You are on page 1of 14

Journal of Youth and Adolescence, Vol. 34, No. 3, June 2005, pp.

185198 (
C 2005)

DOI: 10.1007/s10964-005-4300-9

The Role of Epistemic Cognition in Adolescent


Identity Formation: Further Evidence

Tobias Krettenauer1

Received November 6, 2003; revised April 5, 2004; accepted April 13, 2004

Following Boyes and Chandler (1992), it is investigated how adolescents epistemic understanding
relates to the process of identity formation. In a cross-sectional as well as longitudinal study, identity
status scores of diffusion, foreclosure, moratorium and identity achievement as assessed by the Ex-
tended Objective Measure of Ego Identity Status were analyzed as a function of adolescents epistemic
stance, while simultaneously taking into account individual differences in identity processing styles.
The cross-sectional sample was composed of 200 German adolescents from Grades 7, 9, 11, and
13 (mean age: 16.2 years, SD = 2.41). The longitudinal study consisted of 134 participants (mean
age at retest: 16.5 years, SD = 1.7) and covered a time interval of 18 months. Findings confirmed
the view that the development of epistemic cognition contributes to adolescent identity formation.
Specifically, the development of a multiplistic epistemic stance predicted higher moratorium scores,
whereas the more advanced evaluativist epistemic position contributed to identity achievement over
time. These effects were not due to individual differences in identity processing style.

KEY WORDS: development of epistemological understanding; identity statuses; identity styles.

INTRODUCTION also in their efforts to achieve a sense of personal iden-


tity. As evidenced by a considerable body of research on
According to a Piagetian perspective, adolescence is identity formation, adolescents often go through a period
characterized by a fundamental developmental shift in the of actively questioning and exploring various possibili-
way young persons understand the relation between real- ties before committing themselves to particular goals and
ity and possibility; whereas children up to the age of 10 ideals (Berzonsky and Adams, 1999; Bosma and Kunnen,
or 12 years consider possibilities as a mere extension of 2001; Marcia et al., 1993). Such a period of exploring op-
reality, adolescents with the advent of formal operations, tions and opinions that have not been considered before,
begin to conceive reality as the realization of a particular often referred to as moratorium2 appears to be essential
possibility (Moshman, 1999; Piaget and Inhelder, 1977).
2 Originally
This transformation gives rise to a new intellectual pos- coined by Erikson (1959), the term moratorium was later
ture. Formal operational thinkers spontaneously and sys- adopted by Marcia to refer to a particular status category within the
famous identity status paradigm (for an overview of this approach
tematically generate possibilities and understand reality in
see Marcia et al., 1993). Although the identity status paradigm has
light of these possibilities. As a consequence, factual ar- been criticized repeatedly from both theoretical and empirical grounds
rangements cannot be legitimated simply by pointing out (Blasi, 1988; Cote and Levine, 1988; Meuus et al., 1999; van Hoof,
how things are, but need to be evaluated in view of how 1999), it still can be considered the leading account of research on ado-
things might be. This intellectual posture characterizes lescents and young adults identity development. In this framework, the
process of identity formation is conceptualized along two dimensions:
adolescents thinking in many areas. It is present not only
the extent to which individuals have explored identity issues and seri-
in their attempts to understand nature, self and society, but ously considered alternative possibilities in matters such as vocation,
religion, and political ideology (identity exploration) and the extent
1 Institute
for Psychology, Humboldt University at Berlin, Germany. To to which individuals are committed with regard to such issues (com-
whom correspondence should be addressed at Humboldt University at mitment). The combination of the two dimensions yields four identity
Berlin, Institute for Psychology, Unter den Linden 6, 10099 Berlin; statuses (identity diffusion, identity foreclosure, moratorium, and iden-
e-mail: tobias.krettenauer@rz.hu-berlin.de. tity achievement). Note that these identity statuses are not stages of

185
0047-2891/05/0600-0185/0 
C 2005 Springer Science+Business Media, Inc.
186 Krettenauer

for achieving an ego-identity that provides some sense of p. 250). On this level different ideas and viewpoints can
unity and direction in life. be evaluated or criticized without appealing to an ultimate
The common theme of considering an actual state of truth. Nonetheless, choosing between different viewpoints
affairs in relation to multiple possibilities that runs through is not considered just a matter of personal preference.
important accounts of adolescent cognitive development Rather, one viewpoint may be preferable on grounds that
as well as identity formation, has led some researchers to can be meaningfully debated about. Researchers used dif-
assume that formal operational thinking may be a prereq- ferent labels to characterize these epistemic stances. Kuhn
uisite of identity exploration. However, empirical findings et al. (2000), for instance, coined the term evaluativism to
regarding this hypothesis were equivocal. Some authors describe the third stance, which is neither absolutistic nor
were able to provide evidence for such an association multiplistic, whereas Boyes and Chandler (1992) referred
(Rowe and Marcia, 1980; Leadbeater and Dionne, 1981), to it as post-skeptical rationalism.
others failed in their efforts to do so (Berzonsky et al., In their study, Boyes and Chandler (1992) postulated
1975; Cauble, 1976). In an attempt to clarify this incon- that adolescents who have not yet moved beyond an abso-
clusive situation, Boyes and Chandler (1992) proposed a lutistic epistemic stance and who, as a consequence, are
different account of the relation between cognitive growth not fully aware of the multiplicity of knowledge claims,
in adolescence and identity formation. The authors pos- will score lower on identity exploration and therefore reg-
tulated another line of development that in spite of being ister as either foreclosed or diffuse on identity status mea-
associated with the development of formal operations was sures. By contrast, adolescents who had moved beyond
taken to be more directly linked to the process of identity multiplism and, therefore, envision the possibility of find-
formation; the development of epistemic cognition. ing good reasons for choosing among different alterna-
The term epistemic cognition refers to individuals tives without denying the multiplicity of options were ex-
understanding of the nature of knowledge (i.e. what one pected to be better able to develop personal commitments
believes knowledge is) and the knowing process (i.e. how within a pluralist worldview. It was postulated that these
one comes to know). Research on epistemic cognition adolescents score higher on identity achievement. Finally,
provided ample evidence for three main stances or lev- adolescents who had achieved a multiplistic epistemic
els that represent important milestones in the course of stance but who had not yet moved further were expected
epistemic development (for an overview, see Hofer and to score high on moratorium or identity diffusion. Re-
Pintrich, 1997; Kitchener, 2002, Moshman, 1999). In an sults of the Boyes and Chandler (1992) study largely con-
early period of development, knowledge is understood as firmed these expectations. Hence the authors concluded
a direct copy of reality. The sole source of knowledge is that there is a close connection between epistemic and
the external world. This epistemic stance can be labeled identity development that clarifies the rather elusive rela-
absolutistic (Kuhn et al., 2000). An absolutist thinker as- tion between operativity and the process of identity for-
sumes that there is always one ultimate truth that is either mation (see also Chandler et al., 1990). Other studies,
directly accessible to everyone or known to experts who while based on slightly different conceptions of epistemic
are supposed to be well equipped for finding out truth. This cognition further corroborated this view (Peterson et al.,
conception of knowledge and knowing changes markedly 2004; Stephen et al., 1992).
once individuals attain the subsequent, multiplistic stance. The finding that the development of epistemic cog-
A multiplist understands knowledge as mainly determined nition is associated with the process of identity formation
by ones personal point of view. Thus, the self is the source in adolescence corresponds with another set of findings
of knowledge. This implies that knowledge is nothing but reported in numerous studies: individuals assigned to dif-
personal opinion and that appealing to truth is equivalent ferent identity statuses evidence different styles of cog-
to imposing ones opinion on others. Finally, these polar nitive processing and decision-making (for an overview
extremes of absolutism versus multiplism are brought into see Adams, 1998). In an effort to synthesize these stud-
a more balanced epistemic stance, in which knowing is ies, Berzonsky (1989) defined three identity processing
a joint function of perspective and reality, and thus si- styles that represent different ways of processing self-
multaneously subjective and objective (Moshman, 1994, relevant information and of coping with identity con-
cerns: (1) an information-orientation, (2) a normative-
development, but refer to different phases in the process of identity for- orientation, and (3) a diffuse/avoidant-orientation. The
mation. This process is not uni-directional. The different phases do not information-orientation is defined by actively seeking
form a strict developmental sequence and individuals may go through
out, processing and utilizing self-relevant information.
them repeatedly. Nonetheless, research provided evidence for an over-
all developmental trend towards self-chosen commitments in terms of When confronted with discordant feedback, information-
identity achievement (Waterman, 1999). oriented individuals are willing to test and to revise aspects
Epistemic Cognition and Identity Formation 187

of their self-theory. Normative-oriented individuals, by formation, these findings should be not considered the
contrast, deal with identity questions and decisions mainly final word, but call for further investigation.
by conforming to the expectations and prescriptions of The present study aims at further disentangling the
significant others. They tend to resist information that role of epistemic cognition in adolescent identity forma-
may threaten values and beliefs central to their self- tion. To pursue this goal, associations between adoles-
concept. The diffuse/avoidant-orientation is defined by cents epistemic stance and identity status scores will be
a general reluctance to face identity relevant questions. analyzed while simultaneously taking into account effects
Avoidant individuals, instead, accommodate to life cir- of identity processing style. These analyses are meant
cumstances by ad-hoc adjustments in their self-definition to examine to what extent the impact of epistemic rea-
rather than by stable changes in the identity structure. soning development on adolescents identity formation
Berzonsky (Berzonsky and Adams, 1999; Berzonsky and depends on individual differences in cognitive process-
Neimeyer, 1994) demonstrated that identity processing ing. While Boyes and Chandler (1992) proposed a di-
styles are systematically associated with identity statuses. rect impact of epistemic reasoning development on the
Foreclosures prefer a normative-orientation in processing process of identity formation, this view has been seri-
identity-relevant information, the diffuse/avoidant style ously challenged by the findings of Klaczynski et al.
is associated with identity-diffusion and the information- (1998).
orientation corresponds with identity achievement. Mora- In addition to the studies described above, that
torium, by contrast, was not substantially linked to any were all cross-sectional and thus not suitable for test-
particular identity style. Strikingly, the different identity ing the direction of effects, in the present study, the re-
styles were found to be associated with various episte- lations between epistemic cognition, identity processing
mological assumptions as well. Berzonsky (1994) pro- style and identity status scores will be investigated both
vided evidence that an information-orientation is asso- cross-sectionally and longitudinally. Thus, to investigate
ciated with a constructivist epistemic worldview. The whether epistemic reasoning development actually con-
normative orientation, by contrast, was related to a tributes to the process of identity formation, and not the
formistic epistemic view according to Kramers (1983) other way round, cross-lagged correlations of longitudinal
framework. data will be examined by means of path-analyses while
The finding that identity processing styles are asso- simultaneously taking into account individuals identity
ciated with identity statuses, on the one hand, and with processing style. In these analyses self-report measures
different epistemic assumptions on the other, put the study of identity processing styles are taken as a proxy of in-
conducted by Boyes and Chandler (1992) in a somewhat dividuals cognitive processing when approaching iden-
different perspective. Maybe the development of epis- tity relevant problems and decisions. Self-report measures
temic cognition is not directly linked to the process of of identity styles have been demonstrated to be empir-
identity formation, as Boyes and Chandler (1992) sug- ically associated with actual cognitive processing (e.g.
gested, but largely due to individual differences in cog- Berzonsky, 1999).
nitive processing and, hence, spurious. This view is sup- Based on previous findings reported by Boyes and
ported by a study of Klaczynski et al. (1998) that was Chandler (1992), Berzonsky (1988) and Neimeyer and
designed to clarify the relation between cognitive devel- Berzonsky (1994), the following findings are hypothe-
opment, cognitive processing and identity formation. The sized to emerge. Adolescents who have not yet moved
authors examined adolescents identity status as a function beyond an absolutistic epistemic stance and who, as a
of formal operations, critical thinking beliefs and rational consequence, have not yet developed a full-fledged un-
processing (cf. Epstein et al., 1996). In a series of hierar- derstanding of the multiplicity of knowledge claims are
chical multiple regressions, rational cognitive processing expected to be less inclined to explore identity issues and,
predicted most identity status scores even when control- therefore, should score higher on identity status scores of
ling for formal operations and critical thinking beliefs. diffusion and foreclosure. By contrast, adolescents with a
According to Klaczynski et al. (1998), this finding sug- multiplistic epistemic stance are expected to score higher
gests a moderating role of rational processing which in on moratorium, whereas adolescents with an evaluativist
their interpretation makes it easier for adolescents to ef- epistemic understanding should evidence greater iden-
fectively use their formal operational competencies in the tity achievement. This pattern of findings is expected to
process of identity formation. However, as these findings emerge both cross-sectionally and longitudinally. Thus,
were based on one single age group, and as Klaczynski adolescents epistemic stance at t1 should predict the cor-
et al. (1998) inadvertently may have capitalized on the responding identity status score at t2 , even when control-
rather elusive relation between operativity and identity ling for identity status at t1 .
188 Krettenauer

An analogous pattern of findings is expected for Research participants volunteered in response to an


identity processing styles. Thus, the various identity advertisement distributed in their schools. Each was given
processing styles should promote different development an honorarium of 15 EURO. For adolescents younger than
outcomes in terms of identity statuses. It is assumed 18 years informed consent for participating in the study
that the diffuse/avoidant-orientation is associated with was given by one of the parents. Participants who were
identity diffusion. A normative-orientation should corre- 18 years old or older had provided informed consent by
late positively with foreclosure, whereas an information- themselves.
orientation is expected to be associated with higher scores
on moratorium and identity achievement. Longitudinal Study
Regarding unique and common effects of epistemic
development and identity processing styles as predictors Participants from grade levels 7, 9 and 11 were fol-
of identity status scores, theoretically, two different pat- lowed up 18 months later. This time interval was chosen
terns of findings are possible. Epistemic stances and iden- so as to obtain developmental change in individuals epis-
tity processing styles may predict identity statuses largely temic reasoning in a sufficient number of incidencies. 89%
independent of each other. Such a finding would indi- of 7, 9 and 11 graders were retested, yielding a longitu-
cate that cognitive development and cognitive processsing dinal sample of N = 134. Of the participants in the lon-
variables contribute autonomously to the process of iden- gitudinal sample 59 were male and 75 female. Mean age
tity formation. Conversely, epistemic stances and identity in the sample at the retest was 16.5 years (SD = 1.7). For
processing styles may share most of the variance they the variables used in the following analyses, no systematic
account for in identity status scores. Such a pattern of effects due to sample attrition were found as evidenced by
findings would suggest that both factors in some way de- a MANOVA procedure, F (8, 140) = 1.12, p = .35. Uni-
pend on each other in contributing to the course of iden- variate F-tests conducted separately for each variable did
tity formation. Certainly, these patterns of findings are not not in any instance reach the probability level of p < .10.
mutually exclusive. That is, some identity statuses might
simultaneously depend on both epistemic development MEASURES
and identity processing style, and others not.
Identity Statuses

METHOD To assess participants identity status a German


translation of the Extended Objective Measure of Ego
Participants Identity Statusrevised version (EOMEIS-2) was used.
The EOMEIS is the most frequently employed standard-
Cross-Sectional Study ized measure to assess ego identity status. It has good psy-
chometric properties and had been successfully translated
The cross-sectional sample of the present study was in other languages (for an overview see Adams, 1998).
composed of 200 adolescents equally distributed over four In the present study, the translation of the EOMEIS from
different grade levels (Grade 7, 9, 11 and 13) with equal English to German was done by the author. To check for
numbers of males and females in each age group. Mean consistency in the meaning of items the German transla-
ages for the four groups were as follows: 13.1 years (SD = tion was back-translated to English by a bilingual speaker
0.62) for Grade 7; 15.0 (SD = 0.39) for Grade 9; 17.2 who had been familiarized with the meaning of the differ-
years (SD = 0.52) for Grade 11; and 19.4 (SD = 0.66) ent identity statuses before.
for Grade 13. The sample was drawn from five differ- The EOMEIS assesses participants identity status
ent high schools (Gymnasium) located in socially hetero- with regard to two domains (ideological and interper-
geneous quarters of East- and West-Berlin. All adoles- sonal). Each domain is represented by four issues (the
cents were Caucasian, most of them were born in either ideological domain, for example, includes the issues pol-
West- or East-Germany (42% and 47.5%, respectively). itics, vocation, religion and philosophical lifestyle). Each
The sample was predominantly middle-class according issue is addressed by two items, yielding a 64-item ques-
to Treimans (1977) Standard International Occupational tionnaire (4 statuses by 2 domains by 4 issues by 2 items).
Prestige Scale, with an average occupational status score Identity achievement with regard to vocation, for instance,
of parents of M = 49.4 (SD = 12.2, range = 2178). The is expressed by the following item: It took me a while to
grand mean of the sample was slightly above the German figure it out, but now I really know what I want for a ca-
population average of 43.5, SD = 12.1 (cf. Wolf, 1995). reer. By contrast, identity diffusion with regard to politics
Epistemic Cognition and Identity Formation 189

reads as follows: I havent really considered politics. It with conflicting views about controversial issues and asks
just doesnt excite me much. Foreclosure with regard to them to explain how these viewpoints can be justified, and
religious issues is expressed by the following statement. how, if at all, they may be reconciled. Interview responses
I have never really questioned my religion. What is right are then scored with regard to their epistemic stance and
for my parents must be right for me. Identity moratorium a total-score is derived (e.g. King and Kitchener, 1994).
is represented by items such as In finding an acceptable In the context of the present study, a methodological de-
viewpoint to life itself, I find myself engaging in a lot of vice similar to this commonly used interview technique
discussions with others and some self exploration. Re- was adopted, with one main difference: the studys as-
spondents are asked to rate the statements on a 6-point sessment of respondents epistemic stance was based on
scale ranging from strongly agree to strongly disagree. a recognition rather than production procedure. For this
Usually, responses are averaged across those issues that purpose a German adaptation of the Epistemic Develop-
belong to one domain, thus obtaining separate identity ment Questionnaire (EDQ; Hallett et al., 2002) was used.
status scores for the ideological and the interpersonal The use of a recognition measure for the assessment of
domain. However, it is possible to aggregate responses adolescents epistemic stance appeared to be appropriate
across domains, as well. According to Adams (1998), because in other research areas, notably in the field of
either discrete identity status categories (as obtained by moral reasoning development, recognition measures have
applying specific classification rules to the raw scores) or proven to provide a viable alternative to interview tech-
scores of continuous scales may be used as identity status niques (Rest et al., 1997).
measures. Because this study focuses on the process of The present studys assessment of epistemic stance
identity formation and developmental change rather than is based on the conceptualization of epistemic develop-
stable individual differences related to identity statuses, ment as outlined in the introduction and adopted as a
continuous scales are used as criterion variables. general framework by most of the studies in this field of
In the present study the EOMEIS was slightly short- research. Three epistemic stances were distinguished: ab-
ened. Because it was not the intention to obtain sep- solutism, multiplism and evaluativism. To assess these
arate scores for the ideological and the interpersonal stances, participants were confronted with conflicting
domain only those issues were included that are nor- viewpoints about 14 controversial issues from different
mally addressed in the identity status interview proce- fields of knowledge (natural sciences, social sciences and
dure (cf. Marcia et al., 1993): vocational choice, politics, humanities). All controversies were about different inter-
religion, philosophical lifestyle, friendship, and gender pretations of physical and social facts. After introducing
roles. Responses were averaged across all issues, yield- the conflicting viewpoints adolescents were confronted
ing four scales that represent participants profiles with with a set of three statements. Each statement expressed
regard to the four identity statuses. Each scale com- one of the three epistemic stances (for item examples
prised 12 items. Internal consistencies (Cronbachs-) see Appendix). To familiarize respondents with the items,
for the scales were sufficient, with .71 (diffusion), .79 they were first asked to rate each of the three statements
(foreclosure), .69 (moratorium), and .74 (identity achieve- on a scale ranging from completely agree to completely
ment). With regard to identity diffusion there was a sub- disagree. Second, they had to pick the one statement out
stantial decrease over the 18-month longitudinal inter- of the set of three, which expressed their own viewpoint
val, t(133) = 11.37, p < .01, Mt1 = 3.24 (SD = 0.39), best. These choices reflect individuals preferred epis-
Mt2 = 2.69 (SD = 0.65). The same was true for scores temic stance on the particular issue. Overall, 52% of par-
reflecting identity foreclosure, t(133) = 8.11, p < .01, ticipants choices reflected their modal choice, indicating
Mt1 = 2.61 (SD = 0.60), Mt2 = 2.53 (SD = 0.63). For a sufficient consistency of individuals epistemic prefer-
moratorium there was a significant decrease, t(133) = ences across issues.
2.42, p < .05, Mt1 = 3.55 (SD = 0.58), Mt2 = 3.42 Respondents were assigned to that epistemic stance
(SD = 0.62), whereas the slight increase of identity they had chosen most frequently. This modal score best
achievement over the 18-months longitudinal intervall represents the notion of a stance that is adopted by an
was not significant, t(133) = 1.81, p < .72, Mt1 = 3.79 individual. Moreover, when developmental competencies
(SD = 0.58), Mt2 = 3.89 (SD = 0.64). are assessed by a recognition rather than a production task
individuals cognitive competence is more accurately re-
Epistemic Stance flected by item-choices or item-rankings rather than by
item-ratings, as demonstrated with regard to moral rea-
Individuals epistemic stance is commonly assessed soning development by Rest et al. (1997). Kuhn et al.
by semi-structured interviews that confront participants (2000) proposed a recognition procedure for assessing
190 Krettenauer

individuals level of epistemic reasoning that was solely for instance, is expressed by the following statement: I
based on item choices. think it is better to have a firm set of beliefs than to be
In the present study, 89% of respondents had their open-minded. The diffuse/avoidant-orientation is repre-
modal score on a single epistemic stance, whereas 21 sented by items such as: Sometimes I refuse to believe
of adolescents (11%) had their modal choices on two that a problem will happen, and things manage to work
adjacent stances. In the latter case participants were as- themselves out. The scale assessing the information-
signed to the developmentally more advanced position. orientation includes statements like: When I have a per-
Only in one instance choices were equally distributed sonal problem, I try to analyze the situation in order to
across all three epistemic stances. This participant was ex- understand it. In addition to these scales, the ISI-3 regu-
cluded from further analyses. In the cross-sectional sam- larly comprises items to assess adolescents commitment.
ple, 24.6% of participants were classified as absolutists These items were not included in the present study.
(n = 49), 24.1% as epistemic multiplists (n = 48), and Internal consistencies of the scales assessing the
51.3% as evaluativists (n = 102). At the retest, 21.6% various identity processing styles were sufficient, with
of participants in the longitudinal sample were assigned Cronbachs = .75 for the diffuse/avoidant-orientation,
to the absolutist epistemic stance (n = 29), 29.9% were .70 for the information-orientation, and .65 for the
classified as multiplists (n = 40) and 48.5% as evalu- normative-orientation. Over the 18-months longitudinal
ativists (n = 65). Performing a Wilcoxon-Test on the interval, the normative-orientation as well as the diffuse/
rank-ordered epistemic stances (1 = absolutism, 2 = mul- avoidant-orientation decreased significantly, normative-
tiplism, 3 = evaluativism) indicated a significant devel- orientation: t(133) = 5.01, p < .01, Mt1 = 3.71 (SD =
opmental growth of epistemic cognition over the longitu- 0.66), Mt2 = 3.45 (SD = 0.66), diffuse/avoidant-
dinal interval of 18 months, z = 1.73, ponetailed < .05. orientation: t(133) = 3.69, p < .01, Mt1 = 3.34 (SD =
In the present sample, adolescents epistemic stance 0.66), Mt2 = 3.13 (SD = 0.68). The information-
was substantially correlated with age, Spearmans rho = orientation, by contrast, increased, t(133) = 3.33, p <
.48, p < .01. Moreover, as demonstrated by Krettenauer .01, Mt1 = 4.21 (SD = 0.62), Mt2 = 4.38 (SD =
(2005) high school students, in general, score significantly 0.70).
lower on the measure than graduate students of philosophy
who can be considered experts in matters of epistemology,
U = 871.5, z = 3.96, p < .01. This finding is consistent RESULTS
with the well-documented fact that epistemic stances are
associated with individuals level of education (King and In the introduction, various hypotheses regarding re-
Kitchener, 1994; Kuhn et al., 2000). Finally, a substantial lations between the development of epistemic cognition,
correlation between epistemic stance as assessed by the identity processing style and adolescents identity status
procedure described above and adolescents epistemolog- were formulated. These hypotheses are now addressed
ical understanding of moral beliefs as assessed by a semi- in three consecutive steps: First, bivariate correlations
structured interview was found, Spearmans rho = .57, between the measures involved in the present study are
p < .01 (Krettenauer, 2005). Taken together these find- examined. Second, hierarchical regression analyses are
ings demonstrate that the validity of the present studys run in order to test whether epistemic cognition and
assessment of participants epistemic stance. identity processing styles predict identity status scores
independently of each other. These regression analyses
are based on cross-sectional data. Third, path analyses of
Identity Processing Style longitudinal data are performed in order to investigate to
what extent epistemic cognition and identity processing
The present studys assessment of participants iden- styles predict longitudinal change in identity status
tity processing style was based on a German transla- scores. In all analyses, epistemic development will be
tion of the Identity Style Inventory3rd revision (ISI-3, represented by two dummy-variables that reflect two
Berzonsky, 1992). Again, the English version of the ISI- distinct developmental achievements: (a) having not yet
3 was translated by the author and back-translated by a moved beyond an absolutist epistemic stance (dummy-
bilingual speaker in order to check for the consistency of contrast A: absolutism = 1 vs. multiplism/evaluativism =
item meanings. The reliability and validity of the ISI-3 0) and (b) having achieved an evaluativist epistemic
has been well-documented (Berzonsky, 1989, 1992). In position (dummy-contrast B: absolutism/multiplism = 0
the ISI-3 each identity processing style is assessed by a vs. evaluativism = 1). In case sample sizes differ from
separate scale of 9 to 11 items. The normative orientation, N = 200 for the cross-sectional sample and N = 134
Epistemic Cognition and Identity Formation 191

Table I. Bivariate Correlations Between Identity Statuses, Epistemic Stance, and Identity Processing Styles

(1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Identity status: Diffusion .45


(2) Identity status: Foreclosure .25 .66
(3) Identity status: Moratorium .31 .14 .52
(4) Identity status: Achievement .27 .14 .13 .57
(5) Epistemic stance: Dummy-contrast A .25 .21 .15 .01 .18
(6) Epistemic stance: Dummy-contrast B .27 .25 .09 .01 .59 .32
(7) Identity style: Normative-orientation .09 .51 .12 .52 .17 .23 .59
(8) Identity style: Diffuse/avoidant-orientation .39 .23 .28 .20 .13 .25 .03 .49
(9) Identity style: Information-orientation .32 .07 .11 .44 .32 .20 .12 .21 .58

Note. Entries in normal type (below diagonal) represent correlations as obtained in the cross-sectional sample (N = 199); entries in italics represent
longitudinal stabilities as obtained in the longitudinal sample (N = 133).
p
onetailed < .05; ponetailed < .01.

for the longitudinal study this is due to missing values adolescents with an absolutist understanding of knowing
in one of the variables. All significance tests regarding and knowledge were more norm- and less information-
associations between epistemic stances, identity process- oriented. By contrast, adolescents who had achieved an
ing styles and identity statuses will be one-tailed since evaluativist position were less norm-oriented and evi-
hypotheses always postulated the direction of effects. denced a stronger information-orientation.

Bivariate Correlations Between Epistemic Stances, Regressions of Identity Statuses on Epistemic


Identity Processing Styles and Identity Statuses Stance and Identity Processing Styles

Table I provides a summary of the correlations be- In the following analyses it was tested whether epis-
tween measures involved in the present study. In line temic stances and identity processing styles predict iden-
with the results reported by Adams (1998), identity status tity statuses independent of each other. Four separate hi-
scores were moderately correlated with each other. The erarchical regression analyses were run, each with one of
strongest association between two identity status mea- the four identity status scores as the criterion measure and
sures of r = .31 was obtained for identity diffusion and epistemic stance and identity styles as predictors. In the
moratorium. Longitudinal stabilities for the identity status first step of these analyses, adolescents epistemic stance
measures over the 18 months interval were all substantial. (as represented by the dummy-contrasts A and B) were
A similar picture emerged for measures of adolescents entered in the regression. In the second step, it was tested
identity processing style. whether identity processing style significantly predicted
Correlations between identity processing styles and identity status scores even when epistemic stance was con-
identity status scores were somewhat stronger. The trolled. In a supplementary set of regression analyses, the
normative orientation was associated with foreclosure order of predictor variables was reversed. Results of these
and with identity achievement. The diffuse/avoidant- analyses are summarized in Tables IIV.
orientation correlated substantively with identity diffu- With regard to identity diffusion a significant effect
sion, whereas the information-orientation was linked with of adolescents epistemic stance emerged (see Table II).
identity achievement. This correlational pattern largely Having not yet moved beyond an absolutistic epistemic
corresponds with findings reported by Berzonsky (1989). stance was predictive of higher scores in identity diffusion.
Moreover, significant correlations between identity sta- Conversely, having achieved an evaluativist epistemic po-
tuses and the dummy-contrasts A and B representing dif- sition was associated with lower identity diffusion scores.
ferent epistemic stances were found: having not yet moved Entering identity processing styles in the second step of
beyond absolutism (dummy-contrast A) was correlated the regression analysis improved the prediction substan-
positively with identity diffusion and foreclosure, and neg- tially. The strongest predictor of identity diffusion was the
atively with moratorium. Conversely, having achieved an diffuse/avoidant-orientation. When controlling for iden-
evaluativist position (dummy-contrast B) was negatively tity processing styles, the dummy variables represent-
correlated with identity diffusion and foreclosure. Epis- ing different epistemic stances did not make a significant
temic stances were correlated with identity styles as well: contribution, indicating considerable overlap of both sets
192 Krettenauer

Table II. Summary of Regression Analysis Predicting Identity Table IV. Summary of Regression Analysis Predicting Moratorium
Diffusion
Predictors t R2
Predictors t R2
Step 1 .07a
Step 1 .09a Absolutism vs. multiplism/ .30 3.57
Absolutism vs. multiplism/ .14 1.67 evaluativism (dummy A)
evaluativism (dummy A) Absolutism/multiplism vs. .27 3.14
Absolutism/multiplism vs. .19 2.25 evaluativism (dummy B)
evaluativism (dummy B)
Step 2 .10b
Step 2 .16b Absolutism vs. multiplism/ .24 2.83
Absolutism vs. multiplism/ .10 1.29 evaluativism (dummy A)
evaluativism (dummy A) Absolutism/multiplism vs. .22 2.65
Absolutism/multiplism vs. .13 1.57 evaluativism (dummy B)
evaluativism (dummy B) Normative-orientation .15 2.20
Normative-orientation .13 1.93 Diffuse/avoidant-orientation .28 4.12
Diffuse/avoidant-orientation .30 4.67 Information-orientation .16 2.22
Information-orientation .18 2.67
a F (2, 196) = 7.21, p < .01.
a F (2, 196) = 9.33, p < .01. b F (3, 193) = 7.86, p < .01.
b F (3, 193) = 13.89, p < .01. p
onetailed < .05; ponetailed < .01.
p
onetailed < .05; ponetailed < .01.

of variables as predictors of identity diffusion. Nonethe- cents epistemic stance did not make a significant con-
less, in a supplementary analysis, when identity styles tribution, R 2 = 0.01, F (2, 193) = 0.94, p = .393.
were entered first in the regression, including epistemic Thus, foreclosure status scores mainly depended on iden-
stance yielded a significant increase in the R-squared, tity processing styles, and not on adolescents epistemic
R 2 = 0.04, F (2, 193) = 4.44, p < .05. stance.
For identity foreclosure, again, a significant effect A different pattern of findings emerged with re-
of epistemic stance emerged, with evaluativists scoring gard to the moratorium status (see Table IV). Epistemic
lower on foreclosure (see Table III). However, this effect stance significantly predicted moratorium status scores.
disappeared when entering identity processing styles in Both dummy-contrasts evidenced negative regression co-
the regression. The strongest predictor of identity foreclo- efficients. Thus, higher moratorium status scores were
sure was the normative-orientation. When identity pro- specifically related to the multiplistic epistemic stance.
cessing styles were entered in the regression first, adoles- Entering identity processing styles in the regression

Table III. Summary of Regression Analysis Predicting Identity Table V. Summary of Regression Analysis Predicting Identity
Foreclosure Achievement

Predictors t R2 Predictors t R2

Step 1 .06a Step 1 .00a


Absolutism vs. multiplism/ .09 1.03 Absolutism vs. multiplism/ .01 0.09
evaluativism (dummy A) evaluativism (dummy A)
Absolutism/multiplism vs. .20 2.36 Absolutism/multiplism vs. .01 0.05
evaluativism (dummy B) evaluativism (dummy B)

Step 2 .25b Step 2 .44b


Absolutism vs. multiplism/ .05 0.63 Absolutism vs. multiplism/ .09 1.29
evaluativism (dummy A) evaluativism (dummy A)
Absolutism/multiplism vs. .05 0.69 Absolutism/multiplism vs. .04 0.59
evaluativism (dummy B) evaluativism (dummy B)
Normative-orientation .49 7.92 Normative-orientation .47 8.33
Diffuse/avoidant-orientation .19 2.99 Diffuse/avoidant-orientation .13 2.35
Information-orientation .06 0.98 Information-orientation .38 6.36

a F (2, 196) = 7.25, p < .01. a F (2, 196) = 0.01, p = .99.


b F (3, 193) = 23.94, p < .01. b F (3, 193) = 49.62, p < .01.
p p p
onetailed < .05; ponetailed < .01. onetailed < .05; onetailed < .01.
Epistemic Cognition and Identity Formation 193

improved the regression significantly. All three identity the diffuse/avoidant-orientation and dummy-contrast A
styles were related to moratorium status scores. Nonethe- representing an absolutistic epistemic understanding. By
less, the contribution of both dummy-contrasts represent- contrast, identity achievement was combined with an
ing adolescents epistemic stance remained significant. information-orientation and dummy-contrast B represent-
Correspondingly, in a supplementary regression analy- ing an evaluativist epistemic position. Of special interest
sis, when identity processing styles were entered first, in these analyses were the cross-lagged path coefficients
including epistemic stance in the regression yielded a sig- that indicate whether two measures predict each other re-
nificant increase in the explained variance, R 2 = 0.04, ciprocally or whether there is a predominance of one pre-
F (2, 193) = 4.91, p < .01. dictor over the other. Such a predominance may indicate
Contrary to moratorium status scores, identity a causal effect, though it does not necessarily do so (e.g.
achievement was unrelated to adolescents epistemic Schneider, 1989). Results regarding these path-analyses
stance, whereas identity processing styles were significant are summarized in Table VI.
predictors of identity achievement, with substantial effects Model I included identity diffusion, the diffuse/
of the normative-orientation as well as the information- avoidant-orientation as well as dummy-contrast A (ab-
orientation (see Table V). solutism vs. multiplism/evaluativism). For the variables
in this model significant autoregressive path-coefficients
were found, indicating a considerable longitudinal
Path-Analyses of Longitudinal Data stability of the measures. Two of the nine cross-
lagged path coefficients were significant: the diffuse/
In order to analyze longitudinal change in identity avoidant-orientation at t1 predicted identity diffusion at t2
status scores as a function of epistemic stance and iden- ( = .20). The reverse effect was not significant ( = .09).
tity processing style four different path-analyses were run. At the same time, adopting an absolutist epistemic stance
The generic path-model for these analyses included two at t1 predicted identity diffusion at t2 ( = .14), whereas
repeated measures of an identity status score, two re- the reverse effect was not significant ( = .04). Thus,
peated measures of identity processing style as well as both factors that were expected to be linked to the process
two repeated measures of adolescents epistemic stance of adolescents identity formation evidenced cross-lagged
as represented either by dummy-contrast A or by dummy- effects consistent with this assumption.
contrast B (see Fig. 1). Identity status measures were Model II included identity foreclosure, the
always combined with those identity processing styles normative-orientation as well as dummy contrast A.
and epistemic stances that were expected to be particu- In this model, only one significant cross-lagged coeffi-
larly relevant for the identity status included in the model. cient emerged: identity foreclosure at t1 predicted the
Thus, identity diffusion, for instance, was combined with normative-orientation at t2 ( = .18), the reverse effect

Fig. 1. Generic Path-Model for analyzing longitudinal change in identity status scores as a function of
epistemic stance and identity processing style.
194

Table VI. Path-Analyses of Longitudinal Change in Identity Status Scores as a Function of Epistemic Stance and Identity Processing Style

Cross-lagged path coefficients


rab rbc rac ax by cz ay bx bz cy az cx Rx2 Ry2 Rz2 resxy resyz resxz

Model Ia .41 .27 .15 .45 .34 .18 .20 .09 .04 .14 .01 .08 .26 .26 .03 .40 .00 .03
Model IIb .49 .25 .12 .51 .63 .16 .03 .18 .05 .04 .07 .06 .38 .43 .04 .28 .03 .21
Model IIIc .22 .18 .35 .55 .53 .11 .01 .04 .02 .02 .20 .11 .35 .28 .07 .01 .11 .01
Model IVd .50 .02 .24 .55 .52 .28 .10 .01 .06 .21 .19 .15 .36 .39 .13 .34 .10 .12

Note. All coefficients are indexed according the generic path model as defined in Fig. 1.
The following measures were included in the models:
a Identity status score, diffusion; identity style, diffuse/avoidant-orientation; epistemic stance, dummy-contrast A.
b Identity status score, foreclosure; identity style, normative-orientation; epistemic stance, dummy-contrast A.
c Identity status score, moratorium; identity style, information-orientation; epistemic stance, dummy-contrast A.
d Identity status score, achievement; identity style, information-orientation; epistemic stance, dummy-contrast B.
p
onetailed < .05; ponetailed < .01.
Krettenauer
Epistemic Cognition and Identity Formation 195

was not significant ( = .03). Thus, although identity fusion and foreclosure) cross-sectional and longitudinal
foreclosure and normative-orientation were correlated analyses yielded a similar pattern of results. Regarding
cross-sectionally, a cross-lagged effect of the normative- moratorium and identity achievement scores, however,
orientation on identity foreclosure was not found. Instead, cross-sectional and longitudinal findings diverged. In the
the reverse effect emerged. Over time, identity foreclosure following, results for each identity status are discussed in
appeared to foster the normative identity processing style. turn.
Model III contained moratorium status scores, the Theoretically, identity diffusion was expected to be
information-orientation as well as dummy-contrast A. associated with absolutism and an identity processing
In this model, significant cross-lagged path coefficient style dubbed diffuse/avoidant-orientation. Empirically, a
involving identity status scores did not emerge. There substantial effect of the diffuse/avoidant-orientation on
was only one significant cross-lagged path coefficient identity diffusion was found, both cross-sectionally and
indicating that the information-orientation at t1 was in- longitudinally. The diffuse/avoidant-orientation appeared
versely associated with an absolutistic epistemic po- to foster identity diffusion, whereas the reverse effect
sition 18 months later ( = .20). Thus, adolescents of identity diffusion on identity processing style was in-
information-orientation appeared to contribute to epis- significant. With regard to adolescents epistemic stance
temic development beyond absolutism. cross-sectional as well as longitudinal analyses revealed a
Finally, longitudinal change in identity achievement significant though moderate effect. An absolutistic epis-
as a function of epistemic stance and identity processing temic stance tended to enhance identity diffusion. Again,
style was investigated. In this model, a significant cross- the reverse effect was found to be negligible. Overall,
lagged path coefficient of adolescents epistemic stance theoretical expectations regarding identity diffusion were
emerged; having achieved an evaluativist epistemic posi- largely confirmed, though the impact of epistemic devel-
tion at t1 predicted identity achievement 18 months later opment on identity diffusion was moderate.
at t2 ( = .21). The reverse effect of identity achieve- For identity foreclosure a different pattern of findings
ment on adolescents epistemic stance was not significant emerged. Contrary to our hypothesis, foreclosure was un-
( = .06). At the same time, reciprocal cross-lagged related to absolutism but inversely associated with attain-
effects of adolescents epistemic stance and information- ing the more advanced, evaluativist epistemic position.
orientation were found. Thus, although in the cross- However, this relation was mainly due to individual dif-
sectional analyses the development of epistemic cogni- ferences in identity processing style. When controlling for
tion turned out to be unrelated to identity achievement, a identity styles, epistemic stances were unrelated to iden-
cross-lagged effect emerged, supporting the view that the tity foreclosure. Thus, identity foreclosure appears to be a
development of epistemic cognition fosters adolescents matter of cognitive processing rather than epistemic devel-
identity achievement. opment. In the cross-sectional study, a strong effect of the
normative-orientation emerged, whereas in the longitudi-
nal study, a significant cross-lagged effect of foreclosure
DISCUSSION status scores on the normative identity processing style
was found.
The aim of the present study was to further clar- Identity status scores of moratorium were expected
ify how developmental change in adolescents epistemic to be associated with epistemic multiplism and an infor-
understanding contributes to the process of identity for- mation orientation. In the cross-sectional study, this ex-
mation. To pursue this goal, associations between var- pectation was largely confirmed. Adolescents assigned to
ious epistemic stances, on the one hand, and identity the multiplistic epistemic stance evidenced higher mora-
status scores, on the other, were analyzed while simul- torium scores. At the same time, moratorium status scores
taneously taking into account potential effects of indi- were associated with the information-orientation and the
vidual differences in cognitive processing as represented two other identity processing styles as well. When en-
by different identity processing styles. Associations be- tered simultaneously in the regression identity styles and
tween epistemic stance and identity status scores were epistemic stance were both significant predictors. Thus,
not only examined cross-sectionally, as had usually been both sets of variables independently predicted adoles-
done in previous studies, but longitudinally as well. Thus, cents moratorium scores. However, in the longitudinal
it was possible to test whether or not epistemic devel- study none of these effects emerged. Although there was a
opment predicts longitudinal change in identity status reciprocal longitudinal relation between epistemic stance
scores. The findings were not uniform for the four identity and identity processing style, longitudinal change in mora-
status measures. For two identity statuses (identity dif- torium scores turned out to be unrelated to epistemic
196 Krettenauer

development and adolescents identity style. Perhaps the It should be explicitly noted at this point, that the
discrepancy of cross-sectional and longitudinal findings is conclusions drawn in the present study probably apply
due to the limited time period of the longitudinal study that only to adolescent development, when for the first time
may not have captured the time needed for cross-lagged in their lives young people become fully aware of the
effects to emerge. multiplicity of knowledge claims and, simultaneously,
Identity achievement was hypothesized to be related face the developmental task of constructing their ego iden-
to the attainment of epistemic evaluativism as well as tity. It remains an open question whether or not epistemic
an information orientation. In the cross-sectional study development contributes to the process of identity forma-
significant effects of identity processing styles emerged, tion beyond adolescence.3 It should also be kept in mind
whereas adolescents epistemic stance was unrelated to that the present study did not use categorical measures of
identity achievement. The information-orientation and identity statuses. All analyses were based on continuous
the normative-orientation were both significant predictors identity status scores. The study, thus, predicts tendencies
of identity achievement. Moreover, the diffuse/avoidant- toward the attainment of identity statuses of moratorium
orientation was inversely associated with achievement and achievement but did not investigate whether these
scores. However, in the longitudinal study, a different pat- statuses were actually achieved. Moreover, in the study
tern of findings emerged; there was a significant cross- only self-report measures of individual differences in ap-
lagged effect of epistemic stance. Having attained the proaching identity relevant problems and decisions were
more advanced epistemic stance of evaluativism yielded used. Perhaps, measures of actual cognitive processing
an increase in identity achievement scores 18 months later. would yield results different from those obtained in the
By contrast, identity processing styles did not contribute study. Finally, though cross-lagged analyses of longitudi-
to longitudinal change in identity achievement scores. nal data provide a more powerful method to investigate
Taken together, the results of the present study the direction of effects than cross-sectional correlations,
suggest the following, general conclusion: two identity they still do not reveal causal links between variables,
statuses, moratorium and identity achievement, partly de- as an experimental research design would do. Therefore,
pend on the development of epistemic cognition. In gen- all conclusions drawn in the present study that suggest a
eral, these two statuses are considered developmentally causal role of the development of epistemic cognition in
more advanced than identity diffusion and foreclosure the process of identity formation should still be considered
(e.g. Moshman, 1999; Waterman, 1999). The findings preliminary.
of the present study correspond with this developmental Although it is important to keep these caveats in
view. The impact of the development of epistemic cogni- mind, one may nevertheless speculate about educational
tion on moratorium and identity achievement cannot be at- implications of the present study; since individuals level
tributed to individual differences in cognitive style as rep- of epistemic cognition is generally related to level of ed-
resented by different identity processing styles. Overall, ucation, fostering epistemic reasoning development by
the study, then, confirms basic insights obtained by Boyes educational interventions might be an effective vehicle
and Chandler (1992) and replicates important findings to stimulate the process of identity formation in adoles-
of their study in another cultural context. Moreover, the cence. Perry (1970), one of the founders of research on
present study relativizes conclusions drawn by Klaczynski epistemological beliefs, certainly had this practical im-
et al. (1998) who suggested that individual differences in plication in mind when examining the impact of college
cognitive processing account for the link between cogni- education on students intellectual and ethical develop-
tive development and identity formation. Contrary to the ment. While previous research on students epistemo-
conclusion drawn by these authors, the present study pro- logical beliefs mainly focused on proximal processes
vides evidence that cognitive development and identity that are related to individual differences in epistemic
processing styles contribute independently to the process
of identity formation, in as much as the identity statuses 3 At this point the reader may wonder whether the relations between
of moratorium and identity achievement are concerned. epistemic stances and identity statuses as obtained in the present study
However, since most effect sizes as obtained in the present are moderated by adolescents age. To test for such a moderating role
study were small to moderate epistemic development is of age, in a supplementary set of regression analyses participants age
certainly not the only factor that influences adolescent together with two interaction terms of epistemic stance by age were
identity formation. There are probably many other factors added to those four regression analyses as summarized in Tables IIV.
In none of these regressions entering the interaction terms yielded a
that contribute to the process of adolescent identity for- significant increase in the R-squared. The associations between epis-
mation and that have been left out of consideration in the temic stances and identity statuses as obtained in the present study are
context of the present study. thus unaffected by adolescents age.
Epistemic Cognition and Identity Formation 197

cognition (e.g. scientific reasoning, argumentation skills, Edelstein and Hubert Sydow for their valuable comments
cf. Kuhn 1991) it somewhat neglected the fact that epis- on earlier versions of the manuscript.
temic development also has a bearing for adolescents
identity development. Certainly, much research needs to REFERENCES
be done in order to design appropriate programs that are
able to foster the development of epistemic cognition (cf. Adams, G. (1998). Objective Measure of Ego Identity Status: A Refer-
Kuhn and Udell, 2003). However, if these efforts turn out ence Manual. University of Guelph, Guelph, Ontario, Canada.
to be successful they might provide a valuable tool to pro- Berzonsky, M. D., Weiner, A. S., and Raphael, D. (1975). Interdependece
of formal reasoning. Devel. Psychol. 11: 258.
mote the process of identity formation in adolescence, as Berzonsky, M. D. (1989). Identity style: Conceptualization and mea-
well. surement. J. Adol. Res. 4: 268282.
Berzonsky, M. D. (1992). Identity Style Inventory (ISI3). State University
of New York at Cortland, Cortland, New Yok.
Berzonsky, M. D. (1994). Individual differences in self-construction:
APPENDIX The role of constructivist epistemological assumptions. J. Const.
Psychol. 7: 263281.
Berzonsky, M. D. (1999). Identity styles and hypothesis-testing strate-
Item Examples of the Epistemic Development gies. J. Soc. Psychol. 139: 184789.
Questionnaire (Translated from German) Berzonsky, M. D., and Adams, G. R. (1999). Reevaluating the identity
status paradigm: Still useful after 35 years. Devel. Rev. 19: 557590.
Berzonsky, M. D., and Neimeyer, G. J. (1994). Ego identity status and
Some scientists claim that the electric radiation of identity processing orientation: The moderating role of commit-
cell phones is harmful to peoples health. Other scientists ment. J. Res. Personality 28: 425435.
challenge this view. I think: Blasi, A. (1988). Identity and the development of the self. In Lapsley,
D. K., and Power, F. C. (eds.), Self, Ego, and Identity. Springer-
An exact scientific investigation could settle this Verlag, New York, pp. 226242.
Bosma, H. A., and Kunnen, E. S. (2001). Determinants and mechanisms
dispute once and for all. (absolutist) in identity development: A review and synthesis. Devel. Rev. 21:
Both claims amount to little more than hunches. 3966.
People will never be able to definitely determine Boyes, M. C., and Chandler, M. (1992). Cognitive development, epis-
temic doubt, and identity formation in adolescence. J. Youth Ado-
whether cell phones have harmful effects or not. lescence 21: 737763.
(multiplist) Cauble, M. A. (1976). Formal operations, ego identity, and principled
Though this controversy might never be settled, morality: Are they related?. Devel. Psychol. 12: 363364.
Chandler, M. (1987). The Othello effect. Human Devel. 18: 171180.
both sides may have good reasons to maintain Chandler, M. J., Boyes, M., and Ball, L. (1990). Relativism and stations
their view. (evaluativist) of epistemic doubt. J. Exper. Child Psychol. 50: 370395.
Cote, J. E., and Levine, C. (1988). A critical examination of the ego
Some historians say that the majority of Germans identity status paradigm. Devel. Rev. 8: 147184.
supported the genocide of the Jews during the Third Reich. Epstein, S., Pacini, R., Denes-Ray, V., and Heier, H. (1996). Individual
differences in intuitive-experiential and analytical-rational thinking
Others say, this was the fault of a few men in the Nazi- styles. J. Personality Soc. Psychol. 71: 390405.
government. I think: Erikson, E. (1959). Identity: Youth and Crises. Norton, New York.
Hallett, D., Chandler, M. J., and Krettenauer, T. (2002). Disentangling
Peoples reactions regarding this question will the course of epistemic development: Parsing knowledge by epis-
always differ. Nonetheless some historians may temic content. New Ideas Psychol. 20: 285307.
Hofer, B. K., and Pintrich, P. R. (1997). The development of epistemo-
have a better account than others. (evaluativist) logical theories: Beliefs about knowledge and knowing and their
More than 50 years of historical research should relation to learning. Rev. Educational Res. 67: 88140.
have made the answer clear. (absolutist) King, P. M., and Kitchener, K. S. (1994). Developing Reflective Judg-
ment. Jossey-Bass, San Francisco.
Both views merely reflect personal opinions. It Kitchener, R. F. (2002). Folk epistemology: An introduction. New Ideas
mainly depends on peoples attitudes towards Psychol. 20: 89105.
Germany on which side they are. (multiplist) Klaczynski, P. A., Fauth, J. M., and Swanger, A. (1998). Adolescent
identity: Rational vs. experiential processsing, formal operations,
Note. The epistemic stance as expressed by the statement and critical thinking beliefs. J. Youth Adolescence 27: 185207.
Kramer, D. A. (1983). Post-formal operations? A need for further con-
is indicated in parentheses. ceptualization. Human Devel. 26: 91105.
Krettenauer, T. (2005). Die Erfassung des Entwicklungsniveaus episte-
mologischer Uberzeugungen und das Problem der Ubertragbarkeit
von Interviewverfahren in standardisierte Fragebogenmethoden
ACKNOWLEDGMENTS [Measuring the developmental level of epistemological beliefs and
the problem of transfering interview procedures to standardized
The research reported in this paper was supported questionnaire methods]. Zeitschrift fur Entwicklungspsychologie
und Padagogische Psychologie, 37: 6979.
by a grant of the German Research Foundation (Grant Kuhn, D. (1991). The Skills of Argument. Cambridge University Press,
KR2027/3-1). The author would like to thank Wolfgang Cambridge.
198 Krettenauer

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of Rest, J., Thoma, S., and Edwards, L. (1997). Designing and validating a
epistemological understanding. Cognitive Devel. 15: 309328. measure of moral judgment. J. Edu. Psychol. 89: 528.
Kuhn, D. and Udell, W. (2003). The development of argument skills. Rowe, I., and Marcia, J. (1980). Ego identity status, formal oper-
Child Devel. 74: 12451260. ations, and moral development. J. Youth Adolescence 9: 87
Leadbeater, B. J., and Dionne, J. P. (1981). The adolescents use of 99.
formal operational thinking in solving problems related to identity Schneider, W. (1989). Problems of longitudinal studies with children:
resolution. Adolescence 16: 111121. Practical, conceptual, and methodological issues. In Brambring,
Marcia, J. E., Waterman, A. S., Matteson, D. R., Archer, S. L., and M., Losel, F., and Skowronek, H. (eds.), Children at Risk: Assesse-
Orlofsky, J. L. (1993). Ego identity. Springer, New York. ment, Longitudinal Research and Internvenion, Berlin: De Gruyter,
Meuus, W., Iedema, J., Helsen, M., and Vollebergh, W. (1999). Pat- pp. 313335.
terns of adolescent identity development: Review of literature and Stephen, J., Fraser, E., and Marcia, J. E. (1992). Moratorium-
longitudinal analysis. Devel. Rev. 19: 419461. achievement (MAMA) cycles in life span identity development:
Moshman, D. (1994). Reason, reasons and reasoning: A constructivist Value orientations and reasoning systems. J. Adolescence 15: 283
account of human rationality. Theory Psychol. 4: 245260. 300.
Moshman, D. (1999). Adolescent Psychological Development. Erlbaum, Treiman, D. J. (1977). Occupational Prestige in Comparative Perspec-
Mahwah, NJ. tive. Academic Press, New York.
Perry, W. G. (1970). Forms of Intellectual and Ethical Development in van Hoof, A. (1999). The identity status field re-reviewed: An update
the College Years. Holt, Rinehart, and Winston, New York. of unresolved and neglected issues with a view on some alternative
Peterson, D. M., Marcia, J. E., and Carpendale, J. I. M. (2004). Iden- approaches. Devel. Rev. 497556.
tity: Does thinking make it so? In Lightfoot, C., Lalonde, C., and Waterman, A. S. (1999). Issues of identity formation revisited:
Chandler, M. (eds.), Changing conceptions of psychological life. United States and The Netherlands. Devel. Rev. 19: 462
Mahwah, NJ: Erlbaum. 479.
Piaget, J., and Inhelder, B. (1977). Von der Logik des Kindes zur Logik Wolf, C. (1995). Sozio-okonomischer Status und berufliches Pres-
des Heranwachsenden (From the logic of the child to the logic of tige (Socio-economic status and occupational prestige). ZUMA-
the adolescent). Olten, Freiburg. Nachrichten 19: 102136.

You might also like