Page 1 of 19

87

Advances in Social Sciences Research Journal – Vol.7, No.6

Publication Date: June 25, 2020

DOI:10.14738/assrj.76.7882.

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change.

Advances in Social Sciences Research Journal, 7(6) 87-105.

Epistemic Beliefs Moderate Mediations Among Attitudes,

Prior Misconceptions, and Conceptual Change

James A. Vivian

Department of Educational & Counselling Psychology,

Faculty of Education, McGill University, Montreal, Canada.

Krista R. Muis

Department of Educational & Counselling Psychology,

Faculty of Education, McGill University, Montreal, Canada.

ABSTRACT

We investigated the mediating and moderating roles of attitudes and

epistemic beliefs in conceptual change during learning about genetically

modified foods (GMFs). One hundred twenty undergraduate students

participated. To measure misconceptions about GMFs, students first

completed a prior knowledge test. Students then completed self-report

inventories to measure their attitudes and topic-specific epistemic

beliefs regarding GMFs. Students were then randomly assigned to read

a refutation or expository text about GMFs. Following reading, students

completed a test to assess conceptual change. Results of a repeated

measures ANOVA revealed participants who read a refutation text

changed more misconceptions at post-test than participants who read

an expository text. A moderation mediation analysis revealed attitudes

toward GMFs significantly mediated the relationship between prior

misconceptions and conceptual change, and that this relationship was

moderated by learners’ beliefs regarding the source and justification of

GMFs knowledge. Theoretical and educational implications are

discussed.

Keywords: attitudes; epistemic beliefs; conceptual change; genetically

modified foods

INTRODUCTION

The public’s increasing mistrust in science has become a particularly challenging problem for

learning and science education. An example is the recent increase in measles outbreaks in America

and Europe (World Health Organization, 2018) due to misconceptions that the measles, mumps,

and rubella vaccine (MMR) causes autism (Kata, 2012). The vaccine-autism misconception can be

traced back to a fraudulent study by Wakefield et al. (1998), published in the Lancet medical journal.

wherein they asserted that a temporal association exists between administration of the MMR

vaccine and the onset of autism. Despite its retraction and substantial research evidence to the

contrary, many individuals continue to hold misconceptions about vaccines, and consequently, have

been increasingly refusing to get vaccinated.

Page 2 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 88

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

Misconceptions represent inaccurate knowledge, understanding, or faulty reasoning that departs

from scientific knowledge or understanding of a topic (Heddy, Danielson, Sinatra, & Graham, 2017).

Such errors in reasoning are typically acquired during previous learning and can negatively predict

new learning (Krause, Kelly, Corkins, Tasooji, & Purzer, 2009). Misconceptions can be

counterproductive for learning when individuals are unable to recognize and revise their

misconceptions and discriminate between information based on evidence and information derived

from opinion, including knowledge that has been contorted to meet social, political, and/or

economic gains (Ecker, Hogan, & Lewandowsky, 2017; Sinatra & Seyranian, 2016).

Many individuals hold misconceptions about a variety of complex socioscientific topics, including

vaccines, climate change, stem cell research, and genetically modified foods (GMFs). Indeed, due to

a lack of accurate knowledge (i.e., misconceptions), many individuals believe GMFs are unsafe for

human consumption and, as a result, hold negative attitudes towards them (Heddy et al., 2017). In

addition to negative attitudes, individuals’ epistemic beliefs (i.e., beliefs about the nature of

knowledge and knowing) may lead them to reject scientifically accurate information about GMFs if

they believe their own knowledge or understanding to be just as valid as those of experts. Taken

together, attitudes and epistemic beliefs may play a role in terms of how individuals process,

interpret, and evaluate various sources of knowledge related to a variety of complex socio-scientific

topics of global significance.

Attitudes refer to positive or negative evaluations of objects, people, or events that predispose

individuals to respond to attitudinal objects in preferential ways (Eagly & Chaiken, 1993; Fishbein

& Icek, 1975). Epistemic beliefs, on the other hand, refer to personal theories (or individual doxastic

assumptions) related to the nature of knowledge (structure and certainty) and the nature of

knowing (i.e., source and justification; Hofer & Pintrich, 1997). Attitudes predict the types of

information individuals are likely to select, perceive, process, and encode (Maio & Haddock, 2010),

whereas epistemic beliefs play a role in how individuals interpret and evaluate knowledge claims.

In other words, beliefs are the building blocks of attitudes (Dole & Sinatra, 1998). Indeed, an

individual’s attitude toward an object includes an evaluative component (“I don’t like GMFs...) that

is tied to a cognitive component that may be an unjustified belief (i.e., a misconception, “because

they are not safe to eat”) or a justified true belief (i.e., a correct conception [Sinatra & Syranian,

2016]). Changing individuals’ misconceptions about a topic requires shifting an unjustified belief to

a justified true belief where justification is the central mechanism by which this occurs (Authors,

2012).

However, individuals tend to seek out information that is congruent with their attitudes while

ignoring attitude-incongruent information (Maio & Haddock, 2010). Similarly, individuals are likely

to interpret and evaluate discrepant knowledge claims within the context of their own personal

systems of epistemic beliefs, and subsequently, to reject information that does not conform to their

own understanding or topical knowledge. As such, systems of beliefs (and emergent attitudes) can

reflect judgements of fact or processes of evaluation that may be derivative of conjecture (Rockeach,

1968). Given this link, individuals’ epistemic beliefs may interact with their attitudes to predict

whether conceptual change occurs. That is, misconceptions may be less amenable to change when

tied to individuals’ negative attitudes about that topic (Sinatra & Seyranian, 2016). To date,

researchers have not explored the possibility that attitudes and epistemic beliefs may interact to

predict conceptual change. Additionally, few studies have investigated whether individual learner

Page 3 of 19

89

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

characteristics such as attitudes and epistemic beliefs interact with the quality of a message (i.e.,

message characteristics of an assigned text) to predict conceptual change. Therefore, we included

two types of texts as a treatment measure in our study: a refutation and an expository text. The

refutation text was used as an instructional approach to support conceptual change by directly

confronting a misconception using causal explanations to refute incorrect knowledge. The

expository text served as a control and included the same content as the refutation text, but without

the refutation content. Decades of research in the conceptual change literature has shown refutation

texts to be an effective strategy for supporting conceptual change (see Tippet, 2010 for a full

review). As such, we examined whether conceptual change would vary as a function of learners’

attitudes and epistemic beliefs (i.e., learner characteristics) and type of instructional text (i.e.,

message characteristics).

In a post-truth era when public distrust in science has led to the proliferation of misinformation,

fake news, and erroneous knowledge, expounding a working model of the roles of attitudes and

epistemic beliefs in learning and conceptual change is imperative to understanding how these

factors can be leveraged to enhance learning outcomes for individuals who hold inaccurate

conceptions, negative attitudes, and limited epistemic strategies for interpreting and evaluating

knowledge and knowing related to complex scientific issues (see WHO, 2018 for example).

Therefore, the present study examined the direct, indirect, and conditional indirect effects of prior

misconceptions on conceptual change via learners’ attitudes and epistemic beliefs when learning

about GMFs from refutation and expository texts. Prior to delineating our research questions and

hypotheses, we present relevant theoretical and empirical work.

THEORETICAL FRAMEWORKS

Conceptual Change

Conceptual change involves revising misconceptions and updating inaccurate knowledge to reflect

more accurate representations of knowledge regarding a topic (Kendeou & O'Brien, 2014).

Typically, conceptual change is provoked when a state of cognitive conflict arises between

misconceived prior knowledge and the learning of new, discrepant information (Chan, Burtis, &

Bereiter, 1997). When incorrect prior knowledge comes into conflict with newly acquired, accurate

conceptions of a topic, the ensuing incongruity leads to attempts to reconstruct inaccurate

knowledge to reflect more accurate configurations within a conceptual network (Chan et al., 1997;

Chi, 2008; Lombardi, Nussbaum, & Sinatra, 2016).

Regarding science learning, the notion of conceptual change implies that individuals have pre- existing misconceptions—inaccurate mental representations of a topic—that contradict scientific

understanding of that topic (Lombardi et al., 2016). Science misconceptions are often quite

enduring, resistant to change, and can have particularly deleterious effects on learning and decision- making behaviors (Ecker, Hogan & Lewandowsky, 2017; Vosniadou, 1994). To change science

misconceptions, conceptual change is necessary.

Refutation texts are one method used by researchers to investigate and support conceptual change.

Refutation texts help learners reconcile cognitive conflict regarding a topic by directing

metacognitive awareness away from inaccurate knowledge and toward more accurate conceptions

of a topic via the structured presentation of causal explanations to refute incorrect knowledge.

Although effective, research examining the processes involved in conceptual change has shown that

Page 4 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 90

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

successful revision of misconceptions is not merely a product of refuting inaccurate knowledge (i.e.,

simply stating knowledge to be incorrect) (Ecker et al., 2017; Kendeou & O’Brien, 2014). Rather,

successful conceptual change depends in large part on an interaction between individual learner

characteristics (i.e., prior knowledge, attitudes, beliefs, motivation) and message characteristics

(i.e., comprehensibility, coherence, plausibility, and rhetorical structure of a message) (Dole &

Sinatra, 1998). The dynamic and interdependent relations among learner and message

characteristics in conceptual change are delineated in Dole and Sinatra’s (1998) Cognitive

Reconstruction of Knowledge Model (CRKM).

Cognitive Reconstruction of Knowledge Model (CRKM)

According to Dole and Sinatra (1998), individuals differ in the quantity and quality of their prior

knowledge (i.e., misconceptions), and this prior knowledge can interfere with learning and the

interpretation and evaluation of new information (see also Krause et al., 2009; Sinatra et al., 2008).

In their Cognitive Reconstruction of Knowledge Model (CRKM), Dole and Sinatra (1998) described

how interactions between learner characteristics (i.e., existing knowledge, attitudes, beliefs, and

motivation) and message characteristics (i.e., comprehensibility, plausibility, coherence, and

rhetorical structure) determine the likelihood individuals will cognitively reconstruct or revise

previously acquired, inaccurate knowledge.

Individual learner characteristics, such as the strength and coherence of previously acquired

information (i.e., the richness and explanatory power of a conception), as well as an individual’s

commitment to their existing knowledge (i.e., an individual’s attitudes and beliefs regarding the

value of a conception) determine the likelihood individuals will change pre-existing

(mis)conceptions. The stronger and more conceptually coherent an individual’s prior conceptions,

as well as the degree of commitment an individual has toward previously acquired conceptions, the

less likely conceptual change is to occur (Dole & Sinatra, 1998).

Regarding message characteristics, several factors are also likely to affect whether an individual will

revise their existing knowledge, including (1) the comprehensibility and plausibility of a message;

that is, whether the message is conceptually palpable and individuals view the message as credible

(source evaluations), (2) the coherence of the message and whether the message has explanatory

power and effectively links back to larger conceptual structures (elaborative complexity), and (3)

whether the message is rhetorically compelling; that is, whether the message’s structure, sources

of information, and justification of arguments are sufficiently persuasive. In short, a message that is

elusive, ambiguous, incoherent, or disconnected will not likely facilitate conceptual change.

Conceptual change is an iterative process involving dynamic interactions between both learner

characteristics and message characteristics. For example, a comprehensible and plausible message

may be personally relevant for one individual but not another. Additionally, an individual may have

a strong, coherent prior conception to which he or she is strongly committed despite a plausible and

rhetorically compelling counter message. Further, an individual may be dissatisfied with a

previously acquired conception but not find a new message sufficiently plausible or coherent to

warrant replacement of the existing conception. Finally, a message may be considered rhetorically

compelling, but perceived as implausible. As previously noted, one method researchers have found

to be effective in facilitating conceptual change and promoting deeper engagement while processing

a message is via the use of refutation texts, which are described next.

Page 5 of 19

91

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

Refutation Texts

Refutation texts are rhetorical devices designed to target misconceptions and facilitate conceptual

change (Kendeou et al., 2014). Numerous studies have shown use of refutation texts to be an

effective method for changing misconceptions related to a variety controversial socio-scientific

topics including GMFs (Heddy et al., 2017; Heddy & Sinatra, 2013; Trevors et al., 2017a; Authors,

2016). Unlike expository texts that present information in a descriptive manner, refutation texts

integrate elements of argumentation to draw metacognitive awareness towards misconceptions or

inaccurate beliefs (Tippet, 2010). According to Kendeou et al. (2013, 2014), a refutation text must

include the following three components to be effective. First, it must directly confront a

misconception (i.e., draw metacognitive awareness toward inaccurate knowledge). Second, it must

explicitly state the falseness of the misconception (i.e., create disequilibrium). Third, it must provide

causal explanations based on empirical evidence to refute the misconception (i.e., to make the

message more plausible; Kendeou et al., 2013, 2014). Causal explanations in refutations represent

to-be-learned information and are intended to create a rich tapestry of information to help reduce

cognitive conflict and the activation of previously acquired, incorrect information to aid in the

reconfiguration of inaccurate knowledge structures within a conceptual network (Kendeou et al.,

2013, 2014).

Although refutation texts are designed to be comprehensible, coherent, plausible, and rhetorically

compelling, an individual’s motivation to revise previously acquired misconceptions significantly

determines their relative depth of engagement with a message (Dole & Sinatra, 1998). At the highest

level of engagement, individuals are more likely to critically evaluate the content of a message to

assess the validity of the knowledge claims as well as the quality of its source. At the lowest level of

engagement, individuals are likely to process only that information that is congruent with their

prevailing attitudes or underlying beliefs (Dole & Sinatra, 1998). In other words, learner

characteristics such as attitudes and epistemic beliefs also predict the probability that learners will

conceptually change and revise previously acquired misconceptions.

Attitudes

Attitudes refer to positive or negative evaluations of people, events, or ideas that lead individuals

to respond with some degree of favor of disfavor (Eagly & Chaiken, 1993; Fishbein & Icek, 1975;

Heddy et al., 2017). In lieu of the oft cited tripartite model of attitudes, which presupposes attitudes

to be made up of orthogonally distinct cognitive, affective, and conative components, attitudes

herein are construed as distinct entities existing separately from cognitive, affective, and behavioral

elements, and specifically, to serve an evaluative function in the appraisal of information derived

from these distinct factors (Fabrigar, Macdonald, & Wegener, 2005). In this way, attitudes are

described in terms of simple object-evaluation associations that are typically embedded within

larger semantic networks of associated knowledge structures (i.e., beliefs, affective states, and

representative behavioral schemas).

The structural view of attitudes as simple object-evaluation associations can be concisely described

as follows: an attitudinal object (say, GMFs) represents one node within a semantic network, the

evaluation of the object (say, beliefs about GMFs) represents the other node, and the link between

the two nodes represents the relative strength of the association (Fabrigar, et al., 2005). For

example, GMFs (the object of evaluation) could be evaluated based on a set of beliefs that GMFs are

Page 6 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 92

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

unsafe for human consumptions (attributes of the object), and the strength of these beliefs

associated with the topic of GMFs could result in an overall negative attitudinal appraisal of GMFs.

As previously mentioned, attitudes are commonly embedded (or linked) to larger networks of

associated knowledge structures (Eagly & Chaiken, 1993; Fabrigar et al., 2005). For example,

specific attributes associated with the representational nature of an object (e.g., cognitive, affective,

or conative properties) are also associated with local evaluations of these representational features

in addition to the overall (global) appraisal of the attitudinal object. In other words, the structure of

an attitude comprises not only object-evaluation associations, but also interconnected knowledge

structures (such as systems of epistemic beliefs), that vary as a function of the strength and pattern

of associative links between the attitude and related structures (Fabrigar et al., 2005). Thus,

attitudes are shaped by both general (global) evaluations of a focal object, and more specific (local)

appraisals of an object’s attributes (or interconnected knowledge structures) that situate the

attitudinal object within a larger semantic network, as shown in Figure 1. In theory, any changes in

knowledge (or conditions under which an object of knowledge and knowing is evaluated) at either

the global or local level should lead to changes in the overall object-evaluation association, and

subsequently, to changes in the representational nature of knowledge for which an attitude has

been formed.

Attitudes serve an evaluative function in the processing of information related to a topic. For

example, evaluations related to the credibility of a scientific claim (attributes of the object) could be

biased by pre-existing attitudes (and associated beliefs), and thus, individuals may ignore evidence

that contradicts their own understanding or conceptions (Sinatra, Kienhues, & Hofer, 2014).

Consequently, attitudes can interfere with new learning, including individuals’ interpretations of

information (i.e., perceptions of certainty or complexity), evaluations of source credibility, and

judgements regarding the veracity of scientific information (Sinatra et al., 2014). In other words,

attitudes are likely related to an individuals’ epistemic beliefs. Moreover, research has shown that

epistemic beliefs predict individuals’ understanding, learning, achievement, and conceptual change

(Stathopoulou & Vosniadou, 2007).

Epistemic Beliefs

Epistemic beliefs reflect individuals’ beliefs about knowledge and knowing, are relatively stable

over time, and include doxastic assumptions regarding the nature of knowledge and knowing (Hofer

& Pintrich, 1997; Authors, 2007, 2018; Schommer, 1990; Sinatra & Hofer, 2016). Whether explicit

or implicit, epistemic beliefs play a central role in how individuals reason about knowledge and

knowing (Chinn, Buckland, Samarapungavan, 2011; Sinatra et al., 2014), including evaluations and

judgments regarding the veracity of information obtained from multiple sources (Greene, Yu, &

Copeland, 2014), and are predictive of both learning processes and achievement outcomes (Chinn

et al., 2011; Hofer, 2000; Authors, 2004, 2007).

Based on the work of Hofer and Pintrich (1997), epistemic beliefs are defined as personal theories

related to beliefs about the structure of knowledge, certainty of knowledge, source of knowledge,

and justification for knowing. The first two dimensions (structure and certainty) concern beliefs

about the properties of knowledge, and the second two dimensions (source and justification)

concern beliefs regarding processes of knowing. Beliefs regarding the structure of knowledge

concern whether knowledge is believed to be made up of discrete (simple) facts, or whether

knowledge is multifaceted and composed of highly complex and interrelated concepts. For the

Page 7 of 19

93

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

certainty of knowledge, individuals may view knowledge as static and unchanging (steady state

knowledge) or evolving and perpetually changing (entropic knowledge states). For the source of

knowledge, individuals may believe knowledge is generated and disseminated via authority figures

(externally generated), or personally constructed via reason and logic (internally generated).

Finally, beliefs regarding the justification for knowing refer to whether individuals view knowledge

as justified by expert authority, subjective experience, or via multiple sources of information (Hofer

& Pintrich, 1997; Trevors et al., 2017b).

Research has shown that more constructivist epistemic beliefs (i.e., knowledge is complex, highly

interrelated, uncertain, derived and justified via multiple sources of information) are correlated

with various facets of learning and achievement (Franco, et al, 2012; Stathopoulou & Vosniadou,

2007), including self-regulated learning (see; Authors, 2007, 2010, 2018), conceptual change

(Mason & Gava, 2007; Mason, Gava, & Boldrin, 2008), emotions (Authors, 2015, 2016), complex

problem-solving (Authors, 2008), and digital literacy (Greene et al., 2014). In terms of conceptual

change, numerous studies have found that individuals who hold more constructivist epistemic

beliefs change more misconceptions after reading refutation texts than individuals who hold less

constructivist epistemic beliefs (i.e., knowledge is simple, certain, derived and justified via authority

or personal judgement; see Franco et al., 2012; Authors, 2011; Mason & Gava, 2007; Mason et al.,

2008; Murphy & Alexander, 2016; Trevors et al., 2017).

Epistemic beliefs serve as a focal lens for understanding scientific topics and play an integral role in

how individuals interpret and evaluate scientific knowledge, including discrepant knowledge

claims (Sinatra et al. 2014), and predict the ways in which individuals engage in scientific inquiry,

evaluate theories against evidence, interpret explanatory models of theoretical constructs, and

develop understanding of complex science topics in relation to multiple, often discrepant sources

of information (Stathopoulou & Vosniadou, 2007). Given the important role of epistemic beliefs in

learning and understanding of scientific knowledge (Stathopoulou & Vosniadou, 2007), including

the role attitudes play in how individuals select, interpret, process and encode information (Maoi &

Haddock, 2010), more research is needed to examine how, in what ways, and for whom epistemic

beliefs and attitudes work together to predict conceptual change. Such understanding could provide

deeper knowledge of the structural and functional links between individuals’ underlying systems of

epistemic beliefs and the emergent attitudes by which they appraise objects of knowledge and

knowing. The relationship between attitudes and epistemic beliefs in conceptual change is

discussed next.

Theoretical Relationship Between Attitudes and Epistemic Beliefs

Common assumptions in the research on attitudes is that beliefs (i.e., cognitions and related

knowledge structures) and attitudes are functionally consistent with one another. For example, a

belief that GMFs are harmful for human consumption typically elicits a negative attitude, and

changes to inaccurate beliefs (or misconceptions) can lead to a concomitant shift in the valence of

attitudes (see Heddy et al., 2017 and Sinatra & Seyranian, 2016). According to Ajzen (1989),

however, attitudes and beliefs are not merely consistent with one another; rather, attitudes

systematically vary as a function of individuals’ beliefs such that individuals’ beliefs have a direct or

causal effect on attitudes. In other words, attitudes are conditional upon individuals’ beliefs.

Page 8 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 94

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

While beliefs are typically tacit in exerting their effects on learning and behavior (Ajzen, 1989),

attitudes tend to remain highly accessible to readily guide learning and decision-making behaviors

(Ajzen & Fishbein, 2005). For example, previously formed attitudes regarding GMFs (the object of

evaluation) are automatically activated during encounters with knowledge claims about GMFs

(attributes of the object), and these readily available object-evaluation associations mitigate the

need for individuals to metacognitively process all attributes of an argument vis-a-vis their

predominant underlying epistemic beliefs each time they encounter a knowledge claim about GMFs.

In this way, beliefs moderate attitudinal processing of information during learning and conceptual

change, whereas attitudes serve a heuristic function for learning by enabling individuals to quickly

process topical knowledge and to make quick decisions regarding whether to accept or reject

knowledge claims related to a focal topic.

Although individuals’ beliefs might not necessarily be veridical (i.e., beliefs may be biased or

inaccurate), once an individual has developed a system of beliefs, these beliefs provide the cognitive

substrate from which congruent attitudes are cultivated in a consistent fashion (Ajzen & Fishbein,

2005). Once an attitude is shaped, it can work backwards and influence the development of new

systems of beliefs, or provoke a revision to incompatible systems. In other words, the information

processing qualities of attitudes and epistemic beliefs result in recurring and reciprocal processes

during conceptual change that create feedback loops to enable learners to metacognitively monitor,

adapt, target and revise inaccurate knowledge structures during conceptual change.

THE CURRENT STUDY

Attitudes and epistemic beliefs predict how individuals process, interpret, and evaluate scientific

information, and therefore, play a significant role in learning and conceptual change. Fulmer (2014),

for example, found that students’ attitudes toward science predicted their epistemic beliefs

pertaining to the uncertainty of scientific knowledge, including their subsequent evaluations of

sources of scientific knowledge. Additionally, Kapucu and Bahçivan (2015) found that students’

epistemic beliefs about science positively correlated with their attitudes towards learning physics,

and that students who held more constructivist epistemic beliefs regarding scientific knowledge

tended to self-report more positive attitudes. Although these studies (and others; see Broughton et

al., 2013; Franco et al., 2012) have revealed predictive and directional relationships between

attitudes and epistemic beliefs in science education, few studies have directly investigated how, in

what ways, and for whom these factors function to facilitate or constrain conceptual change while

learning about complex science topics from refutation or expository texts. As previously noted, the

unique rhetorical structure of refutation texts (i.e., message characteristics) has been shown to

effectively support conceptual change (Tippet, 2010). Therefore, we used refutation and expository

texts to examine whether text type (i.e., message characteristics) would interact with individual

learners’ characteristics (i.e., attitudes and beliefs) to predict conceptual change.

Examining relations between attitudes and epistemic beliefs in conceptual change can provide

invaluable insights into how these factors predict socio-scientific reasoning and public

understanding of science, especially regarding whether attitudes systematically vary as a function

of individuals’ epistemic beliefs, and in turn, whether attitudes predict the processing of scientific

information during conceptual change vis-a-vis the evaluative function of learners’ underlying

epistemic beliefs. The current study thus examines potential mediating and moderating effects of

Page 9 of 19

95

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

attitudes and epistemic beliefs in conceptual change while learning about GMFs, a complex socio- scientific topic.

Based on theoretical and empirical considerations, the following hypotheses were generated: (H1)

Participants who read a refutation text will change more misconceptions at post-test than

participants who read an expository text, and (H2) learners’ attitudes towards GMFs will mediate

between their prior misconceptions regarding GMFs and post-test conceptual change, but this effect

will be conditional upon text condition and learners’ epistemic beliefs such that participants with

less constructivist epistemic beliefs will report more negative attitudes toward GMFs, and in turn,

revise fewer misconceptions at post-test than learners with more constructivist epistemic beliefs

and more positive attitudes toward GMFs. However, learners with less constructivist beliefs in the

refutation condition should change more misconceptions that individuals with less constructivist

beliefs in the expository condition. See Figure 2 for the hypothesized model.

METHODS

Participants

One hundred twenty undergraduate students participated in the study (n = 32 males) with a mean

age of 21.29 years (SD = 3.83). Nineteen were in their first year of university, 24 were in their second

year, 37 were in their third year, and 40 were in their fourth year. Students were drawn from an

eclectic range of majors. Participants were recruited from a public university using an

advertisement posted to the university’s online classifieds.

Materials

Demographics Survey

A brief demographics survey was administered to capture participants’ background information

including gender, age, degree major/minor, cumulative GPA, first and second languages spoken and

written, political affiliation.

Prior Knowledge Test

The prior knowledge measure included a 10-item multiple-choice test about GMFs (adapted from

Heddy et al., 2017) and was used to assess participants’ misconceptions related to GMFs.

Participants were required to select the correct answer from a multiple-choice list of four possible

answer choices. The three incorrect options were common misconceptions about GMFs. An example

item included: “Processes used by scientists to modify the genetic makeup of plants and animals

include which of the following?” One point was awarded for each correct answer, and zero points

were awarded for incorrect answers. Participants’ scores were summed, and an overall average

score was calculated. The reliability index for the prior knowledge measure was moderately low

(Cronbach’s = .60; see Nunnally, 1978), yet this level of reliability is expected when participants’

level of prior knowledge is also low, or when the topic being tested is sufficiently complex, as is the

case with the topic of GMFs.

Attitudes Toward GMFs

The attitudes measure included four Likert-type items (adapted from Heddy et al., 2017) and was

used to assess attitudes toward GMFs. An example item included, “I approve of genetically modified

foods.” Participants self-reported their attitudes toward GMFs on a scale ranging from “1 = strongly

Page 10 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 96

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

disagree” to “5 = strongly agree.” The attitudes measure showed very good reliability (Cronbach’s

= .92).

Topic-Specific Epistemic Beliefs Questionnaire

A modified version of the Topic-Specific Epistemic Beliefs Questionnaire (TSEBQ; Strømsø, Bråten,

& Samuelstuen, 2008) was used to assess participants’ epistemic beliefs related to GMFs. The TSEBQ

includes 24 items organized along four belief dimensions including the certainty of knowledge (6

items), complexity of knowledge (6 items), source of knowledge (5 items), and justification for

knowing (7 items). Participants rated each item on a 10-point Likert scale ranging from “1 = strongly

disagree” to “10= strongly agree.” Psychometric analyses of the TSEBQ revealed low to moderate

reliability estimates for each factor, including certainty (Cronbach’s α = 0.68), complexity

(Cronbach’s α = 0.48), source (Cronbach’s α = 0.62), and justification (Cronbach’s α = 0.71).

Experimental Texts

The experimental texts presented information regarding GMFs and included one expository and

one refutation text (Heddy et al., 2017). Both texts were comparatively equivalent in length (617 vs.

624 words, respectively). In terms of ease of readability, each text obtained Flesch-Kincaid ease of

reading scores of 42.1 and 42.2, respectively (Flesch, 1948). Both the refutation and expository texts

included the same information, but the refutation text presented information by directly identifying

a common misconception and refuting it using three empirically validated causal explanations. An

example refutation included the following: “You may think that injecting hormones into a plant or

animal is involved in the production of genetically modified foods. This belief is also incorrect.

Injecting hormones into a plant or animal can increase its growth rate or its size. However, injecting

hormones does not modify the genetic makeup of the plant or animal. In contrast, genetically

modified foods have had some of their characteristics changed at the gene level.” The expository

text presented the same information as the refutation text, but without the refutation content. For

example: “The production of genetically modified foods does not involve injecting hormones into a

plant or animal. Injecting hormones into a plant or animal can increase its growth rate or its size.

However, injecting hormones does not modify the genetic makeup of the plant or animal. In

contrast, genetically modified foods have had some of their characteristics changed at the gene

level.” The refutation text presented participants with a total of four refutations that targeted

common misconceptions related to GMFs.

Post-knowledge Test

To assess post-test learning (i.e., conceptual change), participants completed the same prior

knowledge test. Psychometric reliability for the post-test measure was modest (Cronbach’s α = .71).

Procedure

Participants first provided informed consent and subsequently received instructions on how to

complete the study. Next, participants completed the GMFs prior knowledge test, the attitudes

toward GMFs survey, and an adapted version of the TSEBQ (Strømsø et al., 2008). Following this,

participants were randomly assigned to a refutation (n = 62) or expository (n = 58) text condition.

After reading, participants completed the same GMFs prior knowledge test to assess conceptual

change. Finally, participants completed a demographics survey to obtain basic background

Page 11 of 19

97

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

information. At the end of the session, participants were thanked for their participation and

compensated $10 for their time.

RESULTS

Preliminary Analysis

Each continuous variable was inspected for skewness and kurtosis. Analyses revealed the GMFs

knowledge post-test to be negatively skewed (-4.88). Given the post-test assessment included a true

0-value (answers were tabulated as being either correct or incorrect), no transformation was

applied to normalize the distribution. No issues were observed regarding kurtosis. Descriptive

statistics for all variables can be found in Table 1. Inspection of the data revealed univariate outliers

for the following variables: attitudes (n = 3, z = -2.53), uncertainty of knowledge (n = 2, z = -2.69),

complexity of knowledge (n = 1, z = -2.55), and the post-test knowledge assessment (n = 3, z = -3.29).

In lieu of deletion, all cases were kept given the values were not extreme and equated to less that

2% of n (Cohen, Cohen, West, & Aiken, 2003). There were no multivariate outliers in the data, and

examination of the Pearson correlation matrix revealed no issues of multicollinearity among

variables of interest. See Table 2.

Repeated Measures Analysis of Variance (ANOVA)

A repeated-measures ANOVA was conducted to assess whether there were any significant group

differences in conceptual change from pretest to posttest between learners who read a refutation

text versus learners who read an expository text (H1). Time was used as a within-subjects factor (2

levels, pretest and post-test) and text condition (refutation and expository) as the between-subjects

factor. Results revealed a significant Time by condition interaction, indicating that learners in the

refutation text condition changed significantly more of their misconceptions at post-test than

learners in the expository text condition, Wilks’ Lambda = .850, F(1,118) = 20.76, p < .001, η2 = .15.

Post hoc analyses revealed a significant mean difference (M = .17, SD = .04) in post-test conceptual

change between conditions, t(118) = 4.56, p < .001. Overall, 15% of the variance in post-test

conceptual change was accounted for by text condition whereby learners who read a refutation text

changed significantly more misconceptions regarding GMFs at post-test than learners who read an

expository text.

Moderated Mediation Analyses

To examine whether attitudes mediated between prior misconceptions and post-test conceptual

change and, in turn, whether this relationship was moderated by text type and learners’ epistemic

beliefs (H2), moderation mediation analyses were conducted using Hayes and Preachers’ (2014)

PROCESS macro for SPSS. See Figure 2 for the hypothesized model, and Figure 3 for the final model

with standardized effects.

Results revealed a significant regression of prior misconceptions on attitudes towards GMFs (β =

.38, t(118) = 4.47, p < .001). Findings also revealed a significant regression of prior misconceptions

on post-test conceptual change (β = .20, t(115) = 2.10, p = .04). There was also a statistically

significant regression of attitudes towards GMFs on post-test conceptual change, β = .27, t(115) =

3.08, p = .003. The overall model was significant, F(4,115) = 10.41, p < .001, R2 = .27, with a medium

effect size showing that 27% of the variance in post-test conceptual change could be accounted for

by both the variables in the model.

Page 12 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 98

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

With regard to the moderated mediation analyses, no statistically significant moderated mediation

was detected for type of text. However, results revealed a statistically detectable conditional

indirect effect (Index = -.06, with 95% bootstrap CIs from -.14 to -.01) of prior misconceptions on

post-test conceptual change via learners’ attitudes toward GMFs that varied as a function of their

epistemic beliefs regarding the source of GMFs knowledge at levels of the mean (effect = .12, with

95% bootstrap CIs from .04 to .25) and 1SD below the mean (effect = .18, with 95% bootstrap CIs

from .06 to .36). Overall, findings show that the more learners believed GMFs knowledge to be

internally generated and derived via subjective experience, the less positive their attitudes towards

GMFs, and in turn, the less misconceptions they revised at post-test.

Results also revealed a statistically detectable conditional indirect effect (Index = -.11, with 95%

bootstrap CIsfrom -.23 to -.02) of prior misconceptions on post-test conceptual change via learners’

attitudes that varied as a function of their beliefs regarding the justification of GMFs knowledge at

levels of the mean (effect = .15, with 95% bootstrap CIs from .05 to .26) and 1SD below the mean

(effect = .26, with 95% bootstrap CIs from .09 to .47). Overall, findings showed the more learners

believed GMFs knowledge to be justified via personal judgement, the less positive their attitudes,

and in turn, the less misconceptions they revised at post-test. Implications, future directions, and

limitations of these findings are discussed next.

DISCUSSION

The purpose of our study was to advance knowledge about the roles of attitudes and epistemic

beliefs in conceptual change while learning about GMFs. Based on Dole and Sinatra’s (1998) CRKM

of conceptual change, as well as our hypothesized conceptual model of relations among attitudes

and epistemic beliefs, we examined whether attitudes mediated between prior misconceptions and

post-test conceptual change, and in turn, whether this relationship was moderated by learners’

epistemic beliefs. Additionally, we examined whether refutation texts were an effective intervention

for facilitating conceptual change The results supported our hypotheses, which have theoretical

implications for understanding how individuals interpret, evaluate, and make sense of complex

socio-scientific information.

Refutation Texts and Conceptual Change (Message Characteristics)

Although some findings in the extant literature have shown refutation texts can inadvertently

strengthen learners’ misconceptions (i.e., backfire; Hart & Nisbet, 2012; Nyhan, Reifler, & Ubel,

2013), the clear majority of research indicates that refutation texts are an effective strategy for

facilitating conceptual change (Tippet, 2010). Refutation texts support conceptual change by

directly targeting misconceptions and presenting a rich network of causal explanations based on

scientific evidence to refute inaccurate knowledge (Kendeou et al., 2013, 2014). The causal

explanations in refutation texts compete with previously-acquired misconceptions and begin to

dominate the conceptual network by reducing (or suppressing) activation of previously-acquired

misconceptions, thus facilitating a change in knowledge within a conceptual network. In our study,

we hypothesized (H1) that learners who read a refutation text would change more misconceptions

at posttest than learners who read an expository text. Our results supported this hypothesis, and

provided support for previous studies that have shown refutation texts to be effective rhetorical

devices for facilitating conceptual change (Heddy et al., 2017; Heddy & Sinatra, 2013; Kendeou et

al., 2014; Trevors et al., 2016, 2017b).

Page 13 of 19

99

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

Additionally, our findings provided added support for Dole and Sinatra’s (1998) CRKM by

highlighting the importance of message characteristics for aiding in conceptual change (Dole &

Sinatra, 1998). Building from this, we argue that learner characteristics such as attitudes and

epistemic beliefs serve as focal lenses through which causal explanations in refutation texts are

processed, interpreted, and evaluated during conceptual change. As such, future studies should

augment and organize refutation texts with the goal of scaffolding both attitudinal change and

learners’ development of epistemic strategies for critically evaluating information regarding

complex scientific topics.

Moderating and Mediating Roles of Attitudes and Epistemic Beliefs in Conceptual Change

(Learner Characteristics)

Based on theoretical considerations and previous research, we additionally hypothesized (H2) that

attitudes would play a significate (mediating) role in learners’ processing of socio-scientific

information and while learning about GMFs from refutation texts and, in turn, that epistemic beliefs

would moderate the interceding effects of attitudes on post-test conceptual change. Our findings

supported this hypothesis. Specifically, our findings revealed a learning differential from pre-test to

post-test that was mediated by participants’ attitudes towards GMFs, and this relationship was

moderated by learners’ epistemic beliefs regarding the source and justification of GMFs knowledge.

Accordingly, learners in our study who self-reported more positive attitudes towards GMFs and

who believed GMFs knowledge to be derived and justified via empirical evidence and multiple

sources of information revised more misconceptions at post-test than learners with negative

attitudes who believed GMFs knowledge to be derived and justified via personal judgement, opinion

and anecdotal evidence. Although no effects were observed for beliefs about the complexity or

certainty of GMFs knowledge in our study, it is possible that learners in our study with positive

attitudes and who possessed the skills necessary to critically evaluate and judge the veracity of

diverse sources of information inherently viewed scientific knowledge as sufficiently complex and

uncertain, despite our null results. Whatever the case, our findings nonetheless lend support for the

structural and functional roles of attitudes and epistemic beliefs as vital factors involved in the

processing socio-scientific information, and highlight the importance of both learner and message

characteristics as vital factors in the reconstruction of knowledge.

Educational Implications

Our results underline the importance of considering the role of differential attitude structures in

relation to specific dimensions of epistemic beliefs (rather than general systems) in conceptual

change research. Results from our study show that individuals who believe GMFs knowledge to be

derived and justified via empirical evidence and multiple sources of information also tend to

express more positive attitudes toward GMFs and, in turn, experience greater learning gains in

terms of conceptual change. Our findings suggest that designing refutation texts with persuasive

attitudinal and epistemically-related content (i.e., content designed to change attitudes and improve

epistemic evaluations and judgements regarding the veracity of information) could help equip

learners with more robust strategies for learning and understanding complex socio-scientific topics.

Although attitudes have important implications for how individuals select, perceive, interpret,

encode and retrieve information related to complex science topics, epistemic beliefs play an

important role in how individuals evaluate and judge sources of knowledge. Overall, changing

attitudes (object-evaluation associations) reciprocally influences changes in underlying systems of

Page 14 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 100

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

beliefs (evaluative knowledge structures), and the reciprocal interaction between these factors

predict learning and conceptual change.

Therefore, developing interventions to foster the development of more positive attitudes, as well as

more constructivist epistemic beliefs, has the potential to positively predict learning in relation to

negatively charged, controversial socio-scientific topics that otherwise tend to elicit negative

attitudes and conflict with individuals’ personal, doxastic beliefs. In short, targeting individuals’

sourcing and justification beliefs related to complex socio-scientific topics has the potential for

endowing learners with the skills to critically evaluate and judge the veridicality of complex

scientific information instantiated in diverse media and obtained from multiple sources.

Understanding knowledge and knowing to be comprised of uncertain and complex processes,

learners may better appreciate the diversity of methods for justifying scientific claims, and thus,

become more open to discrepant sources of knowledge and information and, in turn, be more

receptive to revising inaccurate knowledge.

LIMITATIONS AND FUTURE DIRECTIONS

Several caveats should be considered while interpreting the results of our study. First, we used self- report inventories to measure attitudes and epistemic beliefs. Although self-reports of attitudes are

generally reliable indicators of individuals’ attitudes, they can be somewhat unreliable indicators of

epistemic beliefs (Greene et al., 2014). Indeed, for this sample, reliability estimates for epistemic

beliefs were not particularly high. That said, the low to modest reliabilities for the epistemic beliefs

measure could potentially reflect the relative diversity of participants’ epistemic beliefs regarding

GMFs. Alternatively, utilizing think aloud protocols (TAPs) could have provided richer, more

reliable data from which to generalize findings regarding the attitudinal and epistemic processes at

play during conceptual change as TAPs are particularly effective for observing emergent cognitive

and metacognitive processes that arise during learning without interfering with the focal learning

task (Chi, 1997).

Next, no measures of emotions towards GMFs were included in this study. The preponderance of

research in the conceptual change literature has shifted focus away from examining exclusively

‘cold’ cognitive constructs of conceptual change (i.e., information processing) to investigations that

additionally include ‘hot’ constructs, such as emotions and motivation (Broughton et al., 2013;

Pintrich et al., 1993; Sinatra & Seyranian, 2016; Heddy et al., 2017). Although attitudes (a lukewarm

construct) were a primary variable of interest in the present study, the roles of emotions (a hot

topic) and their relationship to attitudes was not included in this research. Including measures of

emotions, however, could have otherwise provided more explanatory power regarding the

processes at play during conceptual change, especially considering that affect serves an evaluative

function in the appraisal of information related to an attitudinal object.

Finally, we did not measure values toward GMFs in the present study. According to Rockeach

(1968), values are an important factor to consider when examining relationships among attitudes

and beliefs because value systems inevitably inform individuals’ systems of beliefs and, in turn, the

emergent attitudes that influence and shape individuals’ processing of information and decision- making behaviors. Arguably, circumventing value systems and their relations to beliefs and

attitudes provides a rather myopic focus on problems of persuasion at the expense of larger issues

related ‘education and re-education’ (Rockeach, 1968). As such, findings from the present study

Page 15 of 19

101

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

provide only a limited account of the potential multivariate factors that may influence conceptual

change. Future research should address this issue by seeking to uncover how and for whom values

predict epistemic beliefs, and in what ways these value-laden belief systems most profoundly

predict subsequent attitudinal appraisals of complex socio-scientific topics. In closing, future

research should focus more explicitly on the dynamic, functional and structural relations among

emotions, values, attitudes, and epistemic beliefs to obtain a more inclusive understanding of how

these factors operate to predict learning, conceptual change, and the processing of socio-scientific

information.

References

Ajzen, I. (1989). Attitude structure and behavior. In A.R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude

structure and function (pp. 241 – 274). Hillsdale, NJ: Erlbaum.

Ajzen, I., & Fishbein, M. (2005). The influence of attitudes on behavior. In D. Albarracín, B. T. Johnson, & M. P. Zanna

(Eds.), The handbook of attitudes (pp. 173-221). Mahwah, NJ: Erlbaum.

Broughton, S. H., Sinatra, G. M., & Nussbaum, E. M. (2013). “Pluto has been a planet my whole life!” Emotions, attitudes,

and conceptual change in elementary students learning about Pluto’s reclassification. Research in Science Education,

43(2), 529–550.

Chan, C., Burtis, J., & Bereiter, C. (1997). Knowledge building as a mediator of conflict in conceptual change. Cognition

and Instruction, 15(1), 1 - 40.

Chi, M. T. H. (1997). Quantifying qualitative analyses of verbal data: A practical guide. Journal of the Learning Sciences,

6(3), 271–315.

Chi, M.T.H. (2008). Three types of conceptual change: Belief revision, mental model transformation, and categorical

shift. In S. Vosniadou (Ed.), Handbook of research on conceptual change (pp. 61-82). Hillsdale, NJ: Erlbaum.

Chinn, C. A., Buckland, L. A., Samarapungavan, A. L. A. (2011). Expanding the dimensions of epistemic cognition:

Arguments from philosophy and psychology. Educational Psychologist, 46(3), 141–167.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral

sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.

Dole, J. A., & Sinatra, G. M. (1998). Reconceptualizing change in the cognitive construction of knowledge. Educational

Psychologist, 33, 109–128.

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Orlando, FL: Harcourt Brace Jovanovich College

Publishers.

Ecker, U. K. H., Hogan, J. L., & Lewandowsky, S. (2017). Reminders and repetition of misinformation: Helping or

hindering its retraction? Journal of Applied Research in Memory and Cognition, 6(2), 185–192.

Fabrigar, L., Macdonald, T., & Wegener, D. (2005). The structure of attitudes. In D. Albarracin, B.T. Johnson, & M.P.

Zanna (Eds.), The handbook of attitudes (pp. 79 – 124). London, UK: Routledge.

Fishbein, M. & Icek, A. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. California,

US: Addison-Wesley.

Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32(3), 221-233.

Greene, J. A., Yu, S., & Copeland, D. Z. (2014). Measuring critical components of digital literacy and their relationships

with learning. Computers & Education, 76, 55-69.

Hart, P. S., & Nisbet, E. C. (2012). Boomerang Effects in Science Communication: How Motivated Reasoning and

Identity Cues Amplify Opinion Polarization About Climate Mitigation Policies. Communication Research, 39(6), 701–

723.

Nyhan, B., Reifler, J., & A Ubel, P. (2012). The Hazards of Correcting Myths About Health Care Reform. Medical care,

51(2), 127-32.

Page 16 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 102

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

Hayes, A. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach.

New York, NY: Guilford.

Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable.

British Journal of Mathematical and Statistical Psychology, 67, 451–470.

Heddy, B. C., Danielson, R. W., Sinatra, G. M., & Graham, J. (2017). Modifying knowledge, emotions, and attitudes about

genetically modified foods. Journal of Experimental Education, 85(3), 515-533.

Heddy, B. C., & Sinatra, G. M. (2013). Transforming misconceptions: Using transformative experience to promote

positive affect and conceptual change in students learning about biological evolution. Science Education, 97(5), 723–

744.

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational

Psychology, 25(4), 378–405.

Hofer, B. K., & Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and

knowing and their relation to learning. Review of Educational Research, 67(1), 88–140.

Kapucu S. & Bahçivan E. (2015). High school students’ scientific epistemological beliefs, self-efficacy in learning

physics and attitudes towards physics: A structural equational model. Research in Science & Technological Education,

33(2), 252-267.

Kata, A. (2012). Anti-vaccine activists, Web 2.0, and the postmodern paradigm: An overview of tactics and tropes used

online by the anti-vaccination movement. Vaccine, 30(25), 3778–3789.

Kendeou, P., & O'Brien, E. J. (2014). The knowledge revision components (KReC) framework: Processes and

mechanisms. In D. Rapp, & J. Braasch (Eds.), Processing inaccurate information: Theoretical and applied perspectives

from cognitive science and the educational sciences (pp. 353-377). Cambridge, MA: MIT Press.

Kendeou, P., Walsh, E. K., Smith, E. R., & O’Brien, E. J. (2014). Knowledge revision processes in refutation texts.

Discourse Processes, 51, 374-397.

Kendeou, P., Smith, E. R., & O’Brien, E. J. (2013). Updating during reading comprehension: Why causality matters.

Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(3), 854–865.

Krause, S., Kelly, J., Corkins, J., Tasooji, A., & Purzer, S. (2009). Using students’ previous experience and prior knowledge

to facilitate conceptual change in an introductory materials course. Paper presented to the Frontiers in Education

Conference, San Antonio, TX.

Lombardi, D., Nussbaum, E. M., & Sinatra, G. M. (2016). Plausibility judgments in conceptual change and epistemic

cognition. Educational Psychologist, 51(1), 35–56.

Maio, G. R. & Haddock, G. (2010). The influence of attitudes on information processing and behavior. In G.R. Maio & G.

Haddock (Eds.), The psychology of attitudes and attitude change (pp. 47-66). London, UK: SAGE Publications Ltd.

Mason, L. & Gava, M. (2007). Effects of epistemological beliefs and learning text structure on conceptual change. In S.

Vosniadou, A. Baltas, & X. Vamvakoussi (Eds.), Reframing the problem of conceptual change in learning and instruction

(pp. 165-196). Oxford, UK: Elsevier Science.

Mason, L., Gava, M., & Boldrin, A. (2008). On warm conceptual change: The interplay of text, epistemological beliefs,

and topic interest. Journal of Educational Psychology, 100, 291- 309.

Murphy, P. K., & Alexander, P. A. (2016). Interrogating the relation between conceptual change and epistemic beliefs.

In J. A. Greene, W. A. Sandoval, & I. Bråten (Eds.), Handbook of epistemic cognition (pp. 439-459). New York, NY:

Routledge.

Nunnally, J.C. (1978). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.

Pintrich, P. R., Marx, R. W., & Boyle, R. A. (1993). Beyond cold conceptual change: The role of motivational beliefs and

classroom contextual factors in the process of conceptual change. Review of Educational Research, 63(2), 167–199.

Rokeach, M. (1968). A theory of organization and change within value-attitude systems. Journal of Social Issues, 24(1),

13–33.

Page 17 of 19

103

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational

Psychology, 82(3), 498-504.

Silverman, J. C. (2007). Epistemological beliefs and attitudes toward inclusion in pre-service teachers. Teacher

Education and Special Education: The Journal of the Teacher Education Division of the Council for Exceptional Children,

30(1), 42–51.

Sinatra, G. M., Brem, S. K., & Evans, E. M. (2008). Changing minds? Implications of conceptual change for teaching and

learning about biological evolution. Evolution: Education and Outreach, 1(2), 189–195.

Sinatra, G. M., & Hofer, B. K. (2016). Public understanding of science. Policy Insights from the Behavioral and Brain

Sciences, 3(2), 245–253.

Sinatra, G. M., Kienhues, D., & Hofer, B. K. (2014). Addressing Challenges to Public Understanding of Science: Epistemic

Cognition, Motivated Reasoning, and Conceptual Change. Educational Psychologist, 49(2), 123–138.

Sinatra, G. M., & Seyranian, V. (2016). Warm change about hot topics: The role of motivation and emotion in attitude

and conceptual change about controversial science topics. In L. Corno & E. Anderman (Eds.), Handbook of educational

psychology (pp. 245–256). Washington, DC: Taylor & Francis.

Stathopoulou, C., & Vosniadou, S. (2007). Conceptual change in physics and physics-related epistemological beliefs: A

relationship under scrutiny. In S. Vosniadou, A. Baltas, & X. Vamvakoussi (Eds.), Advances in learning and instruction

series: Reframing the conceptual change approach in learning and instruction (pp. 145-163). New York, NY: Elsevier

Science.

Strømsø, H. I., Bråten, I., & Samuelstuen, M. S. (2008). Dimensions of topic-specific epistemological beliefs as

predictors of multiple text understanding. Learning and Instruction, 18(6), 513–527.

Tippett, C. (2010). Refutation text in science education: A review of two decades of research. International Journal of

Science and Mathematics Education, 8, 951–970.

Trevors, G. J., Kendeou, P., & Butterfuss, R. (2017a). Emotion processes in knowledge revision. Discourse Processes,

54(5–6), 406–426.

Trevors, G. J., Kendeou, P., Bråten, I., & Braasch, J. L. G. (2017). Adolescents’ epistemic profiles in the service of

knowledge revision. Contemporary Educational Psychology, 49, 107–120.

Vosniadou, S. (1994). Capturing and modeling the process of conceptual change. Learning and Instruction, 4(1), 45–69.

Wakefield, A., Murch, S., Anthony, A., Linnell, J., Casson, D., Malik, M., & ... Walker-Smith, J. (1998). RETRACTED: Ileal- lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. The Lancet,

351(9103), 637-641.

World Health Organization. (2018). Europe observes a 4-fold increase in measles cases in 2017 compared to previous

year. Retrieved March 18, 2018, from http://www.euro.who.int/en/media-centre/sections/press- releases/2018/europe-observes-a-4-fold-increase-in-measles-cases-in-2017-compared-to-previous-year

TABLES

Table 1: Descriptive Statistics for all Variables

Mean SD Skewness Kurtosis

Prior knowledge .44 .20 2.30 .32

Post-test knowledge .77 .20 -4.88 3.08

Attitudes 4.48 1.37 -2.09 .06

Complexity of knowledge 4.12 .70 .42 -0.68

Uncertainty of knowledge 4.71 .82 -1.13 .71

Source of knowledge 4.17 .86 .40 -.13

Justification for knowing 5.36 .70 -1.56 -.47

Note. Prior knowledge and post-test knowledge are reported as proportion of correct conceptions.

Page 18 of 19

URL: http://dx.doi.org/10.14738/assrj.76.7882 104

Vivian, J. A., & Muis, K. R. (2020) Epistemic Beliefs Moderate Mediations Among Attitudes, Prior Misconceptions, and Conceptual Change. Advances in Social

Sciences Research Journal, 7(6) 87-105.

Table 2: Correlations Among All Variables (N = 120)

Variables 1 2 3 4 5 6 7

1. Attitudes — .215* .131 .076 .138 .380** .389**

2. Structure of Knowledge .215* — -.005 .506** .341** .427** .358**

3. Certainty of Knowledge .131 -.005 — .011 .247** .097 .155

4. Source of Knowledge .076 .506** .011 — .350** .351** .108

5. Justification for Knowing .138 .341** .247** .350** — .344** .248**

6. Prior Knowledge .380** .427** .097 .351** .344** — .391**

7. Post-test Knowledge .389** .358** .155 .108 .248** .391** —

*p < .05. **p < .01. Two-tailed.

Figure 1. Object-evaluation association embedded in a semantic network.

Object

(GMFs

Knowledge)

(Epistemic Beliefs about GMFs)

Evaluation of Object Attributes

Strength of

Association

Structure of Knowledge

Certainty of Knowledge

Source of knowledge

Justification for Knowing

Evidence that GMFs

cause health problems,

environmental issues,

and consumer rights

violations.

GMFs research

is conducted

primarily by

biotechnology

companies.

No scientific

consensus

regarding GMFs

knowledge.

The science behind

GMFs is based on easy to

understand facts.

Object Attributes (Knowledge Claims)

Evaluation

(Beliefs about

GMFs)

LOCAL OBJECT-EVALUATION ASSOCIATION

GLOBAL OBJECT-EVALUATION ASSOCIATION

Page 19 of 19

105

Advances in Social Sciences Research Journal (ASSRJ) Vol.7, Issue 6, June-2020

Figure 2. Hypothesized moderation mediation model.

Figure 3. Final moderation mediation model.

Note: All values represent standardized coefficients. Coefficients labelled at the top for the

moderators represent 1SD below the mean, coefficients in the middle represent the mean, and

coefficients at the bottom for the moderators represent 1SD above the mean.

** p < .001 * p < .05.

Prior

Knowledge

Post

Knowledge Attitudes

Justification

for Knowing

Complexity of

Knowledge

Source of

Knowledge

Certainty of

Knowledge

Text

Condition

Prior

Knowledge

b = .20**

b = .38** b = .27**

Post

Knowledge Attitudes

Source of

Knowledge

b = ns

b = .12

b = .18

Justification

for Knowing

b = ns

b = .15

b = .26