities andrespected peers accompanied bydemands for rational choices, the indi-vidual gradually delegitimates the utilityof searching for timeless external truthsaccessible to authorities (dualisms). Atthe levels of development observed

through the MID for this population ofcollege students, standards drawn fromthe uncritical acceptance of externalauthority and absolutes still holds signif-icant, if reduced, sway. Our impressionis that the MID underestimates develop-mental progress, however. A substantivepercentage of students probably aremore fully confronting the implicationsof contextual relativism and its demandsfor self-direction and ways of thinkingthat are capable of coordinating interro-gation of multiple systems of thought.

Perry observed that getting to this devel-opmental threshold may be avoidedthrough various forms of retreat to morelimited ways of thinking. Developmentmight also be arrested through escapethat exploits multiplicity or the intellec-tual tools or contextual relativismthrough a cavalier “anything goes,” or adisassociated value cynicism, or mereintellectual gamesmanship. Self-direc-tion in constructivist theory confrontsthese various forms of relativism pre-cisely because they are prevalent formsof constructed epistemologies in a plu-ralistic society. Habermas (1996)describes a similar kind of escape intorationalistic cynicism. AlthoughKohlberg’s articulation of his stage theo-ry has features that tend toward founda-tionalism, it was also contextually sensi-tive (Rest et al, 1999) and aware of thedevelopmental entanglements with rela-tivism, his stage 4½.21 Nonetheless,what Perry articulated more than otherswas the implications of the individual’sgrowing awareness of relativism to theirpotential for self-direction and an exam-ined commitment to personal and moralideals. Kohlberg’s stage theory articula-tion of broadening social perspectivesprovides another window into self-direc-tion and examined commitments as


defining features of a moral self. Indeed,the integration of Kohlberg’s stage theo-ry with his theories of the developmentof autonomy and responsibility could beusefully pursued.

Alverno College Longitudinal Studyobserved growth on the MJI (seeMentkowski and Associates, 2000) thatgenerally moved from a perspective onnorms as coordinated expectations forthe obligations of interpersonal relation-ships (stage 3) toward an emergingcapacity to take the perspective of a gen-eralized member of society, wherenorms are perceived to promote cooper-ation across members of a society and somust be maintained (stage 4).22 Relatedto Kohlberg’s stage 4, Kegan (1982) hasdescribed the emergence of what he callsthe institutional self. “The institutionalbalance wrests the self from the contextof interpersonalism and brings into abeing an ‘I’ that has, rather than is, itsrelationships. This evolution brings intobeing the self as a form or system. Itsstrength is its capacity for self-regula-tion, its capacity to sustain itself, to par-ent itself—its autonomy” (p. 222). Theindividual’s construction of Kohlberg’ssocial systems morality (stage 4) maytend to pull forward the features ofautonomous morality that Tappan et al.(1987) describe and that overlap withPerry’s existential stance. And, it mayalso tend to pull forward the distinctlynormative criteria for autonomousmorality that Perry did not articulate,such as “universality,” the willingness togeneralize moral rules to everyone. Suchan extension into noncommon criteria issurely speculative, but also acknowl-edges a range of features that Integrationof Self in Context might include as partof a moral self.

We suggest that increasing self-directionis a central feature to the developmentalstructure of a moral self. A number oftheorists have used the related term“autonomy” to point toward the individ-ual’s emergence from embeddedness,whether in interpersonal relationshipsor external absolutes. Historically, thishas been an important term.23 Meyers(1989) argues that the capacity forautonomy involves skill in inquiring intoone’s wants, needs, concerns, and valuesthrough autobiographical retrospection,detection and reconciliation of internalconflicts among them, and identificationwith preferred components of the self.Loevinger has used the term autonomyto point toward an advanced develop-mental stage that features a respect forthe autonomy of the person. Beauchampand Childress (1994) give applied exam-ples of the application of this as a moralprinciple in medical decision-making.But, growth toward autonomy can alsobe more generally defined in terms ofincreasing self-direction and examinedcommitment. Kegan emphasizes the“authorship” of the institutional self andMeyers the centrality of self-definition,which is an active construction ratherthan a rejection of unfavored features ofone’s socialization. We believe this kindof growth in self-direction is compatiblewith increased relatedness to others (cf.Belenky, Clinchy, Golberger, & Tarule,1986; Perry, 1981; Josselson, 1988;Mentkowski & Associates, 2000).

Mentkowski and Associates (2000)found in Perspectives Interview data thatfeatures of independent learning devel-oped through the Alverno curriculumconsolidated into self-directed learning,where students and alumnae came to tac-itly understand their own role in con-structing their experience. Engaging


multiple perspectives or other forms ofbreadth seemed to spur this kind ofgrowth. For example, breadth of alumnaactivities (linked to breadth of prepara-tion in the curriculum) was found to be acause of alumna growth in Integration ofSelf in Context. Thus, we suggest that thedevelopment of a moral self may includenot only self-direction, but also a con-comitantly broadened perspective. Thisconclusion is bolstered not only by theobserved growth in the MJI toward asocial systems perspective, but also byalumna growth on the SCT into the stageof conscientiousness. Conscientiousnessincludes both an orientation toward self-evaluated standards and also long termgoals and ideals.

Moral Self, Moral Ideals, andMoral Identity

The capacity to organize one’s lifearound commitments to abstract idealsseems to be a structural feature of thegrowth of self-direction (Blasi, 1993;Frankfurt, 1993; Loevinger, 1976; Perry,1970). Frankfurt’s (1993) articulation ofthe “unthinkable” is particularly usefulbecause it links the sense of self to thenecessary limits of autonomous volition.“A person’s ideals are the concerns thathe cannot bring himself to betray.” Theyare, he pointed out, similar to the forceof love. Like ideals, “the importance ofloving to us would be lost if we could lovesomething or cease to love it merely bydeciding to do so” (p. 24). The “unthink-able” includes constraints in the self thatcorrespond to an incapacity to allowoneself to even try to not will the ideal.Such ideals are part of a person’s self-definition, without which the personwould not be who he or she is. Suchideals may focus on self-perfection orother personal concerns and still be

moral because they are felt to be obliga-tory. They may also be moral ideals thatare defined interpersonally or in relationto society. For example, Noddings(1984) observed that “an ethic of caring,then, seeks something; it seeks to main-tain caring itself” (p. 107). She observed:

The ethical ideal as I have described itsprings from two sentiments; the naturalsympathy human beings feel for eachother and the longing to maintain, recap-ture, or enhance our most caring and ten-der moments. Both sentiments may bedenied, and so commitment is required toestablish the ethical ideal. We must rec-ognize our longing for relatedness andaccept it, and we must commit ourselvesto the openness that permits us to receivethe other. (p. 104)

Obligatory ideals defined interpersonallyor in relation to societal norms, orsupererogatory ones defined in relationto improving society, are what layper-sons are usually concerned with whenthey, as civic minded members of socie-ty, become concerned with “morals”.Colby and Damon (1992, 1993) studiedthe lives of 23 exemplary moral leaders,individuals who had dedicated their livesto making the world a better place. Colbyand Damon (1992) observed that thesemoral exemplars maintained a dynamicrelationship between openness to con-tinued growth and sustained moral com-mitments. Because the exemplars hadorganized their lives around these moralcommitments, self and morality wereunited in them, such that they did notfeel conflict between their personalinterests and their moral commitments.Or, rather, moral interests were theirinterests. Colby and Damon called thisuniting of the moral and personal the“moral self.” Whereas Colby and Damondescribed the moral self as the develop-


mental achievement of commitment tomoral ideals, we use the term to refer toa dimension of adult growth that reflectsthe structural entanglement of moraland personal development. For us, themoral self describes a development fromunexamined commitments to concretepurposes to reflectively examined com-mitments to ideals. This acknowledgesthat other ideals of self-perfection aremoral, as Colby and Damon (1992, pp.297–298) agree.

Colby and Damon (1993) suggested thatthe degree to which the individual uni-fies personal interests with moral con-cerns about improving society “dependsmore on his or her sense of self than onthe nature of moral beliefs” (p. 151).Here they acknowledge that what theycall the achievement of a moral self maybe organized around different moralcontents. They also argued that“Morality and the self are separate con-ceptual systems that grow somewhatcloser together during the course of nor-mal development but that still tend toremain relatively uncoordinated formost (but not all) individuals” (p.151).We would agree that, for most people,personal ideals do not fully coalesce withmoral ideals for improving society, butalso observe that our findings suggestthat moral development and develop-ment of the self are structurally unified.The capacity to commit to relativelyabstract ideals does seem to be an adultdevelopmental milestone, and the par-ticular content of the ideals may varyfrom the “personal” to the moral andreflect differing kinds of unities. At thesame time, various kinds of moral idealsmay be fertile ground for identity. Blasi(1993) argued:

To experience something akin to theManagement of Identity mode, a personmust see her central concern as ideals tobe pursued. For instance, one’s body orone’s sexuality, as fundamental as theyare for a person’s sense of self, do notseem to provide an adequate basis forgenerating lifelong projects or construct-ing higher identity modes. By contrast,morality seems to offer an ideal groundfor anchoring the successive forms ofexperienced identity. (p. 118)

Like Perry (1970), Blasi (1993) suggestedthat what changes might be more care-fully described, not as now having com-mitment to ideals (or concerns), but,instead, as the manner of one’s relation-ship to the ideals that anchor one’s iden-tity and the meaning that they have.Though individual commitment alwayscarries moral import even when it is nar-rowed to concrete personal interest, thesubjective process of taking personalresponsibility for these is developmen-tally structured and is expressed inideals that are both examined andunthinkable to violate. Cognitively andcontextually constructed ideals are con-joined with volitional necessity.Understood this way, the “moral self”reflects a common developmental struc-turing of increasingly self-determinedcommitment. The total commitment ofmoral exemplars to ideals grounded in aconcern for others might be understoodas a “moral identity” that becomes possi-ble once self-determination becomesconsolidated. Our characterization ofthis increasing self-determination forthis sample includes a broadened per-spective, an enlarged self-process, andan identification with societal norms andideals (connecting motivational trans-formation to Kohlberg’s stage 4). Fromthis point of view, the power of the jus-tice “operations” described by Kohlberg


and of the ideal norms of communicativediscourse described by Habermas isreflected in a psychology of commit-ment. In both cases the commitmentincludes a cognitive basis in sustainedtrains of thought that when suffusedwith belief and existential commitmentbecome cognitively equivalent to thevolitionally unthinkable. Thus, the com-monly observed moral commitments tothe existing norms of society have theirpower because individuals have beencompelled by their own reasoning (andtheir performative stance in dialogue) tointernalize them. But, such an achieve-ment of long-term ideals does not meanthat individuals will organize their lifegoals around a moral identity. Otheridentities are possible, and indeed likely(cf. Nissan, 2000). These identitieswould be connected to cultural framesand social interaction as Colby andDamon suggest. Constructivism is, ofcourse, compatible with interactivesocial processes where ideas and frame-works may have been first appropriatedfrom others and where the developmentfor this reasoning may have been scaf-folded through dialogue (cf. Haste,1993).

Development, Self-Reflection, andIdentity

Mentkowski and Associates (2000)found that structural growth inIntegration of Self in Context (includingits moral aspect) was related to breadthof alumna activities. They also foundthat students and alumnae attributedtheir sense of personal growth as anindependent learner to their experiencesin engaging diverse views and diverseapproaches to learning. The ability-based curriculum, its learning processes,the cultural milieu of the college, and

workplace experiences were related tothe student and alumna’s experience ofgrowth into a related set of identities,such as a learner, an effective performer,a collaborator, a well-rounded person,and a professional. Growth was also tiedto self-reflection. For example, studentsand alumnae said that a particularlypowerful contributor to their personalgrowth was the range of experiences inthe curriculum where they examinedtheir own and others values. Althoughthese various identities seemed linked tothe structural development of the self,Mentkowski and Associates also sawthem as linked to a qualitatively identi-fied domain of self-reflection. Self-reflection as a domain can be under-stood as the products and process of asecond-order reflection on who one is. Inaddition to the developmentally struc-tured subjective experience of self (cf.Blasi, 1988, 1993), self-reflection takesthe self as object (cf. James, 1892; Mead,1913; Mentkowski & Associates, 2000).Self-reflection as a domain includes thenarrative mode of experience that is theground of identity. Mentkowski andAssociates’ description of the domaindrew on social constructionist and othercontextually oriented theories (e.g.,Bruner, 1986; Cole, 1996), but alsoresembles in some ways Noam’s (1990)description of themata, the biographicalstructures of the self. Mentkowski andAssociates described the emergence ofprofessional and collaborative identity inrelation to the college’s educationalprocess and cultural milieu.24 UnlikeNoam, Mentkowski and Associates didnot retrospectively articulate life-spancontinuities in identity, nor did theydescribe biographical constraints indevelopment. These differences lie, inpart, in the differing goals of developingeducational versus therapeutic theory.


But, like Noam, they did articulate theorganizing function of narratively con-structed identity and broached develop-mental connections with structures ofthe person. And, like Noam, they point-ed toward a larger theory of the self thancan be adequately described by develop-mental stage theories.

An Educational Theory of thePerson

Mentkowski and Associates (2000)identified four domains of growth in theperson. We only briefly describe themhere as the domains, the transformativelearning cycles that dynamically connectthem, and the in depth treatment of theempirical evidence in depth are in thebook by Mentkowski and Associates.

The domains of reasoning and develop-ment were empirically distinguished bythe Critical Thinking and Integration ofSelf in Context factors. Performance wasempirically distinguished as a third fac-tor. Five-year alumna performance wascoded through a Behavioral EventInterview methodology originally devel-oped by David McClelland. The fourability factors identified in workplaceperformance were validated in relationto faculty judgments of effectiveness(Rogers & Mentkowski, 1994) and indi-cators of career achievement(Mentkowski & Associates, 2000). Onlyone of the four ability factors correlatedwith the measures tapping the structuraldomains of reasoning and development.This performance factor, CollaborativeOrganizational Thinking and Action,loaded as a third factor distinct fromCritical Thinking and Integration of Selfin Context (see Mentkowski & Asso-ciates, 2000; Rogers, 1994). Like thequalitatively identified domain of self-

reflection, this performance domaininvolves interactions of person and con-text that become central to defining thedomains. The four identified domainsare organized according to two dimen-sions. The domains of reasoning anddevelopment are similar in that theyreflect developmentally structured inter-nal congruences in ways of thinking ormaking meaning. This contrasts with thestructural interaction of person and con-text that constitutes the domains of self-reflection and performance. Alonganother dimension, the domains of rea-soning and performance are similar inthat they have an external focus on com-petence, while the domains of develop-ment and self-reflection are similar inthat they have an internal focus onmeaning. The four domains of growthand their interdependencies define thewhole person.

This enlarged description of the wholeperson has implications for a theory ofmoral development. For example,Developing Others and PerspectiveTaking was one of the four factors in theperformance domain. The abilities inthis factor correspond to prosocial skillssuch as coaching, positive regard, per-ceptual objectivity, relational influence,accurate empathy, and sensitivity toindividual differences. Rather than justcapacities, these are abilities demon-strated in discretionary performance,and so, reflect underlying habit struc-tures.25 Because these dispositional abili-ties are empirically distinct from thestructurally defined factors of CriticalThinking and Integration of Self inContext, theories of the development ofcare will need to be developed thatdynamically relate the multidimensionaland depth aspects of abilities, whichinclude skill, knowledge, disposition,


and self-images. We believe the struc-ture of these abilities is linked to stabili-ties in ongoing learning, including iden-tities they have developed in college(Rogers, 1999). For example, theCollaborative Organizational Thinkingand Action Factor seemed linked to notonly breadth of preparation in theAlverno curriculum, but the develop-ment of an identity as a collaborator dur-ing college and is likely supported by theongoing learning after college that par-ticipative leadership affords.

The educational theory of the personthat we offer can serve the task of broad-ening and integrating a range of moraleducation strategies. It is useful here toreflect on some of the historical originsof developmental theory in relation toeducation. Loevinger, Kohlberg, Perry,and other developmental psychologistshave been influenced by the shadow ofWorld War II. While the political forcesof fascism were defeated, the danger ofits reemergence in our own midst werekeenly felt. Loevinger and Perry devel-oped theories that shared a developmen-tal reinterpretation of the authoritarianpersonality, making it amenable to edu-cational intervention. Kohlberg (1984a)observed that more autonomous partici-pants in Milgram’s studies were muchmore likely to disobey the immoraldirections of an authority figure. Thesepost-war contributions reflect an insightof its time and should be honored assuch. Attention to the development ofthe person’s capacities for self-direction,moral judgment, appreciation of individ-uality, and examined commitmentsshould remain a goal of society.Although developmental sophisticationmay not fully protect us from evil(Noam, 1990; Staub, 1993) it is also agood beyond its political rationale.

Indeed, the political rationale was notexplicitly advanced by Loevinger andPerry. At the same time, a focus ondevelopmental structure leaves a lot out.The education of deeply dispositionalprosocial abilities and identities has itsown contribution to making a better andmore just world.

EPILOGUE

The particular ability factors and identi-ty themes found in Alverno College stu-dents and alumnae reflect broad congru-encies that are to some extent deter-mined by the context of the college andpost-college settings, as well as by broad-er demands in work, personal and civiclife (Mentkowski and Associates, 2000).Alverno’s particular ability-based cur-riculum (Alverno College Faculty,1976/1992), and its student assessmentprocesses (Alverno College Faculty,1979/1994) are relatively distinct inhigher education. The explicit integra-tion of content and abilities (knowingand doing) in the curriculum seems to beimportant to creating a transformativelearning process (metacognitive strate-gies as frameworks for performance)that connects reasoning and perform-ance (Mentkowski and Associates,2000).

Many of the abilities that structure thecurriculum, such as valuing in decision-making (Alverno College Faculty,1992;Earley, Mentkowski, & Schafer, 1980) areof particular relevance to the broad goalsof moral education. The curriculumembedded process of self assessment atAlverno (Alverno College Faculty, 2000)and the valuing ability may be of particu-lar importance to creating a transforma-tive learning cycle (self assessing role per-formance) that connects performance


and self-reflection. The integration of theliberal arts and professions educationmay be important to the third learningcycle identified (engaging diverseapproaches, views, and activities) thatconnects self-reflection with develop-ment. Because the four domains ofgrowth seem to be connected by thesethree transformative cycles of learning,we imagine growth of the whole persondynamically. Growth in one domain cansupport another. And, while the person’sdevelopmentally structured capacities asa meaning-maker may condition howthey generally understand their experi-ence, we do not have to depend on thesometimes distant hope of achieving anadvanced structural integration thatunites postconventional moral reasoning,deep appreciation of human relation-ships, individuality, and autonomy intomoral commitment. Instead, we can cele-brate paths to moral compassion andmoral identity available at conventionalstages of moral development. Here we canencourage the development of prosocial

and interceptive abilities and the forma-tion of prosocial and broadening identi-ties as learner, performer, collaborator,and professional. And we can celebrateand encourage the integration of thesestrands of growth into broader functionalsystems that call forth an enlivenedpotential for commitment to moral idealsas well as one’s of self-perfection and thegood life But, we also can organize cur-riculum for the deep structural develop-ment of the moral self, so that these idealscan take root in a broadened socio-moralperspective on ideals for organizing socie-ty. Here, we can encourage students toengage the kind of breadth of learningthat Mentkowski and Associates founddoes lead to developmental growth of theperson’s ways of making meaning. And,we can imagine and reimagine whatmight be transformative paths for themoral self into ever broader, more just,more sensitive, more caring, more appre-ciative, and more meaningful ways ofbeing in the world.


ENDNOTES

1 Kohlberg (1986) accommodatively wrote “LikeLoevinger, I believe that there is a broaddomain of ego development distinguishablefrom intelligence or cognitive development…and that moral development is one aspect orcomponent of the ego domain” (p. 502).

2 Loevinger (1976) held that her stage conceptionwas “structural; that is, there is an inner logic tothe stages and their progression” (p. 11).

3 Colby and Kohlberg (1987) dropped stage sixfrom the revised scoring manual because ofinsufficient data to define it (see Colby et al.,1987). Only a few individuals have been identi-fied as exhibiting stage six reasoning in theirmoral judgment. The conception of stage sixcontinued to evolve and played a key role inKohlberg’s thinking (Kohlberg, Boyd, &Levine, 1990). In its final articulation, respectfor persons integrated principles of benevo-lence (e.g. care) and justice, enlargingKohlberg’s articulation of the moral domain.Both the existence and theoretical constructionof stage six remains contested (e.g., Puka,1990; Rest et al., 1999). Puka (1990) arguedthat stage five, which focuses on utilitarianprinciples of the greatest good, is highly equi-libriated and more theoretically apt. Stage fivereasoning is itself relatively rare in theKohlbergian interviews, and this itself has the-oretical implications.

4 Kohlberg (1984a; 1986) broadened his rationalreconstruction of the course of the individual’smoral development beyond a Rawlsian per-spective and embraced Jurgen Habermas’ dia-logical or discourse ethics (see Habermas,1990; 1996). The revision of the scoring manu-al noted the concomitant hermeneutical task ofgrasping the first-person perspective of theresearch participants (Colby & Kohlberg, 1987),but was not fundamentally grounded in thelogic of discourse ethics (Habermas, 1996).

5 Kohlberg did emphasize the relation betweenmoral judgment and concrete moral decisionsand actions, holding that moral judgment andaction were more related at the upper levels.Here, the structure of moral judgment affectsthe person (his or her action) rather than theconcrete characteristics of the person affecting

the structure of moral judgment, as Loevingerwould have it.

6 Wren’s (1993) characterization of differencesbetween deontological and ethical theorists isuseful here. “The deontic group, whichincludes not only the juridical conceptions ofmorality but also the proceduralist and intu-itionist conceptions…is so called because itscentral features are keyed to the right action(relatively impersonal features such as justice,criteria of fairness, duties, rights claims, and soon). The ethical group, which includes not onlythe teleological conception but also the variousself actualizing and romantic ones, is so calledbecause its central features are keyed to vari-ous notions of the good (more personallynuanced features such as happiness, self-actu-alization, personal excellence, and other formsof human flourishing), notions that constitutethe ethos or character of one’s culture as wellas of one’s self” (p. 81).

7 Kohlberg was greatly concerned about thoseforms of relativism that denied principledmorality. Kohlberg (1981) observed that “pure-ly conventional people can accept the relativityof the rules of their group because they seeknothing more. Intense awareness of relativity,however, implies a search for, or a dim aware-ness of, universal principles in terms of whichconventional morality seems arbitrary” (p.130).

8 In a more limited sense, all theoretical or expe-rienced conceptions of the self are prescriptiveor normative. As Wren (1993) observes, “theyprovide us with ways of discriminatingbetween successful and unsuccessful attemptsby an individual to live as a human person” (p.87).

9 Nisan (1990) noted that judgment overlapsfrom the personal and moral point of view atKohlberg’s stage 6.

10 “The principle of balanced identity thus allowsfor preference of nonmoral considerations overmoral ones, where under given conditions sucha choice will cause less harm to one’s identitybalance than will the preference from the moralconsiderations. This kind of choice cannot beviewed as ‘alienated,’ since it is the individual’schoice between components of his or her ownidentity” (Nisan, 1993 , p. 264).


11 Noam (1990) observes that Kohlberg’s study ofmoral conduct led him to recognize the need todevelop a theory of moral responsibility as wellas moral judgment. “This interest in moralresponsibility, sensibility, and action led himto develop ideas of the moral self….Kohlbergstated that the subjective experience of moralcommitments forms a unity with moral behav-iors. He defined this unity as the moral self.”(p.364)

12 The factor analyses by Commons et al. (1989)must be considered unstable because of thelow sample size (n =34) and large number ofvariables (n =9). In some ways they differ fromMentkowski and Associates (2000). In partic-ular they found that the SCT loaded on a sepa-rate from factor from the MJI, the Good-LifeInterview (which is a moral developmentalmeasure outside the deontic realm), andCommons and Richards Multisystems Task.The Wechsler Adult Intelligence Scale loadedon a separate factor. In comparison, the sam-ple sizes for the factor analyses of each of thefour times of assessments in Mentkowski andAssociates was 237, 258, 127, and 208, respec-tively, for eight variables.

13 Developmental domains reflect coherences ofthe person across contexts that need to bescaled to scaled to reflect qualitatively changein the overall structure. It is precisely becausethe factor structure was stable across observedadult growth over a ten year period, that wehave confidence in the models of structuralchange articulated in the hierarchical coher-ences scaled in the alternative measures. Inother words, if developmental change is ade-quately scaled in the measures, factor analysisacross measures should show stability acrossstructural growth of the person.

14 The analyses reported in Appendix A corrects adata management error that crept into theanalyses of the first two times of assessment.This error led those who had an exact rawvalue of 1 for meaningless responses to beexcluded, rather than those with 8 or moremeaningless responses. The reanalysis yields asomewhat cleaner factor structure.

15 When the fourth wave of longitudinal data wascollected, an expert scored all of the data inter-leaved from all four waves of assessment usingthe updated Colby et al. (1987) scoring manu-

al. The factor analyses used the weighted aver-age score. Time of assessment was ignored forthe purpose of creating a large enough virtualsample to create potentially stable results.

16 Appendix B expands on the presentation byMentkowski and Associates (2000) by report-ing the loadings for all the variables in theanalysis.

17 All completed DIT protocols were included. Inorder to maximize the sample size, we did notapply criteria for consistency or meaningful-ness. The general desirability of excludingthese protocols is debatable.

18 Because Mentkowski and Associates treatedthe DIT as a single indicator of a distinct con-struct of moral reasoning, they specified itserror as a measure of moral reasoning.

19 The preferred analysis for comparing factorsolutions would be a multi-sample analysis(see Joreskog & Sorbom, 1989). Such an analy-sis would directly test the equivalency of thefactor structures. This has not been conductedat this point because of mundane problems ofgetting a “stacked” data set read by the SPSS-VAX program.

20 Our use of the term “moral reasoning” reflectsnot only historical descriptions of the DIT, butalso mutually supporting causal paths betweengrowth on the DIT and the latent construct ofCritical Thinking. In contrast to the positiveeffect of breadth of alumna activities on alum-na growth in Integration of Self, the achievedgrowth on the DIT at graduation had, if any-thing, a negative affect on the breadth of alum-na activities (see Mentkowski and Associates,2000).

21 Kohlberg’s argument against relativism wasfocused on its skeptical or nihilistic forms(Carter, 1986).

22 This path of growth started in the studentyears continued into the assessment five yearsafter college, where stage 5 was still very rarelyobserved (less than 2 percent). Stage 5 think-ing is rarely scored in an undergraduate popu-lation, and the use of the written interviewmakes it even less likely to be scored.


23 Kant’s use of the term is distinctive andinvolves commitment to universals (Armon,1989). Kant’s use is embedded in a transcen-dental philosophical system, which can createinterpretative discontinuities when it is appro-priated to the discourse of psychologists(Petrovich, 1986).

24 Colby and Damon (1993) theorized that a grad-ual, interactional, and often scaffolded processof socialization caused the transformation ofthe exemplar’s personal goals into moral com-mitments.

25 Davidson and Youniss (1991) propose thatthere is a form of autonomous identity distinctfrom stages of moral development and the

reflective “me” that can be the basis for gener-ating impartial moral judgments. Such anidentity is not the focus of awareness. Rather inthe subjective awareness of the performer (seeDuval & Wicklund, 1972), such a prosocialidentity would be reflected in habit structure.Hoffman (1991) suggests potential interactionsbetween empathy and the internalization ofmoral principles (as normative ideals). Gibbs(1991) proposes an integration of Kohlberg’scognitive developmental theory withHoffman’s moral socialization theory. In theintegration, the theories have distinct roles butare also related.

26 See Mentkowski and Associates (2000) for factoranalyses using Assessments at Times 3 and 4.


Appendix A: Factor Loadings for Human Potential Measures forAssessments at Times 1 and 2 Previously Reported in Mentkowski andAssociates: Exploratory Factor Analysis

Human Potential Measure

Time of AssessmentTime 1 Time 2

Critical Thinking Integration of Selfin Context


Critical Thinking Appraisal,Inference

.53 .11 .55 .18

Critical Thinking Appraisal,Recognition

.55 –.01 .60 .18

Critical Thinking Appraisal,Deduction

.67 .18 .66 .04

Test of Cognitive Development .53 .32 .55 .13

Defining Issues Testa .38 .38 .52 .37

Sentence Completion Testb .08 .41 .20 .28

Measure of IntellectualDevelopment, Essay A

.03 .46 .24 .56

Measure of IntellectualDevelopment, Essay B

.22 .51 .05 .59

Measure of IntellectualDevelopment, Essay C

.11 .64 .11 .63

Note. The analysis reported in this table varies slightly from Mentkowski & Associates (2000). For the prior analysis,a data management error led to excluding those with an exact raw value of 1 on the meaningless index.

This analysis excluded DIT protocols that exceeded the standard criteria for “meaningless” and “inconsistent.” Thus,“meaningless” raw scores with a value of 8 or above are excluded. Orthogonal factors were extracted using principalaxis factoring and rotated through varimax. The two-factor solution explains between 32 and 35 percent of the vari-ance in the battery of measures.a P percent score. b Total protocol rating.

DC# R00120


Appendix B: Factor Loadings for Human Potential Measures Including theMoral Judgment Interview: Exploratory Factor Analysis Ignoring Time ofAssessment

Human Potential Measure

Factor


Critical Thinking Appraisal,Inference

.51 .21

Critical Thinking Appraisal,Recognition

.49 .08

Critical Thinking Appraisal,Deduction

.69 .09

Test of Cognitive Development .56 .07

Defining Issues Testa .44 .49

Sentence Completion Testb .01 .45

Measure of IntellectualDevelopment, Essay A

.18 .65

Measure of IntellectualDevelopment, Essay B

.14 .71

Measure of IntellectualDevelopment, Essay C

.11 .61

Moral Judgment Interviewc .20 .61

Note. This table expands the presentation of this factor analysis, which was report-ed originally in Mentkowski and Associates (2000). Orthogonal factors wereextracted using principal axis factoring and rotated through varimax. The two-fac-tor solution was supported (eigne values = 3.4, 1.6, .88, .76, .55, .47, .32) andexplains 38 percent of the variance in the battery of measures.

The analysis excluded DIT protocols that exceeded traditional standards for“meaningless” and “inconsistent” values. The MJI was not included in the primaryfactor analyses due to limited sample size. For this subanalysis, we ignored time ofassessment in order to create a virtual sample of 175. a P percent score.b Total protocol rating.c Written form of interview. Participants write answers to a structured set of ques-tions that probe their reasoning in relation to moral dilemmas.This analysis usesdata scored using the revised scoring manual for the MJI (Colby et al., 1987).

DC# R0090


Appendix C: Fitted Coefficients for Base Confirmatory Factor Analysis

CriticalThinking

Integrationof Self inContext

DIT

CTAInference

CTAAssumptions

CTADeduction

TCD

MIDBest Class

MIDDecision

MIDCareer

SCT

.77 .41

.43

.57

.63

.82

.68

.60

.53

.66

.54

.44

.72

.57

.71

.81

.26

.64

.38

.80


REFERENCES

Abelson, R., & Nielsen, K. (1967). History ofethics. In P. Edwards (Ed.), The encyclopedia ofphilosophy: Vol. 3, (pp. 81–117). New York:Macmillan Publishing.

Adorno, T. W., Frenkel-Brunswik, E., Levinson,D. J., & Sanford, R. N. (1950). The authoritarianpersonality. New York: HarperCollins.

Alexander, C. N. & Langer, E. J. (Eds.). (1990).Higher stages of human development. NewYork: Oxford University Press.

Alverno College Faculty. (1976/1992). Liberallearning at Alverno College. Milwaukee, WI:Alverno College Institute. (Original work pub-lished 1976, revised 1981, 1985, 1989, and 1992)

Alverno College Faculty. (1979/1994). Studentassessment-as-learning at Alverno College.Milwaukee, WI: Alverno College Institute.(Original work published 1979, revised 1985 and1994)

Alverno College Faculty. (1992). Valuing in deci-sion-making: Theory and practice at AlvernoCollege. Milwaukee, WI: Alverno CollegeInstitute.

Alverno College Faculty. (2000). Self assessmentat Alverno College. (G. Loacker, Ed.).Milwaukee, WI: Alverno College Institute.

Armon, C. (1989). Individuality and autonomy inadult ethical reasoning. In M. L. Commons, J. D.Sinnott, F. A. Richards, & C. Armon (Eds.), Adultdevelopment: Vol. 1 (pp. 179–196). New York:Praeger.

Beauchamp, T. L., & Childress, J. F. (1994).Principles of biomedical ethics (4th ed.). NewYork: Oxford University Press.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R.,& Tarule, J. M. (1986). Women’s ways of know-ing: The development of self, voice, and mind.New York: Basic Books.

Blasi, A. (1976). Concept of development in per-sonality theory. In J. Loevinger, Ego develop-ment: Conceptions and theories (pp. 29–53).San Francisco: Jossey-Bass.

Blasi, A. (1988). Identity and the development ofthe self. In D. K. Lapsley & F. C. Power (Eds.),Self, ego, and identity: Integrative approaches(pp. 226–242). New York: Springer-Verlag.

Blasi, A. (1993). The development of identity:Some implications for moral functioning. In G.G. Noam & T. E. Wren (Eds.), The moral self (pp.99–122). Cambridge, MA: MIT Press.

Boyd, D. R. (1986). The oughts of is: Kohlberg atthe interface between moral philosophy anddevelopmental psychology. In S. Modgil & C.Modgil (Eds.), Lawrence Kohlberg: Consensusand controversy (pp. 4–63). Philadelphia:Falmer Press.

Bruner, J. S. (1986). Actual minds, possibleworlds. Cambridge, MA: Harvard UniversityPress.

Carter R. (1986). Does Kohlberg avoid rela-tivism? In S. Modgil & C. Modgil (Eds.),Lawrence Kohlberg: Consensus and controver-sy, (pp. 9–20). Philadelphia: Falmer Press.

Colby, A., & Damon, W. (1993). The uniting ofself and morality in the development of extraor-dinary moral commitment. In G. G. Noam & T. E.Wren (Eds.), The moral self (pp. 149–174).Cambridge, MA: MIT Press.

Colby, A., & Damon, W. (1992). Some do care:Contemporary lives of moral commitment. NewYork: Free Press.

Colby, A. & Kohlberg, L. (with Abrahami, A.,Gibbs, J., Higgins, A., Kauffman, K., Lieberman,M., Nisan, M., Reimer, J., Schrader, D., Snarey,J., & Tappan, M.). (1987). The measurement ofmoral judgment: Vol. 1. Theoretical foundationsand research validation. New York: CambridgeUniversity Press.

Colby, A., Kohlberg, L., Speicher, B., Hewer, A.,Candee, D., Gibbs, J., & Power, C. (1987). Themeasurement of moral judgment: Vol. 2.Standard issue scoring manual. New York:Cambridge University Press.

Cole, M. (1996). Cultural psychology: A onceand future discipline. Cambridge, MA: BelknapPress.


Commons, M. L., Armon, C., Richards, F. A.,Schrader, D. E. (with Farrell, E. W., Tappan, M.B., & Bauer, N. F.). (1989). A multidomain studyof adult development. In M. L. Commons, J. D.Sinnott, F. A. Richards, & C. Armon (Eds.), Adultdevelopment Vol. 1 (pp. 33–56). New York:Praeger.

Davidson, P., & Youniss, J. (1991). Which comesfirst, morality or identity? In W. M. Kurtines & J.L. Gewirtz (Eds.), Handbook of moral behaviorand development: Vol. 1. Theory, (pp. 105–121).Hillsdale, NJ: Lawrence Erlbaum Associates.

Duval, S., & Wicklund, R. A. (1972). A theory ofobjective self-awareness. New York: AcademicPress.

Earley, M., Mentkowski, M., & Schafer, J. (1980).Valuing at Alverno: The valuing process in lib-eral education. Milwaukee, WI: AlvernoProductions.

Fowler, J. W. (1981). Faith and human develop-ment. New York: Harper & Row.

Frankfurt, H. (1993). On the necessity of ideals.In G. G. Noam & T. E. Wren (Eds.), The moralself (pp. 16–27). Cambridge, MA: MIT Press.

Gibbs, J. C. (1979). Kohlberg’s moral stage theo-ry: A Piagetian revision. Human Development,22, 89–112.

Gibbs, J. C. (1991). Toward an integration ofKohlberg’s and Hoffman’s theories of morality.In W. M. Kurtines & J. L. Gewirtz (Eds.),Handbook of moral behavior and development:Vol. 1. Theory, (pp. 183–222). Hillsdale, NJ:Lawrence Erlbaum Associates.

Gilligan, C. (1982). In a different voice:Psychological theory and women’s development.Cambridge, MA: Harvard University Press.

Gilligan, C., & Murphy, J. (1979). Developmentfrom adolescence to adulthood: The philosopherand the dilemma of the fact. New Directions forChild Development: Intellectual DevelopmentBeyond Childhood, no. 5, 85–99. San Francisco:Jossey-Bass.

Habermas, J. (1990). Justice and solidarity: Onthe discussion concerning stage 6. In T. E. Wren(Ed.), The moral domain: Essays in the ongoing

discussion between philosophy and the social sci-ences (pp. 223–251). Cambridge, MA: MIT Press.

Habermas, J. (1996). Moral consciousness andcommunicative action. (C. Lenhardt & S. WeberNicholsen, Trans.). Cambridge, MA: MIT Press.(Original work published 1983)

Haste, H. (1993). Morality, self, and sociohistor-ical context: The role of lay social theory. In G. G.Noam & T. E. Wren (Eds.), The moral self (pp.175–208). Cambridge, MA: MIT Press.

Hayduk, L. A. (1987). Structural equation mod-eling with LISREL: Essentials and advances.Baltimore, MD: Johns Hopkins University Press.

Heider, F. (1958). The psychology of interper-sonal relations. New York: John Wiley & Sons.

Hoffman, M. L. (1991). Empathy, social cogni-tion, and moral action. In W. M. Kurtines & J. L.Gewirtz (Eds.), Handbook of moral behaviorand development: Vol. 1. Theory, (pp. 275–301).Hillsdale, NJ: Lawrence Erlbaum Associates.

Hy, L. X., & Loevinger, J. (1996). Measuring egodevelopment (2nd ed.). Hillsdale, NJ: Erlbaum.

Inhelder, B., & Piaget, J. (1958). The growth oflogical thinking from childhood to adolescence.New York: Basic Books.

James, W. (1892). Psychology: The briefercourse. New York: Holt, Rinehart & Winston.

Jöreskog, K. G., & Sörbom, D. (1989). LISREL 7:A guide to the program and applications (2nded.). Chicago: Authors/SPSS.

Josselson, R. (1988). The embedded self: I and thourevisited. In D. K. Lapsley & F. C. Power (Eds.), Self,ego, and identity: Integrative approaches (pp.91–106). New York: Springer-Verlag.

Kegan, R. (1982). The evolving self: Problemand process in human development. Cambridge,MA: Harvard University Press.

Kegan, R. (1994). In over our heads: The mentaldemands of modern life. Cambridge, MA:Harvard University Press.

King, P. M., & Kitchener, K. S. (1994).Developing reflective judgment: Understanding


and promoting intellectual growth and criticalthinking in adolescents and adults. SanFrancisco: Jossey-Bass.

King, P. M., Kitchener, K. S., Wood, P. K., &Davison, M. L. (1989). Relationships acrossdevelopmental domains: A longitudinal study ofintellectual, moral, and ego development. In M.L. Commons, J. D. Sinnott, F. A. Richards, & C.Armon (Eds.), Adult development: Vol. 1.Comparisons and applications of developmen-tal models (pp. 57–72). New York: Praeger.

Klemp, G. O., Jr. (circa 1977). Test of CognitiveDevelopment scoring key. Unpublished manu-script, McBer & Company, Boston.

Knefelkamp, L. (1974). Developmental instruc-tion: Fostering intellectual and personal growthin college students. Unpublished doctoral disser-tation, University of Minnesota.

Knefelkamp, L. (1978). Training manual forPerry raters and rater training cue sheets.Unpublished mimeograph, University ofMaryland, College Park.

Knefelkamp, L., & Slepitza, R. (1976). A cogni-tive-developmental model of career develop-ment: An adaptation of the Perry Scheme.Counseling Psychologist, 6(3), 53–58.

Kohlberg, L. (1969). Stage and sequence: Thecognitive-developmental approach to socializa-tion. In D. Goslin (Ed.), Handbook of socializa-tion theory and research (pp. 347–480). Skokie,IL: Rand McNally.

Kohlberg, L. (1976). Moral stages and moraliza-tion: The cognitive-developmental approach. InT. Lickona (Ed.), Moral development and behav-ior: Theory, research, and social issues (pp.31–53). New York: Holt, Rinehart and Winston.

Kohlberg, L. (1981). Essays on moral develop-ment: Vol. 1. The philosophy of moral develop-ment: Moral stages and the idea of justice. SanFrancisco: Harper & Row.

Kohlberg, L. (1984a). Essays on moral develop-ment: Vol. 2. The psychology of moral develop-ment. San Francisco: Harper & Row.

Kohlberg, L. (with Higgins, A.). (1984b).Continuities and discontinuities in childhood

and adult development revisited—again. In L.Kohlberg, Essays on moral development: Vol. 2.The psychology of moral development (pp.426–497). San Francisco: Harper & Row.

Kohlberg, L. (with Levine, C., & Hewer, A.).(1984c). The current formulation of the theory.In L. Kohlberg, Essays on moral development:Vol. 2. The psychology of moral development(pp. 212–319). San Francisco: Harper & Row.

Kohlberg, L. (with Levine, C., & Hewer, A.).(1984d). Synopses and detailed replies to critics.In L. Kohlberg, Essays on moral development:Vol. 2. The psychology of moral development(pp. 320–386). San Francisco: Harper & Row.

Kohlberg, L. (1986). A current statement onsome theoretical issues. In S. Modgil & C. Modgil(Eds.), Lawrence Kohlberg: Consensus and con-troversy, (pp. 485–544). Philadelphia: FalmerPress.

Kohlberg, L., Boyd, D. R., & Levine, C. (1990).The return of Stage 6: Its principle and moralpoint of view. In T. E. Wren (Ed.), The moraldomain: Essays in the ongoing discussionbetween philosophy and the social sciences (pp.151–181). Cambridge, MA: MIT Press.

Kohlberg, L., & Ryncarz, R. A. (1990). Beyondjustice reasoning: Moral development and con-sideration of a seventh stage. In C. N. Alexander& E. J. Langer (Eds.), Higher stages of humandevelopment, (191–207). New York: OxfordUniversity Press.

Lee, L., & Snarey, J. (1988). The relationshipbetween ego and moral development: A theoret-ical review and empirical analysis. In D. K.Lapsley & F. C. Power (Eds.), Self, ego, and iden-tity: Integrative approaches (pp. 151–178). NewYork: Springer-Verlag.

Lind, G. (1985). The theory of moral-cognitivejudgment: A socio-psychological assessment. InG. Lind, H. A. Hartmann & R. Wakenhut (Eds.),Moral development and the social environment:Studies in the philosophy and psychology ofmoral judgment and education (pp. 21–53).Chicago: Precedent Publishing.

Loevinger, J. (1976). Ego development:Conceptions and theories. San Francisco:Jossey-Bass.


Loevinger, J. (1979). Construct validity of theSentence Completion Test of ego development.Applied Psychological Measurement, 3(3),281–311.

Loevinger, J. (1986). On Kohlberg’s contribu-tions to ego development. In S. Modgil & C.Modgil (Eds.), Lawrence Kohlberg: Consensusand controversy (pp. 183–193). Philadelphia:Falmer Press.

Loevinger, J., & Wessler, R. (1970). Measuringego development: Construction and use of a sen-tence completion test (Vol. 1). San Francisco:Jossey-Bass.

Loevinger, J., Wessler, R., & Redmore, C. (1970).Measuring ego development: Scoring manualfor women and girls (Vol. 2). San Francisco:Jossey-Bass.

Mead, G. H. (1913). The social self. The Journalof Philosophy, Psychology and ScientificMethods, 10, 374–380.

Mentkowski, M., & Associates. (2000). Learningthat lasts: Integrating learning, development,and performance in college and beyond. SanFrancisco: Jossey-Bass.

Mentkowski, M., & Doherty, A. (1983).Careering after college: Establishing the validi-ty of abilities learned in college for later career-ing and professional performance (Final Reportto the National Institute of Education).Milwaukee, WI: Alverno Productions. (ERICDocument Reproduction Service No. ED 239 556to ED 239 566)

Mentkowski, M., Moeser, M., & Strait, M. (1983).Using the Perry Scheme of Intellectual andEthical Development as a college outcomesmeasure: A process and criteria for judging stu-dent performance (Vols. 1 and 2). Milwaukee,WI: Alverno Productions.

Meyers, D. (1989). Self, society, and personalchoice. New York: Columbia University Press.

Mines, R. (1982). Student development assess-ment techniques. In D. Hanson (Ed.), Measuringstudent development. New Directions forStudent Services, no. 20, 65–91. San Francisco:Jossey-Bass.

Moore, W. (1983). The Measure of IntellectualDevelopment: A brief review. Unpublished man-uscript, Center for Applications ofDevelopmental Instruction, University ofMaryland at College Park.

Murphy, J. M., & Gilligan, C. (1980). Moraldevelopment in late adolescence and adulthood:A critique and reconstruction of Kohlberg’s theo-ry. Human Development, 23, 77–104.

Nisan, M. (1990). Moral balance: A model of howpeople arrive at moral decisions. In T. E. Wren(Ed.), The moral domain: Essays in the ongoingdiscussion between philosophy and the socialsciences (pp. 283–314). Cambridge, MA: MITPress.

Nisan, M. (1993). Balanced identity: Moralityand other identity values. In G. G. Noam & T. E.Wren (Eds.), The moral self (pp. 239–266).Cambridge, MA: MIT Press.

Nisan, M. (2000). Manifestations of relativismand individualism in moral judgments of indi-viduals: Implications for moral education. In M.Leicester, C. Modgil, & S. Modgil (Vol. Eds.),Education, culture and values: Vol. 4. Moraleducation and pluralism (pp. 71–82).Philadelphia: Falmer Press.

Noam, G. G. (1990). Beyond Freud and Piaget:Biographical worlds-interpersonal self. In T. E.Wren (Ed.), The moral domain: Essays in theongoing discussion between philosophy and thesocial sciences (pp. 360–399). Cambridge, MA:MIT Press.

Noam, G. G., Kohlberg, L., & Snarey, J. (1983).Steps toward a model of the self. In B. Lee & G.G. Noam (Eds.), Developmental approaches tothe self (pp. 59–141). New York: Plenum Press.

Noam, G. G., & Wren, T. E. (Eds.). (1993). Themoral self. Cambridge, MA: MIT Press.

Noddings, N. (1984). Caring: A feminineapproach to ethics and moral education.Berkeley: University of California Press.

Nucci, L. & Lee, J. (1993). Morality and personalautonomy. In G. G. Noam & T. E. Wren (Eds.),The moral self (pp. 123–148). Cambridge, MA:MIT Press.


Perry, W. G., Jr. (1970). Forms of intellectualand ethical development in the college years: Ascheme. Austin, TX: Holt, Rinehart & Winston.

Perry, W. G., Jr. (1981). Cognitive and ethicalgrowth: The making of meaning. In A. W.Chickering & Associates, The modern Americancollege (pp. 76–116). San Francisco: Jossey-Bass.

Petrovich, O. (1986). Moral autonomy and theo-ry of Kohlberg. In S. Modgil & C. Modgil (Eds.),Lawrence Kohlberg: Consensus and controver-sy, (pp. 85–106). Philadelphia: Falmer Press.

Piaget, J. (1972). Intellectual development fromadolescence to adulthood. Human Development,15, 1–12.

Puka, B. (1990). The majesty and mystery ofKohlberg’s Stage 6. In T. E. Wren (Ed.), Themoral domain: Essays in the ongoing discussionbetween philosophy and the social sciences (pp.182–223). Cambridge, MA: MIT Press.

Rawls, J. (1971). A theory of justice. Cambridge,MA: Harvard University Press.

Reisetter Hart, J., Rickards, W. H., &Mentkowski, M. (1995, Apr.). Epistemologicaldevelopment during and after college:Longitudinal growth on the Perry Scheme. Paperpresented at the annual meeting of the AmericanEducational Research Association, San Francisco.Milwaukee, WI: Alverno College Institute.

Renner, J., Fuller, R., Lochhead, J., John, J.,Tomlinson-Keasey, G., & Campbell, T. (1976).Test of Cognitive Development. Norman:University of Oklahoma.

Rest, J. R. (1979a). Development in judgingmoral issues. Minneapolis: University ofMinnesota Press.

Rest, J. R. (1979b). Revised manual for theDefining Issues Test: An objective test of moraljudgment development. Minneapolis: MinnesotaMoral Research Projects.

Rest, J. R. (1986a). Moral development:Advances in research and theory. New York:Praeger.

Rest, J. R. (1986b). Moral research methodology.In S. Modgil & C. Modgil (Eds.), Lawrence

Kohlberg: Consensus and controversy (pp.455–469). Philadelphia: Falmer Press.

Rest, J. R., & Narvaez, D. F. (Eds.). (1994). Moraldevelopment in the professions: Psychology andapplied ethics. Hillsdale, NJ: Erlbaum.

Rest, J. R., Narvaez, D. F., Bebeau, M. J., &Thoma, S. J. (1999). Postconventional moralthinking: A neo-Kohlbergian approach.Hillsdale, NJ: Erlbaum.

Rogers, G. (in collaboration with Kleinman, L.,Wagner, K., & Schwan Minik, K.). (1994, Apr.).Envisioning alumnae abilities. In M. Mentkowski(Chair), Is higher education adequately meetingits responsibility to its alumni and to society?Interactive paper presented as part of an interac-tive symposium conducted at the annual meetingof the American Educational ResearchAssociation, New Orleans.

Rogers, G. (1999, April). Learning through cur-riculum abilities: College learning at work. In M.Eraut (Chair), Professional learning in theworkplace. Paper presented in a symposium atthe annual meeting of the American EducationalResearch Association, Montreal. Milwaukee, WI:Alverno College Institute.

Rogers, G., & Mentkowski, M. (1994). Alvernofaculty validation of abilities scored in five-yearalumna performance. Milwaukee, WI: AlvernoCollege Institute.

Sanders, C. E., Lubinski, D., & Persson Benbow,C. (1995). Does the defining issues test measurepsychological phenomena distinct from verbalability?: An examination of Lykken’s query.Journal of Personality and Social Psychology,69(3), 498–504.

Schwan Minik, K., Rogers, G., & Ben-Ur, T.(1994). Test of Cognitive Development revisedscoring manual. Milwaukee, WI: AlvernoCollege Institute.

Snarey, J., Kohlberg, L., & Noam, G. (1983). Egodevelopment in perspective: Structural stage,functional phase, and cultural age-period mod-els. Developmental Review, 3, 303–338.

Souvaine, E., Lahey, L. L., & Kegan, R. (1990).Life after formal operations: Implications for apsychology of the self. In C. N. Alexander & E. J.


Langer (Eds.), Higher stages of human develop-ment (pp. 229–257). New York: OxfordUniversity Press.

Staub, E. (1993). Individual and group selves:Motivation, morality, and evolution. In G. G.Noam & T. E. Wren (Eds.), The moral self (pp.337–358). Cambridge, MA: MIT Press.

Tappan, M., Kohlberg, L., Schrader, D., Higgins,A., Armon, C., & Lei, T. (1987). Heteronomy andautonomy in moral development: Two types ofmoral judgments. In A. Colby & L. Kohlberg, Themeasurement of moral judgment: Vol. 1.Theoretical foundations and research valida-tion (pp. 315–380). Cambridge, UK: CambridgeUniversity Press.

Tugendhat, E. (1993). The role of identity in theconstitution of morality. In G. G. Noam & T. E.Wren (Eds.), The moral self (pp. 3–15).Cambridge, MA: MIT Press.

Watson, G., & Glaser, E. M. (1964). CriticalThinking Appraisal (Manual for Forms YM andZM). Orlando, FL: Harcourt Brace.

Westenberg, P. M., Blasi, A., & Cohn, L. D. (Eds.)(1998). Personality development. Hillsdale, NJ:Erlbaum.

Widick, C. (1975). An evaluation of developmen-tal instruction in a university setting.Unpublished doctoral dissertation, University ofMinnesota.

Wren, T. E. (1993). The open-textured conceptsof morality and the self. In G. G. Noam & T. E.Wren (Eds.), The moral self (pp. 78–95)


Documents

Disentangling Related Domains of Moral, Cognitive, and Ego Development