Back to Table of Contents








In our quest to explore the factors and beliefs that impact one’s willingness to engage in worldview integration, Chapters Two introduced us to both the concept of worldview and the worldview component referred to as religious worldview.  We also looked at a possible model of the integration process and the available measures for quantifying religious worldview on a theistic continuum.  In Chapter Three, we looked at a passage from the Bible (Romans 1) and the contributions of several classic theologians concerning our initial knowledge of God.  This shed a bit more light into the development of a person’s initial religious worldview.  However, it is also this study’s intent to consider another important worldview component – that of one’s epistemological beliefs, along with issues of motivation and well-being.

In Chapters Four and Five, we will be introduced to the remaining three domains that will be investigated.  These domains are: a) personal epistemology (epistemic worldview – Chapter Four), b) motivation (intrinsic vs. extrinsic) – Chapter Five, and c) subjective well-being – Chapter Five.  While each of these three domains investigates areas of empirical social science, each also depends upon the participant’s all encompassing worldview. 

The domain of personal epistemology is a fairly recent and burgeoning field that requires some preliminary discussion and examines an area of foundational worldview belief—our beliefs concerning truth, knowledge and learning.  It is a field of investigation that hopefully will lead to further understanding of how and whether certain beliefs in this area affect our efforts toward continuing worldview integration. 


Epistemological Worldview


Introduction to Epistemological Worldview

A large component of our worldview concerns our view of truth and human knowledge.  Without some belief in the human capability of acquiring knowledge and the value of engaging in the pursuit, learning becomes meaningless.  This requires formulation of beliefs concerning the very nature and characteristics of truth and knowledge.  Likewise, one must determine if knowledge is either helpful or desired in the pursuit of life’s goals and purposes.  Yet, one’s worldview cannot end with such universal and esoteric concerns.  These beliefs must be distilled down to a personal and practical level.  Are we personally capable of discerning or integrating this knowledge?  How do we discriminate between relevant knowledge and the infinite stream of available knowledge that is irrelevant?  Can our capability to acquire and integrate knowledge be learned or developed?  These beliefs become foundational to the remainder of our worldview and to our ability and willingness to engage in worldview integration.

Thus we discover a paradox of epistemology.  Our epistemology is not only a component part of our worldview, but it also serves as the sentry at the gate of our worldview.  If that sentry allows no knowledge to enter, learning and integration is limited or impossible.  If that sentry allows anything and everything through the gate (for example, by failing to discern the warning signs of dissonance and incoherence), learning and integration loses its utility and purpose.  If that sentry puts up a filter of foundational or basic beliefs, some knowledge may be left at the gate.  Seemingly, part of both learning and worldview integration will involve investigating and experimenting with the sentry itself.  Yet, if the sentry is already indoctrinated with the belief that such investigation and experimentation is dangerous, taboo or inconsequential, worldview integration may be foreclosed from the start.


Personal Epistemology

            Based upon the above, it would seem obvious that some understanding of a learner’s selection and application of epistemology would be crucial both on a content and methodological level.  As might be expected, research has disclosed that students come into the educational system with differing epistemological viewpoints (Perry, 1970; Baxter-Magolda, 1992).  In addition, one’s outlook on both learning and the nature of truth has been shown to have some effect on academic success (Schommer, 1990; Hofer, 2000).  A person’s epistemological beliefs often stimulate, or in the alternative, set ceilings and limitations for learning and the development of critical thinking.  Such beliefs have been shown to directly influence such factors as motivation and the acceptance of mastery goals (Braten and Stromso, 2004), as well as one’s selection of problem solving approaches (Schraw, 2001; Kardash and Scholes, 1996; Schommer, 1994; Schreiber and Shinn, 2003).  Evidence also suggests that as a person’s epistemological beliefs change and become more sophisticated and mature, critical thinking and problem solving skills improve at the same time (Schraw, 2001).  

            This newly emerging field, originally appearing under a number of monikers (intellectual development, cognitive epistemology, reflective judgment, ways of knowing, epistemological development, psycho-epistemological style, and others) seeks to investigate the relationship between one’s adopted epistemological framework and the maturity of their thinking skills.  One author has suggested consolidating these varied descriptions under the label “personal epistemology” (Hofer, 2001).  This term seems to be gathering momentum and will be adopted herein to designate the entire body of research in this area. 

This study will hopefully add additional understanding concerning the relationship between epistemic beliefs, intrinsic motivation to learn and worldview integration.  It is hoped that by understanding these relationships, it may be possible to develop and encourage increased critical thinking and worldview integration among learners.  One researcher reflected that “psycho-epistemological research appears to be able to contribute to the understanding of other concepts in psychological science because psycho-epistemological dimensions represent a personal world-view, which has an enormous impact on all aspects of an individual’s life” (Desimpelaere,, 1999, p. 137).

However, developing the concept and its structure has not been simple or obvious.  As Hofer and Pintrich (1997) note, “Defining the construct is problematic, as there are discrepancies in naming the construct as well as in defining the construct, to the extent that it is sometimes unclear to what degree researcher are discussing the same intellectual territory” (p. 111).  There seem to be four axes of contention that will be considered separately.  First, at what level of cognition does personal epistemology take place … the cognitive or the meta-cognitive?  Second, is personal epistemology based on epigenetic development or is it the product of social, cultural and rational processes?  Third, is personal epistemology a uni-dimensional concept that progresses or regresses in a unified manner and which theoretically may be evaluated using a single standardized measure?  Or, on the other hand, is it multidimensional requiring consideration of a number of distinct components that may progress or regress independently? Finally, does one apply the same personal epistemology to each of the different domains of knowledge?  As set forth in the opening definitions, a domain of knowledge may be described as either a discipline (e.g.  physical science, social sciences, humanities, professions) or different types of knowledge (e.g. personal truths, aesthetic truths, value and ethical truths, facts of the social world, and facts of the physical world).  Is it possible that a person may apply a different personal epistemology to different domains? 

It remains possible, if not likely, that the idea of personal epistemology consists of a complex interaction and integration of these axes requiring a balanced approach.  For example, there may be both developmental and cultural components of our epistemological beliefs; there may be both a general and dimension specific aspects to our beliefs; and there may be different epistemological resources that we develop to be applied differently to distinct disciplines and different types of knowledge.  This may require the development of appropriate integration and balancing strategies.  Hofer (2004a) argues that this requires an awareness of epistemological metacognition.  Schommer-Aikins (2004) suggests the need for an embedded systemic model that deals with the interplay of culture, learning beliefs, knowledge beliefs and metacognitive strategies.  Louca, Elby, Hammer and Kagey (2004) propose that epistemology should be looked at as cognitive resources to be selected from one’s available inventory and applied as appropriate.  These are some of the questions and conceptual issues that the field will need to grapple with as it matures.  Yet despite the existence of different proposed solutions to these questions, the field may still provide invaluable help in investigating the role of epistemology in worldview integration. 


Overview of the Field of Personal Epistemology

Levels of Cognition

King, Kitchener, Davison, Parker & Wood (1983) and Kuhn (2000a) have suggested that one’s personal epistemology is actually an advanced level of general cognition.  Kitchener (1983) has proposed a three step model consisting of a) cognition, b) metacognition, and c) epistemic cognition.  Kuhn’s(2000a) model is quite similar, yet it divides metacognition into two sub-levels – metatask knowledge and metastrategic knowledge.  The simplest level (cognition) deals with understanding of information and includes basic reasoning processes such as computing, reading, logic and perception.  The second level (metacognition) addresses with the strategic learning processes that permit knowledge to be developed from the cognition that has already taken place.  This would include the use of inductive, deductive and evaluative or critical reasoning skills.  According to Kitchener (1983), the most advanced level (epistemic cognition) deals with the nature of knowing and problem solving, including an awareness of the limitations of knowledge.  Essentially, this is not simply knowing or “knowing how to know,” but rather “knowing about knowing.” Hofer (2004a), rather than using a three level hierarchy, would use only two levels – cognition and metacognition.  (See Table 4.1).  However, metacognition would be divided into three components; metacognitive knowledge (which would include beliefs concerning the nature of knowledge itself); metacognitive judgment and monitoring (which would include beliefs about the knowing process); and metacognitive self-regulation and control of cognition (consisting of beliefs concerning value, motivation, and volition).  Thus, concepts such as intrinsic motivation would be anticipated to better correlate with metacognitive self-regulation than with beliefs concerning the nature of knowledge and the process of knowing.

Either way, each model suggests that advanced epistemic cognition can only be expected to emerge after formal operational thinking (Piaget’s highest level of cognition) arrives, normally in late adolescence (Kitchener, 1983).  Thus, epistemological knowing becomes the control center that determines which form of cognition (or metacognition) is most appropriate for the domain or concept under consideration.  As a result, those who have not appropriately developed their epistemic cognition may be those who find themselves unable to deal appropriately with the conflict that occurs between different types of knowing within different domains (Kuhn, 2000a).


Developmental Theories

The field of personal epistemology is generally believed to have commenced with the work of Harvard educator William G. Perry, Jr. (1970) (Kuhn and Weinstock, 2002; Hofer, 2002b).  Perry observed that different students appeared to possess varying

Table 4.1        

Levels of Cognition (Kuhn & Kitchener)

3 Level Model of Cognitive Processing

(Kitchener, 1983)

3 – Level Model of


(Kuhn, 2000)

Hofer (2004a)

Two Level


I.  Cognition


I.  Cognition

II.  Metacognition

I.   Metacognitive knowing

II.  Metacognition


II.  Metastrategic knowing

a.   Metacognitive knowledge


a.   Metatask knowledge

b.   Metacognitive judgment & monitoring


b.   Metastrategic knowledge

c.   Metacognitive self-regulation & control of cognition

III.  Epistemic Cognition

III.  Epistemological Knowing















perceptions of the educational process and the authority and role of their professors.  Some believed that professors are a source of right and wrong answers that simply need to be gathered and understood.  Others recognized that not all issues were so cut and dried and that different professors had equally rational and acceptable answers to the same problem.  For them, there was no right or wrong answer and one need not accept one or the other.  Based upon these observations, Perry conducted a series of longitudinal interviews with over 400 male Harvard students and analyzed the results, finding what appeared to be a logical pattern of “epistemological development” within college students.  Perry’s work builds upon the developmental assumptions of Piaget (1954) and Kohlberg (1987).  As Perry (1970) himself observed:


Our scheme addresses developmental issues to which Piaget's framework does not yet extend. . . One could look on our scheme, then, as adding an advanced "period" to Piaget's outline.  If so, I would call it the "period of responsibility". (p. 229 – 230)


The work of Perry and his followers is decidedly developmental.  It shares many of the same epigenetic, naturalist, interactionist, and constructivist assumptions of previous developmental theorists.  Perry’s (1970) cognitive scheme posited nine “positions” clustered into four basic categories.  (See Table 4.2.)  Perry made it quite clear that these positions, although progressive, neither suggest that a person will necessarily reach any particular position nor that that the positions are equally distant or equally temporally separated.  Unlike Piaget’s theory with its “stages” through which every child is expected to progress, Perry simply suggests descriptive “positions” that a person may reach.  Positions are not stark changes in epistemology, but simply a discernible and describable marker on a potential continuum.  A number of factors including society, culture, and personal volition and agency may affect the positions achieved.  Arguably, even genetics itself may provide limitations. 

The first category labeled Dualism includes positions 1 and 2.  The second category labeled Multiplicity includes positions 3 and 4.  The third category referred to as Relativism encompasses position 5.  Finally, the fourth category is named Commitment in Relativism and includes positions 6 through 9.  It is generally assumed that the higher positions reflect persons who are more developmentally and epistemologically mature.  Perry’s basic categories are as follows:


Table 4.2.       

 Perry’s Epistemic Development Positions



position and description

concept of knowledge


1.   Basic Duality

Assumption of dualistic structure of world taken for granted, unexamined.  Right vs. wrong, we vs. others, good vs. bad, what They want vs. what They don’t want. 

Knowledge is an objective, definite, and organized body of facts that constitute the truth about a subject, to be distinguished from opinion.


2.   Multiplicity: Pre-Legitimate

Truth exists, but not all authorities are knowledgeable.  Multiplicity is perceived, but only as alien or unreal.  As alien it assimilates easily to error and otherness: “others are wrong and confused.”  Assimilated to authority, it leads to opposition.

Knowledge consists of facts, principles, axioms, etc. that can be proved, although it may be difficult to carry out the proof.  Overcoming this difficulty is the expert’s challenge, and some are more expert than others.


3.   Multiplicity Subordinate

Absolute truth has not been discovered, yet.  Multiplicity is perceived with some of its implications.  Authority may not have the answers yet.  But trust in Authority, at least in the ideal, is not threatened

Knowledge consists of facts, principles, axioms, etc. that can be proved, although it may be difficult to carry out the proof.  The coherence and completeness of the system may vary across disciplines, some being more advanced than others.


4.   Multiplicity Correlate

If Authorities don’t know the answer then any opinion is as good as another.  There is more than one approach to a problem.  Relativism is perceived.  However, this is still “how they want us to think” rather than a consequence of the nature of all knowledge

Knowledge is not secure but is any person’s organization and interpretation of available information.  One interpretation is as good as another.  But people with power can assert their interpretations over others.


5.   Relativism Correlate

Relativism is perceived as a way  of analyzing and evaluating.  World is divided into those areas where Authority has the answer and those where relativism must be used.  Relativism is accepted generally, but without any implications for Commitment

Knowledge is always changing or subject to change.  It can be shared but not “measured” or counted upon to remain the same.


6.   Commitment Foreseen

Subjectively choose among alternatives.  Commitment may be perceived as a logical necessity for action in a relative world or as “felt” needed.  This realization may bring various reactions: eagerness, ambivalence, dismay, sturdiness, turmoil or simple acceptance.

Knowledge is not something that is external and definite but something that each individual constructs according to his/her experience, background, etc.


7.   Initial Commitment

First commitments or affirmations.  Acceptance of their origins in self’s experience and choices, some intimations of implications.

Knowledge is the world view one has constructed from learning and experience, along with the ethical implications of this view, synthesized into a consistent philosophy.


8.   Orientations in Implications of Commitment

Some implications of commitment are realized: tensions between feelings of tentativeness and finality, expansion and narrowing, freedom and constraint, action and reflection.  Identity in both content of Commitment and in person style of address to Commitment.

Knowledge is a creative resolution between uncertainty and the need to act, which makes it a dynamic means of transaction between the self, the environment, requiring both stability and flexibility.


9.   Developing Commitment

Reassessment of commitments with new priorities.  Commitments extended or remade in new terms as growth.  Balances are developing in the tensions of qualitative polarities of style, especially alternation of reflection and action.  Acceptance of change of mood and outlook within continuity of identity.  Sense of being “in” one’s life.

Knowledge is the evolution of awareness, best expressed as ascending levels of consciousness, in which the individual must break through to new perspectives and discard those no longer useful.



In Dualism, the person sees the world as consisting of absolutes; “right or wrong” or “true or false.”  Everything is knowable, however only certain authorities have discovered this objective knowledge.  The learner’s job is to simply locate, obtain and absorb that knowledge from the true authorities.  A teacher’s role is to simply dispense these correct truths.  Failure to do so renders the teacher poor and ineffective.  The difficulties we have in solving problems are not caused by epistemological limitations, but rather by our inability to discern true authorities from imposters.  Once knowledge is located, those who deviate from or dispute this authority are simply in error. 

In Multiplicity, one begins to understand that in some areas knowledge is not yet absolute.  The student still trusts that an absolute answer exists and that authorities will eventually discover it.  Yet, in late multiplicity, the learner observes that there truly are some fields and issues where legitimate absolutes will likely never be found or simply do not apply.  In addition, there may be more than one approach or answer to a problem.  Thus some problems can be approached from different angles and resolved using different knowledge.  Imposters will still normally be exposed while those with authority will normally succeed in asserting their opinions over the voices of others lacking knowledge.

Within Relativism, a stark shift in epistemology occurs.  The realization that absolute answers are not easily falling into place causes a crisis in critical thinking.  Here, everything becomes relative, not because authorities like it this way, but simply because this is the ontology and epistemology of the world.  Absolutes (if any still exist) are no longer the norm, but the exception.  Authorities no longer have the answers, and may be unchallenged imposters who are groping along with the rest of the world for answers.  An enlightened authority figure will no longer present claims to knowledge, but will use their authority to promote critical thinking and learning skills. 

Commitment in Relativism is where the first recognition is made that one’s learning, tradition, experiences and tools may be utilized to develop a synthesized and consistent philosophy despite the inability to arrive at absolute and static truth.  Thus, we must recognize the reality of our situation and develop personal and rational positions concerning “how then shall I live.”  The final position within Commitment is where these tensions are worked out and the person has developed a sense of self understanding and peace regarding his choices.  Yet, Perry’s observations show that few, if any, can be expected to reach this position.

Since that time, others have investigated and modified Perry’s initial scheme, often leading to slightly different developmental models (see Table 4.3).  Often, these modifications involve concerns over the generalizability of Perry’s work due to his small sample of predominantly male, well to do, Harvard students.  For example, Belenky, Clinchy, Goldberger & Tarule (1986) investigated a female sample with diverse educational and financial backgrounds.  Students from nine different institutions were included in the study.  They found an epistemological pattern similar to that of Perry’s but added a pre-dualism position which is labeled silence.  Apparently, unlike Harvard males, women occasionally fail to develop any appreciable epistemic style out of the expectation that women are to remain without voice.  Belenky, et al’s theory was labeled “Women’s Ways of Knowing” and included five epistemological categories; a) silence; b) received knowledge – passive learning (similar to Perry’s dualism); c) subjective knowledge – more based on gut feelings to distinguish between conflicting information (similar to Perry’s multiplicity – approaching relativism); d) procedural knowledge – how to process different perspectives (similar to Perry’s relativism – approaching


Table 4.3        

 Parallel Developmental Theories


Perry (1972)

King & Kitchener (1983)

Belenky, et al. (1986)



Baxter-Magolda (1992)

Boyes & Chandler (1992)





(mainly women)



Men – overall more independent & individualistic, Women – more personal and interactionist




Received Knowledge


Absolute Knowing

Naively Realistic











Transitional Knowledge

Defended Realism



Subjective Knowledge








Independent Knowledge

Dogmaticism vs. Skepticism



Procedural Knowledge


separate knowing


connected knowing




Commitment in Relativism




Contextual Knowledge

Postskeptical Rational



Constructive Knowledge




commitment); and e) constructive knowledge – thinking critically, constructively and outside the box (corresponding to Perry’s upper positions of commitment in relativism). 

In addition, Belenky, et al (1986) discovered that those falling within the position of procedural knowledge (approaching commitment) tend to process conflicting information in two separate ways.  Men typically utilize separate knowing where the learning approach is generally socially detached and impersonal.  Here, any knowledge should first be questioned and doubted, with a later decision to be made objectively and without influence from the source of the proposed knowledge.  Women, on the other hand, tend to utilize connected knowing where open and informal conversation is used to understand and insert oneself into the perspective of the other.

Baxter-Magolda (1992) utilized a longitudinal study of both male and female college students to develop her “Epistemological Reflection Model.”  She proposed a progression of four positions, also closely paralleling those offered by Perry; absolute knowing (similar to Perry’s dualism); transitional knowledge (Perry’s multiplicity); independent knowledge (Perry’s relativism); and contextual knowledge (Perry’s commitment in relativism).  In general, gender did not result in a different sequence, but did affect the rate of progression and the methodologies used within each position.  Like Belenky, Baxter-Magolda found patterns of individualistic and impersonal processes in men and more personal and interactionist approaches in women.

Boyes and Chandler (1992) evaluated the epistemic worldview of children (preschool through adolescence) by looking at their beliefs regarding the certainty of knowledge.  Their study theorized that there might be a link between a child’s epistemic worldview and both later cognitive development (Piaget) and the growth of personal identity (Erikson).  They concluded that children develop four levels of “epistemic doubt.”  Level 0 (naively realistic) commences before the child even entertains the notion of uncertainty.  Here, knowledge is based upon personal observation.  Conflicts and disagreements arise due to simply observing different things or not paying attention.  If the other person will simply observe the same thing from the same place, they will arrive at the same answer.  Level 1 (defended realism) begins with an understanding that different perspectives and personal subjectivity is a natural result.  Here again, absolute knowledge exists, but it can be subjectively skewed at times.  Those in Level 2 (dogmaticism – skepticism) acknowledge doubt and the uncertainty of knowledge.  Knowledge carries with it a new found tentativeness that is dealt with in one of two ways.  Some (dogmatic thinking) will fight for the certainty of what is handed down and agreed upon by others.  Others (skeptical thinking) will abandon the hope of certainty and often turn to intuition to solve conflicting knowledge claims.  This often results in resignation and the recognition that each person may “do his own thing.”  Finally, Level 3 (postskeptical rational) arrives normally at adolescence as they become more comfortable with formal operational thinking.  Here, rational decisions can be made despite the fact that knowledge is at best partially certain.  Despite relativism, rational thought can result in choices that are better than others.  Thus, there may be a link between one’s cognitive and task development and their embraced epistemology. 

Chandler, Hallett and Sokol (2002) further point out that the field of personal epistemology typically has little to say about epistemic development outside of the college years.  Indeed, other researchers coming from a theory of mind or cognitive development point of view have discovered traits of subjective appreciation, skepticism and other signs of epistemic development in children as early as their primary school years (Broughton, 1978; Mansfield and Clinchy, 1987; and Walton, 2000).  Yet, how is it that Perry and most of the personal epistemology field observe conflicting evidence that epistemic development is a characteristic of late adolescence and early adulthood?  Chandler et al. (2002) suggest that this implicates a number of possibilities including: a) the recursive (spiral rather than linear) nature of epistemological development; b) the fact that normal education suppresses such development by “encouraging or perpetuating beliefs in the objectivity of knowledge as delivered by true authorities” (p. 160); and/or c) that the true universal and abstract nature of one’s epistemological thoughts do not sink in until one has become firmly established within Piaget’s stage of formal operational thinking.  Clearly more research is necessary in this area.

King and Kitchener (2002) investigated high school students and adults, the development of their epistemological assumptions and how they affected reasoning skills.  The study, first reported in 1981 and completed in 1994, includes both cross-sectional and longitudinal research over a period of nearly 20 years.  While most previous studies analyzed attitudes towards authority, subjectivity and learning, King and Kitchener focused on reasoning processes, including epistemological concerns toward the source and justification of knowledge, as disclosed through scored observation of attempts to solve ill-structured problems (i.e. problems lacking significant research and without the consensus of experts as to how to resolve them). 

King, Kitchener, Davison, Parker & Wood (1983) present seven progressive stages of “reflective judgment” falling within three basic levels where the upper stages (found within the last level) are typically never achieved.  In the pre-reflective level, people generally believe that knowledge is certain and gained by direct observation.  There is no need to justify a belief that directly reflects reality.  The second, quasi-reflective level includes those who believe that knowledge varies with context.  While reality exists, it can never been known with enough certainty to resolve individual subjectivity.  The best we can do is to evaluate each other’s biases.  The final level, reflective, involves the recognition that critical thinking skills can help discern some knowledge as better or more complete than other.  Knowledge claims can be justified by an integrated analysis of all that is available.  This allows for the development of a variety of strategies to help in discerning what knowledge claims to accept and use in governing one’s life. 

King & Kitchener’s (1994) meta-analysis of over 1,500 respondents in 25 different studies showed that a more sophisticated epistemology develops slowly, but steadily, from high school through advanced degree studies.  Median scores (on a 1 to 7 scale reflecting their reflective judgment stages) increased from high school (M=3.2) through college freshmen (M=3.6); college senior (M=4.0); graduate student (M=4.6) and advanced doctoral students (M=5.3).  In addition, adults who had not attended college (M=3.6) showed lower median scores than those who had completed college degrees (M=4.3).  Thus, there seems to be some regular advancement in reflective judgment as one progresses to more advanced studies.

Kuhn (1991) also focused more on the development of specific components of critical thinking skills than on attitudes toward education and subjectivity.  She used a cross-sectional study of four selected age groupings (ranging from late teens to sixties) and presented them with ill-structured problems.  Kuhn categorized her sample into four stages similar to those of Perry: realism where critical thinking is unnecessary (similar to Perry’s dualism); absolutist – critical thinking consists of comparing knowledge to observable reality (also similar to advanced dualism and early multiplicity); multiplist – critical thinking is futile, everything is opinion (Perry’s late multiplicity & relativism); and evaluativist – knowledge claims can be evaluated using critical thinking (Perry’s commitment in relativism).


Epistemological Style Theories

Epistemic style theories either dispute or place less emphasis on the notion that one’s personal epistemology is developmental in nature.  Instead, these theories find more explanation for one’s epistemic beliefs within their environment and personal agency.  These beliefs result from an interconnected web of independent cognitive (rational), social, cultural, psychological and experiential factors that are particular to the individual.  Thus, any effort to coax an individual’s acceptance of more mature epistemological beliefs depends upon recognizing the factors involved and manipulating these factors to produce the most desirable result.  However, unlike the assumed progressive and hierarchical nature of developmental beliefs, style theories have more difficulty establishing whether one set of epistemic beliefs is truly more desirable than another (being non-developmental, any classification as “more mature” no longer applies).  One’s worldview concerning the purpose and goal of epistemology plays a greater role.  Nature no longer makes this decision for us.  There is no longer a set progression of presumably more advanced stages or positions.  Personal epistemology becomes both more malleable and subject to the influence of one’s currently operational (and often culturally bound) worldview.  Theories advocating a style orientation are often less conspicuous as researchers hesitate to tread on the already established groundwork of Perry and others.

Prior to Perry, Royce (1964) detected the existence of three major epistemic styles, each incorporating it own set of value, affective, cognitive and justification systems.  These styles were identified as rationalism (relying more upon cognitive thought); empiricism (relying more upon the senses and empirical observation); and metaphorism (relying more upon narrative and subjective construction).  Similarly, Unger, Draper & Pendergrass (1986) developed an “attitudes about reality” scale that categorized styles along a continuum ranging from logical positivism (knowledge as stable, irreversible and determined) to social-constructivism (knowledge as changeable and constructed from historical and cultural definitions).  See Table 4.4 for a parallel listing of various categorizations of “style.”  Unger, et al. predicted that these styles would eventually be demonstrated to correlate with factors such as age, religion, birth order, and socio-political or socio-economic standing.  However, little has been done to investigate this further.

Wilkinson and Migotsky (1994) evaluated a variety of instruments that had been developed to measure personal epistemology.  Using factor analysis as a tool, they distinguished three factors that were significantly related in each measure.  These factors were labeled as naïve realism (emphasizing authority, literal thinking and an inability to cope with doubt), skeptical subjectivism (emphasizing the acceptance of uncertainty and the use of intuition and reflection in obtaining knowledge), and logical inquiry (emphasizing the use of rational thinking to


Table 4.4        

Parallel Categorizations of Epistemic “Style”

Royce (1964)

Unger, Draper & Pendergrass (1986)

Wilkinson & Migotsky (1994)


Logical Positivism

Logical Inquiry


Naïve Realism


Social Constructivism

Skeptical Subjectivism




resolve knowledge conflicts).  Schommer–Aikens (2002b), although more concerned with the multidimensional nature of personal epistemology, has also suggested that the concept may not be developmental in nature.  More recently, a study out of Belgium (Desimpelaere, et al., 1999) investigated the relationship between psycho-epistemological beliefs and religious beliefs.  Although their statistical research disclosed categories quite similar to Perry’s, they found that the apparent progress of people through the categories over time would not support the idea “that the concepts of Perry (1970) are embedded in a developmental scheme” (p. 135).


Multidimensional Nature of Personal Epistemology

Perry and other developmental schemes tend to assume that one’s epistemological orientation is uni-dimensional and can be measured using a single factor along a linear scale.  Style theorists, seeming to already recognize the numerous outside factors that can influence one’s epistemology, appear more willing to accept a complex and multidimensional nature.  Schommer (1990) noticed that although researchers treated epistemological beliefs as uni-dimensional, their research tended to concentrate on different aspects of epistemology.  This would suggest that personal epistemology is, in fact, multidimensional in nature.  Thus, one’s epistemology might consist of a compilation of one’s “system of more or less independent beliefs” (Schommer-Aikins, 2002, p. 104) within a number of identified, yet distinct, dimensions.  Yet, the existence of multidimensional factors does not preclude developmental theories.  Epistemic dimension theory may be developmental in the sense that there may be a naturally anticipated progression of beliefs within each dimension.  However progression (or regression) within each dimension may simply occur at different times and rates.  Still, one might expect a high degree of interrelatedness between these dimensions in such a way that regress or progress within a given dimension might naturally affect an interrelated dimension.  As Duell and Schommer (2001) reflect upon Schommer’s own work, they suggest that:

… the most important point of Schommer’s theory is that one cannot simply assume that epistemological beliefs are in sync.  This is particularly true when individuals are changing their epistemological beliefs (p. 440).


Schommer (1990) proposes five dimensions (factors) of epistemological belief that should be measured independently; a) simple knowledge (simple answers or complex integration); b) certain knowledge (absolute or subjective / lucid or ambiguous); c) omniscient authority (whether the source of knowledge is authority, critical thought, or social construction); d) quick learning (whether learning is immediate or obtained over time with perseverance); and e) innate ability to learn (genetically fixed or developed over time).  Schommer’s specific categories have generated some criticism due to the failure to separate philosophical epistemic concepts from educational process and learning concepts.  Yet Schommer-Aikins (2002) would suggest that philosophic concepts are not enough.  To get a true look at an epistemology that is operational, it is also important to acknowledge learning beliefs.  Several researchers have suggested regrouping these five factors into two main categories.  See Table 4.5.  Jehng, Johnson & Anderson (1993) designate these two categories as knowledge factors and learning factors.  Hofer (2001) suggests that the dimensions be clustered into two areas; epistemological (including the dimensions of certainty and simplicity) and process of knowing (including the dimensions of source or omniscience, justification, speed and ability).  A philosopher focusing on epistemology might suggest even further separation into three categories: a) nature of truth (certainty, stability and simplicity), b) acquisition of knowledge (source and justification), and c) personal capability (speed and ability).  Despite these concerns, Schommer has developed an Epistemological Questionnaire (EQ) designed to measure these factors and has provided one of the most accepted measures of assessing personal epistemology to date (Hofer, 2001).


Epistemological Domain Theories

Epistemological domain theories, like style theories, propose that learning and critical thinking is influenced by one’s epistemological beliefs.  Similarly, these theories also tend to reject, but do not necessarily disprove, the hierarchical and progressive nature of developmental approaches.  It is possible that there truly is a developmental continuum

Table 4.5. 

Dimensions of Epistemological Belief


Jehng, et al.


Philosophical Dimensions

Fixed Ability

Learning Factors

Process Factors

Personal Capability

Quick Learning

Source of Knowledge (Omniscient Authority)

Knowledge Factors

Acquisition of Knowledge

(Also Dimension of Justification)

Simple Knowledge (Orderly Process)


Epistemological Factors

Nature of Truth

(Also Dimension of Stability of Truth)

Certainty of Knowledge



in which one progresses within epistemological domains.  Domain theories tend to suggest that individuals may employ different epistemologies within different domains.  For example, different epistemic strategies and styles are likely to be used by the same individual in a science class as opposed to an art class or within his social or cultural life.  Thus, a person may validly or invalidly, consciously or subconsciously shift in his/her epistemic beliefs as the domain varies.  One area of contention concerns the classification of domains.  Should they be classified based upon area of study (disciplines or content domains) or based upon the type of knowledge being sought (ie. introspective or personal truths vs. physical facts).  Arguably, different types of truth are typically associated with different disciplines.  Hence, the hard sciences depend heavily on physical facts while business disciplines depend more on personal truths, value and ethical truths, and social facts.

Kuhn (2000b), although generally associated with developmental theories, accepts Schommer’s multidimensional view of knowledge.  However, instead of dividing these dimensions in the manner proposed by Schommer, Kuhn suggests dividing knowledge into five types of judgment domains.  These are: a) personal taste (e.g. music or food preferences); b) aesthetic judgment (e.g. art or music excellence); c) value judgment (both the areas and content of ethics); d) facts about the social world (soft sciences including sociology, education and psychology); and e) facts about the physical world (the hard physical sciences).  This concept, which can be used to propose an individual’s use of different styles among different domains, is then applied by Kuhn to developmental theories.  Here, she suggests that people develop in epistemic maturity at different rates within different judgment domains of knowledge. 

According to Kuhn’s theory, the personal taste domain would be the first to transform from dualism to multiplicity, followed by aesthetic judgment, value judgment, social facts and physical facts in that order.  On the other hand, development of each domain from multiplicity to the evaluativist (Perry’s commitment in relativism) level might occur in reverse order.  Additional studies are needed to evaluate this proposal.  If we recall Kuhn’s earlier proposal that epistemological knowing exists on a cognitive level above that of metacognition, one might hypothesize that the role of epistemological knowing is to develop mature distinctions between the epistemology to be properly applied within each different types of knowledge, together with considering and managing any philosophical incoherence that might exist in such a differential application of epistemological beliefs.

This multidimensional nature becomes an issue when it comes to instruments and measures.  It might be possible to devise a measure that differentiates between a person’s general epistemological beliefs (possibly defined as an average across a wide range of domains or as a total independent construct) and a person’s specific epistemological beliefs in the particular domain that is referenced in the instrument.  For example, if the instrument requests a student’s reaction with regard to a particular class or learning situation, it may not mirror his general or his overall combined epistemological viewpoint (Hofer, 2000; Paulsen and Well, 1998; Jehng, Johnson and Anderson, 1993).  Pintrich (2002) refers to this as the issue of domain generality versus specificity.  Chandler, et al. (2002) suggest that one may develop increasing generality (which he describes as a move from “retail” to “wholesale”) among the content domains over time and with experience.  Schommer (2002) has suggested that there may be some consistency (domain generality) in epistemological beliefs across content domains.  A 2002 study (Schommer-Aikins, Duell and Barker (2002) used Biglan’s classification of academic disciplines (using 2 distinguishing factors: hard vs. soft and pure vs. applied) to investigate whether such factors accounted for the majority of the differences in epistemology between the disciplines, or whether some aspects of domain generality remained.  The results suggest that the “issue of domain specificity/generality goes beyond a simple dichotomy” (p. 364).  Thus, the affect of external factors (such as academic experience) likely has an affect on the breadth of one’s domain applicability.


Survey of Research into Personal Epistemology



Student Epistemology

In general, research into areas of personal epistemology has yielded few univocal and unambiguous discoveries.  While it is clear that epistemology plays a role in one’s racionation and critical thinking, the specific mechanisms behind this influence remain elusive.  Still, there are a number of general areas of consensus among theorists and researchers.  These include the following studies that are not discussed in detail, but which are outlined in the study summary found in Table 4.6.

First, it seems clear that students who attend college are likely to develop more sophisticated epistemological thinking than those who do not (King & Kitchener, 1994).  In addition, more sophisticated beliefs are positively related to intrinsic motivational goals and openness to conceptual change learning (Qian and Pan, 2002), to academic performance (Hofer, 2000), to more sophisticated forms of moral reasoning (Bendixen et al., 1998), and to the adoption of mastery versus achievement goals (Bratan and Stromso, 2004).  Yet, despite Perry’s (1970) initial proposal, it appears questionable whether this improvement is developmental in nature.  Several theorists have disputed Perry’s developmental scheme (Desimpelaere, 1999) and specifically the idea that any such progress is linear (Schommer, 2002).  As a result, a majority of research has now begun to focus on comparing people’s level within multiple dimensions of epistemology.  Schommer’s itemization of five epistemological factors has received more acceptance than attempts to locate a person’s position within the Perry developmental scheme.












            Likewise, it seems clear that one’s personal epistemology is affected by environmental factors.  Research has suggested that one’s level within Schommer’s five epistemological factors is related to a number of societal variables.  Both family upbringing (Schommer, 1990) and years of education (King & Kitchener, 1994) have shown clear signs of being contributing factors.  Yet, there is also some evidence that exogenic thinking may increase or re-emerge with the passage of time unless countered by the intervention of further higher education or professional pursuits involving significant contact with divergent ideas and beliefs (Deopere, 1987; Desimpelaere, 1999).  However, Kuhn and Weinstock (2002) summarize their previous research and suggest that age and experience has a negligible effect on one’s progression toward her evaluativist level of understanding, “except among a small proportion of the population who are exceptionally highly educated” (p. 138).  With these basic observations in mind, I will turn to a discussion of five of the more recent and particularly informative studies concerning student epistemological beliefs.

A study by Jehng, et al. (1993) investigated the relationship between level of education, field of study, and people’s epistemological beliefs.  Here, the researchers’ sample of students at three universities in central Illinois was stratified into academic levels (freshman & sophomores; juniors & seniors, and graduate students) and four academic fields (engineering and natural sciences, arts and humanities, social sciences, and business).  A total of 1000 people were randomly selected by mailing questionnaires to 50 students within each of the 20 possible combinations.  A total of 486 questionnaires were returned, however only 398 were deemed usable.  The questionnaire consisted of a modified version of Schommer’s Epistemological Beliefs Questionnaire that had been reduced to 41 questions as a result of a pilot study and a Cronbach's reliability test of the returned questionnaires. 

Both educational level and field of study showed significant (ά=.05) main effects with graduate students reporting significantly higher scores on the dimensions of certainty of knowledge, omniscient authority and orderly process (knowledge factors) than their undergraduate counterparts.  In this case, higher scores signified a stronger inclination to believe that knowledge is uncertain, acquired from integrative research, and not always orderly.   In each case, the difference between the two undergraduate divisions was insignificant.  There were also no significant differences regarding the dimensions of innate ability or quick process (learning factors).  The data further showed that students in the soft fields (humanities and social science) had significantly higher scores along the dimensions of certainty of knowledge, omniscient authority, and orderly process than those in the hard fields (natural sciences and business).  Once again, no significant differences were found along the dimensions of innate ability or quick process. 

These results suggest that one’s sophistication within the epistemological knowledge factors tends to be higher in graduate students.  In addition, those students who study in the humanities and social sciences are more likely to believe that knowledge is uncertain, are more likely to rely upon their independent reasoning ability, and tend to believe that learning is not an orderly process than students who study in the fields of natural sciences and business.  Interestingly, neither level of education, nor field of study had any affect on the two epistemic “learning” factors.  This differs from Hofer (2000) who found significant differences between field of study (psychology and hard science) in all of her epistemic dimensions.  According to Jehng, et al. (1993):

Three conclusions can be tentatively advanced on the basis of the study: (a) epistemological beliefs are multidimensional, (b) individual beliefs about learning may evolve as one is exposed to more advanced education, and (c) individual epistemological beliefs depend upon a student's academic field. … While additional facets of beliefs about learning and knowledge may remain to be discovered, it is apparent that multidimensional models provide a better approximation to the structure of epistemological beliefs than Perry's (1968) one-dimensional model, which up until now has been the touchstone for research on this topic. (p. 32)


Jehng, et al. (1993) also recognize the limitations of their study.  Due to concerns with sampling, the generalizability only extends to college students attending a campus similar to the one used in the study.  In addition, the instrument used was modified from the original Schommer questionnaire.  Although this new version was analyzed for internal validity and modified using a number of appropriate methods, its external validity and reliability remain unknown.  The authors also recognize that their conclusion might be interpreted differently, and remind us:

Our preferred explanation of these facts is that individual epistemological beliefs are shaped by the surrounding culture, and we have tried to show that this explanation provides an attractive account of the facts.  However, there is an alternate explanation that is also plausible.  It could be that students who choose to go to graduate school have different epistemological beliefs than the typical undergraduate, thus explaining the effect of educational level observed in this study, and that students who choose different academic majors already have different epistemological beliefs, thus explaining the observed effect of field of study. (p. 34)


Youn (2000) wondered whether culture might directly influence epistemological beliefs.  Youn replicated Jehng’s research with both an American (n=496) and a Korean (n=487) sample.  Here, the American data revealed that educational level (graduate vs. undergraduate) continued to show significant positive effects on both the epistemological factors of knowledge and learning.  Although there was a significant difference between graduates and undergraduates (with graduates exhibiting greater degrees of epistemological sophistication), once again there were no significant differences between the undergraduate levels.  Only the knowledge dimensions showed any significant difference based upon academic major.

The Korean data showed that soft field students (academic major) had significantly higher scores (indicating more epistemological sophistication) than hard field students for knowledge; however, there was no difference in learning factors.  For the Koreans, there was no overall effect by educational level.  When comparing American and Korean mean scores, the Korean students showed little change in their scores from undergraduate through graduate programs.  On the other hand, American students showed significant positive changes as they progressed through these programs.  While the limitations may be considerable, these results suggest that domain differences and educational experience may have different affects on epistemological beliefs within different cultures.

Schommer and Walker (1997) sought to determine if epistemological beliefs predicted the attitudes of high school students toward learning and attending college.  Over 1000 students from one Midwestern high school were selected to take the Schommer Epistemological Beliefs Questionnaire.  This correlational study was regressed on the variables of gender, giftedness, year in school, and four of the epistemological belief factors.  The most frequent and significant predictor of perceived educational value was the dimension of fixed ability.  The moderate negative correlation indicated that the less students believed in a fixed ability to learn, the more likely they valued education.  Beliefs in quick learning and certain knowledge were also significant predictors.  The less students believed in quick learning or certain knowledge, the more likely they were to recommend a college education.  Although again limited in its generalizability, the study supports the intuition that students with more sophisticated epistemological beliefs will be more likely to cherish education.

Schreiber and Shinn (2003) investigated the correlation between student epistemology and their selection of information processing methods.  Here, 115 students from a community college were measured using Schommer’s epistemological dimensions and Schmeck & Ribich’s (1978) three distinct processing routes: agentic processing (serial steps with a fact retention focus); elaborative processing (a method that relies heavily on prior knowledge and experience as an ultimate source of reference and information; and deep processing (a more critical process of seeking out, comparing and organizing outside information for the purpose of evaluation).  Students are not categorized solely in one of these three categories, but may possess differing levels (or ratios) of one or all three.  The more critical student will likely be high on all three, while the more challenged student will be low on all or unbalanced in their scores.  The researchers discovered a low to moderate (yet significant) negative correlations between fixed ability and all three processing routes.  Using a path analysis to investigate the direction of causation, it became apparent that a belief that one’s learning capabilities are fixed results in more infrequent use of any processing method.  In addition, there was a significant positive relationship between simple knowledge and agentic processing.  This would suggest, as one might surmise, that those who don’t find much value in integration would show a higher propensity to rely on fact retention methodologies.

Tolhurst (2004) performed a longitudinal study of more than 700 students to investigate whether major changes in course design might impact students’ epistemological beliefs.  In short:

Can we structure our curriculum, courses and learning environments so as to encourage the development of more sophisticated epistemological beliefs in students that lead to greater personal involvement and acceptance of responsibility for learning. (p. 2)


Students were given the Schommer Epistemological Questionnaire and the Hofer Discipline Focused Epistemological Beliefs Questionnaire both before and after the 12 week course.  However, what Tolbert discovered was not unequivocal.  Significant change in a more sophisticated epistemological direction was observed in the area of a) increased belief that we can learn how to learn and b) decreased belief that we should seek out simple answers.  However, interestingly, students seemed to regress in their belief that learning is typical a first time process and seemed to show an increase in their dependency upon authoritative sources of knowledge.  Yet, the results of the Hofer Discipline Focused questionnaire suggested that this might be a domain specific issue given that the class was one in introductory computer information systems.


Teacher Epistemology

Much of the research in the field of personal epistemology has come out of educational psychology.  Therefore, it is only natural that research into the epistemology of teachers would be of interest.  It is important to know if a teacher’s own epistemological beliefs affect their teaching methods and effectiveness.  It might also be interesting to know if the teacher’s epistemological beliefs affect those of his/her students.  Thus, research has recently focused on the personal epistemology of teachers and education students.  How are they developed?  Do these beliefs affect one’s teaching effectiveness?  Are they stable or can they be intentionally modified? 

Prior research suggests that teacher epistemology may be a factor that affects one’s selection of teaching strategies (Hashweh, 1996), the willingness to use problem-centered approaches (Martens, 1992), openness to change (Sinatra and Kardash, 2004), and efforts to adapt curriculum (Benson, 1989).  Here, I will review seven of the most recent and important studies investigating the impact of personal epistemology on teachers and their teaching methodologies (Table 4.7). 

Schraw and Olafson (2002) set out to achieve several goals, including: 1) to establish a measurable typology of epistemological world views; 2) to discover whether these world views correspond to types of teachers; and 3) to link those types of teachers to consistent differences in teaching practice.  In an attempt to do so, the researchers used a sample of 24 urban teachers who were enrolled in a graduate course in curriculum theory.  The teachers completed a group of quantitative tests, including the Epistemic Belief Inventory (Schraw, Bendixen & Dunkle, 2002).  In addition, the students rated three vignettes that described teaching from a realist, contextualist, and relativist world view.  Finally, the researchers reviewed both a number of written assignments prepared during the course and individually interviewed each student.

Unexpectedly, the quantitative findings suggested that years of teaching experience moderately correlated significantly and positively (less sophisticated) with both simple knowledge and omniscient authority, indicating that teachers become more accepting of simple, authoritarian views as they progress in their teaching careers.  Years of teaching also moderately correlated significantly and negatively to both the





contextualist and relativist world views.  Thus, as teachers progress, they tend to adopt a more realist epistemology.  Yet, interestingly, the data showed that age correlated significantly and negatively with omniscient authority, suggesting that teachers became less willing to accept authority as a source of knowledge with age – a seeming contradiction as age and years of teaching experience should also correlate.  In addition, the teacher’s epistemological world view did not appear to correlate with their teaching practices.  Many claimed to be “student centered” but rigorously agreed with the need to utilize the mandated curriculum taught from a more realist viewpoint. 

The Schraw study immediately drew a number of critiques.  Martinez (2002) expressed concerned with the three-category system of world views and with the measurement methods.  Schommer-Aikens (2002) questioned whether teachers can be characterized within the selected three worldviews.  Instead, she suggests, it is likely that individuals can hold multiple views, with some views being subordinate or infrequently applied.  Hofer (2002) added that, "there are obvious generalizability problems in using these findings from a sample of 24 teachers from the same class, trained at the same institution, teaching in the same district, typically with only 3 years of experience" (p. 171).  She seems further concerned with many of the assumptions adopted by Schraw, et al. (2002).  These include the assumption that: a) teachers are “consistent in their epistemological beliefs and can be characterized by only one world view at any particular point in time” (p. 103); and b) “that teacher’s (sic) epistemological world views are consistent across different academic domains” (p. 104). 

Another question concerns whether a teacher’s epistemology is modifiable.  In 1999, forty-one "master teachers" were selected based upon proven teacher excellence from a national pool of applicants submitted to NASA for entry into a graduate technology training program.  For Howard, McGee, Schwartz, & Purcell (2000), this seemed the perfect opportunity to investigate the transformation of teacher epistemology.  Here, a sample of 20 female and 21 male teachers with varying levels of experience (5 years to 20+ years) participated in a concentrated residential program lasting over four weeks where they remained together as a community at all times.  Classes were specifically taught using constructivist techniques.

The researchers used a modified Schommer Epistemological Belief Questionnaire which only examined four of Schommer’s initial five dimensions to compare pre-test and post-test results.  The dimension of source of knowledge was not included.  According to Howard, et. al. (2000), teachers showed significant change in three of four factors in the direction of a more sophisticated (constructivist) epistemology [simple knowledge; quick learning; and certain knowledge].  The factor of fixed ability did not show significant change.  Allegedly, this implied that a) using “constructivist approaches to training teachers may actually produce epistemological change in line with constructivist philosophies;” and b) that “epistemology may be a less stable trait than was previously supposed -- the teachers here demonstrated major changes in only a 4-week period” (p. 459).  According to the authors, the “study provides compelling evidence that the epistemological dimensions hypothesized by Schommer are subject to change” (p. 464).  Yet clearly, significant limitations exist concerning the sampling, generalizability, the disciplinary (hard science) focus, and the short, yet intense, training period.  One might also question what post-test results might show one year later.

White (2000) investigated a) whether pre-service teachers differ in their epistemology, b) whether they move from epistemological category to category in a stage-like fashion, and c) whether there a relationship between the teacher’s level of education and his/her epistemology?  Using an admitted convenience sample of 20 students (17 female, 3 male) in a Human Development class at a mid-size Midwestern university, White conducted interviews probing each student’s thought process in resolving hypothetical classroom situations.  As a secondary measure, White also reviewed students’ written responses to case studies presented during the course.

Instead of using previous categories and instruments, White placed her subjects into the following five categories in order of perceived increasing maturity; Departing Absolutist (n=2; both sophomores); Intuitive Relative (n=2); Selective Relative (n=10, 1 - freshman; 6 sophomores; 3 juniors); Informed Relative (n=2, 1 sophomore, 1 senior); and Reflective Relative (n=3, 1 sophomore, one junior, one senior).  Based on her observations, White concluded that: a) pre-service teachers' views about problematic classroom situations reveal wide differences in epistemology; b) movement from one category to category does not happened in a stage-like fashion; c) teachers’ beliefs appear to interconnect in a generally coherent web; and d) there appears to be no discernible relationship between the teacher’s class level and their epistemology.  Obviously, this study has significant limitations, especially regarding its generalizability.  However, as a pilot qualitative observation; it does lend credence to the fact that teachers are quite different in their epistemological perspectives.  It also gives us pause to question the developmental and hierarchical nature of epistemic beliefs. 

Like White, a study by Many, Howard & Hoge (2002) suggests that teachers see through a number of different epistemological lenses.  Here, 19 education students (15 female, 4 male) enrolled in a literary methods class where they were both interviewed concerning their epistemological beliefs and maintained a reflection journal for all readings from the class.  The author concluded that while several students relied upon a predominant epistemological system, the majority of students exhibited varied epistemological beliefs that were loosely coupled.  There were often conflicting epistemological perspectives creating dissonance that the students did not recognize.  As a rule, “students did not align consistently with specific epistemological orientations, rather, at times they seemed to shift among different epistemological lenses” (p. 307). 

Tsai (2002) investigated natural science instructors in Taiwan and classified them as either traditional, process, or constructivist concerning their epistemology in three areas; a) beliefs in teaching science; b) beliefs in learning science; and c) beliefs concerning the nature of science.  When all three areas were combined, 57 to 59% of the instructors reflected traditional thinking, 27 to 32% reflected process thinking, while 11 to 16% reflected constructivist thought.  When looking at their individual agreement in all three areas, 21 of 37 teachers showed consistent nested beliefs across all three areas.  Of these, 15 showing nested traditional responses; 4 exhibited nested process responses; while only 2 exhibited nested constructivist responses.  For those teachers who did not demonstrate nested epistemologies, most of them still were congruent across two of the three areas.  Interestingly, the degree of nested epistemologies tended to increase with greater teaching experience.

Tsai made several observations.  First, teachers of science tend to be quite traditional.  Second, while he was impressed by the amount of congruence within this domain, he seemed a bit dismayed that the congruence was at the traditional level.  He went on to recommend that science education programs investigate “conflicting conceptions of teaching, learning, and science for pre-service science teachers” (p.779) and to discover whether “changing teachers' beliefs of teaching and learning science may be a prerequisite of changing their beliefs about science, or vice versa" (p. 779).  Interestingly, he seems to assume that it is more important to promote a constructivist epistemology than to develop consistent, congruent and nested beliefs.

Brownlee, Purdie and Boulton-Lewis (2001) performed a longitudinal study to investigate how epistemological beliefs changed over time among a group of pre-service teachers (students training to become teachers) in Australia as result of an intentional program directing students to reflect upon their personal epistemologies.  The study used two groups, a research group (RG) who participated in the program (n=29) and a control group (CG) who did not (n=25).  Both groups took a pre-test and post-test Schommer Epistemological Beliefs Questionnaire.  The control group (CG) provided qualitative data by completing a set of written statements concerning their beliefs about knowledge during the first week of class, and then again at the end of the year.  On the other hand, each student in the research group (RG) turned in regular written assignments and was interviewed both before and after the program.

Brownlee, et al. (2001) found that the research group showed a significantly (p<0.05) larger change over time in the desired (more sophisticated) direction regarding quick learning, however, only within the sub group of "learning is quick".  There was also a similar and significant change in the desired direction regarding certain knowledge, yet only within the subgroup of "avoids ambiguity".  Significant positive changes for the research group were also found within innate ability under the subgroup "cannot learn how to learn" and within omniscient authority within the subcategory "depends on authority." 

The qualitative data was coded into five categories consisting of: CON (construct reasoned truth); CONREC (construct reasoned truths and receive absolute truths); SUBREC (construct subjective truths and receive absolute truths); REC (receive absolute truths); and INCON (inconsistent beliefs).  Interestingly, the largest longitudinal increase in the research group was in the category of inconsistent beliefs.  Yet, Brownlee concluded that, "The RG students have experienced stronger growth in sophistication of epistemological beliefs because of a greater increase in the percentage of students with INCONSISTENT beliefs over the year" (p. 260).  However, even though change was established, one might question the propriety of claiming “increased” epistemic sophistication when one graduates from a consistent position to a set of inconsistent, and arguably incoherent, set of epistemological beliefs.  Thus, the question arises whether the journey through epistemological change necessarily includes periods of confusion, disequilibrium and conflicting inconsistent beliefs.  Brownlee concludes that:

When individuals are encouraged to reflect on and possibly reconstruct their existing beliefs, they may experience confusion or disequilibrium.  In the current study, the increase in the number of students who described INCONSISTENT beliefs may reflect such confusion that emerges from wrestling with discrepancies between pre-existing beliefs and new information.   This would indicate that the teaching programme has helped students to start the process of changing epistemological beliefs. ... Another explanation for the increase in the number of students with INCONSISTENT beliefs may be that they espouse knowledge about epistemological beliefs, rather than actually changing their beliefs. (p. 262)


A later follow-up study (Brownlee, 2003) resulted in third interviews of 11 of the original research group students who had since taught in primary school.  Brownlee used the same qualitative categories from the prior study, except that, without explanation, she eliminated the category of inconsistent.  With the third interview, she found that within the category of “beliefs about knowing,” two participants changed from mixed (either CONREC or SUBREC) to constructivist while four participants decreased from constructivist back to mixed.  In the category of "beliefs about experts," most seemed to remain as constructivists, however one increased from received to constructivist, while one participant went from constructivist back to received.  This seemed to show inconclusive results and change in each direction.

Based upon these studies, it seems clear that education can lead to a teacher’s reflection and consideration of their epistemological beliefs.  It is likewise clear that teachers exhibit a large range of epistemological beliefs and tend to rely on predominately one epistemological style.  Yet, these studies have been unable to establish that any particular philosophy is epistemologically better or more mature—or that even that teachers will recognize one as better than the other.  While several studies seem to presume (and possibly promote) a constructivist epistemology, teachers (especially after many years of service) don’t necessarily follow suit.



Religious Worldview and Personal Epistemology

One of the correlations that this study intends to investigate is whether any relationship exists between one’s religious worldview and their personal epistemology.  There has been little research into this relationship.  Maybe this is because of its controversial nature, maybe it is because the relationship is not interesting to enough researchers, or maybe we simply don’t want to know.  Still, of the occasional studies, four of the six (of which three are dissertations) investigate whether strong religious beliefs correlate with naïve thinking styles.  The remaining two more recent studies investigate the effect of strong religious faith on reflective thinking and whether or not the domain of religion is more susceptible to relativistic and contradictory thinking.  See Table 4.8 for a summary of these studies.

A study in the mid-1980’s at the University of Northern Colorado (Powers, 1985) investigated 40 students of both genders at different stages in their college education.  20 of these students were involved with mainline Christian organizations, while 20 were involved with evangelical Christian organizations.  The study was conducted using recorded interviews graded by a trained Perry scheme rater.  There was no comparison with any normative scores and no rating of any non-religious control group.  Based upon this study, it was concluded that there was no difference in ratings along the Perry scheme between the mainline and evangelical subjects, and no differences based upon gender.

A later dissertation study (Nevard, 1988) at the same University compared a non-religious control group with both a group involved with Campus Crusade for Christ and a group from Colorado Christian College (an evangelical university).  The study,





involving the rating of student interviews by a single paid Perry position “rater,” showed a moderate yet significant correlation between dualism and the two religious groups.  There were no significant differences related to level in school or gender. 

Six years later, another dissertation study (Copeland, 1994) was conducted using 242 students in Christian institutions of higher learning in Western Texas.  The study allegedly found a significant inverse relationship between Protestant Fundamentalism and both intellectual development on the Perry scheme and moral reasoning as defined by Kohlberg.  Fundamentalists, according to Coleman, tend to be dualistic in nature and do not progress much further up the scale.  However, this study exhibited a number of unusual limitations, some of which were not raised by the author.  For example, the study used an objective instrument (Scale of Intellectual Development) (Erwin, 1983) that has been widely criticized as a tool that deficient in measuring any Perry position other than “Dualism” (Stonewater, Stonewater & Hadley, 1986).  Moreover, Copeland’s study used a sample of "Protestant fundamentalists.”  Yet, interestingly, Copeland defines Protestant fundamentalists as follows:

Protestant fundamentalism is a religious ideology characterized by strict adherence to the tenets of Biblical authority, literalism and inerrancy, atonement for sins, the virgin birth of Christ, resurrection and millennial eschatology.  This adherence is enforced through separatism, submission to authority, intolerance for ambiguity and militant exposure of all opposing beliefs and practices. (p. 6)

Yet, by using such a definition, especially concerning the enforcement of adherence, Copeland essentially selects his sample to essentially correspond with Perry’s “Dualism.”  Thus, it is not surprising to find that a sample selected for its dualistic qualities would show a high correlation for “dualism.”  

More recently, a group of professors in Belgium investigated the relationship between psycho-epistemological styles and religion (Desimpelaere, et al., 1999).  Using one of the researcher’s (Hustsebaut, 1996) three categories for religious dimension: orthodoxy, historical relativism and external criticism, a comparison was made to three categories of psycho-epistemological styles (naïve realism, logical inquiry and skeptical subjectivism) (Wilkinson and Migotsky, 1994).  These styles were later found to equate well with Perry’s concepts of dualism, multiplicity and relativism /commitment.  The religious dimensions were defined based upon two continua; a) whether belief is deemed literal or symbolic and b) whether participation in transcendence is deemed inclusive or exclusive.  Orthodoxy included believers who accept that religious questions have one right answer promulgated by religious authority or writings that are immutable.  “This religious style is positively related to anxiety when faced with new questions” (Desimpelaere, et al., 1999, p. 132).  Persons falling within the category of external criticism were those who “fundamentally question the meaning and possibility of religious belief, and reject both the literal and symbolic affirmation of religious statements” (p. 132) or they “feel rebellious toward God and want to be autonomous” (p. 132.)  Historical relativists are those who:


…believe in God, but think about religion in a symbolic way.  They are aware of the fact that their interpretation is but one possibility beside so many other possibilities and that meaning can change over time.  For them, speaking about the absolute is more of a quest.  This religious style is positively related to openness to complex questions (p. 132).

The researchers hypothesized that orthodoxy would correlate positively for dualism and historical relativists would correlate positively for relativism and commitment.  They were not sure what they would find with the external critical group.  The results confirmed the expected correlations for orthodoxy and historical relativists; however they were surprised to discover that the external critical group also had a strong correlation with dualism.  Furthermore, “as age went up, so did dualism as well as orthodoxy” (p. 134).

The researchers first concluded that Wilkinson and Migotsky’s (1994) categories did not factor out as well as Perry’s.  Yet they could not agree, however, “with the fact that the concepts of Perry (1970) are embedded in a developmental scheme” (Desimpelaere, et al., 1999, p. 135), especially when older subjects were found “to think more dualistically, which is opposed to Perry’s ideas” (p. 136).  In addition, “although orthodox subjects make a choice in favour of the church, whereas external critics accept the skepticism of society and science, the underlying thought processes seem to be the same” (p. 137).  In general, the historical relativists were professed believers who “were aware of the relative aspect of believing” (p. 137).

While the studies at the University of Northern Colorado show that there may be a correlation between dualism and Christians whose theology accepts an omniscient authority as the source of knowledge, they say nothing about a number of issues that hopefully this study can investigate.  First, is there any correlation between religious worldview and any other knowledge or learning dimensions of personal epistemology?  Second, do different levels or categories of Christian worldview also correlate in a similar manner?  Additionally, is there any difference between these correlations based upon the level of education that the Christian sample has completed? 

A recent study by Montomery, Sandberg and Zimmerman (2005) examined a small sample of college students to investigate whether students tended use a more relativistic epistemology when it came to religious studies than in their physical and social science domains.  Not only did they find a significant difference between epistemological beliefs among these domains, but they also discovered a significant difference in a student’s willingness to accept clearly contradictory claims within their religious studies than in their other fields of study.

Finally, Dale (2005) used King and Kitchener’s (1994) concept of reflective judgment to see whether immaturity in reflective judgment correlated with a high degree of reference to one’s faith and an espoused epistemology of revealed truth.  Dale used a sample of 38 evangelical students at an eastern seminary.  No such correlation was found.  According to the author, the study “seems to indicate that it is possible for seminary students with an epistemology of revealed truth to develop reflective judgment.  This provides confidence for the seminary to develop students’ reflective judgment” (p. 63).

Here, the studies seem to report conflicting results.  To some, it appears as if those with strong religious faiths tend to have less mature epistemological worldviews.  Dale (2005) seems relieved to find that the promotion of reflective judgment and the related concept of critical thinking is not precluded by an epistemology of revealed truth.  This is an area where this study has the potential to shed much light.  Yet, even if those with strong religious beliefs have the same potential to adopt a mature epistemological worldview – one that hopefully permits and promotes honest worldview integration – the idea that students are willing to accept lower epistemic standards and contradictory evidence in the religious domain is troubling.


Special Assumptions and Limitations of Epistemological Research

As with any field of study, there are certain assumptions and limitations that are inherent in the very nature of the discipline.  These go beyond the particular limitations affecting specific research projects.  For example, most research into epistemology is affected by the limitations of self report.  While self reporting is likely not a problem with instruments investigating personal tastes or physical manifestations of pain, it is unknown how conscious a given individual is at any given time with his epistemological beliefs.  As Schommer-Aikins (2002) suggests, “epistemological beliefs are for the most part unconscious” (p. 115).  Such esoteric concepts may be difficult to honesty report and may be more susceptible to misunderstanding or cultural and other environmental affects.  Direct observations (such as observation of ill-structured problem solving) and qualitative interviews may be more effective, but are subject to their own limitations regarding observer consistency and bias.  In addition to measurement limitations, other limitations include the possible mischaracterization of the concept of epistemological beliefs, the complex and interconnected nature of the concept, and assumptions concerning what are truly “more sophisticated” epistemological beliefs.


Mischaracterizations of Epistemological Beliefs

Occasionally, those engaged in research concerning personal epistemology (and often outsiders without full understanding of the issues involved) equate the more mature or advanced stages with strong or metaphysical (rather than procedural) relativism or constructivism.  Thus, the findings of epistemological research are claimed to establish the relative nature of truth itself.  However, it must be remembered that truth and knowledge are not the same concept.  Insofar as epistemic maturity limits itself to the relative or constructive nature of human knowledge (including issues of humility and the fallible nature of human knowing), this is not a serious problem.  Yet, any extrapolations into philosophical claims concerning the nature or ontology of truth add unwarranted assumptions to the concepts under investigation.  Surely we learn through time and maturity that some things are, and likely will remain, beyond our comprehension.  Human reasoning is imperfect and prone to mistake, irrationality, misapplication of logic, personal bias or, at its worst, volitional manipulation.  As a result, certainty is a level of knowledge that we cannot expect to achieve.  But this does not vitiate the concept of truth.  Instead, it requires the use of all of our available gifts and resources to discern and choose between conflicting claims to truth.  Kuhn and Weinstock (2002) wisely replace Perry’s position of commitment in relativism with the term evaluativist.  As they observe:

At the heart of the evaluativist epistemological position is the view that reasoned argument is worthwhile and the most productive path to knowledge and informed understanding, as well as to the resolution of human conflict.  Competing to some degree with this set of values in modern society are the values of social tolerance and acceptance – reflected in the “live and let live” and “to each his own” adages  ... it is a deceptively simple step down a slippery slope from the belief that everyone has a right to his or her opinion to the belief that all opinions are equally right.  Tolerance of multiple positions, in other words, becomes confused with discriminability among them (p. 138-139).



Sophistication of Epistemological Beliefs

Although there remains some dispute whether one’s personal epistemology consists of simply an acquired style, or occupies a position on a developmental scale of maturing epistemology, nearly all of the studies assume that there are epistemological beliefs that are distinctly more mature and sophisticated.  While all studies assume that exogenic knowledge is less developed than endogenic, it remains undecided what the highest level would look like.  It is here where one should be particularly cognizant of each researcher’s bias and the motivation behind selecting a particular system as the presumed culmination of epistemic growth.  Schommer-Aikins (2002) suggests that most researchers presume epistemological maturity to consist of the “learner’s propensity to believe that knowledge is tentative and complex and that learning is gradual and controllable” (p. 106).  She adds that research trends and observations of expert and novice learners bear this out.

It is clear from work using either research methodology [qualitative or quantitative] that the idea of sophisticated beliefs embraces evolving knowledge, multiple approaches to the justification of knowledge, integration of knowledge, and for those willing to entertain a broader conception of epistemological beliefs, gradual learning, and ever growing ability to learn” (p. 113).


For some (such as Perry, 1970), the pinnacle may be a pragmatic acceptance of relativism.  For others, it consists of balance (Schommer-Aikins, 2004).  Still for others, epistemic sophistication is defined by a constructivist philosophy (Brownlee, 2001).  Yet, for many, it is an ability to discern and critically apply a variety of different philosophies or processes (Hofer, 2002).  For purposes of this study, it might be suggested that the highest level is that level that engages in the most honest and objective worldview integration and analysis.


Measures of Personal Epistemology



Although there are several recognized measures and methods for investigating personal epistemological issues (Duell & Hofer, 2001), there are significant differences in the concepts underlying most.  For the most part, those attempting to measure a unidimensional developmental theory resort to the Measure of Epistemological Reflection (Baxter Magolda, 1992).  This instrument provides a simple score along Perry’s developmental continuum.  However, this instrument requires time consuming qualitative interviews and the use of expert and qualified coders and interpreters.  In addition, there have been several studies suggesting that while the measure is good for identifying those falling within Perry’s dualism position; it remains deficient in its ability to categorize persons in the remaining positions (Stonewater, Stonewater & Hadley, 1986).  King and Kitchener (1994) have also developed a qualitative Reflective Judgment Interview (See also Duell & Hofer, 2001), and have recently begun working with others on a paper and pencil measure tentatively known as the Measure of Epistemic Cognition (Wood, Kitchener & Jensen, 2001).

On the other hand, those arguing for a multidimensional theory of personal epistemology have incorporated paper-and-pencil tests for some time now.  Schommer-Aikin’s Epistemological Questionnaire (EQ) (1990) has been used in numerous studies, often with significant modification (Jehng et al., 1993), however its reliability and validity has occasionally been questioned.  Schommer-Aikins refers to the instrument, which has undergone continual reexamination, as an experimental measure with properties that are still under investigation (Schommer, 1998).  Due to the perceived inability for the EQ to measure domain specific beliefs (Qian and Alvermann, 1995), another domain specific scale has been developed by Hofer (2000).  Recently, another measure known as the Epistemic Belief Inventory (EBI) has been developed by Schraw, Bendixen & Dunkle (2002) using much of the groundwork established by Schommer-Aikins.

Since I have a) embraced a multidimensional outlook, b) have no need for a domain specific measure, and c) desire a paper-and-pencil measure that will yield a numerical and hopefully interval score that may be correlated with other such measures, I have focused on the EQ and the EBI.  Summaries for the studies mentioned in this section are included on Table 4.6 dealing with student research.


Schommer Epistemological Questionnaire (EQ)

Schommer’s questionnaire consists of 63 items within 12 sub-factors.  These 12 sub-factors allegedly load onto her five  dimensions.  However it seems a bit controversial as to which sub-factors load onto which dimension.  In fact, Schommer found “factor analytic evidence for four of the five beliefs [dimensions], but failed to identify an omniscient authority [source of knowledge] factor” (Schraw, Bendixen & Dunkle, 2002, p. 262).  According to the factor analysis of Schraw, et al., the instrument does seem to load onto five factors; however they are slightly different than those described by Schommer.  Schraw et al. label these factors as a) innate ability (α = .74); b) incremental learning (α = .64); c) integrative thinking (α = .61); and d) certain knowledge.  The fifth factor of certain knowledge actually divides into two separate factors described as certain knowledge 1 (pertaining to science’s ability to ultimately discover universal truth) (α = .74); and certain knowledge 2 (pertaining more toward our ability to achieve certainty) (α = .53) (pp. 266-267).

The EQ items utilize a 5 point Likert Scale ranging from strongly agree to strongly disagree and is designed for used with college students or adults.  The test is considered domain general, however a domain specific version is available that instructs students to keep a particular academic domain in mind while taking the test and includes other psychometric modifications (Schommer and Walker, 1995).  Separate tests with modified questions have been developed for middle school, high school (Schommer, 1993; Schommer et al., 1997), and even foreign students (Braten and Stromso, 2004).  The scores are then recoded so that the high scores represent the epistemologically naïve positions.  Schommer (1998) suggests that if one intends to compare only post-test epistemological results to those reported by her, the researcher should derive z scores.  However, this need not be done if the researcher intends to compare pre-test and post-test results within their own study.

Schommer (1993) reports EQ test-retest scores of .74 for college students while no such reliability statistics are stated for middle school or high school samples.  Inter-item correlations are reported as between .63 and .85 for items within each belief dimension.    According to Schommer, external validity is shown by the fact that the items were screened by professionals in the field of educational psychology.  Predictive validity is established by the fact that belief in simple knowledge has been shown to predict comprehension of academic texts and thinking about controversial issues (Schraw, Dunkel and Bendixen, 1995); and that the speed of learning factor has been shown to predict grade point average (Schommer, 1993) and problem solving within structured content (Schraw,, 1995).  Finally, confirmatory factor analysis replicated the four factor structure of the test (Duell and Schommer-Aikins, 2001).


Epistemic Belief Inventory  (EBI)

In 1995, Schraw, Dunkle & Bendixen (1995) began work on a new measure incorporating much of the research commenced by Schommer.  However, they predicted that significant analysis of the items and sub-factors would eventually yield a slightly modified set of Schommer’s initial five factors without the problems inherent in bisecting Schommer’s certain knowledge category.  They initially wrote 25 items for each of the hypothesized factors (125 total questions) and after several combinations of pilot studies and item factor analyses whittled them down to a 32 item test that loaded on five factors labeled a) fixed ability; b) certain knowledge; c) omniscient authority; d) simple knowledge; and e) quick learning.  They later reported their efforts to validate the inventory using significant factor analysis of both the EBI and Schommer’s EQ (Schraw, Bendixen & Dunkle, 2002).

Factor analysis found that the EBI’s five factors explained substantially more sample variation than did the EQ’s factors.  The EQ yielded 17 factors that explained 39% of the sample variation while the five factors for the EBI explained 64% of the sample variation. According to Schraw, et al. (2002), “In addition, the EBI did not yield additional, difficult to interpret factors as did the EQ“ (p. 268).  Upon retest one month later, the test-retest correlations for each of the five factors were as follows:  omniscient authority (.66); certain knowledge (.81); quick learning (.66); simple knowledge (.64); and innate ability (.62).  The coefficient alpha scores were .65, .63, .60, .66, and .63, respectively.

According to Schraw, et al. (2002), their findings suggest several conclusions.  First, ”the two instruments differ with respect to the number of factors they yield and the degree to which these factors match theoretical prediction” (p. 271),  Second, “differences exist with respect to the proportion of the sample variance explained by the two instruments” (p. 271).  Third, the EBI “had considerably better test-retest reliability than the EQ” (p. 271).  The authors further claim that the EBI had better predictive validity through observations that four of the five EBI factors were “modestly, though significantly” (p. 272) correlated to total reading comprehension scores also taken during the testing.  Likewise, Bendixen, et al. (1998), while using an earlier prototype of the EBI, found that epistemic belief scores on all five factors correlated highly with the student’s moral reasoning scores.

Based upon the foregoing, this study will rely upon the EBI as the instrument for measuring the subject’s personal epistemology.  Although the EQ has been the basis of many studies, most have felt the need to modify it.  While the EQ does break down into 12 subfactors, the 5 main factors utilized by the EBI are sufficient for my inquiry.  This study will be analyzing and correlating enough other factors (religious worldview, intrinsic motivation and vitality) that the five present under the concept of personal epistemology will provide plenty to work with.  With this in mind, the EBI seems to provide the best developed and most valid instrument available for our purposes.



The field of personal epistemology, while relatively new, recognizes the relation between one’s epistemological beliefs and their cognitive maturity.  It also seems apparent that epistemological beliefs encompass not only beliefs concerning the nature of truth and its natural attainability, but also beliefs concerning the agent’s capabilities in discerning truth. As such, it is anticipated that those with strong and mature epistemological worldviews will be more willing (and possibly motivated) to consistently review and consider their current operational worldview.  Likewise, those with immature beliefs may find themselves building up filters and blocks to worldview integration.


Copyright 2005 Mark E. Henze -