The Influence of Media on Learning: The

Debate Continues
SLMQ Volume 22, Number 4, Summer 1994

Robert B. Kozma, Director, Center for Technology in Learning, SRI International

Do media influence learning? Perhaps it is time to rephrase the question: How, do media affect
learning? Perhaps it is time to go beyond our concern with “proving” that media “cause” learning
so that we can begin to explore the question in more complex ways. Perhaps we should ask, what
are the actual and potential relationships between media and learning? Can we describe and
understand those relationships? And can we create a strong and compelling influence of media
on learning through improved theories, research, and instructional designs?

There is a certain urgency about this question. In the near future, telephone, cable television, and
digital computer technologies will merge,(5) presenting the prospect of interactive video
integrated with large multimedia databases to be distributed to people in various settings all over
the world. If we do not soon understand the relationship between media and learning—if we
have not forged such a relationship this technology may be used primarily for interactive soap
operas and online purchases of merchandise. Its educational uses may be driven primarily by
benevolent movie moguls who design “edutainment” products whose contribution to learning
may be minimal.

In order to understand the actual and potential relationships between media and learning, we
must first understand why we have thus far failed to establish a causal connection. In large part,
this failure is due to the fact that our theories, research, and designs have been constrained by
vestiges of the behavioral roots of instructional technology.(6) Both traditional instructional
design models and comparative media studies rest on the assumptions of the behaviorist
paradigm: media “stimuli” are described according to the surface features of their technologies,
and their effect on learning is assessed by using “responses” on a test. Missing from this
approach are any descriptions of the cognitive, affective, or social processes by which learning
occurs. Also missing are descriptions of the underlying structures and functions of various media
that influence these processes. Clark’s delivery truck is an apt metaphor for this approach. The
medium itself is only an inert conveyer of an active stimulus to which the learner makes a
behavioral response.

But today we understand that learning is not simply a passive response to instruction’s
“delivery.” Rather, learning is an active, constructive, cognitive, and social process by which the
learner strategically manages available cognitive, physical, and social resources to create new
knowledge by interacting with information in the environment and integrating it with
information stored in memory.(7) From this perspective, knowledge and learning are the result of
a reciprocal interaction between the learner’s cognitive resources and aspects of the external
environment.(8) Moreover, this interaction is strongly influenced by the extent to which internal
and external resources fit together.(9)
Consequently, to understand the role of media in learning, we must fundamentally change our
traditional approach to this issue:

• We must ground a theory of media in the cognitive and social processes by which
knowledge is constructed.
• We must define media in ways that are compatible and complementary with these
processes.
• We must conduct research on the mechanisms by which characteristics of media might
interact with and influence these processes.
• We must design our instruction in ways that embed the use of media in these processes.

A New Look at the Question

Early attempts to review research findings in light of the new assumptions noted above suggest
that “learning with media [is] a complementary process within which representations are
constructed and procedures performed, sometimes by the learner and sometimes by the
medium.” Media embody certain characteristics that “interact with learner and task
characteristics to influence the . . . structure, formation, and modification of mental models.”(10)
Within this framework, particular media formats (e.g., books and magazines, video media,
computer software, and multimedia) possess particular characteristics that make them both more
and less suitable for the accomplishment of certain kinds of learning tasks.

Gavriel Salomon argued that media can be analyzed in terms of their “cognitively relevant”
capabilities—i.e., in terms of those characteristics that affect the ways in which individuals
represent and process information.(11) These capabilities relate to three aspects of each medium:
its technology, symbol system(s), and processing capabilities. “Technology” refers to the
physical, mechanical, or electronic capabilities that determine a medium’s function. “Symbol
systems” are sets of symbolic expressions by which information is communicated(12) according
to specific rules and conventions: spoken language, printed text, pictures, numbers, graphs, and
musical scores exemplify symbol systems. “Processing capabilities” refer to a medium’s abilities
to operate on symbol systems in specified ways-for example, by displaying, receiving, storing,
retrieving, organizing, transforming, or evaluating whatever information is available through a
particular symbol system.

Each medium can be defined and distinguished from others by a profile of these three kinds of
capabilities. Using this profile, a particular medium can be described in terms of how it presents
certain representations and performs certain operations in interaction with learners—who are
simultaneously constructing and operating on mental representations. From this perspective,
then, learning with media is a complementary process within which a learner and a medium
interact to expand or refine the learner’s mental model of a particular phenomenon. The question
then becomes not do media enhance learning but how, do the capabilities of a particular medium
facilitate particular kinds of learning?
Learning with Books
The most common medium encountered in school learning is still the book. As a learning
medium, the book can be characterized by the primary feature of its technology (that is,
stability), by its symbol systems (printed text, pictures, and graphics), and by the way it
influences specific processes (reading).

The primary symbol system used in books and other print media consists of orthographic
symbols that, in Western culture, are words composed of phonemic graphemes, horizontally
arrayed from left to fight. In most printed school media, this arrangement is stable—unlike the
marquee in Times Square, for example, which uses the same symbol system but a different and
transient technology. The stability of the medium has important implications for how learners
process information from books and magazines: it aids in constructing meaning from the text.

In general, reading progresses in a forward direction and at a regular rate as the reader moves
along, readily constructing a mental representation that relates the information in the text to an
existing mental model. But on occasion, reading processes interact with prior knowledge and
skill in a way that relies heavily on the stability of text to aid comprehension and learning. While
poor readers are often thwarted by the effort required to decode the text,(13) fluent readers use
the stability of the text to avoid reading failure: encountering longer or novel words, these
readers will slow their rate, go back to review a word as an aid to recalling a meaning for it, or
review a phrase or sentence to determine the meaning of the word from context.(14) Even
readers with highly developed reading skills and elaborate memory structures rely on the stable
structure of print to process large amounts of text in familiar domains: a study by Charles
Bazerman, for example, revealed a strategy by which seven physicists read selectively and for a
particular purpose by scanning print rapidly and using certain words to trigger decisions either to
skip over familiar information or to move back and forth carefully within a text and across texts
to add to their understanding of their field.(15) Most readers, then, use the stability (technology)
of the printed text to process (read) its content (symbol system) and thereby construct or
elaborate on a mental model.

What happens when pictures or diagrams are introduced into this medium? What is the cognitive
effect of these symbol systems in combination with text? And how does the stability of these
symbols, as presented in books, interact with processing? A large body of traditional research
suggests that using pictures in combination with text generally increases recall, particularly for
poor readers, if the pictures illustrate information central to the text, when they represent new
content that is important to the overall message, or when they depict structural relationships
mentioned in the text.(16) Analyzing this research according to the perspective of this column
suggests that the use of both symbol systems in a stable medium facilitates a particular kind of
processing, particularly for learners who have little prior knowledge of the topic.

Several studies indicate that readers use pictures to create or to evoke preliminary mental models
that guide subsequent reading and assist in the construction of more elaborate and interrelated
models.(17) Other studies suggest that the use and effectiveness of pictures are related to prior
knowledge: more knowledgeable readers tend to build mental models from existing knowledge
and to elaborate on them using information from the text, while less knowledgeable readers tend
to rely more heavily on pictures or diagrams to construct mental representations of new
information.(18) Younger children, who may not have sufficient prior knowledge from which to
generate elaborate mental models, may benefit most from pictures to aid this process.(19 )The
stability of the medium allows the kind of serial, sequential, back-and forth processing between
specific information in the text and components of the pictures that facilitates the construction
and elaboration of mental models.

Learning with Television

Television—or any video medium—differs from books in several ways that may affect cognitive
structures and processes. First, the technology of these media makes both their verbal and visual
symbol systems transient rather than stable. Linguistic information can be orthographic (as in
captioned films); but more often it is oral and, like the images on the screen, disappears quickly.
Because of this transience, and because the two symbol systems are presented simultaneously—
learners may process the information in video media very differently from the way they process
similar information in books and magazines. It is also possible that the symbol systems and their
transient nature affect the mental representations that learners create.

A number of studies have found that viewers allot their attention to television in various and
predictable ways. This research indicates that visual attention increases from very low levels
during infancy to a maximum in the late elementary school years,(20) and that the nature of this
attention is influenced by several factors. Even though they may appear distracted at times,
children continually monitor the audio portion of a television presentation at a superficial level,
and their visual attention is triggered by particular audio cues: women’s and children’s voices,
peculiar voices, sound effects, and auditory changes. Features associated with continued visual
attention are special visual effects, pans, and high physical activity; visual attention frequently
decreases with the use of men’s voices, long zooms, and inactivity.(21) These “formal
features”(22) come to be seen by children as corresponding to the meaningfulness of the content,
and it is this meaningfulness that subsequently guides and maintains visual attention. For
example, children’s viewing experience may lead them to perceive that men’s voices generally
correspond to content that is adult-oriented and therefore less interesting and meaningful to
them.(23)

Several aspects of video media seem to have particular effects on learners’ cognitive
mechanisms: the simultaneous presentation of auditory and visual information, the processing
pace required by transient presentations of information, and the ways in which dynamic qualities
might affect a learner’s mental models.

Most studies of the roles of audio and visual presentations show that the combined use of the two
symbol systems results in more recall than visual-only and audio-only presentations.(24)
Additionally, several studies suggest that each source provides information that retains some of
the characteristics of the original symbol system: children recall sounds and expressive language
from the audio track and visual details from the visual track.(25) It also appears that the
representations derived from the visual symbol systems are more elaborate, making the visual
component of the presentation particularly memorable.(26) Audio may be sufficient for those
who are knowledgeable about a topic and can draw on previous knowledge for their mental
models, but the visual symbol systems supply important situational information for those who
are less informed.

An important aspect of video is its transience. Only limited research has addressed the effect of
pace on comprehension,(27) but this aspect of video presentations clearly distinguishes them
from print and may interact with learning in significant ways. With books, the reader sets his or
her own cognitive pace (i.e., words per unit of time) to accommodate personal requirements for
comprehension. With video, the pace is set by someone other than the learner, and the
presentation (i.e., words or visual elements per unit of time) progresses without regard to
individuals’ cognitive requirements. Analyzing this situation from the perspective of this article
suggests several avenues that researchers might pursue to understand the effect of such pacing on
comprehension: because viewers use their prior knowledge to process words and ideas,
knowledgeable learners are probably able to process information at the pace it is presented partly
because they can use their long-term memories to supplement information they might have
missed. If the topic is unfamiliar, however, and little information exists in long-term memory to
supplement viewing, the pace may exceed the learner’s capacity to process. Further, because the
information is transient, a learner cannot regress over it to refresh short-term memory. The
cascading comprehension failure that has been documented in such situations(28) might well be
explained through research that focuses on these cognitive dimensions.

There is a third aspect of the ways in which video media might affect learning—the ways in
which their dynamic qualities interact with a learner’s mental models. Understanding this aspect
requires some additional information on the nature and uses of mental models. According to
various authors,(29) mental models consist of sets of mental entities that are connected by
relationships and procedures and can be “run” in ways that have some similarities to the running
of a computer program. For example, a mental model of a media center would include
information about the roles of the staff, the nature and location of the collection, the rules and
Procedures for circulation, and so on; the library media specialist trying to solve a problem-for
example, determining how to introduce a new electronic catalog to teachers and students-could
“run” this model to make inferences about what would and should happen in such an
introduction. “Running the model” would thus help the library media specialist design a solution
to the problem at hand.

In this way, mental models are themselves dynamic, since they include mechanisms for moving
from one representation of information to another—for example, from a mental model of the
library media center without an electronic catalog to one of the library media center with such a
resource. Because of this dynamism, the moving, transient nature of video presentations may
help learners build the dynamic properties of their mental models. A film about the Battle of
Gettysburg, for example, might be especially effective in helping students understand not only
specific battlefield conditions but how those conditions changed over the three days of fighting
and transformed what might have been a Confederate triumph into a Union victory. By helping
the student understand the transformation of information (for example, the state of the battle)
from one situation (its state on July 1) to another (its state on July 3), the film might promote the
creation of a mental model that includes a sophisticated awareness of this transformation as well
as of simple facts.
Again, lack of research in this area means that these contentions are only speculative; they
suggest, however, that the technology of video media (dynamic, transient) and their simultaneous
presentation of two symbol systems exert a strong influence on learners’ mental models and the
processes used to construct them.

Learning with Computers

Computers can be distinguished from the two previous formats by what they can do with
information—that is, by their ability to process symbols and symbol systems. The prototypic
“information processors,” computers can transform information in one symbol system to that in
another and they can “proceduralize” information.(30) In its transforming function, a computer
with a voice synthesizer can change typed text (i.e., print) into speech; using an integrated
software package, it can transform numerical values into charts and graphs. In its proceduralizing
function, a computer can operate on symbols according to specified rules: for example, it can
rotate a graphic object on the screen according to the laws of physics. Through both functions, a
computer can help students construct links between symbolic domains—like graphs and
equations—and the real-world phenomena they represent. So it is the processing capabilities of
the computer, rather than its symbol systems per se, that enable this medium to make its primary
contribution to students’ construction of their mental models.

Students are frequently unable to connect their symbolic learning in school to “real world”
situations,(31) but the transformational capabilities of the computer can help them make this
connection. For example, several studies have shown improvement in graph-interpretation ability
for students working in microcomputer-based laboratories.(32) These laboratories use sensors
connected to a computer to collect data (e.g., on temperature and motion); the computer
transforms the data, displaying the information as graphs rather than numbers. The
transformation capabilities of the computer thus make immediate and direct the connection
between the graphic symbols and the world they represent. Seeing this connection aids in the
development of students’ ability to read graphs—that is, to transform a graph into a description
of what it means in the “real world.”

Perhaps even more importantly, the processing capabilities of the computer can help novices
build and refine mental models to be more like those of experts. Much of the research in this area
has involved physics, in which series of studies have established the nature of experts’
knowledge: it is extensive, organized into large chunks that are structured around the laws of
physics, and includes information both about the formal laws of physics themselves and about
how and under what conditions these laws apply.(33) Novices’ knowledge, however, is not only
less extensive but is organized differently: it might include only physical objects like blocks and
pulleys, fragments rather than interrelated sets of concepts, and “laws” that are incomplete or
otherwise incorrect.(34) When trying to solve problems, then, novices often construct mental
models that are incomplete, inaccurate, or otherwise insufficient.

How might the processing capabilities of computers be used by novices to aid them in building
more expert-like models? First, the computer can graphically represent the formal, abstract
entities that novices do not normally include in their models. The computer, for example, can use
an arrow to represent “force”—that is, an influence that changes the movement or shape of an
object—a construct that has no concrete referent in the physical world. Second, the computer can
proceduralize the relationships among these graphic (and other) symbols and display the results
of those procedures. It can change the shape or direction of the arrow to represent what actually
occurs, according to the laws of physics, when force is increased, decreased, or applied from
different directions. Furthermore, the computer allows learners to manipulate these symbols and
observe the consequences, successful or unsuccessful, of their decisions. Through a series of
such experiences, novices may become aware of the inadequacies of their own mental models
and move progressively toward more elaborate, integrated, and accurate ones.(35)

Thus, the processing capabilities of the computer can influence the mental representations and
cognitive processes of learners. Their transformation capabilities can connect symbolic
expressions (such as graphs) to the actual world. Their proceduralizing capabilities can allow
students to manipulate dynamic, symbolic representations of abstract, formal constructs that are
frequently missing from their mental models in order to construct more accurate and complete
mental representations of complex phenomena.

Learning with Multimedia

Little research has been done on learning with multimedia environments, primarily because the
field is still evolving and most efforts within it are focused on development. However,
multimedia present the possibility of combining in a single instructional environment all the
technologies, symbol systems, and processing capabilities of the individual media described
above. Examining how we might use each of these aspects individually and in various
combinations to facilitate learning is an important direction for current and future research.

Computer technology plays a central role in multimedia environments: the computer coordinates
the use of various symbol systems and processes information it receives, collaborating with the
learner to make subsequent selections and decisions. This role is essentially the same whether the
specific multimedia format in use is interactive video or hypermedia.

One of the best known examples of the interactive video environment is the “Jasper Woodbury
Series,”(36) which provides realistic contexts to help middle-school students learn complex
problem solving in mathematics. Each videodisc provides a series of stories about Jasper
Woodbury (who is approximately the same age as the target students) that contain both the
problems to be solved and data that can be used in the solutions. In one story, for example,
Jasper takes a used boat for a “test drive” and decides to buy it. The problem, briefly stated, is
that the boat’s running lights do not work and Jasper must determine if he can get the boat to his
home dock before sunset. The students are left to solve the problem, using major questions
embedded in the story itself: does Jasper have enough time to get home before sunset? enough
gas? enough money to buy the necessary gas?

Students work in groups to determine the solution, encouraged by the teacher to generate
subordinate questions and to identify the information needed to solve them. They review
segments of the videodisc to search for information and to separate relevant from irrelevant facts;
use the facts to solve the subordinate problems; and then relate these partial solutions to the
overall problem. Early research on the influence of “Jasper Woodbury” on learning is
encouraging.(37)

What contribution did the videodisc make to this learning? Several contentions are suggested.
First, the capability of the video to use multiple symbol systems to present complex, dynamic
social contexts and events might have helped students construct rich, dynamic mental models of
the situations. The detailed, dynamic nature of these models might have allowed students to draw
more inferences than they could from mental models constructed from text or still pictures.(38)
As we have already noted, such structures are more memorable than those constructed with
text(39) and rely less on information in students’ heads(40) which is likely to be incomplete or
inaccurate for students with limited prior knowledge. The video also preempts demands on
reading ability, allowing students who have not yet automated their reading skills to focus their
cognitive resources on the problem-solving task.

Second, the videodisc contains a great deal of information crucial to the solution of the problem:
information about distances, available money, and other relevant conditions is embedded in
objects and maps and in what people say, do, and think as the story is enacted. The random
access capabilities of the computer-controlled videodisc allow students to pause, review, and
search for information they may have missed or forgotten. Identifying needed information and
extricating it from a context is an important component of learning to solve problems, and the
ability to do so contributes to successful transfer and performance in actual situations.

Finally, and most important, the visual and social nature of the story, as presented in this
environment, is likely to activate relevant prior knowledge that students can use to solve the
problem. Further, because of the scope of detail and relationships the environment provides,
students are likely to find many ways to connect their new learning to their existing
representations. ‘This, in turn, increases the likelihood that similar situations will evoke the
appropriate solution procedures in the future. Over time and similar experiences, these learned
strategies will become connected to a range of mental representations, promoting transfer of the
strategies to a variety of problem situations.(41)

As a distinct type of multimedia, hypermedia shares the technology and symbol systems of
interactive video environments but embodies processing capabilities that suggest an important
difference for learning. The nonlinearity of hypermedia—that is, the capability of this technology
to allow learners to create associational links within and across text, images, and other symbol
systems—facilitates cognitive flexibility because it allows a topic to be explored in multiple
ways using a number of different concepts and themes.(42) This exploration should result in the
development of integrated, flexible knowledge structures interconnected by crisscrossing
conceptual themes that facilitate the use of this knowledge to solve a wide range of problems.
Each concept can subsequently be used in many different ways, and the same concept can apply
to a variety of situations.

Some hypermedia systems allow learners to add their own information and construct their own
relationships. As Gavriel Salomon points out, such systems can reflect the processes learners use
when constructing interrelationships in their own mental models and thus encourage them to
think not only about ideas but about how they are interrelated and structured.(43) More
important, such systems can provide explicit models of information representation that learners
can use as guidelines for constructing their own internal models.

While there has been only limited research in hypermedia to date, preliminary findings are
encouraging.(44) Despite the appeal of hypermedia, however, it is important to note some
potential disadvantages for learning as well. In hypermedia environments, users are frequently
required to decide what information to select and in what order; building such sequences is likely
to be particularly difficult for novices, who lack the extensive and well-organized mental
representations that would allow them to locate appropriate information and integrate it with
their prior knowledge, experience, and opinions. Getting “lost in hyperspace” and failing to find
or recognize relevant information are other potential problems, particularly for novices, as is
spending inordinate time and cognitive energy processing information that is not relevant to their
purposes.

In summary, the technology of integrated multimedia environments brings together the symbolic
and processing capabilities of all the various media described above. Interactive videodisc
environments may help learners build and analyze mental models of problem situations, while
hypermedia environments may help learners build links across information presented by
different symbol systems and construct meaning based on these links. Plausible rationales have
been given for the expected effectiveness of such environments, but much more research is
needed to understand—let alone forge the relationships that proponents of these environments
hypothesize.

Conclusions
How does the analysis above contribute to a theory of learning with media? Richard Clark would
say it does not. Attributing media effects to their capabilities, or attributes, invokes his criticism
of the media attribute approach.(45) But when we abandon the conception of attributes as single
and discrete causal entities and consider that each medium is defined by its particular cluster of
capabilities, we are perhaps able to broaden the discussion. Further, when we consider the
various ways in which these clusters interact with cognitive processes and structures, perhaps we
can refocus the discussion to explore specific ways in which media capabilities may be used to
influence learning for individual learners performing particular tasks in specific content areas.
Understanding how learners interact with and use the unique capabilities of each medium’s
format is essential to understanding the effect of media on learning.

The field of educational technology is reexamining its foundational assumptions and

questions.(46) This article is part of my ongoing attempt to contribute to that effort. If we move
from “Do media influence learning?” to “In what ways can we use the capabilities of media to
influence learning for particular students, tasks, and situations?” we will both advance the
development of our field and contribute to the improvement of teaching and learning.
References
Note: The first four footnotes are referenced from the introduction to this article by Delia
Neuman.

1. Richard C. Clark, “Reconsidering Research on Learning from Media,” Review of

Educational Research 53 (Winter 1983): 445–59.
2. Robert B. Kozma, “Learning with Media,” Review of Educational Research 61 (Summer
1991): 179–211.
3. See Richard C. Clark, “When Researchers Swim Upstream: Reflections on an Unpopular
Argument About Learning from Media,” Educational Technology 31 (February 1991):
34–40.
4. Robert B. Kozma, “Will Media Influence Learning: Reframing the Debate,” Educational
Technology Research and Development, in press.
5. Information Infrastructure Task Force, The National Information Infrastructure: Agenda
for Action (Washington, D.C.: U.S. Dept. of Commerce, 1993); Gary Stix,
“Domesticating Cyberspace,” Scientific American 269 (August 1993): 100–10.
6. Rita Richey, Designing Instruction for the Adult Learner (London: Kogan, 1992);
William Winn, “Toward a Rationale and Theoretical Basis for Educational Technology,”
Educational Technology Research and Development 37 (Spring 1989): 35–46; William
Winn, “Some Implications of Cognitive Theory for Instructional Design,” Instructional
Science 19 (Winter 1990): 53–69.
7. Thomas Shuell, “The Role of the Student in Learning from Instruction,” Contemporary
Educational Psychology 13 (July 1988): 276–95.
8. Janes Greeno, “Situations, Mental Models, and Generative Knowledge,” in Complex
Information Processing, ed. D. Klahr and K. Kotovsky (Hillsdale, N.J.: Lawrence
Erlbaum,1988); Roy Pea “Practices of Distributed Intelligence and Designs for
Education,” in Distributed Cognitions, ed. Gavriel Salomon (New York: Cambridge
Univ. Pr., 1993); David Perkins, “Person Plus: A DistributedView of Thinking and
Learning,” in Salomon, Distributed Cognitions.
9. Richard Snow, “Aptitude Theory: Yesterday, Today, and Tomorrow,” Educational
Psvchologist 27 (Winter 1992):5–32.
10. Kozma, 1991, p. 179.
11. Gavriel Salomon, Interaction of Media, Cognition, and Learning (San Francisco: Jossey
Bass, 1979).
12. Nelson Goodman, Languages of Art (Indianapolis: Hackett, 1976).
13. David LaBerge and S. Jay Samuels, “Toward a Theory of Automatic Information
Processing in Reading,” Cognitive Psychology 6 (April 1974): 293–323.
14. Evalyn Bayle, “The Nature and Causes of Regressive Movements in Reading,” Journal
of Experimental Education 11 (Sept. 1942): 16–36; Marcel Just and Patricia Carpenter,
The Psychology of Reading and Language Comprehension (Newton,Mass.: Allyn &
Bacon, 1987).
15. Charles Bazerman, “Physicists Reading Physics,” Written Communication 2 (Jan. 1985):
3–23.
16. W. Howard Levie and Richard Lentz, “Effects of Text Illustrations: A Review of the
Research,” Educational Communication and Technology Journal 30 (Winter 1982): 195–
 232; Michael Pressley, “Imagery and Children’s Learning: Putting the Pictures in
Developmental Perspective,” Review of Educational Research 47 (Feb. 1977): 585–622;
Diane Schallert, “The Role of Illustrations in Reading,” in Theoretical Issues in Reading
Comprehension, ed. Rand Spiro, Bertram Bruce, and William Brewer (Hillsdale, N.J.:
Lawrence Erlbaum, 1980).
17. Jennifer Rusted and Veronika Coltheart, “The Effect of Pictures on the Retention of
Novel Words and Prose Passages,” Journal of Experimental Child Psychology 28 (Dec.
1979): 516–24; David Stone and Marvin Glock, “How Do Young Adults Read Directions
With and Without Pictures?” Journal of Educational Psychology 73 (June 198 1): 419–
26.
18. Mary Hegarty and Marcel Just, “Understanding Machines from Text and Diagrams,” in
Knowledge Acquisition from Text and Pictures, ed. Heinz Mandl and Joel Levin
(Amsterdam: Elsevier, 1989); Gunnar Kunz, U. Drewniak, and F. Schott, “On-line and
Off-line Assessment of Self-regulation in Learning from Instructional Text and Pictures,”
paper presented at the annual meeting of the American Educational Research
Association, March 1989.
19. Pressley, “Imagery and Children’s Learning.”
20. Daniel Anderson and others, “Television Viewing at Horne: Age Trends in Visual
Attention and Time with Television,” Child Development 57 (1986): 1024-33.
21. Daniel Anderson and others, “Watching Children Watch Television,” in Attention and
Cognitive Development, ed. Gordon Hale and Michael Lewis (New York: Plenum, 1979);
Daniel Anderson and Diane Field, “Children’s Attention to Television: Implications for
Production,” in Children and the Formal Features of Television, ed. Manfred Meyer
(Munich, Germany: K. G. Saur, 1983).
22. Aletha Huston and John Wright, “Children’s Processing of Television: The Information
Functions of Formal Features,” in Children’s Understanding of Television, ed. Jennings
Bryan and Daniel Anderson (New York: Academic, 1983).
23. Daniel Anderson and Elizabeth Lorch, “Looking at Television: Action or Reaction,” in
Bryan and Anderson, Children’s Understanding of Television.
24. Patricia Baggett and E. Ehrenfeucht, “Information in Content Equivalent Movie and Text
Stories,” Discourse Processes 5 (1982): 73–99; Patricia Baggett and E. Ehrenfeucht,
“Encoding and Retaining Information in the Visuals and Verbals of an Educational
Movie,” Educational Communication and Technology Journal 31 (Winter 1983): 23–32);
Jessica Beagles-Roos and Isabelle Gat, “Specific Impact of Radio and Television on
Children’s Story Comprehension,” Journal of Educational Psychology 75 (Feb. 1983):
128–37; Laurene Meringoff, “What Pictures Can and Can’t Do for Children’s Story
Understanding,” paper presented at the annual meeting of the American Educational
Research Association, March 1982.
25. Beagles-Roos and Gat, “Specific Impact of Radio and Television on Children’s Story
Comprehension.”
26. Patricia Baggett, “Understanding Visual and Verbal Messages,” in Mandl and Levin,
Knowledge Acquisition from Text and Pictures.
27. John Wright and others, “Pace and Continuity of Television Programs: Effects on
Children’s Attention and Comprehension,” Developmental Psychology 20 (July 1984):
653–66.
28. Daniel Anderson and Patricia Collins, “The Impact on Children’s Education: Television’s
Influence on Cognitive Development (Working Paper No. 2).” (Washington, D.C.: U.S.
Office of Educational Research and Improvement, 1988).
29. Greeno, “Situations, Mental Models, and Generative Knowledge”; in John Holland and
others, Induction: Processes of Inference, Learning, and Discovery (Cambridge, Mass.:
MIT Press, 1986); Michael Williams, James Hollan, and Albert Stevens, “Human
Reasoning About a Simple Physical System,” in Mental Models, ed. Dedre Genter and
Albert Stevens (Hillsdale, NJ.: Lawrence Erlbaum, 1983).
30. W. Patrick Dickson, “Thought-provoking Software: Juxtaposing Symbol Systems,”
Educational Researcher 14 (May 1985): 30–38.
31. Lauren Resnick, “Learning in School and Out,” Educational Researcher 16 (Dec. 1987):
13–20.
32. Heather Brasell, “Me Effect of Real-time Laboratory Graphing on Learning Graphic
Representations of Distance and Velocity,” Journal of Research in Science Teaching 24
(April 1987): 385–95; Jan Mokros and Robert Tinker, “The Impact of Microcomputer-
based Labs on Children’s Ability to Interpret Graphs,” Journal of Research in Science
Teaching 24 (April 1987): 369–83.
33. Michelene Chi, Paul Feltovich, and Robert Glaser, “Categorization andRepresentation of
Physics Problems by Experts and Novices,” Cognitive Science 5 (April-June 1981): 121-
52, Mary Hegarty, Michel Just, and I. R. Morrison, “Mental Models of Mechanical
Systems: Individual Differences in Qualitative and Quantitative Reasoning,” Cognitive
Psychology 20 (April 1988): 191-236; Jill Larkin, “The Role of Problem Representation
in Physics,” in Genter and Stevens, Mental Models; Jill Larkin and others, “Expert and
Novice Performance in Solving Physics Problems,” Science 208 (June 1980): 1335-42.
34. J. Clement, “A Conceptual Model Discussed by Galileo and Used Intuitively by Physics
Students,” in Genter and Stevens, Mental Models; Andreadi Sessa, “Knowledge in
Pieces,” in Constructivism in the Computer Age, ed. George Forman and Peter Pufall
(Hillsdale, N.J.: Lawrence Erlbaum, 1988); Larkin, “The Role of Problem Representation
in Physics”; M.McCloskey, “Naïve Theories of Motion,” in Genter and Stevens, Mental
Models.
35. Greeno, “Situations, Mental Models, and Generative Knowledge.”
36. Developed by the Cognition and Technology Group at Vanderbilt University, 1987–
present.
37. Cognition and Technology Group at Vanderbilt, “An Anchored Instruction Approach to
Cognitive Skills Acquisition and Intelligent Tutoring,” in Cognitive Approaches to
Automated Instruction, ed. J. Wesley Regian and Valerie Shute (Hillsdale, NJ.: Lawrence
Erlbaum); James Van Haneghan and others, “The Jasper Series: An Experiment with
New Ways to Enhance Mathematical Thinking,” in Enhancing Thinking Skills in the
Sciences and Mathematics, ed. Diane Halpern (Hillsdale, NJ.: Lawrence Erlbaum, 1992).
38. John Bransford and others, “MOST Environments for Accelerating Literacy
Development,” paper presented at the NATO Advanced Study Institute on Psychological
and Educational Foundations of Technology-Based Learning Environments, Kolymbari,
Crete, Aug. 1992.
39. Baggett, “Understanding Visual and Verbal Messages.”
40. Beagles-Roos and Gat, “Specific Impact of Radio and Television on Children’s Story
Comprehension”; Meringoff, “What Pictures Can and Can’t Do for Children’s Story
Understanding.”
41. Rand Spiro and Jihn-Chang Jehng, “Cognitive Flexibility and Hypertext: Theory and
Technology for the Non-linear and Multidimensional Traversal of Complex Subject
Matter,” Cognition, Education, and Multimedia, ed. Don Nix and Rand Spiro (Hillsdale,
NJ.: Lawrence Erlbaum, 1990).
42. Ibid.
43. Gavriel Salomon, “AI in Reverse: Computer Tools That Turn Cognitive,” Journal of
Educational Computing Research 4 (Spring 1988): 123–34.
44. William Beeman and others, “Hypertext and Pluralism: From Lineal to Non-lineal
Thinking,” in Hypertext ‘87 Proceedings (1987): 67–88; Michael Jacobson and others,
“Hypertext Learning Environments and the Acquisition of Complex Knowledge,” paper
presented at the annual meeting of the American Educational Research Association, April
1994; David Jonassen and Sherwood Wang, “Acquiring Structural Knowledge from
Semantically Structured Hypertext,” Journal of Computer-Based Instruction 20 (Winter
1993): 1–8; Gary Marchionini, “Evaluating Hypermedia-BasedLearning,” in Designing
Hypermedia for Learning, ed. David Jonassen and Heinz Mandl (Berlin: Springer-
Verlag, 1990); Gary Marchionini and Greg Crane, “Evaluating Hypermedia and
Learning: Methods and Results from the Perseus Project,” ACM Transactions on
Information Systems12 (Winter 1994): 5–34; Rand Spiro and others, “Cognitive
Flexibility, Constructivism, and Hypertext: Random Access Instruction for Advanced
Knowledge Acquisition in Ill-structured Domains,” in Constructivism and the
Technology of Instruction: A Conversation, ed. Thomas Duffy and David Jonassen
(Hillsdale, NJ.: Lawrence Erlbaum,1992); Van Haneghan and others, “The Jasper
Series.”
45. Clark, “Reconsidering Research on Learning from Media.”
46. Duffy and Jonassen, Constructivism and the Technology of lnstruction: A
Conversation; Denis Hlynka and John Belland, Paradigms Regained: The Uses of
Illumintive, Semiotic, and Post-modern Criticism as Modes of Inquiry in Educational
Technology (Englewood Cliffs, N.J.: Educational Technology Publications, 1991).