Dual Coding Theory and Education

Author(s): James M. Clark and Allan Paivio

Source: Educational Psychology Review, Vol. 3, No. 3 (September 1991), pp. 149-210
Published by: Springer
Stable URL: https://www.jstor.org/stable/23359208
Accessed: 06-04-2020 08:31 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/23359208?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Educational
Psychology Review

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Educational Psychology Review, Vol. 3, No. 3, 1991

Dual Coding Theory and Education

James M. Clark13 and Allan Paivio2

Dual coding theory (DCT) explains human behavior and. experience in terms
of dynamic associative processes that operate on a rich network of modality
specific verbal and nonverbal (or imagery) representations. We first describe
the underlying premises of the theory and then show how the basic DCT
mechanisms can be used to model diverse educational phenomena. The re
search demonstrates that concreteness, imagery, and verbal associative proc
esses play major roles in various educational domains: the representation and
comprehension of knowledge, learning and memory of school material, effec
tive instruction, individual differences, achievement motivation and test anxiety,
and the learning of motor skills. DCT also has important implications for the
science and practice of educational psychology — specifically, for educational
research and teacher education. We show not only that DCT provides a unified
explanation for diverse topics in education, but also that its mechanistic frame
work accommodates theories cast in terms of strategies and other high-level
psychological processes. Although much additional research needs to be done,
the concrete models that DCT offers for the behavior and experience of stu
dents, teachers, and educational psychologists further our understanding of
educational phenomena and strengthen related pedagogical practices.
KEY WORDS: imagery; verbal processes; unified educational theory.

INTRODUCTION

Both the science and practice of education depend on a firm un

standing of many psychological phenomena, including such cognitive

'Department of Psychology, University of Winnipeg, Winnipeg, Manitoba, Canada R3

department of Psychology, University of Western Ontario, London, Ontario, Canad
5C2.
^To whom all correspondence should be addressed.

149

1040-726X/91/0900-0149$06.50/0
© 1991 Plenum Publishing Corporation

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
150 Clark and Paivio

as the structure of knowledge, study skills, student aptitudes, and eff

instructional practices. Less cognitive psychological topics, such as a
(e.g., student interests and motivation, evaluation anxiety) and perce
motor processes (e.g., handwriting, typing) are also relevant to educa
The range of these phenomena challenges teachers, educational
searchers, and psychologists to develop general psychological theories
can explain diverse facets of human behavior and experience.
One way to achieve such unified theories is by reductionis
proaches that identify molecular psychological mechanisms underlyin
main-specific molar explanations for educational and related cog
phenomena. By molar explanations, we mean strategies, beliefs, and o
high-level psychological processes. For example, a domain-specific m
explanation for effective study skills might describe note-taking, el
tion, and other learning strategies, as well as the metacognitive proc
that control their use. A domain-specific molar explanation for eff
instruction might describe the use of concrete examples, lesson summ
and other teaching methods, along with the executive processes inv
in the selection and implementation of those methods. A reduction
would try to unify the distinct domains of study skills and effective
ing by considering shared mechanisms that underlie the somewhat d
ent molar theories. Both the elaboration of text by students and th
of concrete examples by teachers, for example, may increase the likel
that visual images are evoked by the to-be-learned material, and this
shared imagery mechanism could contribute to the effects on comprehen
sion and memory of both elaboration and examples. Similarly, the meta
cognitive processes of students and teachers involve shared verbal
mechanisms by which self-talk guides and controls behavior (e.g., "Think
of an example"; "Give students a summary"), although the specific con
tents of the statements vary.
The present paper describes one theory of basic psychological mecha
nisms that permits unified explanations for diverse educational phenomena.
Dual Coding Theory (DCT) (Paivio, 1971, 1986) is an empirically well
founded characterization of the mental processes that underlie human be
havior and experience. DCT explains psychological phenomena by the
collective action of nonverbal and verbal mental systems that are special
ized for the processing of imagery and linguistic information, respectively.
DCT theoretical mechanisms and associated empirical phenomena are rele
vant to various aspects of human cognition, as well as emotion, motor skills,
and other psychological domains. This breadth suggests that the theory
could provide a useful foundation for a general psychological model of edu
cation and could strengthen current efforts to explain educational phenom
ena in terms of cognitive mechanisms (e.g., Dillon and Sternberg, 1986;

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 151

Gagne, 1985; Mayer, 1987). We first describe the theory and then show
how DCT mechanisms can help to integrate the extensive research litera
ture on education.

DUAL CODING THEORY

The underlying assumptions of DCT concern basic mental str

and processes: the structures are associative networks of verbal and i
representations, and the processes concern the development and
of those structures, including the effects of context on the spread o
tion among representations. These DCT assumptions are present
fully in many articles and books that describe the overall theo
1971, 1986) and its application to language (e.g., Paivio and Begg
memory (e.g., Clark and Paivio, 1987), and other psychological to

Imagery and Verbal Mental Representations

According to DCT, mental representations are associated wi

retically distinct verbal and nonverbal symbolic modes and ret
erties of the concrete sensorimotor events on which they are b
Fig. 1). The verbal system contains visual, auditory, articula
other modality-specific verbal codes (e.g., representations for su
as book, text, livre, school, teacher, learn, strategy, mathemat
worry). These word-like codes are arbitrary symbols that denote
objects and events, as well as abstract ideas. For example, th
words "book" and "text" and the French word "livre" are arbitra
labels for the same object. Verbal codes retain their separate an
identities even when connected in hierarchies or other associative net
works. That is, a word such as "livre" can be associatively connected to
its English translation "book" and included in the sentence "Livre is th
French translation for book," but the words remain separate entiti
Moreover, verbal representations are generally processed in a serial
sequential manner. Thinking of the sentence "Livre is the French trans
lation for book," for example, means attending successively to the word
A verbal description of some classroom scene or event similarly involv
a sequential description.
Nonverbal representations include modality-specific images for shape
(e.g., a chemical model), environmental sounds (e.g., school bell), action
(e.g., drawing lines or pressing keys), skeletal or visceral sensations relate
to emotion (e.g., clenched jaw, racing heart), and other nonlinguistic objec

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 152 Clark and Paivio

VERBAL STIMULI NONVERBAL STIMULI

SENSORY SYSTEMS

REPRESENTATIONAL CONNECTIONS

V Logogens Imogens
E
R
B REFERENTIAL

>
A CONNECTIONS
L

S
Y
S
"X.
T
E
M

VERBAL RESPONSES NONVERBAL RESPONSES

Fig. 1. Verbal and nonverbal symbolic systems of Dual

Coding Theory. From Mental Representations: A Dual Cod
ing Approach by A. Paivio (1986). Reprinted by permission.
This figure shows the representational units and their ref
erential (between-system) and associative (within-system)
interconnections.

and events. Such imaginai representations are analogous or perceptually

similar to the events that they denote, rather than being arbitrary symbols.
That is, mental images for "book" have visual, tactual, and other perceptual
qualities similar to those evoked by the referent objects on which the im
ages are based. Similarly, mental images evoked by emotionally laden words
or phrases (e.g., "I like my teacher" or "I hate math") have visceral prop
erties similar to those experienced when one is actually in the presence of
the affective object. In contrast to verbal processing, which is sequential,
nonverbal representations can encode information in parallel or simulta
neously. A single mental image of a complex classroom or playground
scene, for example, contains much detailed information; stated colloquially,
"pictures are worth a thousand words," and so are mental images. Because
complex images can integrate the parts of events, objects may "lose" their
separate identities and become spatially embedded or nested in the whole
imaginai structure. For example, the distinct shape of a triangle might
"fade" as imagined lines are added for a proof in geometry. Mental images
are also amenable to dynamic spatial transformations that are not possible
with verbal representations. That is, students can visualize the rotation of
a chemical model, the effect of tilting a container of liquid, an explorer's
ship circumnavigating the globe, or other spatial transformations, but analo
gous operations are not possible within the verbal system.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 153

DCT's other structural assumptions concern the connections that

link verbal and nonverbal representations into a complex associative net
work (see Fig. 1). Links between the two systems are called referential
connections. They join corresponding verbal and imaginai codes and po
tentially allow such operations as imaging to words and naming to pictures.
For example, the word "school" might evoke negative visual images and
unpleasant visceral sensations in children who have formed links between
the word and these nonverbal reactions. Similarly, the word "Canada" spo
ken in a geography class might arouse a clear visual image of the country's
outline, or the word "Principal" might elicit various images in teachers
and students. Referential connections in the opposite direction, from im
age to name, permit students to label pieces of science apparatus (e.g.,
test tube, bunsen burner), biological objects and their parts (e.g., praying
mantis, nucleus), the states on a map of the U.S., and other figurai
information.
A second kind of link, called associative connections in DCT, joins
representations within the verbal and nonverbal systems. On the verbal
side, words are joined to other related words. In the case of a student
who has an aversion to school, the word "school" might elicit such verbal
associations as "hate," "boring," or "afraid." Additional examples of ver
bal associations include: connections between instance and category
names (e.g., "gold," "lead," "iron," and other instance names acquiring
links to the term "metal"), teachers and students encoding a lesson as
an associative chain of key terms and phrases, or the schematic repre
sentation for a lab report as a linked series of labeled parts (e.g., intro
duction, method, results, discussion).
Within the nonverbal system, associative connections join images to
other images in either the same or different sensory modalities. To continue
with the school aversion example, sight of school might evoke visual images
and nonverbal visceral responses reminiscent of unpleasant school experi
ences. Similarly, a visual image of a bunsen burner can be associated with
visual images for other objects in a science experiment, with auditory and
olfactory images for the sound and smell of gas, and with motor images
for adjusting the flow of gas. Students also link successive images for con
crete events reported in novels, history and geography texts, and other con
crete school materials.

Processing Assumptions

The development and activation of verbal and imaginai associative

structures are governed by DCT's processing assumptions. A basic premise

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
154 Clark and Paivio

is that individual verbal and imaginai representations vary in their ac

levels, with some representations highly active and others depresse
any given time. Strong activity may be associated with conscious non
and verbal experiences (e.g., an image of a classroom, or thinking of
words as "lazy" or "intelligent"), although arousal of codes does not a
lead to conscious experiences. Active mental representations can in
activate associatively related nodes in the network, and this spread
tivation results in complex patterns of arousal among the representa
in the network. Metaphorically, one can think of human cognition
constantly changing landscape with elevated areas (peaks) indica
aroused mental codes and low-lying areas (valleys) indicating depres
codes. The particular shape of the landscape depends on comple
subtle associative mechanisms.
As a hypothetical example of this spreading activation, consider how
sight of a particular student would increase the activity in stored images
of the student, and how activation may then spread from those images
to other verbal and nonverbal nodes in the network. If the teacher's im
age of a student is associated with the mental code for the word "lazy,"
then activation of that label would in turn arouse a repertoire of associ
ated images and verbal representations. Nonverbal associates of "lazy"
could include visual or kinesthetic images of patting the child on the
back, frowning, or turning away from the student, as well as codes for
sympathetic or aversive visceral sensations and other mental images.
Verbal associations activated by "lazy" and the image of the boy might
include such covert statements as "Homework probably not done,"
"Needs praise," "What can I do?" or "Just like his brother!" In this man
ner, spreading activation results in complex patterns of arousal in the
network, with each pattern in turn determining succeeding responses of
the mental system.
With respect to the development of mental representations and their
interconnections, DCT emphasizes the central role of past experience. The
word "Scud," for example, had little semantic, visual, or affective meaning
for most people until recent events in the Middle East. One way in which
the importance of experience reveals itself is through idiosyncratic and con
text-dependent reactions to environmental events. Although similar expe
riences promote common mental structures, variability in individual
experiences means that mental representations and their interconnections
can vary from person to person, depending upon people's specific history
with the elements in the network.
Even given common experiences, the development of connections will
vary with the accessibility of existing codes; for example, poor student per
formance may tend to be labeled "laziness" by one teacher and "boredom"

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 155

by another. General predispositions to respond in certain ways are captured

in DCT and other associative theories by differences in the amount of
stimulation required to activate labels in the mental network. In our ex
ample, the net effect of idiosyncratic experiences and dispositional factors
would be that some teachers would acquire strong referential connections
between their images for a student and the mental word "lazy," whereas
others would associate alternative labels (e.g., "bored" or "slow") or would
lack labels because of inadequate experiences, explicit suppression of evalu
ative terms, or other factors.
In addition to the strengths of existing representations and their con
nections, the pattern of activation on any particular occasion is determined
in part by instructions and other moderating contextual influences that tem
porarily enhance activation of some connections and inhibit others. For
example, presenting pictures or telling students to generate images for pairs
of words will prime the imagery system and increase the likelihood that
words will activate mental images. Pictures and instructions to image indeed
do increase reports of imagery in various tasks and produce other effects
consistent with an imagery interpretation (see later discussion).
According to DCT, the relative activation of the nonverbal system is
particularly important for understanding human behavior because the im
agery system has unique theoretical and empirical properties. We noted
earlier, for example, that the imagery system can unify multiple objects into
an integrated image, and we will see later that such integration can in turn
facilitate memory for textbook and other school material. The probability
of imagery processing depends on several classes of variables (Paivio, 1971,
1986), and all have direct educational implications. We have already stated
that instructions and related context effects can influence the arousal of
imagery, so that students and others are more likely to generate mental
images if instructed to do so than if left to their own devices. Such imagery
instructions are incorporated into various memory techniques that facilitate
vocabulary and other school learning.
A second important determinant of imagery processing is the imagery
value or concreteness of the material being studied. Theoretically, imagery
and concreteness reflect the availability and strength of word-to-image ref
erential connections. That is, concrete words such as book, teacher, bunsen
burner, and blackboard denote tangible objects that are more likely than
abstract words to have corresponding images. Abstract words such as abil
ity, success, effort, mass, and learning-disability do not refer to concrete,
tangible objects or events, and are less likely to evoke an image of specific
referents. We show later that concreteness and imagery value are important
attributes of educational material. A poem incorporating concrete events

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
156 Clark and Paivio

and figurative language, for example, will be more likely to evoke im

in students than abstract passages.
A third factor affecting nonverbal processing is variation among peo
in the tendency and capacity to use imagery; that is, individual diffe
Some students and teachers will use imagery easily and spontaneously
many conditions, whereas others will rarely image and only with difficu
These individual differences in imagery abilities and habits have imp
consequences for education. Students who have trouble imaging, for
ple, may fail to remember passages of text that benefit from imagin
essing, may not understand geography or other spatial facts in a con
way, and might do poorly at visualizing the steps in a geometric proof, s
ing difficult words, or even printing letters correctly.
Instructions and related context effects influence not only the relat
activation of verbal and nonverbal systems, but also the patterns of
tion within the nonverbal and verbal systems. Instructions to studen
generate synonyms, for example, will make such responses more likely t
when students free-associate to words (Clark, 1978). Specific context e
can also operate indirectly. Wynne et al. (1965), for example, found
adding antonym-evoking stimuli to the beginning of a free-associatio
increased the frequency of antonym responses for later items. Presum
words such as "black" and "hot," which tend to elicit opposite respo
(e.g., "white," "cold"), activated the word "antonym" or some equiv
term, which then primed antonym responses for later items. In a si
way, a teacher modeling certain classes of response for various object
superordinate names, colors, shapes) will prime that class of respon
subsequent items presented to students.
There are many classroom analogues to selective priming effec
In the case of a teacher's response to a particular student, for exam
the reaction "smart" to John's specific behavior may be more e
aroused if primed by some related event, and the response "Just li
brother!" might be primed by seeing the brother in the playground
ing students to image the events in stories also operates in a sel
manner to increase the likelihood that students experience men
ages. Within the verbal system, asking students to classify names o
mals (e.g., crocodile, rabbit) into categories primes verbal assoc
pathways and responses associated with superordinates (e.g., for rep
mammal), and inhibits other associations not relevant to the task (e
dangerous, Dundee, long ears, bunny, Easter). These effects of cont
can be quite precise; whether the digits 2 and 6 activate 4, 8, 12, or
for example, will depend on the presence or absence of arithmetic
bols (minus, -, plus, +, times, x).

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 157

Discussion of DCT

Multiple verbal and imaginai representations, the complex referent

and associative network, the effects of history and context, and the su
effects of spreading activation all combine to permit powerful models
be built for diverse psychological phenomena. A DCT model, in essen
consists of hypothetical networks of verbal and nonverbal representat
and descriptions of the mediating patterns of activation (i.e., the state
the network) that intervene between stimulus and response events. Le
ing foreign vocabulary, for example, involves successive verbal and non
bal representations that are activated during initial study of the word p
and during later efforts to retrieve the translations. The specific way
which verbal and nonverbal mechanisms contribute to performance w
vary with the task, stimulus characteristics, past and present events,
individual differences.
Although DCT models emphasize basic mechanisms, the theory also
permits and complements explanations cast in terms of such molar proc
esses as strategies and beliefs. Deliberate determinants of teacher and stu
dent behavior, such as strategies, are sometimes contrasted with passive
associative mechanisms, but the true relation between strategies and asso
ciative processes is more complementary than competitive. Ultimately, the
human mind represents and processes strategies in terms of nonstrategy
mechanisms, possibly in terms of associative networks of verbal and non
verbal mental representations. We return to this question in the General
Discussion, but it is important to note that the appropriate questions to
ask during the following analyses of educational phenomena are whether
the proposed mechanisms of DCT are compatible with higher-level expla
nations and further our understanding of how those molar processes work,
rather than whether DCT can replace existing molar theories.
The remainder of this article demonstrates that imagery, concrete
ness, verbal and nonverbal associative networks, and other DCT constructs
can provide unified explanations for educational psychology. A central
theme will be that diverse educational phenomena show the collective con
tribution of imagery and verbal processes to human behavior and experi
ence. The examples considered not only draw on DCT's historical strengths
in the area of human cognition (e.g., the structure of knowledge, learning
and study skills, effective instruction), but also demonstrate that the theory
is well-suited for less cognitive topics (e.g., emotions and motor skills). Be
cause we discuss a broad range of educational phenomena, our coverage
is highly selective, and some controversial and complex issues are men
tioned just briefly.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
158 Clark and Paivio

THE STRUCTURE OF KNOWLEDGE

The primary purpose of education is to transmit knowle

quently, much educational and cognitive research has attempt
the basic psychological mechanisms by which information in text
educational sources is represented. A DCT account of knowle
sizes the contribution of both nonverbal and verbal systems,
sider in turn.

Imagery Processes in the Meaning of Words and Text

The knowledge relevant to education is to a large extent represented

verbally in books, teacher and student notes, and other forms of text. Be
cause of referential connections between the verbal and nonverbal systems
and because images have special properties, DCT maintains that the prob
ability and ease of image arousal plays an important role in the repre
sentation of text meaning. This hypothesis is clearly supported by research
on word meaning, text comprehension, and related phenomena.
DCT predicts that word concreteness and imagery value should be
central variables in cognitive and educational tasks related to meaning. Em
pirically, concreteness and imagery value have been measured by ratings
of the ease with which words, sentences, or larger units of text evoke a
visual, auditory, or other mental picture (imagery value), or the degree to
which they refer to tangible objects with concrete referents (concreteness)
(Paivio et al., 1968). Ratings of these attributes are highly reliable and cor
related, and theoretically reflect differential access to nonverbal perceptual
knowledge. This DCT assumption is supported by a variety of findings (see
Paivio, 1971). Word concreteness and imagery value, for example, correlate
negatively with the latency to generate mental images (e.g., Paivio, 1966,
1975a), and positively with the number of sensory properties that people
can think of for words (Katz, 1976) and with postexperimental reports of
spontaneous imagery during performance of various cognitive tasks (e.g.,
Bugelski, 1970; Clark and Paivio, 1989a). Such findings, and factor analyses
involving other variables (e.g., Paivio, 1968), support the DCT view that
concreteness and imagery values primarily reflect accessibility to visual im
ages and other nonverbal representations. That images for concrete words
can be generated as fast as simple word associations (Clark and Paivio,
1989a; Paivio, 1966) indicates that even conscious imagery occurs quickly
enough to contribute to word and text comprehension.
The importance of imagery value and concreteness for the meaning
and comprehension of individual words has been demonstrated in various

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 159

ways, including correlations with se

words. In one study, for example, un
while they orally defined concrete and
and meaningfulness (Reynolds and P
abstract words contained more hesi
(e.g., "ah" pauses) than definitions for
O'Neill (1972) had university student
define the meaning of 277 nouns. Ease
imagery value and .64 with rated concr
that nonverbal processes play import
A variety of rating and associatio
shown that, relative to abstract wor
superordinate categories (Toglia an
1980), produce more agreement ac
categories to which they belong (Kint
kinds of associative information (Ca
1989; Jones, 1985). Concreteness eff
important given the emphasis on hi
models of meaning. Other research s
words may be less distinctive (i.e., m
In a word association task, for exa
that randomly paired abstract word
sociations (i.e., same response given
paired concrete words. The greater c
firmed by Paivio et al. (1988a), who
mantic relatedness of randomly pai
random abstract pairs received highe
better in a Password-like game (guess
when the targets are concrete word
such effects, Clark and Paivio (1989a)
words are particularly distinct and c
as semantic coherence.

Thus, there are robust effects of concreteness and imagery value on

various measures associated with word meaning. Despite the fact that these
semantic measures have clear relevance to education (e.g., definability, con
fusability, categorizability), little research has systematically looked at the
relation between concreteness and meaning for individual words from
school subjects. In a comparison of several university courses, Donald
(1986) found considerable variation in the average concreteness of the cor
concepts. For example, the concepts in natural science courses were mor
concrete than the concepts in social science courses, including an educa
tional psychology class. DCT and the research reviewed above indicate that

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
160 Clark and Paivio

abstract words from school subjects will be more difficult to defin

categorize, will be more confusable, and will show other theoretically
pedagogically interesting differences from concrete words.
With respect to the meaning of larger verbal units, such as sente
and paragraphs, experimental research has also demonstrated a major
for imagery processes. Paivio and Begg (1971b) showed that comprehe
and imagery latencies are similar in magnitude for concrete sentence
highly correlated. Relative to abstract sentences, concrete sentences
ally are understood faster (e.g., Jorgenson and Kintsch, 1973; Klee an
senck, 1973), although not always significantly so (e.g., Paivio and B
1971b). O'Neill and Paivio (1978) also found that the random exchang
words in sentences was more disruptive to the meaning of concrete
abstract sentences. Specifically, abstract sentences with words random
stituted across unrelated sentences were rated as more comprehensib
sensible than comparable concrete sentences. One interpretation of thi
ing is that the meanings of abstract sentences are more vague (i.e., les
nite) than the meanings of concrete sentences, and therefore less disr
by random substitutions. In addition to research on literal language
creteness and imagery have also been shown to be important in proc
figurative sentences (Katz et al., 1988; Paivio, 1971, 1986; Paivio and B
1981; Paivio and Clark, 1986). Katz et al. (1988), for example, found
the rated imagery value of 484 metaphors correlated with a number
mantic attributes rated by university students (e.g., .79 with ease of com
hension and .80 with ease of interpretation). It has been hypothesized
figurative language may serve to concretize concepts by relating abstract
ics to concrete vehicles (e.g., governments are elephants).
Educational research has confirmed the importance of imagery
concreteness for the comprehension of sentences and larger textual u
Image generation and supplementary pictures generally benefit tex
prehension (see Denis, 1984), although it is not always easy to distin
comprehension effects from closely related memoiy effects discussed
Reading of educational materials seems to elicit substantial amounts
spontaneous (i.e., uninstructed) imagery. Long et al. (1989), for exam
questioned grade 5 students about their thoughts at various points du
the reading of passages from textbooks and obtained indicators of im
on over 60% of the trials. Imagery may be especially common for e
tionally arousing passages, a topic to which we return later (e.g., Sa
1983, 1985; Sadoski and Goetz, 1985). Other findings indirectly supp
the hypothesized role of imagery in reading. Glenberg et al. (1987),
example, showed that university students read text faster if object
would be spatially contiguous in an imaginai representation of the pa
were mentioned together in the text, rather than separated.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 161

The educational importance of concr

ther demonstrated by research on text r
measured text concreteness by the perce
of people, nouns that indicate a speci
-.55 between this measure and the aver
correctly answer half of the test questio
(1952; see Wharton, 1980, 1985) substit
history passages, and observed the eff
To illustrate, the sentence "With Engla
invincible on land the war became an economic contest" was revised to "With
England sweeping the seas and France overrunning the land the war lapsed
into an economic tussle." University students found the high imagery texts
more interesting and scored higher on a comprehension test than with the
originals, even though both texts were equated on traditional aspects of read
ability (e.g., sentence length) and the substituted words were less familiar
than the originals. Educational research has also demonstrated a relation be
tween text readability and counts of abstract or concrete nouns (see Klare,
1974/1975; Morris and Halverson, 1938, cited in Gilliland, 1972). Other read
ability formulae may reflect concreteness indirectly. Some methods, for ex
ample, count the frequency of basic words in the text, but basic word lists
contain few abstract words. Moreover, Cloze measures of readability, which
require subjects to guess deleted words from passages (e.g., McKenna and
Robinson, 1980), implicate various associative correlates of concreteness dis
cussed earlier, such as associative strength (e.g., Cramer, 1968; deGroot, 1989;
Paivio, 1968).
Although the positive effects of imagery and concreteness on compre
hension are consistent with DCT, such effects are not universal (e.g., Long
et ai, 1989). Rather than discrediting DCT, however, these qualifications
often involve interactions with individual differences and other specific ef
fects that demonstrate the strengths of DCT even more clearly than the
simple effects of imagery and concreteness. Consider, for example, the fact
that university students who image narrative passages read the passages
more slowly than students who do not use imagery (Denis, 1982). This find
ing, which only appears to contradict the hypothesized benefits of imagery
for comprehension, is explained by the DCT assumption that concreteness
and related effects result from activation of quasi-perceptual imagery rep
resentations. Being perceptual, images can compete with reading for shared
perceptual processes resulting in a slowed reading rate (Denis, 1982). Teach
ers who instruct their students to image while reading might therefore expect
a decline in reading rate. Conversely, an emphasis on reading speed during
instruction might decrease the likelihood of imaginai processing of text, with
its associated benefits on comprehension and, as we shall see, memoiy.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
162 Clark and Paivio

Other specific phenomena consistent with DCT concern the int

tion between individual differences and the use of imagery in text co
hension. As noted earlier, DCT assumes that people vary in the ease
skill with which they use nonverbal, imaginai processes. Individual d
ences should contribute to comprehension processes and interact with
variables, such as instructions to image. A study by Denis (1982) illus
this line of inquiry. University students were classified as high or lo
agers on the basis of a questionnaire and read descriptive passages u
instructions that emphasized speed, required imagery, or were self-p
All subjects read faster under speed than under image instructions,
there were no differences between high and low imagers in these c
tions. Under self-paced instructions, however, only high imagers rea
slower rate than under speed instructions. This finding is consistent
the hypothesis that high imagers spontaneously imaged the stories,
slowed their reading rate.
The reviewed work confirms a central role for nonverbal proce
in the representation and processing of text, including educational m
als. Nonverbal processes presumably play an even more important r
processing the large amount of educational knowledge that is nonver
nature. Visual information in schools would include maps, graphs, sc
apparatus, geometric shapes and principles, and theoretical mode
many constructs in biology and other sciences (e.g., cells, atoms, che
molecules). Moreover, imagery is implicated in the associative structu
knowledge, to which we now turn.

Associative Structure of Knowledge

According to DCT, words derive meaning from their semantic

tions with other words, as well as from images. Associative relation
tribute to the meaning of all words, and are the primary source of mean
for abstract words that lack object referents (e.g., knowledge, synt
schema, force, strategy). The essential idea is that associative relation
nect words to one another, and activation of this associative structur
tributes to the meaning of the words. Although similar to other mod
semantic memory (e.g., schema theory, hierarchical structure), DCT
phasizes diverse associative relations between verbal representations r
than highly restricted relations (e.g., superordinate, property) betwee
dal abstract representations. Moreover, images can be incorporated
the meaning of concepts, especially concrete concepts, although her
emphasize verbal associative aspects of meaning. In general, assoc
knowledge is viewed as a complex collection of diverse associations be

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 163

verbal representations, with additional referential connections to nonverbal

components of meaning.
Much of the cognitive and educational research on the structure of
semantic memory has emphasized superordinate networks, scripts, and
other hierarchical structures in which instances or component parts con
verge on superordinate nodes (e.g., Bower et al., 1979; Collins and Loftus,
1975; J. M. Mandler, 1984). From a DCT perspective, hierarchical struc
tures reflect associations in which multiple words converge on superordi
nate labels (Clark and Paivio, 1984). Verbal representations for "dog,"
"cat," "lion," and other animal words, for example, become associated with
the verbal representation for "animal." Early research on verbal concept
learning demonstrated that activation of superordinate and related category
information depends on such variables as the strength of the individual
links between instances and concepts, the spacing between presentation of
instances from the same category, and instructions (e.g., Underwood and
Richardson, 1956). Instructions to find what is shared by a set of instances,
for example, increases the likelihood that the shared category or other com
mon feature is identified.
Many facets of educational knowledge can be conceptualized in terms
of verbal associative networks. Indeed, the prototypical examples of a hi
erarchical network are biological and other science taxonomies in which
words are linked into an associative hierarchy with multiple levels that are
themselves labeled (e.g., genus: lion -> family: Felidae -> order: Car
nívora -> class: Mammalia -► phylum: Vertebrata -» kingdom: Animalia).
Hierarchical structures have also been proposed for the representation of
text (e.g., Kintsch and Van Dijk, 1978). Some texts activate preexisting se
mantic structures, but even when prior schemata are not available, com
prehension processes construct an organized hierarchy of statements using
linguistic cues to integrate sentences that vary in importance or centrality
(e.g., Halliday and Hasan 1976; Kintsch and van Dijk, 1978). This research
has produced detailed models of the cognitive processes that underlie men
tal outlines of text, such as identifying both main ideas, which are assigned
central roles, and extraneous material, which can be excluded from the as
sociative network or placed at lower levels (e.g., Williams et al., 1981; van
Dijk and Kintsch, 1983).
Hierarchical and related associative structures relevant to different
school subjects have been measured and studied using a variety of tech
niques (e.g., Preece, 1976; Shavelson, 1974). In an early study, Johnson
(1964) had high school students with different science backgrounds free
associate to physics terms (e.g., mass, momentum, velocity). He found that
the associative structure of the terms varied with the amount and recency
of instruction in science. For example, students with relevant classroom in

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
164 Clark and Paivio

struction were more likely to respond with other terms on the list that
related to the stimulus word by scientific principles, whereas student
ing instruction were more likely to respond with nonscience terms n
the stimulus list. A variety of methods have now been developed to
raw data on cognitive structures (e.g., sorting words, rating similar
Preece (1976) has demonstrated that the different methods lead to si
conclusions about the underlying cognitive structures. Increasingly s
ticated measures are also being developed to extract various propert
associative structures from the raw data. Measures of depth (e.g., nu
of levels), complexity (e.g., number of branches), and other proper
cognitive structure have been related to such educational factors as in
tion, expertise, and age (e.g., Fisher, 1988; Nagy, 1984).
A study by Naveh-Benjamin et al. (1986) illustrates the directio
this research. University students and their instructor sorted 16 terms (e
senile dementia, retrieval, integrity, intimacy) from a Psychology of
course into any order. Student sortings were done at the beginning, m
or end of the course, and on each occasion sortings were done four t
The sortings were analyzed to produce measures for degree of organiz
(lack of randomness), similarity to instructor, and other aspects of cogni
structure. These derived measures were in turn correlated with class g
and showed expected changes from pre- to postinstruction. For exa
associative structures became more complex (less random) and more
lar to the associative structures of the instructor, and these effects w
stronger for students who did well in the course than for students w
poorly.
In their study, Naveh-Benjamin et al. (1986) represented the cogni
tive structures of students and instructors as tree diagrams in which higher
level words are at the uppermost level, specific terms at the lowest level,
and intervening category names between these extremes. Students and
teachers often make similar outlines in the form of diagrams (see later
discussion). This frequent use of spatial methods to represent relations
among the verbal elements in cognitive structures (e.g., hierarchical trees,
multidimensional scaling results) illustrates the intimate relation between
verbal associative knowledge and imagery. Specifically, cognitive maps use
spatial relations to represent the associative links among verbal repre
sentations in a nonverbal way (i.e., as a diagram). In an outline of an
associative hierarchy, for example, lines between words represent class-in
clusion relations and indicate that one element is a subordinate or super
ordinate of the other. These spatial representations for associative
hierarchies are considered again when we examine memory and related
educational phenomena.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 165

The use of imagery to represent ve

onstrates the strengths of DCT's emph
effects of the verbal and nonverbal s
tween words and images is also suppo
tween concreteness and categorizabili
Kintsch, 1974; Toglia and Battig, 19
may somehow facilitate acquisition of
archical structures.
This section has reviewed educational and cognitive phenomena that
are consistent with the DCT characterization of the structure of knowledge.
In short, meaning and cognitive structure result from the separate and col
lective actions of the imagery and verbal associative systems. These same
processes explain learning and memory effects of interest to educational
researchers and teachers.

LEARNING, MEMORY, AND STUDY SKILLS

The successful transmission of new skills and knowledge depends

on student learning and memory processes that have received much at
tention from educational and cognitive researchers. From a DCT per
spective, learning and remembering involve the same imagery and
associative processes discussed in the preceding section on the structure
of knowledge. Indeed, comprehension and memory are investigated with
similar tasks, and the boundary between the two processes is often fuzzy.
For example, it is unclear how much time must pass between reading a
passage and answering questions before the questions tap memory rather
than comprehension.

Imagery Processes in School Learning

One of the first insights associated with DCT was the role of non
verbal, imaginai processes in learning and memoiy tasks. Early DCT re
search demonstrated concreteness and imagery effects in paired-associate
and other experimental learning tasks (Paivio, 1969), including memory for
paragraphs (Yuille and Paivio, 1969). The predicted mnemonic benefits of
imagery were explained by DCT in terms of two imagery-based processes —
elaboration and organization.
The elaborative or "dual coding" explanation for imagery effects es
sentially states that the additive effect of imagery and verbal codes is bet
ter than a verbal code alone (Paivio, 1975b). The benefits of imaginai

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
166 Clark and Paivio

elaboration emerge clearly when subjects are instructed to image to

remembered material. Generating images produces better recall tha
peated encoding conditions (i.e., repeating target words aloud or silen
and even better memory than such deep encoding operations as trans
into another language (Paivio and Lambert, 1981) or generating synon
(Vaid, 1988). Some research also suggests that traditional seman
structions used to invoke deep levels of processing (cf. Craik and Lock
1972) may also involve imagery processes (D'Agostino et al., 1977).
Concreteness and imagery also benefit memory because of the
cial organizational capacities of the imagery system. Specifically, sep
elements can be integrated into a unified or compound image that
sequently permits part of the image to reactivate the whole (Begg,
Bower, 1970; Paivio, 1969). The capacity of a partial cue to react
the entire representation is known as redintegration. For example,
construction of an interactive image of a dog sitting on a chair per
later presentation of the dog or chair alone to redintegrate the ent
image and thereby mediate retrieval of the other object. This integr
or organizational capacity of imagery was first noted in terms of th
ceptual peg hypothesis, which states that concrete words are particu
effective cues for retrieving compound images of stimulus-response
Consistent with this hypothesis, variation in stimulus concreteness
stronger effects in paired-associate tasks than response concreteness
(Paivio, 1963). Integration also provides a plausible explanation for the
effectiveness of interactive images in mnemonic techniques and other
memory strategies.
Marked effects of concreteness and imagery on memory for educa
tional material have been reported, and the results are generally consis
tent with the experimental research and with the DCT analysis. The
interactive images in the keyword mnemonic technique, for example, link
to-be-learned translations with familiar keywords that sound like the un
familiar vocabulary words. Later presentation of the new word, or its
English equivalent, cues retrieval of the keyword, which in turn reinte
grates the image and the translation. This technique has been shown to
facilitate vocabulary learning in various educational domains, including
native- and second-language learning (Atkinson, 1975; Levin and
Pressley, 1985). Keyword and other imagery techniques have considerable
relevance for education (see Bellezza, 1981; Paivio and Desrochers, 1981;
Roediger, 1980) because new vocabulary is a major element in school
learning, perhaps especially in languages and sciences (Eylon and Linn,
1988). Even 60 years ago, a sample of high school biology notebooks
contained over 600 different science terms that students were expected
to know (Baird, 1931).

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 167

Consistent with our earlier analysis of meaning, imagery and con

creteness also play an important role in memory for text (Anderson, 1974;
Begg and Paivio, 1969). Concrete phrases, sentences, and paragraphs are
remembered better than abstract text (e.g., Yuille and Paivio, 1969).
Moreover, subjects receiving imagery instructions remember more than
subjects who do not receive such instructions, and subjects who report
spontaneous images remember more than subjects who do not report
images (see Denis, 1984, for a review). Sadoski (1984), for example, found
that recall scores for a basal reader story correlated .49 with number of
text-specific images reported by grade 3 and 4 students. Further evidence
for the benefits of imageiy come from research on the role of mental
models in memory for text (e.g., Gentner and Stevens, 1983; Johnson
Laird, 1983), inasmuch as mental models are described as percept-like rep
resentations (i.e., images) for situations described by text (Glenberg et al.,
1987). Constructing or processing a supplementary map, one kind of men
tal model or image, improves memory for related text (Dean and Kulhavy,
1981); and, more generally, mental models (e.g., diagrams for scientific
constructs) facilitate memory for ideas and generalization of principles
from explanatory texts (Mayer, 1989).
The preceding studies report effects for concreteness of the target
information, but memory is also influenced by the concreteness of such
supplementary material as advance organizers (Ausubel, 1960). Corkill et
al. (1988), for example, observed that undergraduate education majors re
called more information about 1200- and 5000-word passages of text (e.g.,
chapter on introductory linguistics) when preceded by concrete advance or
ganizers as opposed to abstract advance organizers. Abstract organizers, in
fact, did not facilitate memoiy relative to no organizer. This superiority of
concrete advance organizers is analogous to the early work on concrete
words as conceptual pegs and may similarly result from the special capacity
of concrete words for redintegration of compound images. Imagery may
also contribute to other manipulations that improve sentence memory, such
as precise elaborations. Bransford et al. (see Bransford et al., 1982, for an
overview) demonstrated that precise elaborations of sentences facilitated
associative learning by both university and fifth-grade students. For exam
ple, completing the sentence, "The tall man took the cookies" with an
elaboration relevant to height (e.g., "from the top shelf') produced better
memory for what the tall man did than a completion in which height was
incidental (e.g., "from the counter"). Using a similar task, Pressley et al.
(1987) demonstrated that student-generated elaborations (i.e., asking why
certain facts might be associated with a particular concept) are particularly
effective ways to promote associative learning in undergraduates. Precise
elaborations and other manipulations in the text memory literature may

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
168 Clark and Paivio

help memory in part because they elicit more imagery than do con
conditions.
Another kind of evidence for the important mnemonic role of imageiy
in educational learning comes from research on study skills. A number of
behaviors associated with the quality of study skills can be understood in
terms of imagery processes. An elaborative imagery factor emerged from
a factor analytic study of Kulhavy and Kardash (1988). They obtained self
reports from undergraduate education students about the frequency with
which they engaged in various study activities. The elaborative imagery fac
tor included such imaginai operations as generating mental images and
writing examples for school material, and accounted for much variation in
self-reported study activities. One benefit of generating examples may be
that they evoke images for personal experiences or imagined concrete
events. Items relevant to imaginai processing also occur on standardized
tests and behavioral analyses of study skills and habits. The Study Report
Form (Johnston, 1975), for example, is a behavioral based system devel
oped for university students to record time spent in various study practices.
The form includes items related to elaborative imagery (e.g., writing exam
ples). Schmeck's Inventory of Learning Processes (Schmeck, 1983; Schmeck
et al., 1977; Schmeck and Ribich, 1978) similarly contains an elaborative
processing factor that includes such items as the use of visual imagery.
Schmeck (1983) reviews evidence that elaborative processing scores differ
entiate students of different abilities (e.g., Moss, 1982), and correlate with
grades and experimental learning tasks (e.g., Schmeck et al, 1977). Elabo
rative processing scores also correlate positively with individual differences
in imagery ability (Schmeck and Ribich, 1978), consistent with an imagery
explanation for the memory effects.
Imagery processes are further emphasized in books and programs de
signed to improve learning and study skills. An early text by McMurry
(1909) included a chapter on supplementing thought, which recommended
the use of such imagery-related methods as imagination, elaboration, mak
ing illustrations, experiencing the material, and listing details. Contempo
rary study guides (e.g., Robinson, 1970) and study skills programs stress
similar processes. Dansereau et al. (1979), for example, included imagery
and network construction methods in their program. The network or cog
nitive mapping methods use spatial imagery to represent verbal associative
structures (see sections on associative processes in the structure of knowl
edge and in memory). Imaginai elaboration is also a central component in
the cognitive learning strategies program of Weinstein and her colleagues
(e.g., Weinstein et al., 1979) and in Wittrock's model of generative learning
(e.g., Wittrock and Alesandrini, 1990).

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 169

Much of the research on study sk

elaborative imagery is relevant to s
onstrated to improve memory and co
example, a recent review by Pressle
strategies as constructing internal ima
paring your own life with the text,
elaborative processes are also centra
reading (e.g., Gagne, 1985, Ch. 7) an
signed for use with students of diffe
These diverse findings clearly dem
benefits of imagery for educators.
about the specific conditions under w
about the mechanisms that underlie
fits of mental models vary with studen
1989; Wittrock and Alesandrini, 1990
does not always facilitate memory f
Baird, 1988). As noted earlier, howe
differences and attributes of mater
tions for such findings. Riding and
verbalizer-imager learning styles by
11-year-old children could answer i
prose passage. The children later lea
highly semantic (i.e., verbal) passage
showed no overall differences in th
remembered the highly visual passag
bered the highly semantic passage be
DCT predictions and demonstrate th
tive that accommodates attributes of
ences, and instructions.
Other unresolved questions con
that underlie imagery effects. Some
imagery enhances integration of co
tive results (e.g., Marschark, 1985
ternatives to DCT have been prop
effects in experimental memory ta
Marschark et al., 1987). These issues
note that the verbal and imaginai
and still not fully understood, and,
a central place to individual differen
that complicate simplistic applicatio
negative results is also complicated
to learn about the basic cognitive m

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
170 Clark and Paivio

ory tasks, including mnemonic techniques of interest to educators

Desrochers and Begg, 1987; Paivio and Desrochers, 1981). More
some problematic phenomena, such as integration effects for abstrac
terials, can be explained by considering the role of DCT's verba
ponents in memory.

Associative Processes and Memory

DCT states that imagery and verbal associative processes jointly

termine learning and memory performance, with direct and indirec
ciations between verbal codes influencing storage and retrieva
information. One robust finding that reflects associative processes
effect of organization on learning in various verbal learning paradigm
example, lists of related words from superordinate categories are b
remembered than unrelated words (e.g., Bousfield, 1953), and paired
sociate learning increases with the strength of direct associations bet
the two terms (Murray, 1982) and with the probability that both wor
indirectly linked by a shared associate (Miller, 1970). With respect to
ory for categorized lists of words and related organizational effects,
emphasizes the contribution of indirect verbal associations in which
ple words or phrases converge on a shared associate (Clark and Paivi
1984). The shared associate can subsequently act as a retrieval cue fo
specific instances that originally elicited it. Retrieval of dog, horse
tiger, for example, would be facilitated by the convergent associate
mal" being activated during study and serving as a retrieval cue du
recall.

Similar associative processes contribute to memory for text and other

educationally relevant material. Memory for text, for example, benefits
from preexisting associative structures, such as a familiar schema or script,
and information central to the main theme of the passage is remembered
particularly well (e.g., Anderson, 1977; Bower et al., 1979; J. M. Mandler,
1984). Associative structures also explain effects on learning and memory
of advance organizers (Ausubel, 1960), statements of learning objectives,
and related methods that involve priming the underlying structure of text.
Although much of this research involves use of available semantic struc
tures, readers also impose associative organizations on text material that
lacks an obvious structure. Memory is better, for example, for statements
that are at a higher node in a text's inferred structure than for statements
at lower levels (Kintsch and Keenan, 1973; Meyer, 1977).
The mnemonic benefits of schema, scripts and categories, and related
organizational effects, are consistent with the DCT associative model. At

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory arid Education 171

one time, schematic organization was thought to produce better memory

than categorical organization (Rabinowitz and Mandler, 1983), suggesting
different underlying mechanisms. However, Khan and Paivio (1988) dem
onstrated that lists organized categorically or schematically produce equal
recall when other aspects of the learning situation are equated. Specifically,
university students recalled category and schema lists equally well when
both lists contained labels for the clusters (e.g., animals, living on a farm)
or when neither contained organizing names. Comparisons favored sche
matic organizations only when schematically organized lists with labels were
compared to category lists without labels, as in the Rabinowitz and Mandler
(1983) study. Thus, the findings are consistent with verbal associative views
of memory for organized materials (see also Khan, 1989).
Research on study skills further demonstrates the importance of verbal
associative structures and their spatial representation. Such associative op
erations as outlining, summarizing, and writing important points constitute
one factor in Kulhavy and Kardash's (1988) analysis of study activities. Items
on organization of notes are also included on the Survey of Study Habits
and Attitudes (Brown and Holtzman, 1967) and the Study Skills Test (Ray
gor, 1970). Constructing outlines and other items that implicate verbal and
imaginai associative processes are part of the Study Report Form (Johnston,
1975), and finding related themes and organizing ideas is part of the deep
processing factor on Schmeck's (1983) Inventory of Learning Processes. An
associative explanation for the deep processing factor is supported by evi
dence that it correlates with ability to organize lecture material into asso
ciative hierarchies (Ribich, 1977, cited in Schmeck, 1983). Deep processing
also correlates with student abilities, grades, and performance on other rele
vant tasks (e.g., Moss, 1982; Schmeck, 1983; Schmeck et al., 1977).
The mnemonic benefits of verbal associative processes are empha
sized in books and programs designed to improve study skills. A chapter
in McMurry (1909) on organization of facts included such associative strate
gies as outlines, headings, note-taking, grouping facts, finding relations, sub
ordination, and hierarchies; and similar processes appear in contemporary
study guides (e.g., Robinson, 1970), study programs (e.g., Weinstein et al.,
1979), and theories of effective studying (e.g., Wittrock and Alesandrini,
1990). Study guides also recommend ways to represent verbal associative
structures, including imagery-based methods. Spatial outlines and graphic
organizers are used in schematizing and networking study techniques (e.g.,
Buzan, 1974; Holley and Dansereau, 1984; Novak and Gowin, 1984), and
the network construction methods taught by Dansereau et al. (1979) involve
imaginai representations for verbal associative structures. To illustrate, a
cognitive map of a lesson or chapter on Japan might be summarized as a
tree diagram with Japan as the top heading, such subheadings as geography

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
172 Clark and Paivio

and history at the next level, and so on. Graphic representations for a
ciative structures have also been examined as text supplements (Guri-
enblit, 1989; Simmons et al., 1988). These spatial methods blur the divi
between imagery and verbal associative processes, and may capitalize
the capacity of the nonverbal system to integrate spatially contiguou
events, including verbal labels, in an associative network.
Associative structures and their spatial representations appear
benefit students of all ages. Much of the research reported above was
cerned with older students, but the Pressley et al. (1989b) review conclu
that a number of verbal associative methods were effective with young
dren, including summarizing text and constructing story maps. Associa
components to the structure of text are also central components in mo
of reading (e.g., Gagne, 1985, Ch. 7) and writing (Kellogg, 1988) relev
to students of all ages.
We have reviewed the effects of both imaginai and verbal associat
factors in learning and memory. Although treated here largely as sepa
processes, DCT emphasizes their collective and interactive effects on
havior. We have already seen that verbal associative structures can be r
resented imaginally. It also seems likely that scripts and related associat
structures benefit from spontaneous imagery of script events. That is, usin
scripts to encode episodes (e.g., a visit to the library, a school day, a scie
experiment) may evoke visual images of typical events in the script (e.
setting up the equipment, taking measurements, writing down results).
agery may also play a variety of roles in text cohesion, another aspect
associative meaning (Halliday and Hasan, 1976). Synonyms facilitate te
integration, for example, and concrete synonyms, which share a comm
image, are more strongly connected to one another than are abstract s
nyms (Clark, 1978,1984; Paivio et al., 1988b). These examples involve im
inai effects on associative processing, but verbal associative processes
influence imagery. The ease of generating interactive images, for exam
depends in part on associative relatedness, suggesting that interactive
agery is not a pure nonverbal process (Paivio et ai, 1988a).
Associative processes also explain problematic imagery results discussed
earlier, such as failures to find evidence for the special integrative rol
imagery in memory for sentences and other verbal materials (e.g., Marschar
and Paivio, 1977). Paivio and Walsh (summarized in Paivio, 1991) recen
tested the hypothesis that verbal associations between abstract words provid
integrative retrieval cues that partly mask the integrative benefits of imag
Cued recall of concrete and abstract pairs varying in associative relatedn
was compared to free recall of the same pairs. The superiority of cued r
over free recall is one measure of integration (Begg, 1973). Paivio and W
found that the differences between cued and free recall were substantial and

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 173

similar for concrete-related, concrete-unrelated, and abstract-related pairs,

but that abstract-unrelated pairs showed little evidence for integration. This
finding demonstrates that high imagery is sufficient for integration of concrete
words, even in unrelated pairs, whereas strong verbal associations are neces
saiy for the integration of abstract words.
In conclusion, many findings support the hypothesis that memory for
words and text benefit from elaborative imagery, concreteness, and associa
tive organization, although the collective effect of these processes are com
plex and a number of important questions remain unresolved. Moreover,
the DCT emphasis on verbal and imaginai processes fits well with several
cognitive models of learning and memory in the educational domain (e.g.,
Schmeck, 1983; Weinstein et al., 1979; Wittrock and Alesandrini, 1990).

OTHER COGNITIVE TOPICS IN EDUCATION

The structure of knowledge and the processes that underlie le

and memory are basic to many aspects of educational psychology.
show that DCT provides a unifying perspective on two other cogn
pects of education: the effectiveness of instruction and educationa

DCT and Effective Instruction

Preceding sections have demonstrated the importance of imagery,

concreteness, and verbal associative processes for comprehension and mem
ory of school material. It is expected, therefore, that these same factors
influence the effectiveness of teaching. We consider imagery and verbal
associative factors in turn.
The research demonstrating that imagery and concreteness play cen
tral roles in the representation and acquisition of knowledge is directly rele
vant to instructional practices. The positive effects of concreteness and
imageiy on the readability of texts and on memory, for example, generalize
to oral transmission of information in the classroom. That is, lessons con
taining concrete information and evoking vivid images will be easier to com
prehend and remember than lessons that are abstract and not
image-arousing. Moreover, the same imagery manipulations that benefit
memory for text should also benefit memory for orally presented informa
tion as in classroom lessons. Levin and Berry (1980), for example, asked
fourth graders to recall information from tape-recorded newspaper stories.
Children who listened to the stories while viewing relevant pictures recalled
more than children who only heard the stories.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
174 Clark and Paivio

The instructional benefits of imagery and concreteness also appe

in works on teaching skills and effectiveness. An early and extensive ef
to document teaching activities was the Commonwealth Teacher-Train
Study (Charters and Waples, 1929). In this study, teachers, school adm
istrators, and other participants generated a large number of teachin
traits and activities. The resulting items were subsequently categorized
rated on such dimensions as importance. One large category of activi
was called "Teaching Subject Matter," which included many activities
gesting a direct role for concreteness and imagery in teaching. Pedagog
practices related to imagery included using effective pictures, diagra
models, and other illustrations for lessons, and showing relationships
tween school subjects and events in real life. These recommendations
allel imagery-related suggestions discussed earlier in the context
effective study strategies.
The relevance of imagery and concreteness to teaching is further s
ported by empirical studies of teachers in the classroom. Student rat
of teacher behaviors, for example, produce evidence for factors relevan
imagery-concreteness. At the university level, Murray (1983a, 1985) fo
that such behaviors as "uses concrete examples to explain abstract ide
and "relates subject matter to current events" loaded on a clarity/conc
ness teaching factor. In research on the effectiveness of elementary m
matics teachers, Ebmeier and Good (1979) found that teachers vary in t
degree of abstractness, defined as "using abstract concepts or techniq
or materials with which the students have little familiarity," and that teac
ers who were more educated and secure than other teachers obtained low
scores on this dimension (i.e., they were high on concreteness). Given th
evidence cited earlier for a relation between concreteness and such com
munication variables as fluency and distinctiveness, the relevance of con
creteness to instruction may also be reflected in other correlates of teaching
effectiveness, such as presentation clarity (Frey et al., 1975) and (negatively)
vagueness (Dunkin and Doenau, 1980; Hiller et al., 1969).
This research on teacher effectiveness has led educational psycholo
gists and others concerned with effective teaching to advocate instructional
methods that can be conceptualized in terms of imaginai processes. Wool
folk (1987, pp. 423-425), for example, includes the following recommenda
tions to improve the clarity of teaching and reduce vagueness: "Use concrete
examples or analogies that relate to the students' own lives," "Use models,
examples, and illustrations," and "Use specific (and, if possible, colorful)
names." The importance of concreteness has also been emphasized in writ
ings on effective lecturing at the university level (e.g., Clark and Clark, 1970,
originally published in 1959), although Clark and Clark make the important
point that the concrete illustrations must be relevant to the lecture.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 175

Theoretically, DCT suggests that v

crete and personal examples, and re
hension and retention of lessons b
increasing the arousal of mental ima
expected benefits, ratings of teacher c
with student achievement (e.g., Fr
teacher and course effectiveness (
(1980) also reported that teacher vag
and clarity, had negative effects on
teacher clarity do not always correl
formance or ratings of the amount l
relations are generally positive. Mur
average clarity ratings for 36 introduc
only .16 with average final exam perfo
ratings for amount learned in the co
Similar complexities were observed
imagery on text comprehension and
complexities can be explained by a D
nisms underlying imagery and concr
instructional effectiveness suggests p
and directions for future research. If
by evoking images, for example, th
imagery and see whether effects of
strongly or only for students who
amples. Moreover, the likelihood of
generation during lectures will vary
its concreteness) and individual di
students).
In addition to imagery, DCT states that verbal associative processes
contribute substantially to the effectiveness of instruction. Evidence is gen
erally consistent with this premise. In their early study of teaching activities,
Charters and Waples (1929) found that the "Teaching Subject Matter"
category included many activities related to what we call associative or
ganization. For example, teachers develop outlines for their lessons, arrange
units into orderly sequences, try to summarize the ideas that they are teach
ing, and generally attempt to organize their subject matter knowledge.
Indicators of verbal associative organization have been identified in
research on effective teaching (e.g., Frey et al., 1975; Marsh, 1982), al
though imagery may contribute here as it does to organizational effects in
comprehension and memory. Associative factors on teacher rating scales
include such behaviors as "puts outline of lecture on the board" and "gives
preliminary overview of lecture" (Murray, 1983a, 1985). The explicit use

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
176 Clark and Paivio

of free association and spatial mapping techniques to prepare lecture

lessons (Brown and Atkins, 1988) provides further support for the im
tance of associative structure in teaching. That is, Brown and Atkin
ommend that lectures be prepared by free-associating to the main th
of the lecture and then arranging the ideas into a meaningful organi
using a tree diagram or some other spatial representation for the s
matter. Spatial representations for cognitive structures of teachers ar
obtained in many studies of student cognitive structures that were descri
earlier, presumably on the assumption that the cognitive structure
teachers relate closely to the objectives of instruction.
Research on associative factors in instruction has led to recommen
dations that now appear in educational psychology and related teacher
education courses. Woolfolk (1987, pp. 423-425), for example, includes
the following suggestions to improve teaching: "Organize your lessons
carefully," "Work on an outline with the class," "Break the presentation
into clear steps or stages," and "Signal transitions from one major topic
to another." Hierarchical outlines are also emphasized in characterizations
of effective university instruction. Clark and Clark (1970, p. 13), for ex
ample, suggest that lecturers should "aim at getting home a limited num
ber of points, well defined, properly emphasized and arranged in some
sort of logical order."
From a DCT perspective, verbal associative organization and related
imagery methods can benefit instruction in several ways. One benefit for
student learning is that outlines and related behaviors parallel the verbal
associative structures that underlie knowledge and its acquisition. A teacher
who presents a lesson on Japan in terms of well-identified topic headings
(e.g., geography, history) and subheadings (e.g., major landforms, climate)
is enhancing the likelihood that category names are being acquired, and
that links are being formed between category names and specific pieces of
information. This organization will benefit comprehension and subsequent
retrieval. Such associative strategies as tree diagrams may also benefit
learning because they promote integrative imagery and thereby facilitate
cued retrieval. For example, students who saw a hierarchical outline for a
lesson on Japan or were encouraged to develop their own outlines could
then use a mental image of that outline to study, to retrieve the information
on a test, to organize their answer for an essay question, and to perform
other educational tasks.
This theoretical model for the benefits of associative structures and
mental outlines is consistent with research reviewed earlier on the role of
verbal associative and imaginai organization in text comprehension, mem
ory, and study skills. Such mechanisms explain why the use of effective
associative organization in lessons correlates in some nonexperimental stud

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 177

ies with measures of student achievement (e.g., Frey et al., 1975) and with
ratings of teacher and course effectiveness (e.g., Murray, 1983a). But as
noted earlier for teacher clarity, the relations are not always significant
(e.g., Murray, 1983a; Sullivan and Skanes, 1974). Studies that manipulate
teacher behaviors have also produced support for the cognitive benefits of
lesson organization. In one experimental study, Clark et al. (1979) com
pared sixth-grade teachers using high lesson structure (e.g., stating objec
tives at beginning of lesson, outlining content) with the same teachers using
low lesson structure (absence of structural behaviors). Structuring had posi
tive effects on several measures of student achievement.
DCT would approach these somewhat varied outcomes and the more
general problem of understanding teaching effects by designing studies that
would assess the putative underlying mechanisms. In the case of organiza
tional effects, for example, researchers could examine the quality of stu
dents' notes when teachers did or did not use outlines (e.g., presence or
absence of schematic outlines in notebooks), measure student cognitive
structures using sorting and other methods discussed earlier (e.g., degree
of structure, similarity to instructor), and assess interactions between lesson
organization and relevant individual differences (e.g., verbal and imagery
abilities, study skills). Such research would reveal the cognitive mechanisms
underlying instructional effects, including the relative contributions of ver
bal and nonverbal processes. An analytic approach based on DCT may
complement other efforts in educational and cognitive psychology to ex
plain instructional effects by basic cognitive processes (e.g., Gliessman et
al., 1988; Rosenshine and Martin, 1974; Winne and Marx, 1977).

DCT and Educational Testing

We have already observed that individual differences can qualify ef

fects of instructions, materials, and other variables in cognitive and educa
tional research. Testing of individual differences is a major activity in
education and serves many purposes, including student selection into spe
cial programs and differential instruction within the classroom (Salvia and
Ysseldyke, 1988). A large part of educational assessment involves tests of
intelligence, achievement, and related cognitive processes. DCT provides a
useful framework for thinking about cognitive tests. In particular, the DCT
hypothesis of distinct verbal and nonverbal cognitive systems corresponds
to analogous distinctions in most tests and theories of intelligence.
With respect to the nonverbal or imagery system, intelligence tests
usually include subscales that measure nonverbal abilities, and factor analy
ses of general test batteries identify a perceptual-spatial ability factor that

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
178 Clark and Paivio

is distinct from verbal abilities (Anastasi, 1988). Nonverbal or imagin

processes contribute to the Performance scales of Weschler's test (Wes
ler, 1974) and of Jackson's (1984) Multidimensional Aptitude Battery,
to the simultaneous processing dimension (e.g., Das et al, 1975; see Pai
1975c) that underlies the Kaufman Assessment Battery for Children (K
man and Kaufman, 1983). More specifically, the Wechsler tests includ
such imaginally laden tasks as object assembly and block design. Jackso
test also has a paper and pencil version of the mental rotation task, wh
has been used in many experimental studies of imagery processing and
volves the comparison of figures (e.g., shapes or letters) at different
entations (e.g., tilted 45 degrees left or right). Models of intelligence ot
than two-factor theories also assign a major role to nonverbal abilities. W
respect to Guilford's (1967) Structure of Intellect, for example, imag
and spatial abilities underlie the Figurai Transformations component, w
includes block rotation, Minnesota Paper Formboard, and other specif
measures of imagery abilities (see Paivio, 1986, Ch. 6).
The verbal side of DCT is reflected in language measures on in
ligence and related tests, including the Verbal scales of the Wechsler t
(Wechsler, 1974) and of Jackson's battery (Jackson, 1984), the success
processing dimension of the Kaufman (Das et al, 1975; Kaufman a
Kaufman, 1983; see Paivio, 1975c), and various components in Guilfor
model (Guilford, 1967). Factor analytic studies have produced evidenc
for a distinct verbal ability factor (Anastasi, 1988). Examination of ver
tasks demonstrates the central role for verbal representations and rela
associative processes. Similarities items on the Wechsler tests, for exam
require the identification of commonalities between two words (e.g.,
and boat) and measure convergent associative processes. That is, corre
responses require the activation of a shared associate of the two word
(e.g., transportation). Verbal tests that require the production or iden
fication of synonyms, antonyms, and other related terms similarly inv
verbal associative processes.
In addition to its correspondence with general taxonomies of abilit
DCT's theoretical mechanisms permit detailed models for a wide variet
cognitive tasks used in educational assessment. Models for different ta
are constructed from hypothetical networks of associatively and referentia
related mental representations for verbal and nonverbal components of
task. This analytic approach is consistent with other efforts in educati
and cognitive psychology to explain test performance in terms of underlyin
cognitive processes (e.g., Glaser, 1981; Pellegrino, 1987; Ronning et al., 1
Sternberg, 1981). To illustrate, the Similarities task discussed earlier can
modeled in terms of verbal representations for instance and superordi
names, and associative relations between the instances and the supero

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 179

nates. Failure on the task can result from

of the model, such as missing represe
tation"), absent connections (e.g., "boat
or failure to suppress interfering asso
A second illustration of how DCT
tasks is provided by picture vocabula
assess child language abilities and dys
education and clinical settings, and in
et al., 1981; Snowling et al., 1988; van
production tests that require active n
Word Picture Vocabulary Test (Gardn
are associated with a DCT model of pic
with a single label (e.g., apple, scissor
with multiple labels (e.g., cat, purse). T
(i.e., number of different names) has
errors (Johnson and Clark, 1988) and
research on naming (e.g., Lachman, 1
hypothesized that pictures with many
nections and are difficult or slow to name either because activation is dif
fused across the alternative pathways, or because the alternative responses
that are activated compete with and inhibit one another (Clark, 1988;
Clark and Johnson, unpublished; Paivio et al., 1989). A similar mechanism,
namely interference from competing instance names, contributes to the
exceptional difficulty young children have on naming tests that require
the retrieval of superordinate names (Johnson and Clark, 1988; see also
Clark and Johnson, unpublished). For example, young children might have
difficulty calling a picture of an apple "fruit" because they cannot suppress
the more available label "apple."
DCT also provides analytical models for nonverbal tests and for edu
cational correlates of imagery ability. Nonverbal tasks involve the special
properties of the imaginai system described earlier, such as the capacity to
integrate and redintegrate information. One property that has been par
ticularly important in the individual difference domain is the capacity of
images for spatial transformations. Many spatial tests, such as mental ro
tation tasks, involve dynamic transformations of spatial stimuli (e.g., Loh
man and Kyllonen, 1983). Gender differences in imagery ability are
particularly robust on such dynamic imagery tasks (e.g., Linn and Hyde,
1989; Maccoby and Jacklin, 1974; Paivio and Clark, 1991). Understanding
how imaginai transformations are performed could therefore shed light on
some controversial issues associated with individual differences in imagery
ability and mathematics education, including controversial relations be
tween gender, mathematics, and imagery ability (e.g., Benbow and Stanley,

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
180 Clark and Paivio

1983; Burnett et al., 1979; Hyde, 1981; Linn and Hyde, 1989), and the
periential or biological origins of any differences (e.g., Baenninger an
Newcombe, 1988; Kimura, 1987). The mechanisms underlying imag
tasks, mathematics, and the relation between imagery ability and ma
matics-science (e.g., Paivio, 1983a; Shepard, 1978) are also relevant to ot
educational phenomena, such as mathematics-specific learning disabili
(Ozols and Rourke, 1988). Moreover, imagery and other DCT mechanis
have been used to develop a general theory of number processing (Cla
and Campbell, 1991).
Individual differences in imagery abilities are also relevant to man
of the educational phenomena discussed in earlier sections of the pape
Reaction time studies have demonstrated that people who score high
measures of imagery ability and habits image more quickly than peop
who score low on such measures, perhaps especially for abstract word
(Ernest and Paivio, 1971). This finding explains why individuals of hi
imagery ability are more likely to use imagery when reading normal
whereas individuals of low imagery ability only image when instruct
(Denis, 1982). Imagery latencies also explain individual difference
memory effects of imagery (e.g., Schmeck and Ribich, 1978). Specifica
people who find it effortful and slow to generate images are not likely
spontaneously image to help their comprehension and memory, and i
deed forced use of imagery might have a negative effect on their perfo
ance. Analytic studies of the role of word attributes in image genera
tasks (e.g., Paivio et al., 1989) also further our understanding of indivi
differences, by revealing the mechanisms involved in imaging (e.g., wo
identification, activation of referential connections to images) and by
viding precise information about the kinds of items that differentiate h
and low imagers.
A final contribution of DCT to individual difference research follows
from its emphasis on the role of experience in human cognition. Because
experience and practice determine the ease and likelihood that imagery
and other cognitive processes are activated, performance depends on men
tal habits or preferences as well as abilities. The Individual Differences
Questionnaire (IDQ; Paivio, 1971; Paivio & Harshman, 1983) was devel
oped to assess tendencies or preferences for processing information either
verbally or imaginally, Respondents are asked, for example, to what extent
statements such as the following are true of them: "When remembering a
scene, I use verbal descriptions rather than mental pictures" and "I often
use mental pictures to solve problems." Other questions ask about how
well people think they can perform various verbal and nonverbal opera
tions. IDQ scores correlate somewhat with relevant measures of ability
(e.g., Ernest and Paivio, 1971), but also show some unique effects. For ex

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 181

ample, scores for verbal and imagery habits tend to be uncorrelated,

whereas verbal and nonverbal abilities are positively correlated.
In this section we have shown that DCT is relevant to educational
research on instruction and assessment. Imagery and verbal associative
processes provide a unifying framework that helps organize much r
search on effective teaching strategies (e.g., use of examples, outlines)
and educational testing (e.g., verbal vs. nonverbal abilities, picture v
cabulary tests). Moreover, DCT permits the development and testing of
detailed models for the effects of instructional variables and for perform
ance on specific tasks used in educational assessment. The usefulness of
DCT for describing instructional and testing phenomena demonstrates
how general its basic theoretical mechanisms are, at least with respect
to cognitive domains.

DCT AND "NONCOGNITIVE" EDUCATIONAL TOPICS

The preceding sections of this article have concentrated on t

that are primarily cognitive in nature, but DCT also provides a fo
for understanding less cognitive phenomena, such as emotions and
skills. This is not surprising, given the heavy emphasis on cognit
ments in contemporary models for both emotions and motor ski
therefore something of a misnomer to label these topics as "nonco
hence the quotes.

A DCT View of Emotion and Education

DCT conceptualizes emotion as a complex pattern of activati

among interconnected nonverbal and verbal representations. The nonver
components of emotion include visual images for affect-related objec
(e.g., school, teachers, notebooks), kinesthetic images involving facial a
other skeletal muscles (e.g., smile, frown, clenched fist), and somatic
autonomic images associated with visceral reactions (e.g., queasy stom
heart rate). Much research has demonstrated the importance for emotio
experience of facial expressions (e.g., Schwartz et al., 1979) and viscer
reactions (e.g., G. Mandler, 1958, 1984; Schwartz et al.,. 1978). Imagery
also a central element in several theoretical and clinical models of emotion
(e.g., Bandura, 1969; Craighead et al., 1976, pp. 145-151; Lang, 1979a,b).
The DCT model of separate verbal and nonverbal systems leads to
the inference that visual images for objects may be more directly connected
to other nonverbal components of emotion (e.g., somatic and kinesthetic

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
182 Clark and Paivio

images) than are verbal representations for objects. Consistent with t

expectation, people are faster to choose the more pleasant of two obj
when pictures are shown than when object names are shown (Paivio, 19
Paivio and Marschark, 1980), and pictures elicit higher ratings of emot
ality than do words and fewer reports of no emotion when subjects
asked to generate emotional reactions to stimuli (Clark and Paivio, 1989
Preferential access to affect by nonverbal stimuli is also suggested by
finding that spatial questions, such as "Picture and describe the last s
ation in which you laughed," elicit stronger emotion ratings and more facia
muscle activity from undergraduate subjects than do verbal questions,
as "Give me a synonym for the word happy" (Schwartz et al., 1979). M
over, high imagers demonstrate stronger physiological reactions to im
ined emotional situations than do low imagers (Lang, 1984).
The central role of imagery in emotion is further supported by
search on emotions relevant to education, such as achievement motivatio
and evaluation anxiety. With respect to achievement motivation, goal
lated images (e.g., receiving a diploma at graduation, a teacher smi
while returning a paper) motivate students and others to persist with t
that have long-term benefits (e.g., Singer, 1966). Projective measures
achievement motivation are based on verbally described images elicited
pictures (McClelland et al., 1953). Imagery methods have also been use
in many studies to enhance motivation in athletes (see Paivio, 1985, for
overview), an approach that could promote motivation and persistenc
students. For example, students might be explicitly trained to mentally
age success outcomes for their academic efforts. A positive correlation
tween student interest and imagery has also been reported in research
text processing (Sadoski and Goetz, 1985; Wharton, 1980). That is, conc
text that evokes imagery tends to be more interesting than abstract t
Analogous effects may contribute to the positive effects of imagery
concreteness on teaching effectiveness (cf. Clark and Clark, 1970).
Another emotion with considerable educational influence is anxiety,
particular, evaluation anxiety. One component of evaluation anxiety, se
tivity to audiences or "stage fright" (Paivio, 1965; Paivio and Lambert, 1959)
has been measured in high school and university students using imagery-ba
projective measures similar to those used to measure achievement motivatio
That is, respondents describe imaginaiy stories for pictures that depict
ence situations. Imagery is also involved in many psychological treatme
for such affective disorders as evaluation anxiety (Hembree, 1988). Systemat
desensitization, for example, requires students to maintain a relaxed bod
state while mentally imaging a graduated series for the feared object (
teacher announcing a test in two weeks, studying the night before, wal
in to take the test, trying to answer a difficult question). The relaxed s

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 183

in systematic desensitization and other

duced using imagery techniques (e.g., i
feel the muscular relaxation flowing thr
As in the cognitive domains that w
of emotion includes verbal elements,
bored), general evaluative terms (e.g.,
(e.g., tense, queasy stomach), and nam
school, test, mathematics) and events
Similar verbal codes or their abstract
theories of emotion (e.g., Bandura, 19
151; Lang, 1979a,b).
Verbal labels relevant to emotion a
tions to one another and by referentia
Free association, sorting, and related
emotion terms may be organized in a h
found in cognitive domains, and severa
associative structures (e.g., Bower, 19
and Storm (1987), for example, develo
with three categories of negative ter
ated, discouraged), anxiety and fear (
(e.g., anger, rebellious); two categories
admiration, concern) and not interpe
category of relatively neutral terms i
lient (e.g., alienated, absorbed). Items
the examples given, are relevant to ed
fuller understanding of the associative
The relevance to education of emotion terms such as these and other
verbal aspects of emotion is easy to demonstrate, although the area is a
complex one. Persistent activation of positive (e.g., pride) or negative (e.g.,
disappointed) emotion terms in the context of school will affect students'
attitudes toward school and various actions that depend on students' emo
tional responses to school (e.g., attending classes, studying). Moreover, if
school-related terms form an associative structure, as we have suggested,
positive or negative experiences with one component of the network could
spread to other components. A particularly good experience with a specific
teacher, for example, could have widespread consequences for many ele
ments in the school-related semantic network. Verbal processes are also
involved in student and teacher self-talk related to emotion. Positive self
statements (e.g., I should try harder, Good work) are motivating and pro
mote student persistence (Chapin and Dyck, 1976), and the cognitive or
woriy components of anxiety (e.g., I cannot do this; What if I fail?) are
probably verbal (Schwartz et al., 1978). Negative self-statements also

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
184 Clark and Paivio

contribute to dysfunctional cognitive processing in test-anxious stud

(e.g., Benjamin et al., 1981; Bruch et al., 1986) and play a central role
students with poor academic self-concepts. On the positive side, appr
ate self-talk has been used in cognitive behavior-modification techniq
to reduce test anxiety (e.g., Don't worry, Relax) (Craighead et al., 197
and it contributes to positive academic self-concepts.
The preceding examples demonstrate that DCT provides a use
framework for thinking about the role of emotion in education, but ba
scratches the surface of an immense topic. Student and teacher emot
serve as both predictor and criterion variables in education. As predict
student attitudes, interests, motivation, expectations, self-concepts,
anxiety contribute to school attendance, course selection, and achievem
(e.g., Gauld and Hukins, 1980; Hembree, 1988; Rumberger, 1987; Schib
and McGaw, 1981; Vollmer, 1984). On the teaching side, interests
motivation determine who enters teaching (Jackson, 1977). Teachers
have attitudes toward such educational topics as mainstreaming (Guer
1979), and they have anxieties about teaching such subjects as scie
(Coates and Thoresen, 1976; Westerback and Primavera, 1988). Moreov
teachers' attitudes about courses contribute to student achievement (e
Moore, 1988). With respect to emotions as criterion or output variabl
affective constructs have a central place in several taxonomies of edu
tional objectives (e.g., Gagne, 1977; McAshan, 1974), and are influence
by a variety of educational variables, such as school quality (Sosniak a
Ethington, 1989), courses (e.g., science — Gauld and Hukins, 1980
Schibeci, 1984; Schibeci and McGaw, 1981), and specific teaching meth
(e.g., mastery learning — Block and Tierney, 1974). Education program
similarly attempt to modify attitudes of education students (e.g., Yee
1969).
Despite the substantial evidence that student attitudes and other emo
tional constructs are of great educational importance, teachers may lack
the theoretical tools to deal with affective topics in a systematic way. Sci
ence teachers, for example, rate attitudinal objectives as less important than
cognitive objectives, do not systematically measure attitudes, and believe
that attitudes cannot be changed (Schibeci, 1981). One strength of DCT
is that the same constructs that explain cognitive phenomena can be used
to build concrete models of the verbal and imagery mechanisms that un
derlie student and teacher affect, and its modification. This unified ap
proach to cognition and emotion may help teachers to better understand
emotional phenomena and to achieve affective goals in their classrooms.
The same mechanisms can also be applied to other noncognitive domains,
such as motor skills.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 185

Motor Skills and Education

By now readers will anticipate (correctly) that DCT conceptualizes

motor skills in terms of images for perceptual patterns and movements,
relevant verbal codes, and associative and referential processes that govern
activation of the interconnected codes; that is, motor skills involve the same
basic mechanisms that account for cognition and affect. Nonverbal com
ponents of the motor system include both kinesthetic and visual images.
Kinesthetic imagery refers to the "feel" of the action from an internal per
spective. Examples are what it feels like to print or type particular letters
and numbers, to pour a beaker of liquid into a test tube, to play a musical
instrument, or to articulate a unique sound in a foreign language. Visual
imagery refers to the appearance of actions from the perspective of an
external viewer; for example, what the preceding actions would look like
if another person (or yourself) performed the action. There are many theo
retical links between imagery and motor skills, both past and present. His
torically, for example, Piaget equated imagery with internalized imitation.
More recently, Paivio (1971) has emphasized the role of motor components
in transformational imagery, and the imagery system in DCT has counter
parts in the nonverbal components of other theories for skilled movement
(e.g., Bandura, 1969; Shallice, 1978). Motor schema, for example are hy
pothesized to represent response parameters and sensory consequences of
movement, presumably in a nonverbal form (Schmidt, 1975).
Empirically, research has demonstrated a number of important rela
tions between action, imagery, and other cognitive phenomena (e.g., Engel
kamp, 1990). One sort of evidence for concrete imaginai codes comes from
modality-specific interference and priming effects. Klatzky et al. (1989), for
example, found that priming a specific motor action (e.g., clenching the
hand) in undergraduates facilitated subsequent judgments for statements
about actions related to the prime (e.g., crumple a newspaper), whereas
verbal primes (e.g., saying "clench") did not produce priming. A close re
lation between motor and imagery processing is further supported by evi
dence that individual differences in imagery ability correlate positively with
performance on experimental motor tasks. Across 219 subjects ranging
from 10—40 years of age, for example, Goss et al. (1986) found that novel
movements of the hand were learned more easily by people with good vis
ual and kinesthetic imagery than by people with poor imagery ability. Other
research suggests that visual imagery ability contributes more to early stages
of learning, whereas kinesthetic imagery ability contributes to later stages
(Fleishman and Rich, 1963). The central role of imagery in movement is
also consistent with the widespread use of imagery methods by sports psy
chologists (Paivio, 1985; Suinn, 1983; Weinberg, 1982). In a comprehensive

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
186 Clark and Paivio

review of research on the benefits of mental rehearsal for teaching

skills, Feltz and Landers (1983) concluded that imaginai practice imp
performance relative to a control group, especially for motor skills
cognitive components.
With respect to the role of imagery in motor skills, there are m
potential educational benefits of experimental research and relevant
ries, such as DCT. The work on mental rehearsal of motor skills sug
that students can use mental imageiy to perform and practice new ac
Children learning to print letters, for example, might be instructed to th
of a picture for each letter before starting to draw, or to mentally pr
printing between actual printing sessions. Although none of the 60 m
rehearsal studies reviewed by Feltz and Landers involved an academ
tor skill, educators have considered the role of images in motor skill
used remedial techniques that involve imaginai processes. Mercer and
cer (1981, p. 348), for example, attributed shape and size errors in pri
to incorrect mental images and described fading techniques that invo
copying or tracing models that were gradually faded out (p. 355; se
Sisson et al., 1988). Theoretically, such fading methods should foste
generation of images as models prior to action, a strategy that was
gested by Thomassen and Teulings (1983, p. 195). Imagery mnemoni
also be used to facilitate perceptual and motor learning of letter sh
and other novel patterns. One mnemonic method relates the shape
ters to images for familiar objects that begin with the sound of the
(see Clark and Paivio, 1987). The letter /, for example, can be ment
imaged as a flower with leaves at the cross-bar and a blossom at its
This mental image serves as a visual pattern for decoding and produ
the letter, and as a mnemonic for its sound.
Verbal processes also contribute to the DCT view of motor sk
including such verbal representations as names for specific acts (e.g.,
grasp, and other action labels), words for general qualities of acts (e
slowly, soft), and names for particular patterns that movements can prod
(e.g., curve, vertical line, circle). Moreover, all words inherently invol
tor skills, inasmuch as words are spoken, written by literate people, t
by those with keyboarding skills, and represented manually in sign
guage. Indeed, articulatory codes may form the basis for word r
sentations, since speaking is generally acquired before reading. V
components or their abstract equivalents also appear in other theor
models of movement (e.g., Bandura, 1969; Shallice, 1978).
Verbal and imaginai representations relevant to action appear to
organized in associative networks analogous to those seen in the cog
and emotional domains. The priming studies of Klatzky and her colle
(e.g., Klatzky et al., 1989), for example, suggest that spreading activ

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 187

is similar for motor and semantic rep

al. (1987) demonstrated that many ha
gories (e.g., push, poke) on the basis o
involved in the action (e.g., pencil, sta
similarity between motor and semant
gested by the robust effect of uncerta
naming slows as the number of alter
1989), the speed of motor responses s
(e.g., Fitts and Switzer, 1962; Hyman,
trieval also both involve inhibitory p
ple, used inhibition to explain the fact
fingers on the same hand (e.g., typin
middle fingers) is slower than respond
(e.g., typing; and d with the right index
hypothesized that making one respo
sponses, and that it was easier to sup
hand than to suppress an adjacent fin
Associative structures in the motor
component (Kimura and Archibald, 19
ticularly with verbal aspect of movem
ample, note that serial motor tasks (e.g
verbal component than do continuous
or football). Verbal tasks also disrupt
as tapping, especially for the right h
This relation between verbalization and
with the DCT assumption that langua
essing (Paivio, 1971, 1986).
These verbal and related associative
teaching. Deaf children, for example, d
abilities relative to hearing children,
difficulties (Ewing, 1957, cited in Law
components are also used to teach new
the initial description and mental rehear
a, for example, might be guided by su
touching the line" and "draw a straigh
side." The associative differences between within-hand and between-hand
actions suggest that teachers of music, typing, and related motor skills
should anticipate particular problems with within-hand sequences, might
instruct students to practice them more often, and could develop special
exercises or feedback to enhance suppression of interfering responses on
the same hand.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
188 Clark and Paivio

The DCT view of motor skills and associated research have numerous

other applications in education. Motor skills are involved in many regular

classroom activities, such as printing and writing, articulation of speech
sounds, typing, physical education, art and music, and manual skills involved
in the use of equipment in vocational and laboratory classes (e.g., Brown
& Atkins, 1988, p. 91; Kirschner and Meester, 1988). With respect to speci
education, some school-related dysfunctions are primarily motoric (e.g., stut
tering, illegible writing). Moreover, one consequence of mainstreaming ma
be that teachers will have to teach increasingly heterogeneous motor skill
since the mentally retarded and other special populations can require in
struction in dressing, brushing their teeth, cooking, and other adaptive motor
skills (Robinson and Robinson, 1976).
Motor skills are also relevant to research on individual differences and
associated work on the prediction of early school success. Some studies have
concluded that measures of motor skill are relatively independent of meas
ures of intelligence (e.g., Terrasi and Airasian, 1989), but other work sug
gests some correspondence between the two domains. Jensen (1987) and
others, for example, report that people with higher scores on intelligence
tests respond more quickly on choice reaction time tasks and are less af
fected by increases in uncertainty than are people with lower intelligence.
The DCT view that common mechanisms (e.g., imagery, verbal associations,
inhibition) underlie motor and cognitive domains could help to explain their
correlation. Motor skills are also used as predictors of later academic
achievement (e.g., Belka and Williams, 1979; Butler et al., 1982), which ex
plains the interest of early educational researchers in children's drawing
skills (e.g., Childs, 1915). Motor skills are major components of the McCar
thy scales and other contemporary tests of ability and school readiness in
young children. Simner (1989) has developed a test that predicts subsequent
academic performance from the frequency of form errors in copying letters
and numbers (i.e., errors that involve addition, deletion, or misalignment of
parts, such as copying F as E or 2 as Z). Form error scores of 16 or more
in preschool, for example, identify over 70% of the students who will have
academic difficulties at the end of first grade. Particular motor disturbances
are also sometimes found in children with speech and language disorders,
again suggesting a close relation between the verbal and motor domains
(e.g., Cromer, 1983; Noterdaeme et al., 1988).
To conclude this section, DCT provides theoretical mechanisms that
are useful in understanding such noncognitive domains as emotion and mo
tor skills. Images, verbal representations, and associative and referential
processes are implicated in many experimental studies of emotion and mo
tor skills, and permit the development of specific models relevant to a wide

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 189

variety of educational topics (e.g., achievement motivation, mental rehears

al of motor skills).

THE SCIENCE AND PRACTICE OF EDUCATIONAL

PSYCHOLOGY

A comprehensive theory of education must explain not only th

chology of students and teachers, but also the psychological process
underlie the science and practice of educational psychology. Here w
onstrate that DCT can further our understanding of the epistemol
foundations of educational and psychological research, and also has
to say about such practice issues as teacher education.

DCT and the Science of Educational Psychology

A fundamental question about educational psychology concerns

epistemological foundations for its standing as a science. Tradition
proaches to educational and psychological research acknowledge the
sity of empirical testing of tentative hypotheses, and the greater cer
of observation terms relative to theoretical terms. These traditional views
of science are reflected in educational psychology texts, which tend to en
dorse empirical approaches to knowledge, justify their conclusions by em
pirical research, and base educational research methods on the distinction
between observation and theory (e.g., Ash and Love-Clark, 1985). Moreover,
educational researchers in several areas have explicitly contrasted constructs
at the observational and theoretical levels. Researchers who study teaching,
for example, make a distinction between low- and high-inference behaviors,
such as "uses outlines" vs. "is organized" (e.g., Murray, 1983a,b; Rosenshine
and Furst, 1971).
In recent decades, some educators have questioned this norm of ob
jectivity in empirical social science (Westkott, 1979), and have become dis
enchanted "with the underlying assumptions and the methods of traditional
social science approaches" (Heshusias, 1988, p. 62). Such challenges are
often based on philosophical analyses of science that give decreased weight
to empirical factors because data are themselves thought to be "theory
laden" (e.g., Brown, 1977; Kuhn, 1962). Criticisms of present approaches
to social science often include recommendations for a shift to less empirical
research methods in education and related disciplines (e.g., Bloland, 1989;
Eastwood, 1988). Similar criticisms of empirical approaches to knowledge
are found in psychology (e.g., Koch, 1981; Royce, 1982) and science edu

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
190 Clark and Paivio

cation (e.g., Hodson, 1985). Finley (1983), for example, states that cu
science programs give too much emphasis to observation, laboratory
ties, and other empirical practices. He proposes that science educatio
based instead on less empirical philosophies of science.
DCT provides a well-founded perspective on these challenges to
traditional epistemological foundations of educational psychology and rela
disciplines. Despite the criticisms and some complex issues (see Clark
Paivio, 1989a; Paivio, 1986), the empiricist separation of observation an
ory is consistent with the DCT distinction between concrete and ab
terms, and we have already seen that concreteness has robust effects on c
prehension, meaning, the structure of knowledge, and other attribute
vant to scientific discourse. To generalize the natural language resul
scientific terms, Clark and Paivio (1989a) had independent groups of
chologists rate the observability and concreteness of 72 psychological
on 7-point scales. High ratings on observability, for example, were to be g
to terms designating observable entities and whose meanings were bas
marily on direct or instrument-mediated activation of the senses. Me
ings were calculated for each term, and across the 72 terms, observab
and concreteness correlated .89, suggesting that the two ratings mea
common construct. Sample terms (and their mean observational rating
image (2.11), schema(ta) (2.33), attribution (2.89), anxiety (3.78), intellig
(3.78), comprehension (4.11), peer group (4.89), reward (5.67), recall (
crying (6.33), and heart rate (7.00). In a subsequent task Clark and P
measured graduate students' reaction times to generate images or ver
sociations for the items. Observability correlated -.52 with imagery re
times for the items and .65 with the frequency of spontaneous image
the verbal association task, as indicated on postexperimental questionn
These results demonstrate that the observability (i.e., concreteness) of
tific terms can be measured in reliable and valid ways, contrary to philoso
cal assertions that the distinction lacks a clear basis.

To evaluate whether observational terms enjoy the benefits claimed

for them by empirical views of science (e.g., greater intersubjective agree
ment), we had academic psychologists independently rate the terms for
their consistency and distinctiveness of meanings. Terms rated high on con
sistency were to have stable and reliable meanings that would not vary from
one person or time to another. Mean observability (and concreteness) cor
related very highly with mean ratings for consistency (.81) and distinctive
ness (.78) of meaning. Such results demonstrate that observable
psychological terms are more likely to mean the same thing to different
researchers than are theoretical terms, which weakens the idea that obser
vational and theoretical terms are equally theory-laden. The study also
demonstrates that epistemological issues are amenable to empirical inves

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 191

tigation and that researchers in psychology and education need not depend
entirely on philosophers of science for the foundations of their disciplines.
Such findings support traditional empirical approaches to science and
educational research. Moreover, this work and other research on scientific
language and thought will eventually permit the development of a well
founded cognitive theory for educational psychology as an empirical sci
ence. Empirical and theoretical considerations suggest that the DCT
constructs of concreteness and imagery (cf. Paivio, 1983a) will emerge as
important elements in models of science, and will have important conse
quences for the science of educational psychology and for its practice.

DCT and Teacher Education

The practice side of educational psychology includes its major r

in teacher education. Pedagogical aspects of educational psychology co
and texts can be conceptualized in terms of the DCT constructs of c
creteness, imagery, and verbal associative processes. Indeed, everything
we have discussed in this article about the structure of knowledge, effect
instruction, individual differences, and so on is relevant to educationa
chology as a subject in teacher education and other university progr
However, we will concentrate on some special phenomena relevant t
creteness and imagery effects.
The importance of concreteness and imagery in DCT raises the
tion of the relative concreteness of educational psychology, the use o
agery, and the relation between concrete and abstract concepts. Alth
empirical data are presently inadequate to permit strong conclusions,
are some indications that educational psychology may be overly abs
and theoretical, or may not connect theory and practice as strongly
sible. In discussing theories in educational psychology, for example, Br
(1979, p. 738) criticized the "overly abstract and grandiose theories
have plagued the field to date." An educational psychology course st
by Donald (1986) also emphasized abstract concepts, likening the cou
more to philosophy than to science (see also Goldschmid, 1967).
Evidence that educational psychology may emphasize the theoret
and abstract, rather than concrete phenomena, also comes from obs
tions that data are seldom presented in educational psychology text
examined the percent of pages in four contemporary educational psy
ogy texts that actually presented data in a table or a figure. The me
percent was only 3.69%, with one text presenting actual data on only
842 pages. Although a lack of data in tables and figures might be com
sated for by detailed verbal descriptions of the results of studies, the lim

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
192 Clark and Paivio

presentation of data suggests a possible underemphasis of concrete

nomena. The shortage of data is particularly striking when one con
that, over 60 years ago, Starch (1928) wrote an educational psychology
that contained figures or tables on 43.1% of its 536 pages. Moreover, S
(1930) provided a manual of laboratory research activities relevant t
cation. If concrete phenomena provide the conceptual pegs on which
abstract theoretical ideas are hung, then Starch's emphasis on facts
concrete research activities seems well-motivated.
Textbooks are only one way that information is transmitted, but other
aspects of teacher education lead to some concern about the extent to which
concrete data and research methods are discussed and related to theory.
Champion (1984) observed that education faculty often mention but less
often model or demonstrate research practices in their courses, again sug
gesting an overly abstract approach to educational research and theory. This
lack of emphasis on research activities and data may occur because education
faculty tend to be more committed to teaching and practice than to research
(Nystrom et al., 1984; Stark et al., 1986; Troyer, 1986). Perhaps these and
related phenomena contribute to the historical gap that has existed between
educational research and practice (e.g., Clifford, 1973; Resnick, 1981), despite
the extensive research literature available on many educational topics.
Another potential source of concrete experiences to which abstract
theories can be tied is the classroom. Data once again do not permit strong
conclusions about the extent to which theory and actual classroom experi
ences are linked, but several studies suggest that instructional theory and
its relation to practice are seldom the focus of supervisor conferences with
education students. In her review, Kagan (1988, p. 9) concluded that super
visors "almost exclusively discussed narrow, particularistic concerns rather
than attempting to embed immediate problems in larger theoretical con
texts." This finding suggests that teacher education students are not being
explicitly shown how to relate educational theories to their classroom prac
tices, nor receiving feedback that would strengthen the cognitive processes
that underlie translation between abstract theory and concrete experience.
Another way in which abstract theories can be concretized is through
the development of mechanistic models for educational phenomena. We
have seen that images of nonverbal objects or events, verbal associative
links, and other DCT constructs permit the development of concrete,
mechanistic models in diverse areas relevant to education. For example,
activation of the imagery system can explain the instructional benefits of
concrete examples, and converging activation of shared verbal associations
can explain the effects of organization on memory for text. Research on
the benefits of concrete mental models for comprehension and generaliza
tion of principles (Mayer, 1989) suggests that these mechanistic models pro

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 193

vided by DCT could make the case f

to teachers and could lead to more s
niques. If concrete examples functio
teachers who understand that princ
who lack the experiences necessary t
their imagery abilities strengthened.
As emphasized throughout this ar
plex and subtle interplay between c
and instructional factors. If theory
guments that educational psychology
nomena and abstract theory need st
changes will require considerable sen
education students and faculty. Educ
be predisposed to think about educa
interest and personality patterns are d
(e.g., Goldschmid, 1967; Jackson, 19
be lower than students in science an
although the findings on ability are
to general aptitude, specific individua
relevant to certain aspects of thinki
as interpreting graphs and conceptu
methods. Moreover, we have already
ulty may be somewhat incompatible
search on the thought processes of
provide more information relevant t
In this section, we have shown tha
ings are relevant to both the scienc
chology. Concreteness effects in n
consistent with the special epistemolog
terms, and with traditional approache
to practice, DCT and research on the
suggest that teacher education might
crete phenomena and models. Much r
determine the validity of these claim

GENERAL DISCUSSION

We have demonstrated that an integrated set of DCT principle

commodates cognitive, affective, and sensorimotor phenomena of i
to teachers and educational psychologists. Moreover, DCT and relat
search suggest possible roles for concreteness and imagery in the

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
194 Clark and Paivio

and practice of educational psychology itself. In this general discussio

return to the question of DCT and molar theories for educationa
nomena, describe some general benefits of a unified DCT approach t
cation, and briefly note some unresolved challenges.

Compatible with Higher Levels of Explanation

DCT accommodates both basic associative phenomena and suc

higher-order cognitive processes as strategies, expectations, decisions
attributions. Molar explanatory constructs are paradigmatic features o
temporary educational and cognitive psychology, underlying many
tional models and much research on cognitive process instruction. I
a strategy orientation is the dominant approach to understandin
teaching comprehension, writing, learning, memory, thinking, problem-s
ing, and other general cognitive skills (e.g., Pressley et al. 1989a,c). Str
or process approaches are also advocated for teaching science (e.g., G
1966; Reif and Heller, 1982) and other specific school subjects, fo
moting affective change, and for enhancing motor skills learning.
DCT incorporates strategies and related theoretical constructs wi
the general associative framework described in this article, with verbal pr
esses playing a central role at the present time. The importance of lan
in strategic thinking is consistent with much research and theory on
regulation and cognitive process instruction. Early work by Luria (1
demonstrated that verbal self-instructions were a fundamental basis for self
control, and cognitive modification methods continue to emphasize self-talk
(e.g., Meichenbaum, 1976). Despite the importance of verbal processes,
however, strategies and associated knowledge can also be represented imag
inally. Spatial networks, for example, can be used to represent an associa
tive hierarchy of different strategies (e.g., Dansereau et al., 1979); some
strategies might be represented imaginally rather than verbally (e.g., draw
ing circles freehand using a cookie-cutter in one practice task described by
Pressley et ai, 1985); and affective information stored about strategies (e.g.,
their positive or negative valence) could be nonverbal.
To illustrate the DCT approach, consider a single strategy in which
an impulsive boy learns to say to himself, "Stop and think," before acting
(e.g., Meichenbaum, 1976). A DCT model for the effects of this treatment
will describe the verbal representation of the "stop and think" instruction,
how it is evoked by various verbal and nonverbal stimulus patterns (e.g.,
teacher asking a question, doing seat work, specific verbal prompts), and
how it moderates subsequent cognitive processing by the student (e.g., in
hibiting impulsive reactions). An explanation at this mechanistic level will

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 195

necessarily involve associative mechanisms, including the inhibition that

was discussed in the context of cognition and motor skills. It is already
recognized that inhibitory mechanisms play important roles in the meta
cognitive processes that underlie self-regulation (e.g., Maccoby, 1980).
That is, such self-instructions as "Stop and think," "Don't use rote re
hearsal," or "Pay attention to my work" entail suppression of the specific
cognitive, affective, or motor actions. Moreover, activation (i.e., choice)
of one class of cognitive operation generally involves inhibition of com
peting operations, especially when those competing responses are the
more familiar and habitual reaction of students. Generating mental im
ages, for example, may require the suppression of rote repetition strategies
that students have used in the past.
In addition to explaining single strategies, DCT provides a framework
by which to represent more elaborate and sophisticated theories of meta
cognition, such as the Pressley et al. (1985) model for metamemory about
strategies (see their Fig. 1). From a DCT perspective, expert strategy users
would have verbal labels for the various strategies (e.g., SQ3R, keyword,
imagery), and these verbal labels could be organized as an associative hi
erarchy in which similar strategies are clustered into superordinate catego
ries (e.g., rehearsal, elaboration, and other classes). The entire network
could be subsumed under a general label such as memory strategies or
methods, and might be represented imaginally as a cognitive map. Each
strategy label would be associated with mental representations that con
trolled the sequence of acts necessary to implement the strategy. The key
word strategy, for example, might be represented by a linked sequence of
phrases such as, "generate similar sounding keyword" and "construct in
teractive image." Moreover, a verbal associative network could instantiate
other knowledge about specific strategies, such as the situations under
which they work (e.g., learning strange words) and evaluative information
(e.g., works well). As in most domains, students would differ from one an
other with respect to the scope and organization of their strategy network
because of different experiences (e.g., differential teaching of strategies)
and differences in their capacity to benefit from experience (e.g., the use
of mental imagery to store strategy information).
Educational and cognitive researchers have become increasingly
skilled at eliciting knowledge from both expert and novice cognitive proc
essors, and much empirical data on the verbal and nonverbal elements of
student and teacher strategies is now being accumulated. Although we can
not yet instantiate all sophisticated strategies in terms of basic DCT mecha
nisms, we believe that all higher-level constructs, even those related to
abstract metamemory acquisition procedures (Pressley et al., 1985), must
ultimately be reducible to primitive associative mechanisms. Stated another

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
196 Clark and Paivio

way, strategies, expectations, and similar high-level constructs are

sented by the brain somehow and mediated by mechanistic processes
are fundamentally associative in nature. This does not deny the impor
of strategies and deliberate behavior in explaining student and teach
havior, but simply acknowledges that ultimately a nonstrategic level
planation must be achieved. The higher-level models are necessary bo
guides for practice and also to direct researchers as they piece toge
the complex network of basic mechanisms that underlie strategies and ot
higher-level phenomena. For their part, lower-level mechanisms conc
abstract theories about mental processes; permit more precise predi
and control of those processes and their effects; and, as we have tr
show in this article, provide unified explanations for diverse phenom
In short, top-down and bottom-up approaches to educational researc
strengthen one another and are mutually dependent, and DCT is
suited to this cooperative approach.

An Integrated Theory

The particular strength of DCT that we have tried to emphasize

this article is its unified approach to psychological phenomena releva
education. Integrated theories overcome a major barrier to the develo
of sound scientific foundations for educational practices, namely the
plexity of human behavior and the seemingly fragmented nature of
psychological research and theory. Piecemeal views of education comp
the learning and application of psychological principles because, as w
seen in this article, educators must consider many facets of human behav
and each of the domains is very complex. It is impossible for educa
psychologists and teachers to become masters of all these areas with
some unifying theoretical framework. Moreover, applied researcher
practitioners must deal with people as whole units and do not have th
ury of neatly dividing human behavior into pure cognitive and affective
ponents. Some of the disenchantment with basic psychological resea
expressed by applied researchers and practitioners may result from t
jointed nature of psychology and from domain-specific theories.
We can only speculate about the psychological mechanisms that
derlie such fragmentation because our understanding of scientific a
other scholarly behaviors is itself incomplete. One contributing fac
however, may be a tendency to use abstract constructs to explain p
logical phenomena rather than considering basic underlying mechan
Different labels for abstract constructs imply distinct theoretic
structs, but the differences between tasks, phenomena, and processe

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 197

be superficial and mask common und

this article, consideration of the conc
performance can reveal commonalit
nominally distinct processes. Clark
strated that convergent associative m
diverse cognitive tasks.
The present article does not reflect
tegration of educational topics becaus
tures of individual school subjects an
education. Each of these domains enta
esses and models that can be captured
briefly alluded to DCT research and t
but these need considerable elaboration
search on figurative language (Katz et
on number processing (Clark and Camp
in imagery ability, science from resear
tific language and thought (Clark and
from research on motor skills learnin
seling from the role of imagery in anx
With respect to special education, non
enhance learning of syntax in deaf ch
and DCT can explain various phenomen
bols to teach language to cerebral pals
language dysfunctions (Yovetich and P
provides a unified perspective on spec
ena; but these and many other topics

Theoretical Challenges to DC

This article has emphasized the pos

general perspective on educational and
of the topics that we have discussed i
theoretical issues that remain unresol
concrete sentences do not always dem
stract sentences on memory tasks, and
does not always correlate with learning.
the basic premises of DCT, however
standing of the role of nonverbal and
have tried to show that deeper theor
tence memory and other tasks permi
DCT. Deeper analysis includes consider

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
198 Clark and Paivio

differences, and instructions, and the joint contribution of imagery and ve

bal processes to psychological phenomena.
In addition to specific challenges, some basic assumptions of D
remain controversial. The DCT assumption of specific mental codes co
trasts with the view that written and spoken words, images, and other for
of a concept elicit a common amodal representation of its abstract mean
In DCT, semantically equivalent words (e.g., "book," "text," and "livre
and their corresponding objects are encoded by distinct verbal and imag
representations that do not converge on a single, generic, abstract cod
shared image may exist for different concrete words, but not some uni
and abstract meaning code (Clark, 1978; Paivio and Desrochers, 1980). T
issue is complex, and we just note here that DCT excludes inferred abst
codes because of much evidence for the importance of modality-speci
representations in human cognition (see Paivio, 1983b), and beca
specific imaginai and verbal codes can explain findings used to justify
notion of abstract codes. Clark (1987), for example, argued that the D
assumption of referential connections between verbal and imagery syst
can explain the processing of sentences with embedded pictures (cf., Po
et al., 1986), without the mediating abstract code that Potter et al. proposed
Moreover, evidence for the benefits of concrete psychological terms
earlier review and Clark and Paivio, 1989a) makes us wary of exces
abstraction in psychological (and educational) theory.
DCT's empirical nature and theoretical flexibility have also been cr
cized for promoting ad hoc explanations (e.g., Potter and Kroll, 1987). B
we think that the associative complexity of DCT and its probabili
mechanisms are well-suited to the richness and variability of human
havior, which tend to be oversimplified in more formal models. In addit
the empirical approach of DCT to basic underlying mechanisms (e.g.,
search on the effects of concreteness on image generation) has produ
elegant explanations for diverse phenomena, as we have tried to demo
strate using examples from both experimental and educational psychol
In conclusion, the basic mechanisms of DCT permit unified expla
tions for a wide variety of educational phenomena, although unresolv
issues remain. We have seen that concreteness, imagery, and verbal a
ciative mechanisms lead to concrete models for specific phenomen
many areas of education and at many levels: comprehension, learning a
memory, effective instruction, educational assessment, affect, motor sk
and the science and practice of educational psychology. The successfu
plications of the theory and its ability to accommodate strategic proce
increase our confidence that the interconnected verbal and imaginai r
resentations of DCT provide powerful mechanisms with which to appro
the understanding of human behavior and experience. A coherent and

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
Dual Coding Theory and Education 199

fied set of principles provides detailed and specific explanations for quite
diverse phenomena; consequently, DCT can help to integrate the complex
field of educational psychology and to advance the science and practice of
education.

ACKNOWLEDGMENTS

Preparation of this article was supported by the Natural Scie

Engineering Research Council of Canada through grants OGP004
James M. Clark and A0087 to A. Paivio. We appreciate comment
lier drafts of the article from Stephen Benton, Carla Johnson,
Pressley, Mary Walsh, and an anonymous reviewer, and also appr
original invitation for the article from John Glover.

REFERENCES

Anatasi, A. (1988). Psychological Testing (6th ed.), Macmillan, New York.

Anderson, R. C. (1974). Concretization and sentence learning. Journal of E
Psychology, 66: 179-183.
Anderson, R. C. (1977). The notion of schemata and the educational enterprise. In
R. C., Spiro, R. J., and Montague, W. E. (eds.), Schooling and the Acquisit
Knowledge, Erlbaum, Hillsdale, New Jersey, pp. 415-431.
Ash, M. J., and Love-Clark, P. (1985). An historical analysis of the content of Ed
Psychology textbooks 1954-1983. Educational Psychologist, 20: 47-55.
Atkinson, R. C. (1975). Mnemotechnics in second-language learning. American Psy
30: 821-828.
Ausubel, D. P. (1960). The use of advance organizers in the learning and retention o
meaningful verbal material. Journal of Educational Psychology 51: 267-272.
Baenninger, M., and Newcombe, N. (1988). The role of experience in spatial test performance
A meta-analysis. Sex Roles 20: 327-344.
Baird, D. O. (1931). A vocabulary study of biology notebooks of 50 representative secondary
schools in New York State. Journal of Educational Psychology 22: 512-516.
Bandura, A. (1969). Principles of Behavior Modification, Holt, Rinehart and Winston, New York
Begg, I. (1973). Imagery and integration in the recall of words. Canadian Journal of Psychology,
27: 159-167.
Begg, I., and Paivio, A. (1969). Concreteness and imagery in sentence meaning. Journal of
Verbal Learning and Verbal Behavior, 8: 821-827.
Begg, I., Upfold, D., and Wilton, T. D. (1978). Imagery in verbal communication. Journal of
Mental Imagery 2: 165-186.
Belka, D. E., and Williams, H. G. (1979). Prediction of later cognitive behavior from early
school perceptual-motor, perceptual, and cognitive performances. Perceptual and Motor
Skills, 49: 131-141.
Bellezza, F. S. (1981). Mnemonic devices: Classification, characteristics, and criteria. Review
of Educational Research, 51: 247-275.
Benbow, C. P., and Stanley, J. C. (1983). Sex differences in mathematical reasoning ability:
More facts. Science, 222: 1029-1031.
Benjamin, M., McKeachie, W. J., Lin, Y., and Holinger, D. P. (1981). Test anxiety: Deficits
in information processing. Journal of Educational Psychology, 73: 816-824.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 200 Clark and Paivio

Block, J. H., and T

in mastery learnin
962-967.
Bloland, H. G. (1989). Higher education and higher anxiety: Objectivism, relativism, and irony.
The Journal of Higher Education, 60: 519-543.
Bousfield, W. K. (1953). The occurrence of clustering in the recall of randomly arranged
associates. Journal of General Psychology, 49: 229-240.
Bower, G. H. (1970). Imagery as a relational organizer in associative learning. Journal of Verbal
Learning and Verbal Behavior, 9: 529-533.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36: 129-148.
Bower, G. H., Black, J. B., and Turner, T. J. (1979). Scripts in memory for text. Cognitive
Psychology 11: 177-200.
Bransford J. D., Stein, B. S., Vye, N. J., Franks, J. J., Auble, P. M., Mezynski, K. J., and
Perfetto, G. A. (1982). Differences in approaches to learning: An overview. Journal of
Experimental Psychology: General, 111: 390-398.
Brophy, J. E. (1979). Teacher behavior and its effects. Journal of Educational Psychology, 71:
733-750.
Brown, G., and Atkins, M. (1988). Effective Teaching in Higher Education, Methuen, London.
Brown, H. I. (1977). Perception, Theory and Commitment: The New Philosophy of Science,
University of Chicago Press, Chicago.
Brown, W. F., and Holtzman, W. H. (1967). Survey of Study Habits and Attitudes: Manual,
The Psychological Corporation, New York.
Bruch, M. A., Pearl, L., and Giordano, S. (1986). Differences in the cognitive processes of
academically successful and unsuccessful test-anxious students. Journal of Counseling
Psychology, 33: 217-219.
Bugelski, B. R. (1970). Words and things and images. American Psychologist 25: 1002-1012.
Burnett, S. A., Lane, D. M., and Dratt, L. M. (1979). Spatial visualization and sex differences
in quantitative ability. Intelligence, 3: 345-354.
Butler, S. R., Marsh, H. W., Sheppard, M. J., and Sheppard, J. L. (1982). Early prediction
of reading achievement with the Sheppard School Entry Screening Test: A four-year
longitudinal study. Journal of Educational Psychology, 74: 280-290.
Buzan, T. (1974). Use Your Head, BBC Books, London.
Cattell, J. M. (1889). Mental association investigated by experiment. Mind, 14: 230-250.
Champion, R. H. (1984). Faculty reported use of research in teacher preparation courses: Six
instructional scenarios. Journal of Teacher Education, 35: 9-12.
Chapin, M., and Dyck, D. G. (1976). Persistence in children's reading behavior as a function
of N length and attribution retraining. Journal of Abnormal Psychology, 85: 511-515.
Charters W. W., and Waples, D. (1929). The Commonwealth Teacher-Training Study,
University of Chicago Press, Chicago.
Childs, H. G. (1915). Measurement of the drawing ability of two thousand one hundred and
seventy-seven children in Indiana city school systems by a supplemented Thorndike scale.
Journal of Educational Psychology, 6: 391-408.
Clark, C. M., Gage, N. L., Marx, R. W., Peterson, P. L., Stayrook, N. G., and Winne, P. H.
(1979). A factorial experiment on teacher structuring, soliciting, and reacting. Journal of
Educational Psychology, 71: 534-552.
Clark, G. K., and Clark, E. B. (1970). The Art of Lecturing: Some Practical Suggestions, W.
Heffer and Sons, Cambridge, England.
Clark, J. M. (1978). Synonymity and concreteness effects on free recall and free association:
Implications for a theory of semantic memory. Unpublished doctoral dissertation,
University of Western Ontario, London, Canada.
Clark, J. M. (1984). Concreteness and semantic repetition effects in free recall: Evidence for
dual-coding theory. Canadian Journal of Psychology, 38: 591-598.
Clark, J. M. (1987). Understanding pictures and words: Comment on Potter, Kroll, Yachzel,
Carpenter, and Sherman (1986). Journal of Experimental Psychology: General, 116:
307-309.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 201

Clark, J. M. (1988, June). Accessing knowledge

the Annual Meeting of the Canadian Psychol
Clark, J. M. (1989). Reconciliation of data on a
teachers. Journal of Teacher Education, 40: 49
Clark, J. M., and Campbell, J. I. D. (1991). Inte
skills and acalculia. Brain and Cognition (in p
Clark, J. M., and Johnson, C. J. (unpublished)
and superordinate naming of pictures.
Clark, J. M., and Paivio, A. (1984, May). Associati
at the Annual Meeting of the Canadian Psych
Clark, J. M., and Paivio, A. (1987). A dual co
McDaniel, M. A., and Pressley, M. (eds.), Im
Theories, Individual Differences, and Applicatio
Clark, J. M., and Paivio, A. (1989). Observatio
cognitive perspective on scientific language. Am
Clark, J. M., and Paivio, A. (1989, June). Aro
presented at the Annual Meeting of the Can
Canada.
Clifford, G. J. (1973). A history of the impact of research on teaching. In Travers, R. M. W
(ed.), Second Handbook of Research on Teaching, Rand McNally, Chicago, pp. 1-46.
Coates, T. J., and Thoresen, C. E. (1976). Teacher anxiety: A review with recommendation
Review of Educational Research, 46: 159-184.
Collins, A. M., and Loftus, E. F. (1975). A spreading-activation theory of semantic processin
Psychological Review, 82: 407-428.
Corkill, A. J., Glover, J. A., and Bruning, R. H. (1988). Advance organizers: Concrete versu
abstract. Journal of Educational Research, 82: 76-81.
Craighead, W. E., Kazdin, A. E., and Mahoney, M. J. (1976). Behavior Modification: Principle
Issues, and Applications, Houghton Mifflin, Boston.
Craik, F., and Lockhart, R. (1972). Levels of processing: A framework for memory researc
Journal of Verbal Learning and Verbal Behavior, 11: 671-684.
Cramer, P. (1968). Word Association, Academic Press, New York.
Cromer, R. F. (1983). Hierarchical planning disability in the drawings and constructions of
special group of severely aphasie children. Brain and Cognition, 2: 144-164.
DAgostino, P. R., O'Neill, B. J., and Paivio, A. (1977). Memory for pictures and words
a function of levels of processing: Depth or dual coding? Memory & Cognition, 5: 252-256.
Dansereau, D. F., Collins, K. W., McDonald, B. A., Holley, C. D., Garland, J., Diekhoff, G.,
and Evans, S. H. (1979). Development and evaluation of a learning strategy training
program. Journal of Educational Psychology, 71: 64-73.
Das, J. P., Kirby, J., and Jarman, R. F. (1975). Simultaneous and successive syntheses: An
alternative model for cognitive abilities. Psychological Bulletin, 82: 87-103.
Davitz, J. R. (1969). The Language of Emotion, Academic Press, New York.
Dean, R. S., and Kulhavy, R. W. (1981). Influence of spatial organization in prose learning
Journal of Educational Psychology, 73: 57-64.
deGroot, A. M. B. (1989). Representational aspects of word imageability and word frequenc
as assessed through word association. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 15: 824-845.
Denckla, M. B., Rudel, R. G., and Broman, M. (1981). Tests that discriminate between
dyslexic and other learning-disabled boys. Brain and Language, 13: 118-129.
Denis, M. (1982). Imaging while reading text: A study of individual differences. Memory
Cognition, 10: 540-545.
Denis, M. (1984). Imagery and prose: A critical review of research on adults and children.
Text, 4: 381-401.
Desrochers, A., and Begg, I. (1987). A theoretical account of encoding and retrieval process
in the use of imagery-based mnemonic techniques: The special case of the keywor
method. In McDaniel, M. A., and Pressley, M. (eds.), Imagery and Related Mnemonic

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 202 Clark and Paivio

Processes: Theories,
pp. 56-77.
Dillon, R. F., and Sternberg, R. J. (1986). Cognition and Instruction, Academic Press, New
York.
Donald, J. G. (1986). Knowledge and the university curriculum. Higher Education, 15: 267-282.
Dunkin, M. J., and Doenau, S. J. (1980). A replication study of unique and joint contributions
to variance in student achievement. Journal of Educational Psychology 72: 394-403.
Eastwood, J. (1988). Qualitative research: An additional research methodology for speech
pathology? British Journal of Disorders of Communication, 23: 171-184.
Ebmeier, H., and Good, T. L. (1979). The effects of instructing teachers about good teaching
on the mathematics achievement of fourth grade students. American Educational Research
Journal, 16: 1-16.
Engelkamp, J. (1990). Memory of action events: Some implications for memory theory and
for imagery. In Cornoldi, C., and McDaniel, M. A. (eds.), Imagery and Cognition,
Springer-Verlag, New York, pp. 181-219.
Ernest, C. H., and Paivio, A. (1971). Imagery and verbal associative latencies as a function
of imagery ability. Canadian Journal of Psychology, 25: 83-90.
Eylon, B., and Linn, M. C. (1988). Learning and instruction: An examination of four research
perspectives in science education. Review of Educational Research, 58: 251-301.
Feltz, D. L., and Landers, D. M. (1983). The effects of mental practice on motor skill learning
and performance: A meta-analysis. Journal of Sport Psychology, 5: 25-57.
Finley, F. N. (1983). Science processes. Journal of Research in Science Teaching, 20: 47-54.
Fischman, M. G., Christina, R. W., and Vercruyssen, M. J. (1981). Retention and transfer of
motor skills: A review for the practitioner. Quest, 33: 181-194.
Fisher, K. M. (1988). Depression in free-word-association responses at adolescence. Journal
of Educational Psychology, 80: 247-250.
Fitts, P. M., and Switzer, G. (1962). Cognitive aspects of information processing: I. The
familiarity of S-R sets and subsets. Journal of Experimental Psychology 63: 321-329.
Fleishman, E. A., and Rich, S. (1963). Role of kinesthetic and spatial-visual abilities in
perceptual-motor learning. Journal of Experimental Psychology, 66: 6-11.
Flesch, R. (1950). Measuring the level of abstraction. Journal of Applied Psychology, 34:
384-390.
Frey, P. W., Leonard, D. W., and Beatty, W. W. (1975). Student ratings of instruction:
Validation research. American Educational Research Journal, 12: 435-447.
Gagne, E. D. (1985). The Cognitive Psychology of School Learning, Little, Brown and Company,
Boston.
Gagne, R. M. (1966). Elementary science: A new scheme of instruction. Science, 151: 49-53.
Gagne, R. M. (1977). The Conditions of Learning (3d ed), Holt, Rinehart and Winston, New
York.
Gardner, M. F. (1979). Expressive One-Word Picture Vocabulary Test, Academic Therapy
Publications, Novato, California.
Gauld, C. F., and Hukins, A. A. (1980). Scientific attitudes: A review. Studies in Science
Education, 7: 129-161.
Gentner, D., and Stevens, A. L. (eds.) (1983). Mental Models, Erlbaum, Hillsdale, New Jersey.
Gilliland, J. (1972). Readability, Hodder and Stoughton, London.
Glaser, R. (1981). The future of testing: A research agenda for cognitive psychology and
psychometrics. American Psychologist, 36: 923-936.
Glenberg, A. M., Meyer, M., and Lindem, K. (1987). Mental models contribute to
foregrounding during text comprehension. Journal of Memory and Language 26: 69-83.
Gliessman, D. H., Pugh, R. C., Dowden, D. E., and Hutchins, T. F. (1988). Variables
influencing the acquisition of a generic teaching skill. Review of Educational Research,
58: 25-46.
Goldschmid, M. L. (1967). Prediction of college majors by personality tests. Journal of
Counseling Psychology, 14: 302-308.
Goss, S., Hall, C., Buckolz, E., and Fishburne, G. (1986). Imagery ability and the acquisition
and retention of movements. Memory & Cognition, 14: 469-477.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 203

Guerin, G. R. (1979). Regular teacher concern

children. Psychology in the Schools, 16: 543-545
Guilford, J. P. (1967). The Nature of Human In
Guri-Rozenblit, S. (1989). Effects of a tree diagr
in an expository text with multiple themes. Re
Halliday, M. A. K., and Hasan, R. (1976). Cohes
Hembree, R. (1988). Correlates, causes, effect
Educational Research 58: 47-77.
Heshusias, L. (1988). The arts, science, and the study of exceptionality. Exceptional Children,
55: 60-65.

Hidi, S., and Baird, W. (1988). Strategies for increasing text-based interest and students' recall
of expository texts. Reading Research Quarterly, 23: 465-483.
Hiller, J. H., Fischer, G. A., and Kaess, W. (1969). A computer investigation of verbal
characteristics of effective classroom lecturing. American Educational Research Journal, 6:
661-675.
Hodson, D. (1985). Philosophy of science, science, and science education. Studies in Science
Education, 12: 25-57.
Holley, C. D., and Dansereau, D. F. (1984). Spatial Learning Strategies: Techniques,
Applications, and Related Issues, Academic Press, New York.
Hyde, J. S. (1981). How large are cognitive gender differences? A meta-analysis using w2 and
d. American Psychologist, 36: 892-901.
Hyman, R. (1953). Stimulus information as a determinant of reaction time. Journal of
Experimental Psychology, 45: 188-196.
Jackson, D. N. (1977). Jackson Vocational Interest Survey: Manual, Research Psychologists
Press, Port Huron, Michigan.
Jackson, D. N. (1984). Multidimensional Aptitude Battery (MAB): Manual, Research
Psychologists Press, Port Huron, Michigan.
Jensen, A. R. (1987). Individual differences in the Hick paradigm. In Vernon, P.A. (ed.),
Speed of Information-Processing and Intelligence, Ablex, Norwood, New Jersey, pp.
101-175.
Johnson, C. J., and Clark, J. M. (1988). Children's picture naming difficulty and errors: Effects
of age of acquisition, uncertainty, and name generality. Applied Psycholinguistics, 9:
351-365.
Johnson, P. E. (1964). Associative meaning of concepts in physics. Journal of Educational
Psychology, 55: 84-88.
Johnson-Laird, P. N. (1983). Mental Models: Toward a Cognitive Science of Language, Inference,
and Consciousness, Harvard University Press, Cambridge, Massachusetts.
Johnston, J. M. (1975). Behavior Research and Technology in Higher Education, Charles C.
Thomas, Springfield, Illinois.
Jones, G. V. (1985). Deep dyslexia, imageability, and ease of predication. Brain and Language,
24: 1-19.
Jorgenson, C. C., and Kintsch, W. (1973). The role of imagery in the evaluation of sentences.
Cognitive Psychology, 4: 110-116.
Kagan, D. M. (1988). Research on the supervision of counselors- and teachers-in-training:
Linking two bodies of literature. Review of Educational Research, 58: 1-24.
Katz, A. N. (1976). Verbal concept identification: Disentangling the dominance effect.
Unpublished doctoral dissertation, University of Western Ontario.
Katz, A. N., Paivio, A., Marschark, M., and Clark, J. M. (1988). Norms for 204 literary and
260 nonliterary metaphors on 10 psychological dimensions. Metaphor and Symbolic
Activity, 3: 191-214.
Kaufman, A. S., and Kaufman, N. L. (1983). Kaufman Assessment Battery for Children,
American Guidance Service, Circle Pines, Minnesota.
Kellogg, R. T. (1988). Attentional overload and writing performance: Effects of rough draft
and outline strategies. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 14: 355-365.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 204 Clark and Paivio

Khan, M. (1989. Mem

categorical organiz
Western Ontario, L
Khan, M., and Paiv
replication and ext
Psychology: Learnin
Kimura, D. (1987). A
133-147.
Kimura, D., and Archibald, Y. (1974). Motor functions of the left hemisphere. Brain, 97: 337-350.
Kintsch, W. (1974). The Representation of Meaning in Memory, Erlbaum, Hillsdale, New Jersey.
Kintsch, W., and Keenan, J. (1973). Reading rate and retention as a function of the number
of propositions in the base structure of sentences. Cognitive Psychology, 5: 257-274.
Kintsch, W., and van Dijk, T. A. (1978). Toward a model of text comprehension and
production. Psychological Review, 85: 363-394.
Kirschner, P. A., and Meester, M. A. M. (1988). The laboratory in higher science education:
Problems, premises, and objectives. Higher Education, 17: 81-98.
Klare, G. R. (1963). The Measurement of Readability, Iowa State University Press, Ames, Iowa.
Klare, G. R. (1974/1975). Assessing readability. Reading Research Quarterly, 10: 62-102.
Klatzky, R. L., McCloskey, B., Doherty, S., Pellegrino, J., and Smith, T. (1987). Knowledge
about hand shaping and knowledge about objects. Journal of Motor Behavior, 19: 187-213.
Klatzky, R. L., Pellegrino, J. W., McCloskey, B. P., and Doherty, S. (1989). Can you squeeze
a tomato? The role of motor representations in semantic sensibility judgments. Journal
of Memory and Language, 28: 56-77.
Klee, H., and Eysenck, M. W. (1973). Comprehension of abstract and concrete sentences.
Journal of Verbal Learning and Verbal Behavior, 12: 522-529.
Koch, S. (1981). The nature and limits of psychological knowledge: Lessons of a century qua
"science." American Psychologist 36: 257-269.
Kornblum, S. (1965). Response competition and/or inhibition in two-choice reaction time.
Psychonomic Science, 2: 55-56.
Kuhn, T. S. (1962). The Structure of Scientific Revolutions, University of Chicago Press,
Chicago.
Kulhavy, R. W., and Kardash, C. A. M. (1988). Studying study: An analysis of instructional
encoding behavior. Educational and Psychological Research, 8: 25-38.
Lachman, R. (1973). Uncertainty effects on time to access the internal lexicon. Journal of
Experimental Psychology, 99: 199-208.
Lang, P. J. (1979a). Language, image, and emotion. In Pliner, P., Blankstein, K. R., and Spigel,
I. M. (eds.), Perception of Emotion in Self and Others, Plenum, New York, pp. 107-117.
Lang, P. J. (1979b). A bio-informational theory of emotional imagery. Psychophysiology, 16:
495-512.
Lang, P. J. (1984). Cognition in emotion: Concept and action. In Izard, C. E., Kagan, J., and
Zajonc, R. B. (eds.), Emotions, Cognition, and Behavior, Cambridge University Press, New
York, pp. 192-226.
Lawton, D. (1968). Social Class, Language and Education, Routledge & Kegan Paul, London.
Levin, J. R., and Berry, J. K. (1980). Children's learning of all the news that's fit to picture.
Educational Communication and Technology, 28: 177-185.
Levin, J. R., and Pressley, M. (1985). Mnemonic vocabulary instruction: What's fact, what's
fiction. In Dillon, R. F. (ed.), Individual Differences in Cognition, Vol. 2, Academic Press,
New York, pp. 145-172.
Lewis, R. S., and Christiansen, L. (1989). Intrahemispheric sex differences in the functional
representation of language and praxic function in normal individuals. Brain and Cognition,
9: 238-243.
Linn, M. C., and Hyde, J. S. (1989). Gender, mathematics, and science. Educational Researcher,
18: 17-19, 22-27.
Lohman, D. F., and Kyllonen, P. C. (1983). Individual differences in solution strategy on
spatial tasks. In Dillon, R. F., and Schmeck, R. R. (eds.), Individual Differences in
Cognition, Vol. 1, Academic Press, New York, pp. 105-135.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 205

Long, S. A., Winograd, P. N., and Bridge, C.

characteristics on imagery reported during and
24: 353-372.
Luria, A. R. (1961). The Role of Speech in the Regulation of Normal and Abnormal Behavio
Pergamon, London.
Maccoby, E. E. (1980). Social Development: Psychological Development and the Parent-Chil
Relationship, Harcourt Brace Jovanovich, New York.
Maccoby, E. E., and Jacklin, C. N. (1974). The Psychology of Sex Differences, Stanfor
University Press, Stanford, California.
Mandler, G. (1958). Autonomic feedback: The perception of autonomic activity. Journal o
Abnormal and Social Psychology, 56: 367-373.
Mandler, G. (1975). Mind and Emotion, Wiley, New York.
Mandler, G. (1984). Mind and Body: Psychology of Emotion and Stress, W. W. Norton, Ne
York.
Mandler, J. M. (1984). Stories, Scripts, and Scenes: Aspects of Schema Theory, Erlbaum,
Hillsdale, New Jersey.
Marschark, M. (1985). Imagery and organization in the recall of prose. Journal of Memory
and Language, 24: 734-745.
Marschark, M., and Hunt, R. R. (1989). A reexamination of the role of imagery in learning
and memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15:
710-720.
Marschark, M., and Paivio, A. (1977). Integrative processing of concrete and abstract
sentences. Journal of Verbal Learning and Verbal Behavior, 16: 217-231.
Marschark, M., Richman, C. L., Yuille, J. C., and Hunt, R. R. (1987). The role of imagery
in memory: On shared and distinctive information. Psychological Bulletin, 102: 28-41.
Marsh, H. W. (1982). Validity of students' evaluations of college teaching: A multitrait
multimethod analysis. Journal of Educational Psychology, 74: 264-279.
Mayer, R. E. (1987). Educational Psychology: A Cognitive Approach, Little, Brown, Boston.
Mayer, R. E. (1989). Models for understanding. Review of Educational Research, 59: 43-64.
McAshan, H. H. (1974). The goals Approach to Performance Objectives, W. B. Saunders,
Philadelphia.
McClelland, D. C., Atkinson, J. W., Clark, R. W., and Lowell, E. L. (1953). The Achievement
Motive, Appleton-Century-Crofts, New York.
McKenna, M. C., and Robinson, R. D. (1980). An Introduction to the Cloze Procedure: An
Annotated Bibliography, International Reading Association, Newark, Delaware.
McMurry, F. M. (1909). How to Study and Teaching How to Study, Houghton Mifflin, New
York.
Meichenbaum, D. (1976). Toward a cognitive theory of self-control. In Schwartz, G. E., and
Shapiro, D. (eds.), Consciousness of Self-Regulation: Advances in Research, Vol. 1, Plenum,
New York, pp. 223-260.
Mercer, C. D., and Mercer, A. R. (1981). Teaching Students with Learning Problems, Charles
E. Merrill, Columbus, Ohio.
Meyer, B. J. F. (1977). What is remembered from prose: A function of passage structure. In
Freedle, R. O. (ed.), Advances in Discourse Production (Vol. 1): Discourse Production and
Comprehension, Ablex, Norwood, New Jersey, pp. 307-336.
Miller, S. (1970). Prediction of mediated paired-associate learning. Journal of Experimental
Psychology, 86: 131-132.
Moore, B. M. (1988). Achievement in basic math skills for low performing students: A study
of teachers' affect and CAI. Journal of Experimental Education, 57: 38-44.
Moss, C. J. (1982). Academic achievement and individual differences in the learning processes
of basic skills students in the university. Applied Psychological Measurement, 6: 291-296.
Murray, D. J. (1982). Rated associability and episodic memory. Canadian Journal of
Psychology, 36: 420-434.
Murray, H. G. (1983a). Low-inference classroom teaching behaviors in relation to six measures
of college teaching effectiveness. In Proceedings of the Conference on the Evaluation and
Improvement of University Teaching: The Canadian Experience, Montebello, Quebec.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 206 Clark and Paivio

Murray, H. G. (198
college teaching eff
Murray, H. G. (1985
In Donald, J. G.,
Jossey-Bass, San F
Nagy, P. (ed.) (1984)
in Education, Toronto, Canada.
Naveh-Benjamin, M., McKeachie, W. J., Lin, Y., and Tucker, D. G. (1986). Inferring students'
cognitive structures and their development using the "Ordered Tree Technique." Journal
of Educational Psychology, 78: 130-140.
Noterdaeme, M., Amorosa, H., Ploog, M., and Scheimann, G. (1988). Quantitative and
qualitative aspects of associated movements in children with specific developmental
speech and language disorders and in normal pre-school children. Journal of Human
Movement Studies, 15: 151-169.
Novak, J. D., and Gowin, D. B. (1984). Learning How to Learn,, Cambridge University Press,
Cambridge.
Nystrom, N. J., King, J. A., and Wimpelberg, R. K. (1984). Teacher education in private
colleges and universities: Scholarship and teaching. Journal of Teacher Education, 35:
42-45.
O'Neill, B. J. (1972). Defineability as an index of word meaning. Journal of Psycholinguistic
Research, 1: 287-298.
O'Neill, B. J., and Paivio, A. (1978). Some consequences of violating selection restrictions in
concrete and abstract sentences. Canadian Journal of Psychology, 32: 3-18.
Ozols, E. J., and Rourke, B. P. (1988). Characteristics of young learning-disabled children
classified according to patterns of academic achievement: Auditory-perceptual and
visual-perceptual abilities. Journal of Clinical Child Psychology, 17: 44-52.
Paivio, A. (1963). Learning of adjective-noun paired-associates as a function of adjective-noun
word order and noun abstractness. Canadian Journal of Psychology, 17: 370-379.
Paivio, A. (1965). Personality and audience influence. In Maher, B. A. (ed.), Progress in
Experimental Personality Research, Vol. 2, Academic Press, New York, pp. 127-173.
Paivio, A. (1966). Latency of verbal associations and imagery to noun stimuli as a function
of abstractness and generality. Canadian Journal of Psychology, 20: 378-387.
Paivio, A. (1968). A factor-analytic study of word attributes and verbal learning. Journal of
Verbal Learning and Verbal Behavior, 7: 41-49.
Paivio, A. (1969). Mental imagery in associative learning and memory. Psychological Review,
76: 241-263.
Paivio, A. (1971). Imagery and Verbal Processes, Holt, Rinehart, and Winston, New York
(Reprinted 1979, Erlbaum, Hillsdale, New Jersey).
Paivio, A. (1975a). Neomentalism. Canadian Journal of Psychology, 29: 263-291.
Paivio, A. (1975b). Coding distinctions and repetition effects in memory. In Bower, G. H.
(ed.), The Psychology of Learning and Motivation, Vol. 9, Academic Press, New York, pp.
179-214.
Paivio, A. (1975c). Imagery and synchronic thinking. Canadian Psychological Review, 16:
147-163.
Paivio, A. (1978). Mental comparisons involving abstract attributes. Memory & Cognition, 6:
199-208.
Paivio, A. (1983a). The minds eye in arts and science. Poetics, 12: 1-18.
Paivio, A. (1983b). The empirical case for dual coding theory. In Yuille, J. C. (ed.), Imagery,
Memory and Cognition: Essays in Honor of Allan Paivio, Erlbaum, Hillsdale, New Jersey,
pp. 307-332.
Paivio, A. (1985). Cognitive and motivational functions of imagery in human performance.
Canadian Journal of Applied Sport Sciences, 10: 22-28.
Paivio, A. (1986). Mental Representations: A Dual-Coding Approach, Oxford University Press,
New York.
Paivio, A. (1991). Dual coding theory: Retrospect and current status. Canadian Journal of
Psychology.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 207

Paivio, A., and Begg, I. (1971a). Imagery and

Journal of Experimental Psychology, 89: 40-4
Paivio, A., and Begg, I. (1971b). Imagery an
sentence concreteness and structure. Percep
Paivio, A., and Begg, I. (1981). Psychology of
Jersey.
Paivio, A., and Clark, J. M. (1986). The role of topic and vehicle imagery in metaphor
comprehension. Communication & Cognition 19: 367-388.
Paivio, A., and Clark, J. M. (1991). Static versus dynamic imagery. In Cornoldi, C., and
McDaniel, M. (eds.), Imagery and Cognition, Springer-Verlag, New York, pp. 221-245.
Paivio, A., Clark, J. M., Digdon, N., and Bons, T. (1989). Referential processing: Reciprocity
and correlates of naming and imaging. Memory & Cognition, 17: 163-174.
Paivio, A., Clark, J. M., and Khan, M. (1988a). Effects of concreteness and semantic
relatedness on composite imagery and cued recall. Memory & Cognition, 16: 422-430.
Paivio, A., Clark, J. M., and Lambert, W. E. (1988b). Bilingual dual coding theory and
semantic repetition effects on recall. Journal of Experimental Psychology: Human Learning
and Memory, 14: 163-172.
Paivio, A., and Desrochers, A. (1980). A dual-coding approach to bilingual memory. Canadian
Journal of Psychology, 34: 390-401.
Paivio, A., and Desrochers, A. (1981). Mnemonic techniques in second-language learning.
Journal of Educational Psychology, 73: 780-795.
Paivio, A., and Harshman, R. A. (1983). Factor analysis of a questionnaire on imagery and
verbal habits and skills. Canadian Journal of Psychology, 37: 461-483.
Paivio, A., and Lambert, W. E. (1959). Measures and correlates of audience anxiety ("stage
fright"). Journal of Personality, 27: 1-17.
Paivio, A., and Lambert, W. (1981). Dual coding and bilingual memory. Journal of Verbal
Learning and Verbal Behavior, 20: 532-539.
Paivio, A., and Marschark, M. (1980). Comparative judgments of animal intelligence and
pleasantness. Memory & Cognition, 8: 39-48.
Paivio, A., Yuille, J. C., and Madigan, S. A. (1968). Concreteness, imagery, and meaningfulness
values for 925 nouns. Journal of Experimental Psychology Monograph Supplement, 76 (1,
Pt. 2): 1-25.
Pellegrino, J. (1987). Measuring versus understanding individual differences in cognitive
abilities. In Modgil, S., and Modgil, C. (eds.), Arthur Jensen: Consensus and Controversy,
Falmer Press, New York, pp. 267-277.
Potter, M. C., Kroll, J. F., Yachzel, B., Carpenter E., and Sherman, J. (1986). Pictures in
sentences: Understanding without words. Journal of Experimental Psychology: General, 115:
281-294.
Potter, M. C., and Kroll, J. F. (1987). Conceptual representation of pictures and words: Reply
to Clark. Journal of Experimental Psychology: General, 116: 310-311.
Preece, P. F. W. (1976). Mapping cognitive structure: A comparison of methods. Journal of
Educational Psychology, 68: 1-8.
Pressley, M., Borkowski, J. G., and O'Sullivan, J. (1985). Children's metamemory and the
teaching of memory strategies. In Forrest-Pressley, D. L., MacKinnon, G. E., and Waller,
T. G. (eds.), Metacognition, Cognition, and Human Performance, Academic Press, New
York, pp. 111-153.
Pressley, M., Goodchild, F., Fleet, J., Zajchowski, R., and Evans, E. D. (1989a). The challenges
of classroom strategy instruction. The Elementary School Journal, 89: 301-342.
Pressley, M., Johnson, C. J., Symons, S., McGoldrick, J., and Kurita, J. (1989b). Strategies
that improve children's memory and comprehension of what is read. Elementary School
Journal, 90: 3-32.
Pressley, M., McDaniel, M. A., Turnure, J. E., Wood, E., and Ahmad, M. (1987). Generation
and precision of elaboration: Effects on intentional and incidental learning. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 13: 291-300.
Pressley, M., Symons, S., Snyder, B. L., and Cariglia-Bull, T. (1989c). Strategy instruction
research comes of age. Learning Disability Quarterly, 12: 16-30.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 208 Clark and Paivio

Rabinowitz, M., and

of Experimental Ps
Raygor, A. L. (197
California.
Reif, F-, and Heller, J. I. (1982). Knowledge structure and problem solving in physics.
Educational Psychologist, 17: 102-127.
Resnick, L. B. (1981). Social assumptions as a context for science: Some reflections on
psychology and education. Educational Psychologist, 16: 1-10.
Reynolds, A., and Paivio, A. (1968). Cognitive and emotional determinants of speech.
Canadian Journal of Psychology, 22: 164-175.
Richardson, J. T. E. (1980). Concreteness, imagery, and semantic categorization. Journal of
Mental Imagery, 4: 51-58.
Riding, R. J., and Calvey, I. (1981). The assessment of verbal-imagery learning styles and
their effect on the recall of concrete and abstract prose passages by 11-year-old children.
British Journal of Psychology, 72: 59-64.
Robinson, F. P. (1970). Effective Study (4th ed.), Harper & Row, New York.
Robinson, N. M., and Robinson, H. B. (1976). The Mentally Retarded Child: A Psychological
Approach (2nd ed), McGraw-Hill, New York.
Roediger, H. L. (1980). The effectiveness of four mnemonics in ordering recall. Journal of
Experimental Psychology: Human Learning and Memory, 6: 558-567.
Ronning, R. R., Conoley, J. C., and Glover, J. A. (1987). Introduction: The implications of
cognitive psychology for testing. In Ronning, R. R., Conoley, J. C., Glover, J. A., and
Witt, J. C. (eds.), The Influence of Cognitive Psychology on Testing: Buros-Nebraska
Symposium on Measurement and Testing, Vol. 3, Erlbaum, Hillsdale, New Jersey, pp. 1-8.
Rosenshine, B., and Furst, N. F. (1971). Research on teacher performance criteria. In Smith,
B. O. (ed.), Research in Teacher Education: A Symposium, Prentice Hall, Englewood Cliffs,
New Jersey, pp. 37-72.
Rosenshine, B., and Martin, M. (1974). Teacher education and teacher behavior: Comments
on the state-of-the-research. Educational Researcher 3: 11-12.
Royce, J. R. (1982). Philosophical issues, division 24, and the future. American Psychologis
37: 258-266.

Rumberger, R. W. (1987). High school dropouts: A review of issues and evidence. Review o
Educational Research, 57: 101-121.
Sadoski, M. (1983). An exploratory study of the relationships between reported imagery an
the comprehension and recall of a story. Reading Research Quarterly, 19: 110-123.
Sadoski, M. (1984). Text structure, imagery, and affect in the recall of a story by children. In
Niles, J. A., and Harris, L. A. (eds.), Changing Perspectives in Research in
Reading/Language Processing and Instruction, 33d Yearbook of the National Reading
Conference, National Reading Conference, Washington, D.C., pp. 48-53.
Sadoski, M. (1985). The natural use of imagery in story comprehension and recall: Replication
and extension. Reading Research Quarterly, 20: 658-667.
Sadoski, M., and Goetz, E. T. (1985). Relationships between affect, imagery, and importance
ratings for segments of a story. In Niles, J. A., and Lalik, R. (eds.), Issues in Literacy: A
Research Perspective, 34th Yearbook of the National Reading Conference, National
Reading Conference, Washington, D.C., pp. 180-185.
Salvia, J., and Ysseldyke, J. E. (1988). Assessment in Special and Remedial Education (4th
ed.), Houghton-Mifflin, Boston.
Schibeci, R. A. (1981). Do teachers rate science attitude objectives as highly as cognitive
objectives? Journal of Research in Science Teaching, 18: 69-72.
Schibeci, R. A. (1984). Attitudes to science: An update. Studies in Science Education, 11: 26-59.
Schibeci, R. A., and McGaw, B. (1981). Empirical validation of the conceptual structure of
a test of science-related attitudes. Educational and Psychological Measurement, 41:
1195-1201.
Schmeck, R. R. (1983). Learning styles of college students. In Dillon, R. F., and Schmeck, R. R.
(eds.), Individual Differences in Cognition: Vol. 1, Academic Press, New York, pp. 233-279.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 Dual Coding Theory and Education 209

Schmeck, R. R., and Ribich, F. D. (1978). Const

processes. Applied Psychological Measurement
Schmeck, R. R., Ribich, F. D., and Ramanaiah
inventory for assessing individual differences
Measurement, 1: 413-431.
Schmidt, R. A. (1975). A schema theory of discre
82: 225-260.
Schwartz, G. E., Ahern, G. L., and Brown, S. (1979). Lateralized facial muscle response to
positive and negative emotional stimuli. Psychophysiology 16: 561-571.
Schwartz, G. E., Davidson, R. J., and Goleman, D. J. (1978). Patterning of cognitive and
somatic processes in the self-regulation of anxiety: Effects of meditation versus exercise.
Psychosomatic Medicine, 40: 321-328.
Shallice, T. (1978). The dominant action system: An information processing approach to
consciousness. In Pope, K. S., and Singer, J. L. (eds.), The Stream of Consciousness:
Scientific Investigations into the Flow of Human Experience, Plenum, New York, pp.
117-157.
Shavelson, R. J. (1974). Methods for examining representations of a subject-matter structure
in a student's memory. Journal of Research in Science Teaching, 11: 231-249.
Shepard, R. N. (1978). Externalization of mental images and the act of creation. In Randhawa,
B. S., and Coffman, W. E. (eds.), Visual Learning, Thinking, and Communication,
Academic Press, New York, pp. 133-189.
Simmons, D. C., Griffin, C. C., and Kameenui, E. J. (1988a). Effects of teacher-constructed
pre- and post-graphic organizer instruction on sixth-grade science students' compre
hension and recall. Journal of Educational Research, 82: 15-21.
Simner, M. L. (1989). Predictive validity of an abbreviated version of the Printing Performance
School Readiness Test. Journal of School Psychology, 27: 189-195.
Singer, J. L. (1966). Daydreaming: An Introduction to the Experimental Study of Inner
Experience, Random House, New York.
Sisson, L. A., Kilwein, M. L., and van Hasselt, V. B. (1988). A graduated guidance procedure
for teaching self-dressing skills to multihandicapped children. Research in Developmental
Disabilities, 9: 419-432.
Snowling, M., Van Wagtendonk, B., and Stafford, C. (1988). Object naming deficits in
developmental dyslexia. Journal of Research in Reading, 11: 67-85.
Sosniak, L. A., and Ethington, C. A. (1989). School quality and high school seniors' attitudes
toward academic coursework. Journal of Educational Research, 82: 113-119.
Starch, D. (1928). Educational Psychology (rev. ed.), Macmillan, New York.
Starch, D. (1930). Experiments in Educational Psychology (revised by author and Harry F.
Latshaw), Macmillan, New York.
Stark, J. S., Lowther, M. A., and Hagerty, B. M. K. (1986). Faculty roles and role preferences
in ten fields of professional study. Research in Higher Education, 25: 3-30.
Sternberg, R. J. (1981). Testing and cognitive psychology. American Psychologist, 34: 1181-1189.
Storm, C., and Storm, T. (1987). A taxonomic study of the vocabulary of emotions. Journal
of Personality and Social Psychology, 53: 805-816.
Str0mnes, F. J., and Iivonen, L. (1985). The teaching of the syntax of written language to
deaf children knowing no syntax. Human Learning, 4: 251-265.
Suinn, R. M. (1983). Imagery and sports. In Sheikh, A. A. (ed.), Imagery: Current Theory,
Research, and Applications, Wiley, New York, pp. 507-534.
Sullivan, A. M., and Skanes, G. R. (1974). Validity of student evaluation of teaching and the
characteristics of successful instructors. Journal of Educational Psychology, 66: 584-590.
Terrasi, S., and Airasian, P. W. (1989). The relationship between adaptive behavior and
intelligence for special needs students. Psychology in the Schools, 26: 202-208.
Thomassen, A. J. W. M., and Teulings, H. H. M. (1983). The development of handwriting.
In Martlew, M. (ed.), The Psychology of Written Language: Developmental and Educational
Perspectives, Wiley, New York, pp. 179-213.
Toglia, M. P., and Battig, W. F. (1978). Handbook of Semantic Word Norms, Erlbaum,
Hillsdale, New Jersey.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms
 210 Clark and Paivio

Troyer, M. B. (1986
Journal of Teacher
Underwood, B. J.,
instructions and dom
Vaid, J. (1988b). Bil
Canadian Journal of
van der Wissel, A. (1
Journal of Learning
van Dijk, T. A., and
Press, New York.
Vollmer, F. (1984). E
Weaver, W. T. (1983
New York.
Weinberg, R. S. (1982). The relationship between mental preparation strategies and motor
performance: A review and critique. Quest, 33: 195-213.
Weinstein, C. E., Underwood, V. L., Wicker, F. W., and Cubberly, W. E. (1979). Cognitive
learning strategies: Verbal and imaginai elaboration. In O'Neill, H. F., and Spielberger,
C. D. (eds.), Cognitive and Affective Learning Strategies, Academic Press, New York, pp.
45-75.
Weschler, D. (1974). Manual for the Wechsler Intelligence Scale for Children —Revised, The
Psychological Corporation, New York.
Westerback, M. E., and Primavera, L. H. (1988). Anxiety about science and science teaching.
In Spielberger, C. D., and Butcher, J. N. (eds.), Advances in Personality Assessment,
Vol. 7, Erlbaum, Hillsdale, New Jersey, pp. 175-202.
Westkott, M. (1979). Feminist criticism of the social sciences. Harvard Educational Review,
49: 422-430.
Wharton, W. P. (1980). Higher-imagery words and the readability of college history texts.
Journal of Mental Imagery, 4: 129-147.
Wharton, W. P. (1985). Imagery and the readability of textbooks. In Sheikh, A. A., and Sheikh,
K. S. (eds.), Imagery in Education, Baywood, Farmingdale, New York, pp. 145-160.
Williams, J. P., Taylor, M. B., and Ganger, S. (1981). Text variations at the level of the
individual sentence and the comprehension of simple expository paragraphs. Journal of
Educational Psychology, 73: 851-865.
Winne, P. W., and Marx, R. W. (1977). Reconceptualizing research on teaching. Journal of
Educational Psychology, 69: 668-678.
Wittrock, M. C., and Alesandrini, K. (1990). Generation of summaries and analogies and
analytic and holistic abilities. American Educational Research Journal, 27: 489-502.
Woolfolk, A. E. (1987). Educational Psychology (3d ed.), Prentice-Hall, Englewood Cliffs, New Jersey.
Wynne, R. D., Gerjuoy, H., and Schiffman, H. (1965). Association test antonym-response set.
Journal of Verbal Learning and Verbal Behavior, 4: 354-359.
Yee, A. H. (1969). Do cooperating teachers influence the attitudes of student teachers?
Journal of Educational Psychology, 60: 327-332.
Yovetich, W. S., and Paivio, A. (1981). Cognitive processing of Biiss-like symbols by normal
populations: A report on four studies. Report of EASE 80 Communication and Handicap,
The Finnish Association for Special Education.
Yuille, J. C., and Paivio, A. (1969). Abstractness and the recall of connected discourse. Journal
of Educational Psychology, 82: 467-471.

This content downloaded from 159.237.12.142 on Mon, 06 Apr 2020 08:31:19 UTC
All use subject to https://about.jstor.org/terms