Abstract
Research has demonstrated the importance of beliefs about knowledge and knowing, or epistemic and ontological cognition (EOC), to learning and achievement; however, little research has examined occupational therapy students’ EOC or determined whether occupational therapy programs promote its development. This study examined changes in EOC over 18 mo of didactic coursework in an occupational therapy program. Thirty-one students completed the Epistemic Beliefs Inventory at the beginning, middle, and end of 18 mo of didactic coursework. Results indicated no difference in ontological cognition. However, change occurred in epistemic cognition, with students demonstrating statistically significantly weaker beliefs in justification of knowledge by an omniscient authority by the end of the didactic portion of the program. Although causal inferences cannot be made from this small study of one cohort of occupational therapy students, it is possible that intensive study in the discipline of occupational therapy contributed to change in these students’ epistemic cognition.
Epistemic and ontological cognition (EOC) was the term used by Greene, Azevedo, and Torney-Purta (2008) to refer to beliefs about knowledge and knowing. These beliefs involve the nature and characteristics of knowledge, what “counts” as knowledge, and where knowledge originates (e.g., from authority figures, personal experience, or reason and logic). Research evidence has shown that EOC can influence learning and achievement (e.g., Bendixen & Hartley, 2003; Bråten & Strømsø, 2005; Dutton, 2003; Hofer, 2004; King & Kitchener, 2004), including learning the contextual application of the ways of knowing used by occupational therapy practitioners to solve complex occupational performance problems (Dutton, 2003). Mitchell (2013b) described how EOC may also influence occupational therapy–specific critical thinking and evidence-based practice. For example, people who believe that knowledge is certain and simple may tend to reason with a focus on factual aspects of knowledge such as the client’s diagnosis and procedures used in intervention rather than considering contextual variables and the meaning of the disability to the client. The belief that knowledge is static and unchanging may also pose a challenge to the application of new research findings that contradict traditional methods.
Despite the importance of understanding EOC, it has received little attention in the occupational therapy literature. This longitudinal study sought to examine EOC during the didactic portion of an occupational therapy program. The purpose of this study is to provide evidence to help determine whether occupational therapy students’ EOC changes in response to the typical methods used over the course of an occupational therapy curriculum. If occupational therapy curricula do not effectively facilitate the development of the EOC needed for successful practice, it will be incumbent on occupational therapy educators to develop and test new methods and techniques to promote more sophisticated EOC.
Theoretical Background
Concepts related to beliefs about knowledge and knowing have been discussed using various labels, including epistemological beliefs (Schommer-Aikins, 2002), personal epistemology (Hofer & Pintrich, 1997), epistemological reflection (King & Kitchener, 2004), and EOC (Greene et al., 2008). Educational psychologists have developed various theories and models to further the understanding of people’s beliefs about what constitutes knowledge, the characteristics of knowledge, and the sources of knowledge. They have described developmental progressions of beliefs about knowledge and knowing (e.g., Baxter Magolda, 2002; King & Kitchener, 2004; Perry, 1970) as well as different dimensions of beliefs (e.g., Hofer & Pintrich, 1997; Schommer-Aikins, 2002). The reader is referred to Mitchell (2013a) for an overview of many of these theories and models.
This study was based on Greene et al.’s (2008) Model of EOC Development (EOCD model), which describes a progression of beliefs about knowledge and knowing, based on changes in different dimensions of beliefs. In their model, Greene et al. (2008; Greene, Torney-Purta, & Azevedo, 2010) referred to the nature and characteristics of knowledge as ontological cognition. Within ontological cognition, they focused on the dimensions of certainty (i.e., knowledge is definite and fixed vs. fluid and changeable) and simplicity of knowledge (i.e., knowledge is straightforward and discrete vs. complex and integrated). Epistemic cognition, in their view, relates to the nature of knowing. It includes justification of knowledge by authority (i.e., accepting the word of teachers, experts, or other authority figures) and personal justification of knowledge (i.e., relying on personal experience as the source of knowledge).
Greene et al. (2008, 2010) used these dimensions of EOC to classify a person’s beliefs according to one of four developmental positions: realist, dogmatist, skeptic, or rationalist. The most naïve beliefs are held by the realist, who strongly believes that knowledge is simple and certain and may use either authority figures or personal experience for justification of knowledge. In Greene and colleagues’ model, people progress from realist beliefs only by relinquishing their strong adherence to simple and certain knowledge beliefs. Occupational therapy practitoners must recognize the complex and changeable nature of knowledge to incorporate changes in practice in response to research findings. Using research evidence to shift from more traditional to contemporary task-oriented approaches to treatment of neurological impairments is one example.
Once the person begins to accept the complex and changeable nature of knowledge, he or she progresses to either a dogmatist or a skeptic position (Greene et al., 2008, 2010). Although Greene and colleagues (2008, 2010) described these positions as developmentally equivalent, and both demonstrate weak beliefs in certain and simple knowledge, they differ in their means of justifying knowledge (i.e., epistemic cognition). Dogmatists use authority figures to justify knowledge, whereas skeptics use personal experience. Personal justification leads the skeptic to believe that knowledge is specific to the individual. Students in the skeptic position may challenge instructors, textbooks, and research evidence unless their personal experiences are consistent with the presented information. Students in the dogmatist position may press instructors to provide the one correct answer to a problem and experience frustration when instructors respond to questions with “It depends.” People with dogmatist or skeptic stances toward knowledge may also struggle with evidence-based practice. On one hand, the dogmatist may seek one definitive, authoritative source on which to base practice decisions and be frustrated and confused by equivocal findings from various research studies. The skeptic, on the other hand, may be reluctant to incorporate research evidence that runs counter to his or her clinical experience.
Like the dogmatists and skeptics, rationalists have weak beliefs in simple and certain knowledge; however, their view of knowledge is more contextual. They may justify knowledge through authority, personal experience, or a combination of the two sources, depending on the context (Greene et al., 2008, 2010). Rationalists consider multiple sources of evidence, evaluate its quality, and use logic and reason to reach a conclusion. Skilled therapists demonstrate these qualities when devising individualized treatment plans using evidence from research, past experience, and continuing education courses and tailoring the approach to the client’s specific desires, needs, and contexts.
Importance of Epistemic and Ontological Cognition
Educational psychologists have demonstrated how EOC can influence metacognitive skills, learning, and academic performance (e.g., Bendixen & Hartley, 2003; Bråten & Strømsø, 2005; Dutton, 2003; Hofer, 2004; King & Kitchener, 2004). The widespread influence of EOC could also affect an occupational therapy student’s approach to practice. Naïve EOC such as that of the realist, dogmatist, or skeptic could contribute to difficulty evaluating the available evidence, considering contextual factors, and determining when to accept the recommendations of authority figures and when to apply knowledge gained from experience. Thus, people with naïve EOC could have difficulty learning to solve the complex problems addressed by occupational therapy practitioners and could struggle in the occupational therapy program. Occupational therapy educators need to gain a better understanding of the development of EOC and how it affects the achievement of occupational therapy students to provide more effective educational experiences and prepare competent practitioners.
Educational experiences have been identified as facilitators of EOCD (Greene et al., 2008, 2010); therefore, participation in an occupational therapy program could be expected to foster EOCD. Two studies have provided cross-sectional data to investigate differences in novice and more experienced occupational therapy students’ EOC. Dutton (2003) found differences between the sophistication of 1st-yr and 2nd-yr students’ EOC. She found that 1st-yr students were more likely to perceive cases as protocols that could be duplicated and used in similar situations. Second-year students, however, were more likely to view case-based learning as a way to learn to apply and synthesize knowledge and think critically—skills they would need to apply flexibly in future practice situations.
Similarly, Mitchell (2014) described differences in entering and postdidactic occupational therapy students’ EOC. Using the Epistemic Beliefs Inventory (EBI; Schraw, Bendixen, & Dunkle, 2002), a modified version of Schraw and Olafson’s (2008) Four-Quadrant Scale, and open-ended written questions, she found no differences in the two groups’ general knowledge beliefs. However, there was evidence of more sophisticated occupational therapy–specific knowledge beliefs in postdidactic than entering students. Entering students tended to take more dogmatist and skeptic stances, with minimal evidence of a rationalist position. Evidence from the postdidactic students suggested primarily skeptic stances, although rationalist views of occupational therapy–specific knowledge seemed to be emerging.
Several researchers have called for longitudinal studies to assess the processes that influence the development of EOC (Bråten & Strømsø, 2005; Greene et al., 2010; Hofer & Pintrich, 1997). As Hofer (2000) argued, there is a need to study EOC as students develop disciplinary expertise. A limited number of longitudinal studies have been conducted to date, and no longitudinal studies have been published in the discipline of occupational therapy. Longitudinal research is needed to examine whether occupational therapy students’ EOC changes over the course of an occupational therapy program. If change does occur, future research could seek to identify the specific methods that facilitate this change and how to further capitalize on them. If change does not occur, future research will be needed to develop and test new methods for delivering occupational therapy education and facilitating the EOC needed for effective occupational therapy practice. The specific research question examined in this study was, “How does occupational therapy students’ EOC change over the course of the didactic portion of an occupational therapy program?”
Method
Research Design
This study used a longitudinal repeated-measures design. The university’s institutional review board granted approval for the study.
Participants and Setting
Participants were entry-level MOT students enrolled on a health science center campus in the midsouth region of the United States. Volunteers were recruited from among all occupational therapy students in the program. Thirty-five students completed the first EBI. Two of these students chose not to volunteer for the second testing, and 2 additional students chose not to volunteer for the third testing, resulting in a total of 31 participants who completed the study.
All students had completed at least 90 credit hr of prerequisite coursework. Because an undergraduate degree is optional for entry into the program, students may or may not have earned a bachelor’s degree. During the study, the students completed 66 credit hr of basic science and occupational therapy coursework, including three 2-wk Level 1 fieldwork experiences. Students were not involved in Level 2 fieldwork during the study because Level 2 fieldwork occurs after the didactic portion of the program.
Instruments and Procedures
Schraw et al.’s (2002) EBI is a 32-item instrument representing the dimensions of certain knowledge, quick learning, simple knowledge, omniscient authority, and fixed ability. Participants use a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) to rate the strength of their beliefs on items such as “Too many theories just complicate things” (Simple Knowledge factor) and “When someone in authority tells me what to do, I usually do it” (Omniscient Authority [OA] factor). In this study, the Simple and Certain Knowledge (SCK) factors were combined and used as dependent variables, in addition to the OA factor. EBI internal consistency reliabilities range from .50 to .65. Test–retest reliability ranges from .62 to .81. Studies have shown that the EBI explains around 40% of sample variance. Construct validity for the five factors of the EBI has been demonstrated, and research has shown modest but significant predictive validity for reading comprehension (Schraw et al., 2002).
The EBI was administered to all students in a classroom setting during the 1st wk of the occupational therapy program, via Blackboard Academic Suite™ midway through the didactic coursework (i.e., after 9 mo in the program), and in a classroom setting within 2 wk of the end of the didactic coursework (i.e., after 18 mo in the program, before Level 2 fieldwork). There were no time limits, and students were assured both verbally and in the written instructions that there were no right or wrong answers.
Data Analysis
The data were first checked for accuracy using frequencies. Internal consistency reliability (Cronbach’s α) of the scales, means, standard deviations, and effect sizes (η p 2) were calculated. Repeated-measures multivariate analysis of variance was used to examine changes in mean scores on the SCK and OA factors from the EBI at the beginning, middle, and end of the didactic portion of the program.
Results
Demographics
The 31 participants were 2 men and 29 women with an average age of 22.7 yr (range = 21–36) at the beginning of the study and 24.0 yr (range = 22–38) at the end of the study. Eighty-eight percent were White, 9% were African-American, and 3% were Asian-American.
Internal Consistency Reliability of the Epistemic Beliefs Inventory Factor Scales
Because of low initial internal consistency reliability, Items 11, 18, and 22 (item-total correlations of −.09, .025, and .002, respectively) from the SCK scale and Item 7 (item-total correlation of .129) from the OA scale were eliminated. Therefore, we used a 12-item SCK scale and a 4-item OA scale; Cronbach’s α for each scale was .623. These amended scales were used in the subsequent analyses.
Repeated-Measures Multivariate Analysis of Variance
Time in the program (0 mo, 9 mo, and 18 mo) was the repeated variable, and EBI factor scores (SCK and OA) were the dependent variables. Thirty-one students completed the EBI at each time point. Preliminary analyses of all dependent variables using skewness and kurtosis statistics indicated no serious problems; none departed significantly from normal distribution curves. No extreme outliers were detected for scores on either of the variables. The assumption of sphericity was met for both variables: SCK, W = 0.968, χ2(2) = 0.931, p = .628; OA, W = 0.983, χ2 = 0.499, p = .779.
The multivariate test was statistically significant, Wilks’s Λ = .775, F(4, 118) = 4.006, p = .004, η p 2 = .120, indicating that change was not the same across time for the set of dependent variables. The univariate tests showed that the OA variable contributed to the multivariate significance, F(2, 60) = 5.671, p = .006, η p 2 = .159, but the SCK variable did not, F(2, 60) = 2.572, p = .085. Tukey’s post hoc tests indicated statistically significant differences between OA scores at entry into the program and OA scores after completion of the didactic portion of the program, but no statistically significant differences between OA scores at entry and midway through the didactic portion of the program or between mid-didactic and postdidactic scores (Figure 1).
Omniscient Authority (OA) and Simple and Certain Knowledge (SCK) factor scale mean scores at three time points in the program.
Dependent t tests showed statistically significant differences between mean OA and SCK scores at each point in the didactic portion of the program, indicating more sophisticated ontological cognition than epistemic cognition at all three points in time (Table 1). For this group of students, there was a statistically significant moderate positive correlation between the mean OA and SCK scores, but only at the end of the didactic portion of the program (Table 2).
Dependent t-Test Comparisons of EBI OA and SCK Factor Mean Scores at Each Time Point (N = 31)
Note. EBI = Epistemic Beliefs Inventory; OA = Omniscient Authority; SCK = Simple and Certain Knowledge.
Correlations Between the EBI OA and SCK Factor Mean Scores at Each Time Point (N = 31)
Note. EBI = Epistemic Beliefs Inventory; OA = Omniscient Authority; SCK = Simple and Certain Knowledge.
Discussion
Many researchers have studied changes in beliefs about knowledge and knowing (e.g., Baxter Magolda, 2002; Cano, 2005; Kienhues, Bromme, & Stahl, 2008; Valanides & Angeli, 2005). These authors have theorized that exposure to disparate beliefs and tentative information during educational experiences can trigger the development of more sophisticated EOC. Consistent with this idea, the occupational therapy students in this longitudinal study demonstrated more sophisticated general epistemic cognition after 18 mo in the program than they did at entry into the program. These findings are not surprising, because occupational therapy programs use techniques and approaches such as reflection, authentic fieldwork experiences, and case-based methods incorporating analysis of problems with more than one potential solution (American Occupational Therapy Association [AOTA], 2007; Dutton, 2003; Falk-Kessler & Ciaravino, 2006; Hooper, 2006; Kramer et al., 2007), all of which have been recommended as methods to facilitate the development of EOC (Cano, 2005; King & Kitchener, 2004; Moore, 2002; Schommer-Aikins, 2002). It does, however, run counter to the cross-sectional study by Mitchell (2014), suggesting that intense study in a specific discipline may facilitate the development of general EOC.
According to developmental theorists (e.g., Baxter Magolda, 2002; King & Kitchener, 2004; Perry, 1970), EOC progresses from a view of knowledge as objective and derived from an external source to a view of knowledge as more subjective and originating from an internal source. The results of the current study indicate that a shift away from reliance on an external source occurred for these students during the 18 mo of the occupational therapy program. However, the EOCD model (Greene et al., 2008, 2010) describes two positions with equally sophisticated ontological cognition: the dogmatist, who relies on justification by authority, and the skeptic, who relies on personal justification of knowledge. For occupational therapy students, the progression away from reliance on authority figures as a source of definitive information may facilitate evidence-based practice; however, relying on clinical experience for making practice decisions, as in the skeptic position, could hinder evidence-based practice. It seems clear that effective evidence-based practice requires a rationalist position, an approach that relies on reason and logic for making decisions, considering sources of knowledge such as research, clinical experience, the client, and the context.
Although this group of occupational therapy students demonstrated a decrease in the strength of their belief in an omniscient authority as the source of knowledge, the EBI does not include a personal justification factor that could assist in determining whether the students transitioned from a more dogmatist to a more skeptic position or whether their epistemic cognition progressed to that of a rationalist. The fact that their scores remained slightly above the median score of 3 on the OA factor of the EBI suggests that they had not discarded the belief in authority figures as a source of knowledge, but rather that this belief had been tempered. It seems possible that by the end of the didactic portion of the program, the students had begun to rely on authority figures as one of many sources of knowledge, as in the rationalist position and consistent with occupational therapy epistemic beliefs, but that conclusion cannot be confirmed by these data. It is evident, however, that similar to previous research (e.g., Baxter Magolda, 2002; Cano, 2005; Kienhues et al., 2008; Valanides & Angeli, 2005), rigorous study in the discipline of occupational therapy was associated with development of epistemic cognition.
Despite changes in epistemic cognition, there was no difference in the students’ ontological cognition as measured by the SCK factor of the EBI. This seems consistent with the EOCD model (Greene et al., 2008, 2010), in that only the most naïve position, realist, demonstrates strong beliefs in SCK. It seems likely that students who enter the occupational therapy program after completing at least 90 credit hr of prerequisite coursework would have progressed beyond beliefs in SCK. In fact, these students’ scores on the SCK factor of the EBI were below the median of 3 on the 5-point Likert scale, suggesting weak beliefs in SCK and development beyond the realist position. Their scores on the SCK scale were also statistically significantly lower than their scores on the OA scale, indicating significantly weaker beliefs in SCK than in justification by authority at all three points in time. Weak beliefs in SCK do not differentiate among the dogmatist, skeptic, and rationalist, however, because more sophisticated ontological cognition is characteristic of all three of these positions in the EOCD model.
Greene et al.’s (2008, 2010) model predicts that naïve ontological cognition will be incompatible with sophisticated epistemic cognition, and Schraw and Olafson’s (2008) study of practicing teachers supported this view. In the current study, the correlation between entering students’ OA and SCK scores approached significance; however, only at the end of the didactic portion of the program was there a statistically significant correlation between the two. This change over time in the relationship between epistemic cognition and ontological cognition may reflect a developmental process during which students are experiencing the dissonance that is believed to be necessary for change to occur in EOC (Baxter Magolda, 2002; Hofer & Pintrich, 1997). That is, in the occupational therapy program, students may encounter situations that challenge naïve EOC and lead to more sophisticated stances. The lack of significant mid-didactic correlation between OA and SCK scores could reflect a period of reorganization or development of EOC. By the end of the didactic program, both epistemic and ontological cognition may have matured and thus become more related. The fact that the participants in the current study were students rather than practicing clinicians may explain the inconsistency between Schraw and Olfason’s findings and those of the current study and shed light on the developmental process involved in the maturation of EOC.
The correlation between the postdidactic OA and SCK scores was moderate and positive. A moderate positive correlation would indicate that the direction of fluctuation in the strength of beliefs in justification by authority would tend to correspond with the direction of fluctuation in the strength of beliefs in SCK. Because the dogmatist holds weak beliefs in SCK and strong beliefs in justification by authority (Greene et al., 2008, 2010), the weakening of OA scores observed in the current study, in conjunction with weak SCK scores, suggests movement away from a dogmatist position. Again, without a measure of personal justification, it is difficult to determine whether the students were moving toward a skeptic or a rationalist position. According to Greene and colleagues’ (2008, 2010) model, the dogmatist and skeptic are at the same level of sophistication, and development beyond these positions would result in rationalist EOC; however, Mitchell (2014) found evidence of both skeptic and rationalist stances in postdidactic occupational therapy students. Further research that includes a measure of personal justification is needed to clarify whether, contrary to the EOCD model, students transition from dogmatism to skepticism before reaching the rationalist position.
In their study of middle through graduate school students, Greene et al. (2010) found four profiles of latent class factor means that did not match those proposed in the EOCD model. On the basis of these results, they suggested that the dogmatist may be more accurately described as holding strong beliefs in SCK, along with strong beliefs in justification by authority and weak beliefs in personal justification. They conceded that their conceptual model may need to be modified to include additional profiles. It is possible that the profiles discovered in the current study are not accurately represented in the current version of the EOCD model. Further research is needed to clarify this possible difference, however.
Implications for Occupational Therapy Education and Practice
The results of this study have the following implications for occupational therapy education and practice:
This study suggests that changes occur in occupational therapy students’ epistemic cognition over the course of the didactic portion of an occupational therapy program. Although causal inferences cannot be made on the basis of this study, it seems reasonable to conclude that the occupational therapy program contributed to this change.
This study did not consider the relationship between academic achievement and EOC; however, it seems logical that courses incorporating case-based problems with more than one potential solution and requiring decisions based on a synthesis of evidence could be challenging for students with a more naïve EOC position.
Exploration of students’ beliefs about knowledge and knowing could prove fruitful when attempting to remediate learning difficulties.
A better understanding of students’ EOC may assist educators in developing more effective learning experiences, improving learning outcomes, and preparing skilled evidence-based practitioners.
Limitations
The EBI does not include a Personal Justification factor; therefore, it did not allow for direct measurement of this aspect of epistemic cognition; however, no instrument based on the EOCD model and including a Personal Justification factor is currently available. The results of this study should also be considered in light of the fact that most of the participants were women. Although there may be differences between the EOC of men and women (Baxter Magolda, 2002), the majority of occupational therapy practitioners are women. Response bias may also have affected the results.
This study did not include a control group; therefore, causal inferences cannot be made from these data. There are many potential explanations for the changes that occurred over the 18-mo period of the study. Nevertheless, researchers have suggested that intense study in a discipline (e.g., learning experiences in the occupational therapy program) may provide the environment needed to facilitate change in EOC (Baxter Magolda, 2002; Cano, 2005; Kienhues et al., 2008; Valanides & Angeli, 2005). Use of a small convenience sample also precludes generalization of this study’s results to other disciplines or to other occupational therapy students, and replication using larger samples of occupational therapy students in various locations across the country is recommended.
Recommendations for Future Research
Additional longitudinal studies of occupational therapy–specific EOC are needed. Although this study found changes in general EOC during the didactic portion of the occupational therapy program, greater change in occupational therapy–specific EOC might be expected. Longitudinal studies that include measures of both domain-specific and domain-general EOC could provide additional information about the relationships between the two and inform theory related to EOC.
Qualitative studies may provide deeper understanding of changes in occupational therapy students’ EOC over time. Future research that includes an objective measure of personal justification could also help clarify occupational therapy students’ epistemic cognition and how it changes during an occupational therapy program. Replication of this study with larger samples is also recommended.
Future research could also examine the relationships among grades, fieldwork scores, and EOC. Understanding the contribution of fieldwork experiences to the development of EOC could help educators determine the most optimal timing and the most appropriate types of fieldwork experiences to incorporate at different points in the curriculum. Further research to examine changes in EOC during the fieldwork portion of the program might help direct occupational therapy programs and accrediting bodies as they seek to determine the appropriate balance between didactic coursework and fieldwork experiences.
Currently, little is known about the effectiveness of occupational therapy programs in facilitating the development of EOC, although methods commonly used by occupational therapy educators (AOTA, 2007; Dutton, 2003; Falk-Kessler & Ciaravino, 2006; Hooper, 2006; Kramer et al., 2007) are consistent with techniques that have been recommended to promote growth in EOC (Cano, 2005; King & Kitchener, 2004; Moore, 2002; Schommer-Aikins, 2002). Nonetheless, research is needed to assist occupational therapy educators in identifying, developing, and testing methods for promoting sophisticated EOC within the domain of occupational therapy. The effectiveness of various methods could depend on student characteristics, the timing of implementation of such approaches, or both. It seems possible that entering students might respond to different methods or techniques than students at the midpoint of didactic instruction. Additional research could also help clarify parameters for the effective use of particular methods and improve occupational therapy programs’ ability to prepare competent practitioners.
Conclusion
Understanding students’ EOC could facilitate effective instruction, assist educators in understanding and interpreting a student’s learning difficulties, and help instructors identify effective remedial approaches. This study suggests that an occupational therapy program may contribute to the development of students’ general EOC, but there is still much to be learned about the influence of EOC on occupational therapy students’ achievement, learning, and transformation into competent practitioners. As students come to understand the complexity and contextual application of knowledge and consider a variety of credible sources of knowledge, they may be more likely to make reasoned and effective practice decisions.
Footnotes
Acknowledgments
I thank the students who participated in this study and Denise Winsor for her assistance.