ArticlePDF Available

Dynamic Media in Computer Science Education; Content Complexity and Learning Performance: Is Less More?

Article

Dynamic Media in Computer Science Education; Content Complexity and Learning Performance: Is Less More?

Abstract and Figures

With the increasing use of dynamic media in multimedia learning material, it is important to consider not only the technological but also the cognitive aspects of its application. A large amount of previous research does not provide preference to either static or dynamic media for educational purposes and a considerable number of studies found positive, negative or even no effects of dynamic media on learning performance. Consequently, it is still necessary to discern which factors contribute to the success or failure of static or dynamic media. The study presented here can be seen as another brick in the wall of understanding students' learning supported by dynamic media. In this study, aspects of cognitive load and the ability to generate mental representations for the purpose of appropriate animation design and development are considered. The learning performance of static versus dynamic media amongst a total of 129 Computer Science students, including a control group, was investigated. The results showed that learning performance using dynamic media was significantly higher than those of the static textbook lesson when the learning material had a certain level of complexity; the more complex the learning material, the larger the benefit of using animations. The results were examined for possible factors that contributed to the success or failure of dynamic media in education. In conclusion, this study has successfully confirmed the theory that dynamic media can support learning when cognitive load and learners' mental representations are taken into account during the design and development of learning material containing dynamic media.
Content may be subject to copyright.
Holzinger, A., Kickmeier-Rust, M., & Albert, D. (2008). Dynamic Media in Computer Science Education; Content Complexity
and Learning Performance: Is Less More? Educational Technology & Society, 11 (1), 279-290.
279
ISSN 1436-4522 (online) and 1176-3647 (print). © International Forum of Educational Technology & Society (IFETS). The authors and the forum jointly retain the
copyright of the articles. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies
are not made or distributed for profit or commercial advantage and that copies bear the full citation on the first page. Copyrights for components of this work owned by
others than IFETS must be honoured. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior
specific permission and/or a fee. Request permissions from the editors at kinshuk@ieee.org.
Dynamic Media in Computer Science Education; Content Complexity and
Learning Performance: Is Less More?
Andreas Holzinger
Institute of Software Technology and Interactive Systems, Vienna University of Technology, Austria, Tel: +43 676
3122 673 // a.holzinger@computer.org
Michael Kickmeier-Rust
Department of Psychology, University of Graz, Austria // Tel: +43 316 380 8549 // michael.kickmeier@uni-graz.at
Dietrich Albert
Department of Psychology, University of Graz, Austria // Tel: +43 316 380 5118 // dietrich.albert@uni-graz.at
ABSTRACT
With the increasing use of dynamic media in multimedia learning material, it is important to consider not only
the technological but also the cognitive aspects of its application. A large amount of previous research does not
provide preference to either static or dynamic media for educational purposes and a considerable number of
studies found positive, negative or even no effects of dynamic media on learning performance. Consequently, it
is still necessary to discern which factors contribute to the success or failure of static or dynamic media. The
study presented here can be seen as another brick in the wall of understanding students’ learning supported by
dynamic media. In this study, aspects of cognitive load and the ability to generate mental representations for the
purpose of appropriate animation design and development are considered. The learning performance of static
versus dynamic media amongst a total of 129 Computer Science students, including a control group, was
investigated. The results showed that learning performance using dynamic media was significantly higher than
those of the static textbook lesson when the learning material had a certain level of complexity; the more
complex the learning material, the larger the benefit of using animations. The results were examined for possible
factors that contributed to the success or failure of dynamic media in education. In conclusion, this study has
successfully confirmed the theory that dynamic media can support learning when cognitive load and learners’
mental representations are taken into account during the design and development of learning material containing
dynamic media.
Keywords
Static media, Dynamic media, Animations, Learning performance, Cognitive load
Introduction and Motivation for Research
A common distinction of learning material refers to static (e.g., texts or images) and dynamic (e.g., animations or
simulations) media. Dynamic media are very popular and are almost omnipresent in today’s information and media
society. Dynamic media, moreover, is an increasingly important factor in educating the so called twitch speed
generation.
The term twitch speed generation was introduced by Prensky (2001) and basically means the under-30 generation,
whom he presumes to be capable of processing information faster and more in parallel than the generations before.
Prensky sees some reasons for this change in information processing in the increasing influence on children and
adolescents of speeding media such as MTV (music television) or computer games.
While initiated by such manifest considerations, the following research also resulted from a practical question: The
publisher (Vogel Wuerzburg, Germany) of the first author’s student textbooks (Holzinger, 2002) for computer
science students wanted to know, whether offering additional electronic material in the form of small multimedia
Learning Objects (LO) with these textbooks would benefit the students and, if so, to what extent. Naturally, the
cost/benefit ratio of such electronic learning material is also an issue for a publisher. However, in this study we
concentrate on aspects of learning performance of generic learning material, although we investigated the
technological aspects of such Learning Objects, their interoperability, reusability, and packaging on learning
platforms (see e.g. Holzinger, Nischelwitzer & Kickmeier-Rust (2006)).
280
Consequently, three main questions emerged: 1) Is there, principally, a discernable difference in learning
performance between electronic learning material, shown here in the form of minimal (“spartanic”) electronic
learning material (containing dynamic media), and printed matter (static media, including diagrams and pictures);
and 2) if there is a difference, what are these differences and how far do they extend? Finally, 3) what can we learn
from research, which can be applied to the design and development of such multimedia learning material? From this,
one can conclude the necessity of combining research in Human–Computer interaction (HCI), which delivers the
necessary findings, with Usability Engineering (UE), which ensures the appropriate practical application of the
results in development at systemic level (Holzinger, 2005).
A survey of the literature failed to provide clear guidelines as to which types of animations are suitable for which
information. Moreover, past research relied primarily on certain standard type animations (e.g., on how lightning
works, how a toilet tank works, or how brakes work etc.) and most of the experiments took place in a laboratory
setting (cf. Mayer, 2005), rather than in real world environments. In our opinion, contributing to the investigation of
these questions, research must be more context-orientated and both the psychological and developmental aspects
must be taken into account (Holzinger & Ebner, 2003, 2005).
Past Research
A considerable part of past research on the impact of different types of media on learning performance is based on
the simple distinction in dynamic and static media. Dynamic media can be divided into interactive (e.g., computer
simulation within which the learner can manipulate certain variables) and non-interactive media (e.g., animations or
movies). Ainsworth & van Labeke (2004) argued that the classification as either static or dynamic representation is
probably an insufficient level of granularity. A more detailed classification refers to aspects such as modality (text or
images), abstraction level (for example iconic or symbolic), sensory channel (auditory or visual), dimension (i.e., 2D
or 3D), and dynamism (static media versus dynamic media). Lowe (2004), additionally, identified three
characteristics of animations, transformations (which alter properties of objects such as size, shape and color),
translations (which move objects from one location to another), and transitions (which make objects disappear or
appear). Besides static representations, Ainsworth & van Labeke (2004), proposed three types of dynamic
representations.
The first type is a Time Persistent Representation (TPR), which is similar to static media; the only difference
between a representation within a textbook and an animation is that the dynamic representation displays information
incrementally rather than presenting the whole information from the beginning. Taking, for example, a time-series
graph and making it dynamic does not add any new information, although it may make certain features of that
information more pronounced.
The second type is Time Implicit Representation (TIR), which shows the relationship between variables over a period
of time. Unlike the TPR, the timescale of this information is only perceivable when presented dynamically. The rate
of change of the values being graphed is only visible when the animation is running and the representation is being
adjusted dynamically. Consequently, when a simulation is stopped, viewers must invoke internal representations to
compare current values to a particular previous state in order to answer questions about the timeline.
The third type, Time Singular Representation (TSR), displays one or more variables at a single instance of time. This
is the classical case of animation which is transitory. As TSRs display one state of the system, they contain less
information for each dimension of information in comparison to TPR and TIR. As a result, they are often used when
the information to be conveyed is highly complex and involves many interacting elements. Consequently, when an
animation does not move, the external representation contains only limited information, placing higher demands on
cognitive ability. For example, to compare current values to previous ones, viewers must hold previous states in their
memory in order to integrate them with new information; such activities are known to place much effort on working
memory (Hagafors & Brehmer, 1980).
Research on the impact of dynamic and static media on learning performance has a long tradition and is
consequently rich in empirical results. Generally, past research revealed that the mode of presenting learning
contents significantly affects learning processes and, therefore, learning performance (cf. Mayer, 2001). Often it was
assumed that dynamic media might be the most successful method for presenting learning content about complex
281
dynamic systems and that such dynamic media might significantly facilitate learning (e.g., Davies, 2002; Park &
Gittelman, 1992).
Past research, however, found little support for this assumption and research results showed that learning with static
media is equal or even superior to learning with dynamic media (cf. Mayer, Hegarty, Mayer, & Campbell, 2005;
Tversky, Morrison, & Betrancourt, 2002). For example, Mayer et al. (2005), reported that in a series of experiments,
learning with static media resulted in significantly higher, or at least equal, learning performance than with
animations.
On the basis of these results, one needs to consider the advisability of using dynamic media for displaying
information and for educational purposes. Past researchers, however, also argued that dynamic media might be
superior to static media for learners with low spatial abilities or with increasing complexity of learning content.
Furthermore, Narayanan & Hegarty (2002) argued that dynamic media might be superior when the learning content
deals with processes, which are not observable in the real world (in computer science, for example, computer
algorithms). The question as to which types of animations are suitable in which situations; for which contents and for
which learners still remains a matter of debate.
Theoretical Background
As mentioned above, previous research provided inconsistent and sometimes contradictory results on effects of
different types of media and their learning performance. A number of studies found dynamic media advantageous
(e.g., Kaiser, Proffitt, Whelan, & Hecht, 1992; Rieber, 1991), other studies found disadvantages in dynamic media
(e.g., Mayer et al., 2005; Rieber, 1990; Schnotz, Böckheler, & Grzondziel, 1999) and often no differences in learning
performance were found (e.g., Mayer et al., 2005; Price, 2002; Pane, Corbett, & John, 1996).
According to Ainsworth & van Labeke (2004), research on dynamic media effectiveness has produced those mixed
results for a variety of reasons: First, methods and measures used to investigate the impact of different types of
media varied (cf. also with Price, 2002); second, the types of animation used for research were mostly the same (as
mentioned within the introduction), and third, learners’ individual differences may contribute to different learning
performance with different media. For example, some learners tend to focus on the more obvious perceptual events
rather than on those that are of most conceptual interest.
Lowe (2003) argued that processing dynamic media may be problematic in two ways. First, learners may tend to
view animations only superficially and under-process them due to a lack of cognitive challenge or engagement.
Second, learners may be intensively occupied with animations, however, they may be unable to process the presented
material satisfactorily due to an excessive cognitive challenge or cognitive demands imposed. This excessive strain is
associated with aspects of animations such as high information load (particularly if the subject matter is complex)
and the temporally distributed nature of the presentation. Consequently, the different information (e.g., graphical
elements) may cause split-attention effects, that is, an increase of extraneous load when presenting related
information spatially or temporarily separated. Such effects have been investigated, for example, between pictures
and text (cf. Mayer & Moreno, 1998).
Ferguson & Hegarty (1995) characterize static diagrams as often being underspecified, a characteristic that can result
in problems for learners in the proper compartmentalization of the display into appropriate graphic elements
(features). Animations present the additional challenge of dividing the temporal sequence into events involving
features in the display that are thematically relevant (feature events). Moreover, in traditional animation, the material
is presented in a predetermined manner that plays from start to finish without the possibility of viewer control over
such aspects as the length of the animated sequence viewed; its pace and its direction. However, learners could be at
a disadvantage if their speed of comprehension cannot keep pace with the speed at which an animation presents its
information (Hegarty, Narayanan, & Freitas, 2002).
A further reason for inconsistent results may lie in the research methods. Often static media and dynamic media did
not display the same information or the same amount of information, dynamic media images were superior to those
of their static counterparts, maximum learning effects were reached and, in some cases, the dynamic media learning
282
procedures were superior to those of the static media (cf. Betrancourt & Tversky; 2000; Hegarty, Kriz, & Cate, 2003;
Lowe, 1999; Mayer et al., 2005; Tversky, Morrison, & Betrancourt, 2002). Even in cases where static media and
dynamic media are comparable both in content and visually, it does not necessarily follow that dynamic media is
superior to static media in terms of perceptual or cognitive demands. From a theoretical perspective, the most
prominent theoretical frameworks for explaining the effects of different learning modalities are the Theory of
Multimedia Learning (Mayer, 2001, 2005) and the Cognitive Load Theory (Paas, Renkl, & Sweller, 2003; Sweller,
1988, 1999, 2005; Valcke, 2002).
Mayer’s theory is based on three assumptions: First, it is assumed that visual and auditory information is processed
via different information channels. This assumption goes originally back to ideas of the Dual Coding Theory (Clark
& Paivio, 1991; Paivio, 1991; Paivio & Csapo, 1973) which states that human working memory consists of two
separate but interrelated channels for processing information: verbal channel and pictorial channel. Whilst these
channels can be activated independently, there are interconnections between the two systems allowing dual, and
therefore more efficient, coding of information (Rieber, 1994). The second assumption of Mayer’s theory is that the
processing capacity of each channel is limited. Only a small portion of information can be processed at one time.
Finally, the third assumption interprets learning as an active process, for example by constructing mental
representations of learning material and integrating it into existing previous knowledge (Holzinger, 2000).
Cognitive Load Theory (CLT) basically states that a learner’s attention and working memory is limited. This limited
amount of attention can be directed towards intrinsic, germane, or extraneous processing. Intrinsic processing
describes a learner’s focus on the learning content and its key features; it is determined by the intellectual demands
of learning content (the complexity of the content). Germane processing describes a deeper processing of the content
by its organization to cognitive representations and its integration into existing representations (integrating previous
knowledge). Finally, extraneous processing describes cognitive demands during learning, which do not foster the
actual objectives of the learning material, for example cross-references or navigation elements.
The problem with Cognitive Load Theory is, that it is rather difficult to quantify cognitive load of different
representations or events. However, on the basis of this theoretical framework, Mayer (2001) defined his static and
dynamic media hypotheses.
In brief, the static media hypothesis assumes that static media such as text or images facilitate learning processes by
shifting attention and processing capacities from extraneous processing to germane processing by providing only
relevant information (e.g., only important steps of a dynamic process) and by encouraging learners to construct
mental representations of the learning material. Moreover, static media allow learners to control the pace and order
of attended learning content. The dynamic media hypothesis, on the other hand, assumes that dynamic media such as
animations might facilitate learning by reducing extraneous load and encouraging germane processing by reducing
the efforts of constructing mental representations and by attracting interest and increasing motivation.
The Study
Initiated by our questions (see introduction) and on the basis of the inconsistent results of previous research (see
theoretical background) and with regard to the influence of static and (non-interactive) dynamic media on learning
performance, we compared the learning performance of a static textbook lesson (including static images), with the
learning performance of animations on a very basic level. The aim of our study was primarily to identify factors that
make dynamic representations of learning material beneficial for learning and to gain insight and gather knowledge
for design and development of such learning material. We used small textbook lessons and developed corresponding
animations, in order to provide a maximum of comparability between both representations and an isolated and
controlled variation of experimental variables. Moreover, we realized three different levels of complexity of learning
material.
Based on previous research we assumed that animations providing equal or lower cognitive load, dynamic media are
superior to static media. Measuring cognitive load, as mentioned before, is difficult. Therefore, we attempted to
operationalize the cognitive load concept by the simplicity of learning material design.
283
We hypothesized that replacing static pictorial information, which can only be understood when reading the related
text and constructing a mental representation of the concepts, with dynamic information representing a reasonable
mental model of the concepts, leads to higher learning performance. Additionally, we assumed that the more
complex the learning material is, the bigger is the advantage of learning with dynamic media in comparison to static
media. To investigate these hypotheses, we realized a 3 x 3 factorial design with three experimental conditions (static
media S, dynamic media D, and a control group C) and three levels of complexity (L1 – L3). The experiment was
conducted in a real-life classroom scenario.
Participants
The participants of the present study were N=129 undergraduate students of Computer Science from both Graz and
Vienna University of Technology. In total, there were 27 female and 102 male students. The average age was 21.60
years (SD = 2.14), the youngest participant was 18, the oldest 28 years of age. The first part of the participants (in
total 49) was recruited during the winter term 2002/03 and the last part (in total 80) during summer term 2007.
Materials
For the experiments, we used a small and clearly limited amount of generic learning content. We selected a one-page
textbook lesson (from Holzinger, 2002) on various topics in basic Information Technology, particularly we selected
signal transmission, which explains the fundamental principles of simplex, semi-duplex, and duplex transmission of
signals (Figure 1).
This example consisted of short, simply organized text paragraphs and three static images. This material also
included three different levels of complexity (i.e., simplex – low complexity, duplex – medium complexity, semi-
duplex – high complexity), whilst the general topic and the visual representation of these levels were the same.
Figure 1. The left panel illustrates the static textbook lesson used in the present experiment. The right panel
illustrates the corresponding dynamic representation
In addition we developed a time implicit animation (Ainsworth & van Labeke, 2004) using Macromedia Flash MX,
representing exactly the same content. To realize animations that are as comparable as possible to the static textbook
lesson we used the same terminology, the same information, and the same type of visualization. In the dynamic
284
representation we skipped only verbal and pictorial information (i.e., arrows in the static images) which directly
referred to signal transmission paths and replaced it with animated dots displaying the transmission paths (Figure 1,
right panel). The dots moved according the transmission direction at a pace of approximately 100 pixel per second
and the animation was looped during the whole presentation. In contrast to a number of previous similar studies, we
attempted to design the animation with a maximum of simplicity, meaning that we avoided designing appealing visual
gadgets (e.g., transmitters etc.).
To assess conceptual knowledge in this particular domain and to record learning performance, we created a multiple
choice test, which included three main topics, each assigned to each of the signal transmission modes. The three
items were chosen in order to represent three different levels of complexity, high (L1), medium (L2), and low (L3).
Each item had four answer alternatives, whereby none, one, or more alternatives could be correct. The score for each
item ranged from 0 to 4 and the guessing probability for each item, consequently, was 6.25 percent. This test was
used to assess previous knowledge in a pre-test and learning success in a post-test after the learning sessions. In the
pre-test, we additionally asked for biographic data (for example age or sex) and we recorded a subjective rating
regarding the participants’ expertise in the domain of communication technologies and three ratings regarding
learning preferences (“I like learning from textbooks”, “I like learning from computer animations”, and “I am rather
a visual learner type”).
All ratings were based on a five-category scale, ranging from “I absolutely agree” (1) to “I absolutely disagree” (5).
This scale type is appropriate for students in Austria because it corresponds to school grades (where 1 is the best and
5 the worst grade).
Procedure
At first, the participants were informed about the aim of the study and asked to rate four questions about their
experience in communication technologies and their learning preferences. In addition, the participants completed the
pre-test in order to assess previous knowledge. In the following, the participants were randomly assigned to one of
the three experimental conditions (static textbook lesson, dynamic animations, and control group).
Participants in the static image condition S were presented copies of the textbook paragraph (see Figure 1) taken
from Holzinger (2002). Participants in the dynamic animation condition D were seated in front of Personal
Computers (Standard IBM-PC’s, 17’’ screen, 1280 x 1024 pixel) and presented the learning material, containing the
Flash animations (Figure 1, right panel). The additional control group C received a one-page text on computer
systems, which was irrelevant for the domain of signal transmission. The participants were instructed to learn as
much as possible within a fixed time of eight minutes. The eight minutes interval was selected based on a pilot run
with experts and students in order to determine an average benchmark time sufficient to go through the learning
material. Moreover, this interval acknowledged real-life learning scenarios within which the time students can spend
on learning content is most often limited. After a short break of 15 minutes (to make sure that students do not have
the material in short term memory), all participants of the three experimental conditions were asked to complete the
post-test in order to assess their learning performance. Finally, the participants were debriefed. The complete
procedure took about 40 minutes.
Results
The main focus of the present study was learning performance with static (S) and dynamic (D) learning material. The
learning performance of these groups, in addition, was compared with the results of the control group (C). As
expected, in the pre-test scores no differences between the experimental conditions were found. In condition S
participants achieved an average total score (i.e., the sum of scores in L1, L2, and L3) of 7.66 (SD = 2.07), in
condition D a score of 7.34 (SD = 1.89), and in condition C a score of 6.86 (SD = 1.86).
As illustrated in Figure 2, equivalent results were found when the total score was divided into the three levels of
complexity (Figure 2). The pre-test mean score in condition S was 2.02 (SD = 1.31) for L1, 2.59 (SD = .81) for L2,
and 2.68 (SD = .76) for L3. In condition D the mean scores were 1.95 (SD = 1.08) for L1, 2.50 (SD = .88) for L2,
and 2.70 (SD = .82) for L3. Finally in the control group C the mean scores were 2.02 (SD = 1.23) for L1, 2.25 (SD =
285
.92) for L2, and 2.59 (SD = .82) for L3. An analysis of variance (ANOVA) yielded non-significant differences
between the levels of complexity L1 (F(2, 128) = .047, p = .954), L2 (F(2, 128) = 1.729, p = .182), and L3 (F(2, 128)
= .249, p = .780). A post-hoc Scheffé test, moreover, revealed that no significant differences were found for the three
experimental conditions (with p-values ranging from .210 to 1.000).
Confirming our initial hypotheses, in the post-tests different results were found. Condition D resulted in a higher
average total score in the knowledge test (10.80, SD = 1.09) than condition S (9.39, SD = 1.56) and, in turn,
condition S in a higher average total score than the control group C (6.95, SD = 1.76).
As illustrated in Figure 2, equivalent results were obtained for the three levels of complexity. For further analysis we
computed the learning performance (i.e., the differences between post-test scores and pre-test scores for each level of
complexity). In condition S the average learning performance was .85 (SD = 1.24) for L1, .54 (SD = 1.12) for L2,
and .46 (SD = 1.16) for L3. In condition D the average learning performance was 1.68 (SD = 1.27) for L1, 1.05 (SD
= 1.01) for L2, and 1.02 (SD = .95) for L3. Finally, in the control group C the average learning performance was .05
(SD = 1.29) for L1, .02 (SD = 1.15) for L2, and .02 (SD = .93) for L3.
An ANOVA yielded on the 5%-level significant differences in learning performance for all three levels of
complexity (L1: F(2, 126) = 18.321, p < .001; L2: F(2, 126) = 9.60, p < .001; L3: F(2, 126) = 10.70, p< .001).
A post-hoc Scheffé test revealed significant differences between conditions D and S (p = .013), D and C (p = .015),
and S and C (p < .001) for complexity level L1. For L2, significant differences were found between S and C (p <
.001), however, not between D and S (p = .105) as well as D and C (p = .101). Finally for lowest level of complexity
L3, significant differences were found between conditions S and D (p = .044) as well as D and C, between conditions
S and C (p = .140) no significant differences were found.
Summarizing these results, the more complex the learning material was (and therefore the complexity of the test
items), the larger were the advantages of learning with dynamic media (condition D) in comparison to condition S.
In the control group C no learning occurred (see Figure 2), as was expected. In order to verify the reliability of the
test items, supposed to assess the conceptual knowledge in this particular domain of signal transmission, we
computed Guttman’s split-half coefficient for the test scores in the control group C in pre-test and post-test, which
was .62 (α = 1.00 in both parts). This result indicates a reasonable reliability of the test items.
Figure 2. Mean test scores in pre-test and post-test for the three learning conditions (S, D, C) and three levels of
complexity (L1 – L3). The whiskers indicate the 95% confidence intervals of the means.
Besides the knowledge tests, the participants were required to rate the following four questions on a scale from 1 to 5
(where a rating of 1 means complete agreement and 5 means complete disagreement):
286
Question Q1 referred to the participants’ experience in communication technologies. The average rating was 2.75
(SD = 0.97).
Question Q2 referred to preferences for learning with static textbook lessons. The average rating was 2.78 (SD =
1.08).
Question Q3 referred to preferences for learning with multi-media material. Here the average rating was 2.70 (SD =
1.18).
Question Q4 referred to visual learning preferences. Here, the average rating was 2.29 (SD = .95).
Figure 3 illustrates the median (bold line), the range between 25th and 75th percentile (grey boxes), and minimum and
maximum ratings (whiskers) for each question.
Figure 3. Box plots of the ratings in the four questions regarding experience in communication technologies (Q1)
and preferences for static (Q2) and dynamic (Q3) learning material as well as regarding visual input preferences (Q4)
To analyze the relationships between the answers in these questions and test scores, we computed the Spearman
correlations between the ratings and the total learning performance. As summarized in Table 1, we found only small
to moderate correlations throughout; distinct correlations were significant on the 5% and the 1% level. These
correlations demonstrated that the ratings in the four questions are confounded to a certain extent. However, the
ratings in the four questions only moderately affected learning performance; higher experience in communication
technologies (a lower rating) resulted – quite consistently – in higher learning performance and higher preferences
for multi-media learning material (a lower rating) resulted in higher learning performance.
Table 1. Correlations between ratings in the questions regarding experience and preferences (Q1 – Q4) and learning
performance
Q1 Q2 Q3 Q4 Learning
Performance
Q1 1 -.022 .0125 .266** -.241**
Q2 -.022 1 -.338** -.099 .057
Q3 .125 -.338** 1 .330** -.191*
Q4 .266** -.099 .330** 1 -.062
Learning Performance -.241** .057 -.191* -.062 1
* Significant on the 5% level, ** significant on the 1% level
Discussion
Of course this study can only be a small part of a series of work aiming at finding another brick in the wall of
understanding human learning supported by dynamic media. Our goal was to identify factors (in the present study we
focused on simplicity of design and the complexity of learning material) that contribute to learning benefits through
287
dynamic media in order to gain insight on what to consider when designing and develop learning material including
such dynamic media.
However, a large number of studies reported that learning with dynamic representations does not improve learning
success per se; a few, contradictory, results were found which did not favor either static or dynamic media (cf.
Betrancourt & Tversky, 2000; Hegarty, et al., 2003; Lowe, 1999; Mayer et al., 2005; Tversky, et al., 2002). Within
our study, we focused on the isolated variation of content complexity and on reducing cognitive load with simple
animations. We hypothesized that when animations provide more simplicity in design (and consequently lower
cognitive load) they are superior to static media in terms of learning performance. Additionally, we assumed that the
more complex learning material the larger the difference in learning performance between static and dynamic media.
The results of our experiment provide some evidence for these assumptions.
In the present study, we found a considerable amount of previous knowledge in the relevant domain of Information
Technology (i.e., signal transmission). Across the three levels of complexity, performance in the pre-tests was above
50% (i.e., above a score of 2 in each multiple choice item). This performance is clearly above the guessing
probability. As shown in Figure 2, although a limited time interval was provided for learning and although
participants had a relatively high level of previous knowledge, both experimental conditions (S and D) resulted in a
significant gain of conceptual knowledge. Moreover, the present results confirm our assumptions about the impact of
static and dynamic media on learning. While, as expected, no differences were found in previous knowledge on the
topic of signal transmission in the pre-tests, the post-tests showed that animations resulted in significantly higher
learning performance than the static textbook lesson. The differences in learning performance increased with an
increasing level of complexity (Figure 2).
These results indicate that the complexity of learning material is an important factor for the success of learning
material containing dynamic media. The more complex the learning content, the higher the benefits of using
animations may be expected to be. This effect can be explained with the Cognitive Load Theory: In the case of
simple content, only little efforts have to be spent on intrinsic and germane processing and, therefore, a certain
amount of limited cognitive resources are available to process the verbal and/or visual representation of the content
(i.e., extraneous processing). However, with increasing intrinsic load, the available resources for extraneous
processing decrease and, consequently, the importance of appropriate representations increase. These results also
provide some further evidence for the assumptions of the Dynamic Media Hypotheses. Mayer (2001) argued that
dynamic media might facilitate the generation of mental representations of a process or a system and, additionally,
might attract a learner’s attention and interest. The animations developed for the present experiment offered a
representation of signal transmission modes which likely corresponds to a meaningful mental model of that
processes, because the animated dots display the continuous movement of a piece of information (a bit) from a
sender to a receiver.
However, what factors in the present experiment made the animations the more appropriate form of representation?
We argue that this, again, can be explained on the basis of Cognitive Load Theory. On the one hand, animated
sequences require more extraneous processing because, instead of a single image, a time-frame of a series of
individual images has to be perceived and processed. On the other hand, in the present experiment, we reduced
cognitive load by avoiding text paragraphs explaining signal transmission paths, which is probably more demanding
than just processing the animations of transmission directions.
Moreover, dynamic media are only appropriate and facilitate learning when they represent a meaningful mental
model of a process or a system. This representation must also be within the limits of the cognitive system, and it must
build upon learners’ previous knowledge and expertise.
One argument for the inconsistent results regarding static versus dynamic media was individual differences (see for
example Ainsworth & van Labeke, 2004; Price, 2002). In the present experiment we did not find indications for such
assumptions. As shown in Table 1, only small correlations were found between individual experience, preferences
for static or dynamic material, or visual input preferences and learning performance. Although the present work has a
limited scope and focuses on a specific domain and a specific form of representation, we are of the opinion that our
results and conclusions can be generalized.
288
Nonetheless, future research must extend beyond the limitations of this study, for example: the isolated and
controlled investigation of different types of animations and the complexity of content must be investigated. An
interesting issue for further research concerns the variations of the visualization complexity of a certain concept;
also, it will be necessary to provide designers and developers of multimedia learning material with rapid methods to
check – from the viewpoint of psychology – whether and to what extend, their material is appropriately designed for
the target end-users. This means that per se Psychology and Informatics must closely cooperate in order to integrate
findings from research into development at systemic level. Further work must also address interactive dynamic
media (e.g., interactive simulations) to find out whether interactivity increases the complexity of the representation
of learning material.
Conclusion
Although dynamic media and multiple media have become very present and popular in education, the existing
community of research in this domain recommends a cautious use of such media (Mayer et al., 2005; Rieber, 1990;
Schnotz, Böckheler, & Grzondziel, 1999). Dynamic media is only successful in facilitating learning in comparison to
traditional static media such as texts or images, when they are able to (1) reduce the cognitive load, which is
necessary to comprehend them, (2) serve to generate mental models of a concept and, consequently (3), offer
visualizations that correspond to a meaningful mental model.
Moreover, dynamic media must be attuned to learners’ experience, expertise, and most of all previous knowledge.
Therefore, material containing dynamic media must avoid information, animations, and elements, which are not
necessary to comprehend a concept. As an example, the signal transmission visualization, as in the present
experiment, could have been made far more complex by using appealing graphics, more attractive signal
representations, or more appealing background images, colors etc. Such information, however, does not serve the
comprehension and the learning process and, not necessarily, the generation of appropriate mental models. The more
complex and difficult the learning content is, the more important it is to direct cognitive and perceptual resources to
intrinsic and germane processing. Finally, this study has successfully confirmed the theory that dynamic media can
support learning when limited cognitive resources, cognitive load, and learners’ mental representations are taken into
account during the design and development of learning material containing dynamic media.
Acknowledgement
We are grateful to all students who participated in this study. The first author expresses his gratitude to Professors
Rudi Freund & A Min Tjoa, who made this research possible during a Visiting Professorship at Vienna University of
Technology, Faculty of Informatics, Institute of Software Technology & Interactive Systems and for the financial
support of this Professorship by SIEMENS Austria. In addition we highly appreciated the suggestions by the
anonymous reviewers.
References
Ainsworth, S., & van Labeke, N. (2004). Multiple forms of dynamic representation. Learning and Instruction, 14,
241–255.
Betrancourt, M., & Tversky, B. (2000). Effect of computer animation on users' performance: A review. Travail
Humain, 63 (4), 311–329.
Clark, J.M., & Paivio, A. (1991). Dual coding theory and education. Educational Psychology Review, 71, 64–73.
Davies, C.H.J. (2002). Student engagement with simulations: a case study. Computers & Education, 39 (3), 271–
282.
289
Ferguson, E.L., & Hegarty, M. (1995). Learning with real machines or diagrams: Application of knowledge to real-
world problems. Cognition and Instruction, 13,129–160.
Hagafors, R. & Brehmer, B. (1980) Effects of Information Presentation Mode and Learning-Paradigm on the
Learning of Probabilistic Inference Tasks under Different Levels of Memory Strain. Scandinavian Journal of
Psychology, 21 (4), 249–255.
Hegarty, M., Kriz, S., & Cate, C. (2003). The roles of mental animations and external animations in understanding
mechanical systems. Cognition and Instruction, 21, 325–360.
Hegarty, M., Narayanan, N.H., & Freitas, P. (2002). Understanding machines from multimedia and hypermedia
presentations. In J. Otero, A.C. Graesser, & J. Leon (Eds.), The psychology of science text comprehension, Mahwah,
NJ: Lawrence Erlbaum, 357–384.
Holzinger, A. (2002). Basiswissen IT/Informatik, Band 1: Informationstechnik. Das Basiswissen für die
Informationsgesellschaft des 21, Jahrhunderts, Wuerzburg, Germany: Vogel, retrieved January 15, 2008 from
http://www.basiswissen-it.at.
Holzinger, A. (2000). Basiswissen Multimedia Band 1: Technik. Technologische Grundlagen multimedialer
Informationssysteme, 2
nd Ed., Würzburg, Germany: Vogel Buchverlag, retrieved January 15, 2008 from
http://www.basiswissen-multimedia.at.
Holzinger, A. (2005). Usability engineering for software developers. Communications of the ACM, 48 (1), 71–74.
Holzinger, A., & Ebner, M. (2003). Interaction and usability of simulations & animations: A case study of the Flash
technology. In Rauterberg, M., Menozzi, M. & Wesson, J. (Eds.), Human-Computer Interaction. IFIP Interact 2003,
Amsterdam: IOS Press, 777–780.
Holzinger, A., & Ebner, M. (2005). Visualization, animation and simulation for mobile computers: Experiences from
prototypes. In Zara, J. & Sloup, J (Eds.), Proceedings of the 2nd Central European Multimedia and Virtual Reality
Conference. Prague: Czech Technical University, 37–41.
Holzinger, A., Nischelwitzer, A. K., & Kickmeier-Rust, M. (2006), Pervasive E-Education supports Life Long
Learning: Some Examples of X-Media Learning Objects. Paper presented at the 10th IACEE World Conference on
Continuing Engineering Education (WCCEE), retrieved January 15, 2008 from
http://www.wccee2006.org/papers/445.pdf.
Kaiser, M.K., Proffitt, D.R., Whelan, S.M., & Hecht, H. (1992). The influence of animation on dynamical
judgments. Journal of Experimental Psychology: Human Perception and Performance, 18, 669–690.
Lowe, R. K. (1999). Extracting information from an animation during complex visual learning. European Journal of
Psychology of Education, 14 (2), 225–244.
Lowe, R.K. (2003). Animation and learning: selective processing of information in dynamic graphics. Learning and
Instruction, 13, 157–176.
Lowe, R.K. (2004). Interrogation of a dynamic visualization during learning. Learning and Instruction, 14 (3), 257–
274.
Mayer, R.E. (2001). Multimedia learning, Cambridge (MA): Cambridge University Press.
Mayer, R.E. (2005). The Cambridge handbook of multimedia learning, Cambridge: Cambridge University Press.
Mayer, R.E., Hegarty, M., Mayer, S., & Campbell, J. (2005). When static media promote active learning: Annotated
illustrations versus narrated animations in multimedia instruction. Journal of Experimental Psychology, 11 (4), 256–
265.
290
Mayer, R.E., & Moreno, R. (1998). A split-attention effect in multimedia learning: Evidence for dual processing
systems in working memory. Journal of Educational Psychology, 90 (2), 312–320.
Narayanan, N.H., & Hegarty, M. (2002). Multimedia design for communication of dynamic information.
International Journal of Human-Computer Studies, 57 (4), 279–315.
Paas, F., Renkl, A., & Swell, J. (2003). Cognitive load theory and instructional design: Recent developments.
Educational Psychologist, 38, 1–4.
Paivio, A. (1991). Dual coding theory: retrospect and current status. Canadian Journal of Psychology, 45 (3), 255–
287.
Paivio, A., & Csapo, K. (1973). Picture superiority in free recall: Imagery or dual coding? Cognitive Psychology, 5
(2), 176–206.
Pane, J.F., Corbett, A.T., & John, B.E. (1996). Assessing dynamics in computer-based instruction. In M.J. Tauber
(Ed.), Proceedings of the ACM Conference on Human Factors in Computing Systems, Vancouver: ACM, 797–804.
Park, O.C., & Gittelman, S.S. (1992). Selective use of animation and feedback in computer-based instruction.
Educational Technology, 40, 125–167.
Prensky, M. (2001). Digital game-based learning, New York: McGraw-Hill.
Price, S.J. (2002). Diagram representation: The cognitive basis for understanding animation in education (TR 553),
School of Computing & Cognitive Sciences, University of Sussex, UK, retrieved January 15, 2008 from
http://www.informatics.susx.ac.uk/interact/papers/pdfs/csrp553.doc.
Rieber, L.P. (1990). Using computer animated graphics in science instruction with children. Journal of Educational
Psychology, 82 (1), 135–140.
Rieber, L.P. (1991). Animation, incidental learning, and continuing motivation. Journal of Educational Psychology,
83 (3), 318–328.
Rieber, L.P. (1994). Computers, graphics and learning, Madison, WI: WCB Brown & Benchmark.
Schnotz, W., Böckheler, J., & Grzondziel, H. (1999). Individual and co-operative learning with interactive animated
pictures. European Journal of Psychology of Education, 14, 245–265.
Sweller, J. (1988) Cognitive load during problem solving: Effects on learning. Cognitive Science, 12 (2), 257–285.
Sweller, J. (1999). Instructional design in technical areas, Camberwell: ACER Press.
Sweller, J. (2005). Implications of cognitive load theory for multimedia learning. In R.E. Mayer (Ed.), Cambridge
handbook of multimedia learning, Cambridge: Cambridge University Press, 19–30.
Tversky, B., Morrison, J.B., & Betrancourt, M. (2002). Animation: can it facilitate? International Journal of Human-
Computer Studies, 57 (4), 247–262.
Valcke, M. (2002). Cognitive load: updating the theory? Learning and Instruction, 12 (1), 147–154.
... This can be attributed to the field's emphasis on hands-on practice [1][2][3][4] and requirement for students to constantly solve problems through programming. These tasks require students to bridge the intellectual gap between the abstract concepts they learn in class and their concrete implementations, [5,6], a task with which novices in particular have difficulty [7,8]. Presenting both abstract and concrete information in a tightly integrated, interactive form is done with the intent of allowing students to learn these relationships through experimentation with interactive elements. ...
... Much of the prior work on digital textbooks and interactive visualizations are relevant to investigations on interactive computing textbooks. Some notable points of focus include comparing the mediums of textbooks (e.g digital vs paper), modeling textbook knowledge concepts, modeling student reading patterns, and investigating student engagement with interactive visualizations [5,6,[13][14][15][16]. These efforts contribute to our understanding of how to build adaptive learning modules, promote personalized learning, and examine techniques that model student reading. ...
Conference Paper
Full-text available
Interactive textbooks have great potential to increase student engagement with the course content which is critical to effective learning in computing education. Prior research on digital textbooks and interactive visualizations contributes to our understanding of student interactions with visualizations and modeling textbook knowledge concepts. However, research investigating student usage of interactive computing textbooks is still lacking. This study seeks to fill this gap by modeling student engagement with a Jupyter-notebook-based interactive textbook. Our findings suggest that students' active interactions with the presented interactive textbook , including changing, adding, and executing code in addition to manipulating visualizations, are significantly stronger in predicting student performance than conventional reading metrics. Our findings contribute to a deeper understanding of student interactions with interactive textbooks and provide guidance on the effective usage of said textbooks in computing education.
... Namun kenyataannya, masih banyak siswa yang beranggapan matematika sebagai pelajaran yang tidak menyenangkan, sulit, bahkan ditakuti, dan dampaknya matematika menjadi kurang diminati oleh siswa (Liu & Lin, 2019). Untuk mewujudkan pendidikan Matematika yang baik guru memerlukan media pembelajaran yang menarik dan kreatif untuk meningkatkan minat belajar dan konsentrasi siswa di dalam kelas (Holzinger et al., 2008). Tidak ada media pembelajaran yang menjadi penghambat proses pembelajaran, Maka dituntutnya untuk melakukan pengembangan sebuah media pembelajaran yang dapat mendukung dalam proses pembelajaran dan meningkatkan kualitas pembelajaran (Noh et al., 2015). ...
Article
Materi matematika yang selalu menjadi mata pelajaran yang menakutkan bagi siswa belum ada yang dikemas dalam media pembelajaran Motion Graphic. Padahal proses pembelajaran dengan menggunakan video Motion Graphic lebih menarik perhatian siswa, dapat meningkatkan fokus dan semangat belajar siswa. Oleh karena itu, peneliti memilih mengembangkan media pembelajaran Motion Graphic berbasis Android sebagai solusi permasalahan yang ada. Melalui pengembangan media pembelajaran video animasi Motion Graphic berbasis Android ini peneliti bertujuan untuk membuat peserta didik akan lebih semangat lagi untuk belajar matematika. Penelitian ini bertujuan untuk mengetahui pengaruh penggunaan media pembelajaran Motion Graphic matematika berbasis android pada pelajaran matematika. Metode yang dipakai metode Research and Development (penelitian dan pengembangan), tapi lebih difokuskan pada proses pengembangan menggunakan metode ADDIE. Tahapan pengembangan dalam penelitian ini diawali dengan menganalisis kebutuhan, merancang, membuat media pembelajaran Motion Graphic berbasis Android, dan dilanjutkan dengan uji pakar. Hasil penelitian menunjukkan bahwa media yang dikembangkan layak dan memenuhi syarat untuk digunakan sebagai media pembelajaran Motion Graphic matematika berbasis Android pada siswa kelas V SD semester 1. Hal ini dilihat dari hasil validasi oleh ahli media sebesar 87,5% dinyatakan baik. Lalu hasil validasi oleh ahli materi sebesar 87,5 % dinyatakan baik. Setelah dilakukan uji validasi oleh para ahli kita lakukan uji coba kepada kelompok kecil mendapatkan hasil 95% dengan kategori sangat baik. Berdasarkan hasil tersebut dapat disimpulkan bahwa media pembelajaran Motion Graphic matematika berbasis Android ini layak digunakan dalam pembelajaran.
... Multimedia animation can contain words, pictures, sounds, images, and moving pictures [49]. Previous studies indicated that animation improves a learner's ability to remember facts and process information [50,51]. With sound, action, and images, animation can elucidate complex abstract concepts for learners' pictures [49]. ...
Article
Full-text available
This study examines the process of creating successful, engaging, interactive, and activity-based online educational materials, while taking the cognitive aspects of learners into account. The quality of online educational materials has become increasingly important in the recent period, and it is crucial that content is created that allows our students to learn effectively and enjoyably. In this paper, we present the milestones of curriculum creation and the resulting model, the criteria of selecting online learning environments, technical requirements, and the content of educational videos, interactive contents, and other methodological solutions. In addition, we also introduce some principles of instructional design, as well as a self-developed model that can be used to create effective online learning materials and online courses. There was a need for a self-developed, milestone-based, practice-oriented model because the models examined so far were too general and inadequate to meet the needs of a decentralized developer team, who work on different schedules, with significant geographical distances between them and do not place enough emphasis on taking cognitive factors into account. In these processes, special attention should be paid to having a clear and user-friendly interface, support for individual learning styles, effective multimedia, ongoing assistance and tracking of students’ progress, as well as interactivity and responsive appearance.
... The Simplicity principle means that a good Gestalt organization is a simple organization (Stickel, Ebner, and Holzinger, 2010). Visual complexity can be considered as extraneous load (Schmutz et al., 2009), influencing the cognitive load and mental effort of the viewer (Holzinger, Kickmeier-Rust, and Albert, 2008;Harper, Michailidou, and Stevens, 2009) and his or her emotions (Tsai et al., 2008). Applying the new approach in database courses offers students a strategy for better coping with complex textual descriptions of organizational requirements and constraints. ...
Article
Conceptual modeling of databases is a complex cognitive activity, particularly for novice database designers. The current research empirically tests a new pedagogy for this activity. It examines an instructional approach that stresses visualizing gradual transitions between levels of abstraction in different hierarchic levels of a relational database schema. The new approach builds on a four-level TSSL model from the field of human-computer interaction. TSSL, an acronym for the Task, Semantics, Syntax, and Lexical levels, is applied here to describe the levels of conceptual database modeling and to explain how improved instructional design can help minimize extraneous cognitive load during the design of database schemas. We tested the effectiveness of the proposed instructional approach via a controlled experiment carried out on IS students. We divided students into two groups, those exposed to a visual emphasis on the syntax of gradual transitions in a schema structure and those not exposed to it. We then measured performance in terms of errors in students' solutions while also recording their perceptions and attitudes toward the instructional approach and the activity of database modeling. Our results show that the new approach is an effective tool for teaching database modeling.
... An implicit hypothesis here is that, for example, a versatile use of learning materials and social study habits are beneficial and productive for studying mathematics in university, cf. Holzinger, Kickmeier-Rust, & Albert (2008) and Blasco-Arcas, Buil, Hernández-Ortega, & Sese (2013). ...
Article
Full-text available
In this article, we report from a group (N=98) of students from two campuses of one Finnish university, on their study habits, and to what extent they use different kinds of learning materials in university mathematics courses. Our results show that the older students are more communicative with their teachers, whereas the younger students ask for help more often from fellow students. The sociomathematical norms that constitute the local study culture have a significant impact on the study habits and on the use of learning materials. For example, the use of videos and studying lecture materials before the lectures were clearly more usual at one campus than at the other. We also found some significant differences between the groups that are based on the study programmes. The students of mathematics without an intention to become a teacher were most traditional in their study habits, whereas the students of applied physics were most active to participate in teaching. The student teachers most often lie in the middle in the issues where the other groups differ from one another. Quite unexpectedly, students' previous performance in upper secondary school does not explain the differences in the study habits in the university.
Article
Full-text available
E-textbooks and e-learning technologies have become ubiquitous in college and university courses as faculty seek out ways to provide more engaging, flexible, and customizable learning opportunities for students. However, the same technologies that support learning can serve as a source of frustration. Research on frustration with technology is limited, especially in educational settings. This study examined student frustration with e-textbooks and the factors contributing to the frustration within undergraduate general biology courses through the development of an E-Text Frustration scale. Exploratory factor analysis of the E-Text Frustration scale revealed a three-factor structure that provides quantified support for frustration with (1) e-textbook interactions on the screen, (2) problems with technology, and (3) e-text curriculum integration. This structure was supported by a confirmatory factor analysis. The construct validity of the scale was established using a correlation analysis that revealed significant relationships among the three e-text frustration measures, cognitive load, and motivation variables. Furthermore, the measurement invariance analyses indicated that the scale measures the same construct in the same way in males and females. Overall, the study findings suggest that the E-text Frustration scale is a useful instrument with high reliability and validity evidence that can be used by researchers and practitioners. Implications for future research on frustration with e-learning technologies are discussed.
Article
Spatial thinking skills are crucial for success in any of the STEM (science, technology, engineering, and mathematics) domains, and they are malleable. One approach to support the development of student spatial skills is through the use of innovative technologies, like the augmented reality (AR) sandbox, that can also effectively teach geoscience content in the process. In this study, we aimed to create a student-informed spatial topographic map activity designed to emphasize mental rotation, spatial orientation, and spatial visualization skills using the AR sandbox that incorporated elements such as drawing topographic profiles and recognizing stream flow direction. Furthermore, this study explored the spatial reasoning beliefs and challenges of undergraduate students at a large-enrollment Southeastern US university. Both quantitative and qualitative measures were employed to better understand student performance on and challenges with the topographic map and spatial tasks. Overall, the students found spatial visualization tasks in the activity to be the most challenging, and they were least confident in their spatial visualization skills. However, they believed that their spatial visualization skills were most improved after completing the topographic map activity, and those activities were reported to be the most effective at getting them to think in spatial terms. These results highlight that multi-step mental manipulations required to perform spatial visualization tasks are of great interest to instructors when developing topographic map activities using the AR sandbox. With more investigation, the AR sandbox has the potential to aid in the development of students’ spatial visualization skills while simultaneously teaching geological content. Supplemental data for this article is available online at https://doi.org/10.1080/10899995.2021.1969862
Article
Full-text available
Animation films were used frequently during online didactic activities during the COVID-19 pandemic. We have not identified research that focuses on the problems that teachers face when using animated films to enable students’ learning of Natural Sciences and on the training models that can be organized based on this resource. Therefore, we analysed the problems faced by teachers and, to solve them, we identified, analysed, and described several instruction models based on animation films: illustrative, expositive, interactive, flipped classroom, learning through discovery, and learning through simulation. We analysed the way to integrate them into the didactic activity and, finally, we realized some recommendations.
Chapter
In conventional classroom, when a student’s progress is stymied by the lack of direct contact with his or her instructor and peers, mobile devices offer new forms of communication that break down the traditional barriers to education. Personal digital assistant (PDA) has been a popular choice – more so than cell phone – for implementing m-learning in the last few years. However, how to design a collaborative learning system based on PDA to create an appropriate interface for content presentation on small screen in order to improve performance of learning is the major objective of this study. The authors used principles of human-computer interaction (HCI) and utilized task analysis, literature review, experimental design, small-screen design to construct a collaborative learning system based on PDA, and addressed potential issues with appearance in order to design a friendly PDA interface whose performance was analyzed with a post-experiment user satisfaction questionnaire. The results of this experiment show that the principles of Interface design based on HCI used in design of the learning system would help students to improve performance of learning and become capable, self-reliant, self-motivated and independent. The findings also demonstrate that variations among students’ preference or content are associated with differences in the display presentation mode in which the landscape interface is recommended for use.
Article
Full-text available
Computer-based learning environments provide the possibility to present interactive animated pictures, which can be manipulated for active exploratory learning and which allow to display the dynamic behavior of a complex subject matter. Due to the large range of possibilities of exploratory interaction, such learning environments seem to be well suited for cooperative learning, where different learners analyse a subject matter from different perspectives. Knowledge acquisition from interactive animated pictures was compared with knowledge acquisition from static pictures in two empirical studies under the conditions of individual learning (Study I) and of cooperative learning (Study II). In Study I, learning with interactive animated pictures resulted in a better encoding of detail information, but did not have positive effects on performance in mental simulation tasks. In Study II, learning with interactive animated pictures resulted both in lower encoding of detail information and poorer results in mental simulations. These findings and the analysis of discourse protocols of the co-operation suggest that exploratory learning with interactive animated pictures is associated with extraneous cognitive load, which can be further increased by the co-ordination demands of co-operative learning. Although animated pictures provide external support for mental simulations, they seem to be not generally beneficial for learning, as they can prevent individuals from performing relevant cognitive processes.
Book
In recent years, multimedia learning, or learning from words and images, has developed into a coherent discipline with a significant research base. The Cambridge Handbook of Multimedia Learning is unique in offering a comprehensive, up-to-date analysis of research and theory in the field, with a focus on computer-based learning. Since the first edition appeared in 2005, it has shaped the field and become the primary reference work for multimedia learning. Multimedia environments, including online presentations, e-courses, interactive lessons, simulation games, slideshows, and even textbooks, play a crucial role in education. This revised second edition incorporates the latest developments in multimedia learning and contains new chapters on topics such as drawing, video, feedback, working memory, learner control, and intelligent tutoring systems. It examines research-based principles to determine the most effective methods of multimedia instruction and considers research findings in the context of cognitive theory to explain how these methods work.
Article
The effects of computer animations and mental animation on people's mental models of a mechanical system are examined. In 3 experiments, students learned how a mechanical system works from various instructional treatments including viewing a static diagram of the machine, predicting motion from static diagrams, viewing computer animations, and viewing static and animated diagrams accompanied by verbal commentaries. Although students' understanding of the system was improved by viewing both static and animated diagrams, there was no evidence that animated diagrams led to superior understanding of dynamic processes compared to static diagrams. Comprehension of diagrams was enhanced by asking students questions that required them to predict the behavior of the machine from static diagrams and by providing them with a verbal description of the dynamic processes. This article proposes that predicting motion from static diagrams engages students' mental animation processes, including spatial visualization, and provides them with information about what they do and do not understand about how the machine works. Verbal instruction provides information that is not easily communicated in graphics and directs students' attention to the relevant information in static and animated diagrams. The research suggests that an understanding of students' mental animation abilities is an important component of a theory of learning from external animations.
Article
The effects of animated presentations and practice were studied in a computer-based science lesson involving fourth-and fifth-grade students. Three levels of visual elaboration (static graphics, animated graphics, and no graphics) were crossed with three levels of practice (behavioral, cognitive, and no practice). Behavioral practice consisted of traditional questioning and cognitive practice consisted of a structured simulation. Animated graphics were superior to static graphics and no graphics so long as practice was provided. Behavioral practice was effective only when paired with lessons containing animated graphics. Cognitive practice was generally superior to the other practice conditions and did not appear dependent on visual elaboration. These results suggest that animated presentations can promote learning under certain conditions, and they also demonstrate a successful application of interactive graphics in the design of cognitively based practice activities.
Article
Free verbal recall is generally higher for items presented as pictures than for items presented as words. Possible interpretations of this effect include differential verbal elaboration, superiority of nonverbal imagery as a memory code, and dual encoding favoring pictures. A series of experiments investigated the relative contributions of imaginal and verbal memory codes using incidental recall tasks in which the orienting task was designed to control the way items are encoded during input. Three experiments required subjects to encode words and pictures verbally, by writing or pronouncing the words or picture labels; or imaginally, by drawing or imaging the picture or object suggested by the word. Two further experiments involved a probability learning task which required no encoding reaction to pictures or words, but an analogue of imaginal and verbal coding was provided in one experiment using picture-picture, picture-word (or word-picture) and word-word repetitions. Recall tests following these manipulations consistently yielded much higher recall for pictures than for words under all conditions except when subjects imaged to words. Strong support was also found for the hypothesis that image and verbal memory codes are independent and additive in their effect on recall. In addition, the contribution of imagery appeared to be substantially higher than that of the verbal code. Thus the usual superiority of pictures in free recall is best explained by dual encoding, or a combination of image superiority and dual coding, both of which are ordinarily favored when items are presented as pictures.
Article
Students viewed a computer-generated animation depicting the process of lightning formation (Experiment 1) or the operation of a car's braking system (Experiment 2). In each experiment, students received either concurrent narration describing the major steps (Group AN) or concurrent on-screen text involving the same words and presentation timing (Group AT). Across both experiments, students in Group AN outperformed students in Group AT in recalling the steps in the process on a retention test, in finding named elements in an illustration on a matching test, and in generating correct solutions to problems on a transfer test. Multimedia learners can integrate words and pictures more easily when the words are presented auditorily rather than visually. This split-attention effect is consistent with a dual-processing model of working memory consisting of separate visual and auditory channels.