J. EDUCATIONAL COMPUTING RESEARCH, Vol.

28(1) 15-30, 2003

SUCCESSFUL LEARNING WITH HYPERMEDIA:

THE ROLE OF EPISTEMOLOGICAL BELIEFS AND
METACOGNITIVE AWARENESS

LISA D. BENDIXEN
KENDALL HARTLEY
University of Nevada

ABSTRACT

This study examined the relationship between epistemological beliefs, meta-

cognition, and student achievement in a hypermedia learning environment.
Epistemological beliefs refer to beliefs about the nature of knowledge and
knowing (see review by Hofer & Pintrich, 1997) and metacognition refers
to the ability to reflect upon, understand, and control ones learning (Schraw
& Dennison, 1994). We predicted that the epistemological beliefs and meta-
cognitive awareness of 116 preservice teachers would be significantly related
to achievement in a hypermedia tutorial. Achievement was measured by
a posttest based on the tutorial content. Results from a forced-order,
hierarchical regression analysis indicated that reading comprehension, GPA,
and three of the five epistemological beliefs (i.e., fixed ability, omniscient
authority, and quick learning) significantly predicted posttest performance.
Belief in omniscient authority and fixed ability being related to lower
achievement supports previous research. Contrary to our expectations, a
belief in quick learning was positively correlated with achievement. Educa-
tional and instructional design implications are discussed.

This study examined the relationship between epistemological beliefs, meta-

cognition, and student success in a hypermedia learning environment. We used the
term epistemological beliefs to refer to beliefs about the nature of knowledge
and knowing (for a review, see Hofer & Pintrich, 1997). Metacognition refers to
the ability to reflect upon, understand, and control ones learning (Schraw &

15
2003, Baywood Publishing Co., Inc.
16 / BENDIXEN AND HARTLEY

Dennison, 1994). We predicted that epistemological beliefs and metacognitive

awareness would be significantly related to achievement in a hypermedia tutorial.
For example, students with more complex beliefs about knowledge and more
advanced metacognitive awareness would fare better and gain more knowledge
in a hypermedia learning environment. In general terms, a student who believes
more effort is unnecessary (an epistemological belief) and lacks comprehension
monitoring skills (an example of metacognitive awareness) will be less likely
to utilize the hypermedia format effectively. The increasing use of hypermedia
for instructional purposes demands that we better understand how it affects
all learners.

THEORETICAL FRAMEWORK
A review by Dillon and Gabbard (1998) explored the hypermedia literature
concerning learner comprehension, learner control, and learner style (i.e., learner
characteristics). They concluded that the benefits of hypermedia, over other
methods of instruction, are limited. There is, however, evidence that individual
characteristics play a role in learning from hypermedia environments (Dillon &
Gabbard, 1998). The preponderance of existing hypermedia research has investi-
gated differences in user interfaces or instructional methods. Less research has
been conducted that explores learner characteristics and their impact on learning
in a hypermedia environment. Some of the attributes that have been studied
in a hypermedia environment include ability (Repman, Willer, & Lan, 1993),
passive/active learners (Lee & Lehman, 1993), field independence/dependence
(Jonassen & Wang, 1993), deep vs. shallow processors (Shute, 1993), and learning
strategy use (Hartley, 2001). Dillon and Gabbards review concluded that
cognitive characteristics might offer the beginning of an explanation for the
generally conflicting results in the literature comparing hypermedia and non-
hypermedia learning environments (p. 344).
Self-regulated learning theory may provide the overriding framework that
will help address issues of individual characteristics and the use of hypermedia
(Hartley & Bendixen, 2001). Self-regulated learning (SRL) has been conceptu-
alized as including metacognitive, motivational, and behavioral processes related
to the active participation of individuals in their own learning (Zimmerman, 1990).
Using a broader definition of metacognition (i.e., knowing about knowing) from
Flavell (1979), Hofer (2001) proposes that dimensions of personal epistemologies
include aspects of metacognitive processes. Specifically, Hofer (2001) suggests
that the dimensions of fixed ability (nature of intelligence) and quick learning
(the speed of knowledge acquisition), often associated with epistemological
beliefs (Schommer-Aikins, 2002), seem to be more metacognitive than epistemic
in nature. Although the current study chooses to take a more traditional view of the
dimensionality of epistemological beliefs (i.e., Schommer, 1990), their relation-
ship to metacognition and the more general contribution that epistemological
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 17

beliefs and metacognition make to self-regulated learning are important

considerations.
The learner attributes of epistemological beliefs and metacognition are almost
completely unexplored in empirical studies of hypermedia. These areas of
research have established that beliefs and metacognitive skills impact learning,
in general. The current study addressed how they are related, specifically to
learning with hypermedia.

EPISTEMOLOGICAL BELIEFS

Epistemology refers to the study of the nature of knowledge and knowing.

Educators have been interested in epistemological beliefs since the seminal
work of Perry (1970). Perry and a number of other theorists believe that indi-
viduals pass through a predictable sequence of epistemological growth. In early
stages, individuals hold simple, dichotomous views of knowledge; reasoning
then becomes increasingly more complex and relativistic. As students epistemo-
logical beliefs further develop, views about knowledge become post-relativistic
with a focus on the evaluation of different viewpoints (Hofer, 2001; King &
Kitchener, 1994).
More recently, researchers have investigated how epistemological beliefs
influence comprehension, strategy use, and academic performance. For example,
expanding on the work of Perry (1970) and King and Kitchener (1994),
Schommer-Aikins (2002; Schommer, 1990) proposed five independent epistemo-
logical dimensions corresponding to beliefs about knowledge. Each dimension
is based on a continuum. The following lists the naive end of the continuum
for each dimension: 1) certain knowledge (i.e., absolute knowledge exists and
will eventually be known); 2) simple knowledge (i.e., knowledge consists of
discrete facts); 3) omniscient authority (i.e., authorities have access to otherwise
inaccessible knowledge); 4) quick learning (i.e., learning occurs in a quick or
not-at-all fashion); and 5) fixed ability (i.e., the ability to acquire knowledge
is static).
Researchers have shown that the epistemological beliefs students hold have
important influences on thinking and problem solving. For example, research
conducted with traditional instructional materials indicates that certain epistemo-
logical beliefs correlate with achievement. Schommer (1990) found that when
asked to read a passage of text and supply a concluding paragraph, undergraduate
students holding beliefs in quick learning and certain knowledge gave over-
simplified and inappropriately absolute conclusions, respectively. Schommer,
Crouse, and Rhodes (1992) reported that beliefs in simple knowledge negatively
affected complex problem solving. Schoenfeld (1983) investigated some of the
consequences of a belief in quick learning. He reported that even experienced
students who were asked to solve math problems gave up after five or ten minutes
18 / BENDIXEN AND HARTLEY

on the assumption that if they failed to solve the problem during this time, the
problem could not be solved.
There is some evidence that these beliefs will also impact learning in a hyper-
media environment. For example, a study involving hypermedia by Jacobson
and Spiro (1995) briefly addressed epistemological beliefs. They found students
with simple epistemological beliefs had difficulty with the non-linear and multi-
dimensional nature of an ill-defined hypertext system. The measurements used
in this study were exploratory in nature and epistemological beliefs were not
the focus of the study. As indicated by Jacobson and Spiro (1995) in their article,
more research is necessary, but this study represents an important beginning to
the application of what is known about epistemological beliefs to hypermedia
learning.
Why would these differences exist and be particularly important in a hyper-
media learning environment? As discussed previously, particular epistemological
beliefs coincide with a less adequate approach to learning. The belief in fixed
ability as a primary determinant of success leads students to believe that more
effort does not coincide with more learning. As a consequence, the additional
hypermedia tools available, such as links to definitions, diagrams, self-check
materials, objectives, and advanced organizers, may have little positive impact.
The effort required to use the hypermedia tools is less than in traditional settings
(e.g., clicking on a term for a definition is easier than looking it up in the glossary).
However, it still requires mental effort and the choice to use them; therefore, it
could be expected that a belief in fixed ability would be negatively correlated with
use of hypermedia tools and subsequent successful learning. Similarly, a student
who takes a quick learning approach to hypermedia instruction will be less likely
to take the extra time involved in taking advantage of some of the useful tools
available in hypermedia environments and would not reap the benefits of them.
The aforementioned examples point to the potential impact particular epistemo-
logical beliefs may have on learning in a hypermedia environment.

METACOGNITION
Like epistemological beliefs, students metacognitive awareness will signifi-
cantly mediate success in most learning environments. Metacognition refers to
the ability to reflect upon, understand, and control ones learning (Schraw &
Dennison, 1994). Previous accounts of metacognition distinguish between two
major components: 1) knowledge about cognition; and 2) regulation of cognition.
Knowledge of cognition refers to a learners understanding of his or her own
thought processes (Schraw, 1998). Knowledge of cognition has been further
divided into declarative (about), procedural (how), and conditional (when) knowl-
edge (Brown, 1987; Jacobs & Paris, 1987; Schraw, 1998). The use of a strategy is
dependent on the students awareness of the strategy (declarative), understanding
of how the strategy works (procedural), and knowing when to use the strategy
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 19

(conditional). Regulation of cognition is understood as those behaviors that

demonstrate control of, and/or utility with, the knowledge of cognition. Examples
would include a students ability to monitor, evaluate, and plan for their learning
(Schraw, 1998). Metacognition thus relies on ones knowledge of their abilities
and strategies that may improve their own learning. Metacognition also depends
on the students ability to regulate their thinking in a way that takes advantage
of their knowledge of cognition.
For example, if a student knows that they have a difficult time remembering
dates, they may develop a timeline to better facilitate memorization. The under-
standing of this difficulty with dates, and their understanding of when and
how to use timeliness are examples of the students knowledge of cognition.
The decision to use the strategy is an example of regulation of cognition.
Therefore, the effective use of strategies is dependent on several factors, not
the least of which is the ultimate decision on the part of the learner whether or
not to use a strategy. The learners capacity in each of the previously mentioned
areas is vital to success in any learning environmentincluding hypermedia
instruction. The learners repertoire of reading strategies like summarization
(knowledge of cognition), and their willingness to invoke such strategies
(regulation of cognition) when appropriate, will have a dramatic effect on their
understanding.
Most studies that have investigated metacognition and hypermedia were con-
cerned with teaching self-regulatory strategies within the hypermedia environ-
ment (e.g., Puntambekar & duBoulay, 1997). Other studies have investigated
strategies used within the hypermedia environment (Hill & Hannafin, 1997).
No studies that we are aware of have investigated the relationship between
metacognitive skills, such as monitoring for understanding, and learning with
hypermedia.
It would seem logical that more metacognitively-aware students are better
equipped to take advantage of the hypermedia environment. The skilled learner
recognizes when their understanding is lacking and then how to remedy the
situation. For example, the metacognitively-aware learner will seek out definitions
or clues for the meaning of a new word. In a hypermedia environment, many of the
more difficult words can be hyperlinked to definitions or related text passages,
thus giving the learner another way to learn the word. The less aware student will
likely continue to read. The point is that the more aware the learner is the better
equipped they are to take advantage of the resources provided by this new
environment.
The goal of this study was to expand what we know about epistemological
beliefs and metacognition to include the role they play in learning with hyper-
media. More specifically, we were interested in how these two learner charac-
teristics would correlate with learning in a hypermedia environment. Previous
research indicates that more complex epistemological beliefs and higher meta-
cognitive awareness are positively related to achievement in traditional learning
20 / BENDIXEN AND HARTLEY

environments. In this study, we examined if such a correlation exists when using

hypermedia environments.

METHOD

Participants
One hundred sixteen undergraduate teacher candidates (88 females, 28 males)
from a large Southwestern university participated in the study as part of their
educational psychology course assignment. Participants average age was 28.12.

Materials
Materials included a packet containing: a) a 32-item Epistemological Beliefs
Inventory (EBI); b) a 52-item Metacognitive Awareness Inventory (MAI);
c) a 36-item reading comprehension test (Nelson-Denney); d) a 16-item test of
syllogistic reasoning; and e) a brief demographic information sheet.
Epistemological beliefs were measured using the EBI developed by Schraw,
Bendixen, and Dunkle (2002). This inventory is based on Schommers (1990)
five dimensions of epistemological beliefs which include: 1) certain knowl-
edge; 2) simple knowledge; 3) omniscient authority; 4) fixed ability; and
5) quick learning. All items were written using a 5-point Likert-type scale.
Cronbachs a for the EBI was .71 indicating good internal consistency. See
Table 1 for sample items.
Students metacognitive awareness was measured using the Metacognitive
Awareness Inventory (MAI) developed by Schraw and Dennison (1994) in which
items are classified into two categories of metacognition (i.e., knowledge of
cognition and regulation of cognition). All items were written using a 5-point
Likert-type scale. Cronbachs a for the MAI was .86 indicating excellent internal
consistency. See Table 1 for sample items.
To best isolate variance resulting from the MAI and EBI, reading compre-
hension and reasoning measures were included. Reading comprehension was
assessed using the Nelson-Denny reading comprehension test, Form E (Brown,
Bennett, & Hanna, 1981). This test contains eight passages on various topics
and five-option multiple choice questions for each passage.
Syllogisms were used to provide a general measure of logical reasoning
skill. The 16 syllogisms consisted of two-premise statements. Individuals
were asked to choose a valid conclusion from four possible responses, only
one of which was correct. The syllogisms have been normed on a similar
group and ranged in difficulty from 94 percent to 31 percent correct (Schraw,
Dunkle, & Bendixen, 1995). The demographic variable sheet included infor-
mation about age, gender, year in school, academic major, and estimated grade
point average.
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 21

Table 1. Factors and Sample Items from the EBI and MAI

Instrument
(No. items) Factor Item

EBI (8) Certain knowledge What is true today will be true tomorrow.

EBI (7) Simple knowledge Too many theories just complicate things.

EBI (5) Omniscient authority People should always obey the law.

EBI (7) Fixed ability How well you do in school depends on

how smart you are.

EBI (5) Quick learning Working on a problem with no quick

solution is a waste of time.

MAI (20) Regulation of I organize my time to best accomplish my

cognition goals.

MAI (32) Knowledge of I understand my intellectual strengths

cognition and weaknesses.

The hypermedia instructional materials were developed by the second author to

simulate instructional materials now commonly used. Specifically, the materials
included tools like links to definitions, advanced organizers, and self tests. The
tutorial screen was partitioned into four areas (see Figure 1). The top 10 percent of
the screen, the banner, contained links to an advanced organizer, objectives,
glossary, site map, and self-check questions. The banner was always visible. The
other 90 percent of the screen was divided into three columns, the menu, content,
and supplementary information. The menu contained links to the four major topics
in the tutorial. The content contained the text of the tutorial, which included links
to relevant terms, images, and other text pages. The supplementary information
section displayed definitions and images when selected from the content section.
For example, if while reading about Yugoslavia the user clicked on the term
Serbia, a map would appear in the supplementary information section. The
descriptive unit was developed on the subject of the former Yugoslavia, which
was timely and thus of interest to the students.
To assess student learning in the hypermedia tutorial, 11 short-answer/
completion items based on information about the former Yugoslavia were
developed. Because the items pertained to factual information found throughout
the tutorial, we felt that short answer items were most appropriate (Linn &
Gronlund, 2000). For example, some of the items asked students to list several
facts about the former Yugoslavia.
 22
/
BENDIXEN AND HARTLEY

Figure 1. Screen shot of a portion of the hypermedia tutorial.

 SUCCESSFUL LEARNING WITH HYPERMEDIA / 23

Procedure
Students completed the packet of measures (i.e., EBI, MAI, Nelson-Denney,
syllogisms, and demographic sheet) before the tutorial began. After a demon-
stration of the particulars of the tutorial, participants were instructed to Study the
material as if it were for a class and that they had 30 minutes in which to do it. The
instructions and timeframe were chosen to encourage the students to approach the
materials as they would naturally if preparing for a class (e.g., study a chapter from
the text). After 30 minutes the objective exam was given to measure students
achievement in the unit.

Scoring
All exams were scored by the authors and two trained graduate assistants
using an answer key that was developed. Each participant was given a score out of
37 points possible. Scores ranged from 5 percent correct to 81 percent correct.
After a practice session, each of the four raters scored each exam independently.
All disputes in scoring were settled in conference following a thorough discussion.

RESULTS
Given that our primary question was whether epistemic beliefs and meta-
cognitive awareness were related to achievement in the hypermedia tutorial, we
performed a conceptually-driven forced-order hierarchical regression to examine
whether epistemological beliefs and metacognitive awareness would predict
achievement after accounting for the variance associated with students logical
reasoning (syllogisms), GPA, and reading ability.

Variables in the Analysis

The criterion measure was the number of correct short-answer/completion
posttest items. We specified the order to enter the variables into the equation to
systematically remove variance associated with several learner characteristics. By
proceeding in this fashion, we were able to remove some of the variance from
individual differences in logical reasoning, GPA, and reading ability. Students
scores on the syllogisms measure were entered into the equation to remove
variance associated with students logical reasoning ability. Students reported
GPA was used as a general achievement variable. Scores on the Nelson-Denney
reading comprehension measure were entered into the equation to remove
variance associated with students general reading ability. The three aforemen-
tioned variables were entered as block one.
This allowed us to examine the effects of metacognitive awareness (i.e.,
block two included the variables regulation of cognition and knowledge of
cognition) to see if they accounted for additional variance beyond the logical
24 / BENDIXEN AND HARTLEY

reasoning, GPA, and reading ability variables. The five dimensions associated
with epistemological beliefs (i.e., fixed ability, omniscient authority, simple
knowledge, certain knowledge, and quick learning) were entered as a block in
the final step of the regression analysis (see Table 2 for a description of the
variables used in the analysis). Gender did not contribute significant variance so
it was dropped from the analysis.

Regression Analysis
Table 3 shows the results of the forced-order regression. As expected, students
reading comprehension and GPA were strong predictors of achievement in the
hypermedia tutorial and accounted for a considerable proportion of the variance.
Contrary to our predictions, the two variables associated with metacognitive
awareness did not reach significance.
Even after removing variance from the variables in blocks one and two, the
epistemic beliefs variables accounted for additional variance. Three of the five
variables associated with epistemic beliefs reached significance, including
omniscient authority, fixed ability, and quick learning. Omniscient authority and
fixed ability were negatively correlated with achievement in the hypermedia
tutorial indicating that the less students believed that authorities have access to
knowledge that others do not (i.e., omniscient authority) and that the ability
to acquire knowledge is innate (i.e., fixed ability), the better they did in the
hypermedia tutorial. Interestingly, the quick learning variable was positively

Table 2. Means and Standard Deviations for All Variables

Range

Variable M SD Minimum Maximum

Achievement measure 12.70 6.17 2.0 30.0

GPA 3.22 .42 2.30 4.0
Reading comprehension 23.22 5.52 9.0 36.0
Syllogisms 10.50 2.50 5.0 15.0
Regulation of cognition 3.42 .42 2.56 4.66
Knowledge of cognition 3.67 .39 2.61 4.78
Simple knowledge 3.10 .51 1.50 4.50
Certain knowledge 2.32 .51 1.10 3.70
Omniscient authority 3.16 .70 1.00 5.00
Quick learning 1.92 .55 1.00 3.80
Fixed ability 2.69 .65 1.43 4.71
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 25

Table 3. Results of Forced-Order Hierarchical Regression Analysis

Cumulative

Predictor R R2 change B

GPA .184*
Reading Comprehension .367***
Syllogisms .054
Block One .473 .224
Regulation of Cognition .118
Knowledge of Cognition .097
Block Two .483 .010
Simple Knowledge .005
Certain Knowledge .014
Omniscient Authority .197*
Quick Learning .286*
Fixed Ability .280*
Block Three .557 .077
Note: Block One contains the variables GPA, Reading Comprehension, and Syllogisms.
Block Two contains the variables Regulation of Cognition and Knowledge of Cognition.
Block Three contains the variables Simple Knowledge, Certain Knowledge, Omnisicient
Authority, Quick Learning, and Fixed Ability.
*p < .05. **p < .01. ***p < .001.

correlated with achievement in the tutorial, meaning that the stronger the students
belief in quick learning, the higher their achievement scores on the tutorial
assessment. This contradicts previous research (e.g., Schoenfeld, 1983;
Schommer, 1990) that has found a negative relationship between a belief in quick
learning and achievement in various problem solving situations. Neither the
simple knowledge or the certain knowledge variables reached significance.

DISCUSSION

Currently, very little research has been done to investigate epistemological

beliefs and metacognition and their relationship with learning in a hypermedia
context. It is imperative that these relationships be investigated as the frequency
of learning from hypermedia environments is rapidly increasing (Hartley &
Bendixen, 2001).
26 / BENDIXEN AND HARTLEY

The present study indicated that students reading ability plays an important
role in learning within a hypermedia context. This is, essentially, a replication of
numerous studies examining the influence that reading skills have on achievement
(e.g., Royer, Abranovic, & Sinatra, 1987; Royer, Marchant, Sinatra, & Lovejoy,
1990). The crux of the tutorial was a timed reading for comprehension task. This
study serves as an important reminder that if reading is significantly involved in a
learning task, whether it be in a more traditional or in a hypermedia environment,
its influence cannot be taken for granted.
Certain epistemological beliefs were related to success in the hypermedia
task. This finding is supported by previous research looking at epistemological
beliefs and achievement (Schommer et al., 1992). In particular, students less
likely to believe in omniscient authority and fixed ability did better in terms of
achievement in the hypermedia tutorial. An important aspect of the tutorial on the
former Yugoslavia was discussion from various, sometimes opposing, experts.
Students less likely to see experts as having exclusive access to information may
be able to appreciate and understand conflicting views on a subject. Similar
to previous research (Dweck & Leggett, 1988; Schommer, 1990) a belief that
more effort does not coincide with more learning (i.e., fixed ability) was associated
with lower achievement in the hypermedia task. It may very well be that the
use of additional hypermedia tools available, such as links to definitions,
diagrams, self-check materials, objectives, and advanced organizers were viewed
as unnecessary mental effort by these participants and, therefore, not utilized.
Unlike previous research, a belief in quick learning was associated with success
in the hypermedia tutorial. For example, Schommer (1990) found that a belief
in quick learning was related to poor performance in a complex problem solving
task. Current results may be an artifact of the task itself. Students had 30 minutes
to study the material. Under strict time limitations, a belief that learning
happens quickly or not-at-all may have actually aided some students in navi-
gating through the tutorial in a more timely manner. In addition, the objective
assessment at the end of the tutorial may not have required higher levels of
thinking; therefore, this particular belief in quick (potentially low-level) learning
may not have been a detriment.
Finally, a broader question remains regarding the current studys findings:
Why didnt beliefs in simple and certain knowledge, and metacognition play a
significant role in the hypermedia tutorial? The answer to this question may, again,
lie in the structure of the hypermedia task and the subsequent assessment used.
A growing body of research has supported the idea that epistemological beliefs
are more related to ill-defined problem solving, as opposed to well-defined
problem solving (Bendixen & Schraw, 2001; Bendixen, Schraw, & Dunkle,
1998; Kitchener, 1983; Schommer, 1990). Well-defined problems are those in
which there is one agreed-upon solution and, generally, one way of getting there.
This is in contrast to ill-defined problems, which may contain several solutions
and various means of achieving them. Kitchener (1983), for example, proposed a
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 27

three-level hierarchical model of cognitive processing. According to this model,

well-defined problems can be solved without making epistemic assumptions
because they lead to certain, guaranteed solutions (her first and second levels). In
contrast, ill-defined problems cannot be solved without epistemic assumptions
because they do not have certain, guaranteed solutions (her third level). A study
by Schraw, Dunkle, and Bendixen (1995) investigated this model and found
that self-reported epistemic beliefs were indeed only significantly related to an
ill-defined problem solving task and not to a well-defined problem solving task.
Another study by Bendixen and Schraw (2001) investigated how epistemological
beliefs affect written arguments about an ill-defined problem.
The hypermedia task used in the current study could be considered a well-
defined task and beliefs in certain and simple knowledge, therefore, would not
play a part. In addition, the form of assessment was well-defined in that it
focused on factual information. It could very well be that these epistemological
beliefs were not needed to solve the current studys task.
The ill-defined problem solving influence may also help explain why the
dimensions of metacognition (i.e., knowledge of cognition and regulation of
cognition) did not show a significant relationship with the hypermedia learning
environment. For example, students awareness of their own understanding may
not be an important influence in a well-defined task.
The short duration of the intervention (i.e., students were given 30 minutes to
read the 12 hypermedia pages) and the nature of the assessment may also explain
the lack of a relationship between metacognitive awareness and achievement.
Given more time and a more ill-defined task, students may have had more
opportunity to take advantage of some of the available resources. As it was, some
only had time to the read the text and/or they didnt deem it necessary to do
anything but read the text.
Combined, the previous explanations are in support of the current literature
on metacognition. It has been well-established that learning processes that have
become more automated and require less cognitive resources (e.g., reading for
comprehension and the recalling of factual information) would not necessarily
need to tap into metacognitive skills (Pressley & McCormick, 1995).

Implications

This study has important information for those who are designing and assessing
hypermedia environments and for educators involved in using technology to
enhance learning. Just because the task itself may be defined as hypermedia
(i.e., links to various forms of media) does not guarantee that it is necessarily
as non-linear and ill-defined a system as many have considered it (Jacobson &
Spiro, 1995). A cursory examination of the pathways participants took as they
navigated through the tutorial leads us to believe this may be the case. From this
preliminary analysis of log files, for example, it is clear that several of the
28 / BENDIXEN AND HARTLEY

participants read through the text only and did not use any additional hypermedia
tools available for the entire 30 minutes. Such information regarding hyper-
media environments should not be taken lightly when instructional design and
educational implications are being considered.
There are a number of educational implications stemming from the current
study. This study provides further evidence that what the student brings to the
learning task by way of abilities and epistemological beliefs have a significant
effect on learning outcomes, and vice versa. In other words, current evidence also
points to the idea that the types of learning and assessment tasks we offer students
may require certain types of abilities and beliefs for them to be successful.

Future Research
A number of questions have been raised by the current study that future research
could examine. In general, future research considering the unique aspects of
learning in a hypermedia context needs to further tease apart individual learner
differences, such as epistemological beliefs and metacognitive awareness, and
their role in learning. Specifically, future investigations could examine how
hypermedia environments can be presented in more ill-defined ways, the role of
epistemological beliefs and metacognition in them, and the subsequent impact
on learning.
Future research could also take a more detailed look at the pathways students
take while navigating through hypermedia learning environments. For instance:
What resources/hypermedia tools do students actually use?; For how long do they
use them?; and, more importantly, What do they gain from them? Along more
causal lines, future research could examine how epistemic beliefs impact problem
solving in hypermedia learning environments. For example: What deeper cogni-
tive processes are being tapped while students are engaged in learning with hyper-
media and how do they impact learning? Studies such as these could help us begin
answering important questions related to the educational potential of hypermedia.

REFERENCES
Bendixen, L. D., & Schraw, G. (2001, April). Why do epistemological beliefs affect
ill-defined problem solving? In L. D. Bendixen (Chair), Epistemological beliefs and
learning: What do we know and how do we know it? Symposium conducted at the American
Educational Research Association Annual Meeting, Seattle, WA.
Bendixen, L. D., Schraw, G., & Dunkle, M. E. (1998). Epistemic beliefs and moral
reasoning. The Journal of Psychology, 132:2, 187-200.
Brown, A. (1987). Metacognition, executive control, self-regulation, and other more
mysterious mechanisms. In F. Weinert & R. Kluwe (Eds.), Metacognition, motivation, and
understanding (pp. 65-116). Hillsdale, NJ: Erlbaum.
Brown, J. E., Bennett, J. M., & Hanna, G. (1981). The Nelson-Denny Reading Test,
Form E. Chicago, IL: The Riverside Publishing Company.
 SUCCESSFUL LEARNING WITH HYPERMEDIA / 29

Dillon, A., & Gabbard, R. (1998). Hypermedia as an educational technology: A review

of the quantitative research literature on learner comprehension, control, and style. Review
of Educational Research, 68, 322-349.
Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and
personality. Psychological Review, 95:2, 256-273.
Flavell, J. H. (1979). Metacognition and cognitive monitoring. American Psychologist,
34, 906-911.
Hartley, K. (2001). Learning strategies and hypermedia instruction. Journal of
Educational Multimedia and Hypermedia, 10, 285-305.
Hartley, K., & Bendixen, L. D. (2001). Educational research in the Internet age:
Examining the role of individual characteristics. Educational Researcher, 30:9,
22-26.
Hill, J. R., & Hannafin, M. J. (1997). Cognitive strategies and learning from the World
Wide Web. Educational Technology Research and Development,45:4, 37-64.
Hofer, B. K. (2001). Personal epistemology research: Implications for learning and
teaching. Journal of Educational Psychology Review, 13:4, 353-382.
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.
Jacobs, J. E., & Paris, S. G. (1987). Childrens metacognition about reading: Issues in
definition, measurement, and instruction. Educational Psychologist, 22, 255-278.
Jacobson, M. J., & Spiro, R. J. (1995). Hypertext learning environments, cognitive flexi-
bility, and the transfer of complex knowledge: An empirical investigation. Journal of
Educational Computing Research, 12, 301-333.
Jonassen, D., & Wang, S. (1993). Acquiring structural knowledge from semantically
structured hypertext. Journal of Computer Based Instruction, 20, 1-8.
King, P. M., & Kitchener, K. S. (1994). Developing reflective judgment: Understanding
and promoting intellectual growth and critical thinking in adolescents and adults.
San Francisco: Jossey-Bass.
Kitchener, K. S. (1983). Cognition, metacognition, and epistemic cognition. Human
Development, 26, 222-232.
Lee, Y., & Lehman, J. (1993). Instructional cueing in hypermedia: A study with active
and passive learners. Journal of Educational Media and Hypermedia, 2, 25-37.
Linn, R. L., & Gronlund, N. E. (2000). Measurement and assessment in teaching
(8th ed.). New Jersey: Prentice-Hall, Inc.
Perry, W. G. (1970). Forms of intellectual and ethical development in the college years:
A scheme. New York: Holt, Rinehart, and Winston.
Pressley, M., & McCormick, C. B. (1995). Advanced educational psychology for
educators, researchers, and policymakers. New York: HarperCollins College Publishers.
Puntambekar, D., & duBoulay, B. (1997). Design and development of MIST: A
system to help students develop metacognition. Journal of Educational Computing
Research, 16:1, 1-35.
Repman, J., Willer, H., & Lan, W. (1993). The impact of social context on learning
in hypermedia-based instruction. Journal of Educational Multimedia and Hypermedia,
2, 283-298.
Royer, J. M., Abranovic, W. A., & Sinatra, G. M. (1987). Using entering reading com-
prehension performance in college classes. Journal of Educational Psychology, 79, 19-26.
30 / BENDIXEN AND HARTLEY

Royer, J. M., Marchant, H. G. III, Sinatra, G. M., & Lovejoy, D. A. (1990). The
prediction of college course performance from reading comprehension performance:
Evidence for general and specific prediction factors. American Educational Research
Journal, 27, 158-179.
Schoenfeld, A. (1983). Beyond the purely cognitive: Belief systems, social cognitions,
and metacognitions as driving sources in intellectual performance. Cognitive Science, 7:4,
329-363.
Schommer, M. (1990). Effects of beliefs about the nature of knowledge on com-
prehension. Journal of Educational Psychology, 82, 498-504.
Schommer-Aikins, M. (2002). An evolving theoretical framework for an epistemo-
logical belief system. In B. K. Hofer & P. R. Pintrich (Eds.), Personal epistemology:
The psychology of beliefs about knowledge and knowing (pp. 261-275). Mahwah, NJ:
Erlbaum.
Schommer, M., Crouse, A., & Rhodes, N. (1992). Epistemological beliefs and mathe-
matical text comprehension: Believing it is simple does not make it so. Journal of
Educational Psychology, 82, 435-443.
Schraw, G. (1998). Promoting general metacognitive awareness. Instructional Science,
26:(1-2), 113-125.
Schraw, G., Bendixen, L. D., & Dunkle, M. E. (2002). Development and validation of
the Epistemic Belief Inventory (EBI). In B. K. Hofer & P. R. Pintrich (Eds.), Personal
epistemology: The psychology of beliefs about knowledge and knowing (pp. 261-275).
Mahwah, NJ: Erlbaum.
Schraw, G., & Dennison, R. S. (1994). Assessing metacognitive awareness. Contem-
porary Educational Psychology, 19, 460-475.
Schraw, G., Dunkle, M. E., & Bendixen, L. D. (1995). Cognitive processes in ill-defined
and well-defined problem solving. Applied Cognitive Psychology, 9, 523-538.
Shute, V. (1993). A comparison of learning environments: All that glitters. In S. Lajoie
& S. Derry (Eds.), Computers as cognitive tools (pp. 47-73). Hillsdale, NJ: LEA.
Zimmerman, J. (1990). Self-regulated learning and academic achievement: An over-
view. Educational Psychologist, 25:1, 3-17.

Direct reprint requests to:

Dr. Lisa D. Bendixen
Department of Educational Psychology
Box 453003
4505 Maryland Parkway
University of Nevada, Las Vegas
Las Vegas, NV 89154-3003
e-mail: libendixen@ccmail.nevada.edu