This article was downloaded by: [University of Chicago]

On: 05 July 2012, At: 11:00

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Educational Psychology: An
International Journal of Experimental
Educational Psychology
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/cedp20

Cognitive Styles—an overview and

integration
Richard Riding a & Indra Cheema a

a
Assessment Research Unit, University of Birmingham, England

Version of record first published: 25 Nov 2010

To cite this article: Richard Riding & Indra Cheema (1991): Cognitive Styles—an overview and
integration, Educational Psychology: An International Journal of Experimental Educational
Psychology, 11:3-4, 193-215

To link to this article: http://dx.doi.org/10.1080/0144341910110301

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-

conditions

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any
representation that the contents will be complete or accurate or up to date. The
accuracy of any instructions, formulae, and drug doses should be independently
verified with primary sources. The publisher shall not be liable for any loss, actions,
claims, proceedings, demand, or costs or damages whatsoever or howsoever caused
arising directly or indirectly in connection with or arising out of the use of this
material.
 Educational Psychology, Vol. 11, Nos. 3 and 4, 1991 193

EDITORIAL ARTICLE

Gognitiue Styles—an oreroiew and integration

Downloaded by [University of Chicago] at 11:00 05 July 2012

RICHARD RIDING & INDRA CHEEMA, Assessmenc Research Unit, UniT›ersity

of Birmingham, England

xnswcr This review article considered the nature of styles and si:rategies and then
surc›eyed work on cognitic e s(yfes. Different researchers have used a variety of labels for
the styles they have investigated. Analysis of the sray in srhich they assessed style, its
effect on behaviour and perfomiance and studies of the relationship to other labels,
suggested thai che labels may be grouped into no prinGipol fogritrire styles. These mere
labelled the Wliolist-Analytic and Verbialiser-Imager dimensions. A computer presented
method of assessing the position of an individual on these dimensions was described.

Introduction
The cognitive style construct has been elusive; this is partly due to the fact that many
researchers working within the learning/cognitive style research, fail to mention the
existence of other types of styles. As a result, like the blind man and the elephant,
different theorists have been working with different concepts and have referred to
them as a ’cognitive/learning style’. Indeed, attempts to unite these scattered schools
of thought have been extremely rare.
Lewis (1976, pp. 304—5) remarked that
Different groups of researchers seem determined to pursue their own pet
distinctions in cheerful disregard of one another ... There is the impulsive
versus re8ective distinction, which seems to indicate something about the
tempo of learning. There is the field dependent versus field independent
distinction, the serialist and the holist, and a lot more ... In my opinion, the
right thing to do is to focus ... on the search for individual differences which
are basic, in the sense that they underlie (and to that extent, explain), a
whole range of more readily observable differences.
 194 R. Riding ‹Sr I. Gheema

Curry (1983) too, recognised the root causes preventing significant progress in
application of cognitive styles to training and education. She claimed that “chief
among these difficulties is the bewildering confusion of definitions surrounding
learning style conceptualizations and the concomitant wide variation in scale or scope
of behaviour claimed to be predicted by learning style models”.
Smdies in learning styles/cognitive styles initially developed as a result of interest in
individual differences. These issues were very much in fashion during the 1960’s,
enjoyed a continuing, popularity during the early 1970’s, but have since tended to
decline. This waning of interest left the whole field of investigation fragmented and
incomplete, and without clear usefulness for the central concerns of education. Despite
attracting little interest in the last two decades, the authors feel that cognitive/learning
styles are once again coming into prominence, and are now being more seriously
considered by the teaching and training world.

The Nature of Styles and Strategies

Downloaded by [University of Chicago] at 11:00 05 July 2012

The Need for Clarification

The term ‘cognitive style’, was used by Allport (1937), and has been described as a
person’s typical or habitual mode of problem solving, thinking, perceiving and
remembering.
The terms cognitive style and learning style have been much used by theorists, but
what they mean still remains very much up to its author. Whilst some, e.g. Entwistle
(1981) believe the two terms to mean the same and so have used the terms
interchangeably; others, e.g. Das (1988a) consider the two to be different and attempt
to define them as separate concepts.
‘Learning style’, seems to have emerged as a more common term or a replacement
term for cogrtitive style in the 1970’s. Indeed, the impression that is formulated in the
usage of these terms is that those worJxing under the umbrella of ’learning style’, take
cognitive style into consideration, but would probably describe themselves as inter-
ested in more practical, educational or training applications and are thus more ’action-
orientated’, while the term cognitive style has been reserved for theoretical, academic
descriptions. Therefore, if cognitive style actually underlies learning style, it has quite
practical implications.
One main difference between cognitive and learning style, is the number of style
elements considered. That is, whilst cognitive style is a bipolar dimension, learning
style entails many elements and are usually not ’either-or’ extremes. One either has or
does not have the element in one’s style, similarly, the absence of one element does not
necessarily imply the presence of the opposite element.
Cognitive/learning style measures traditionally occupy a middle ground between
aptitude measures and personality measures. As such, their status has been unclear and
this lack of clarity has meant that regular attempts to distinguish them from aptitude
and personality measures have been made (Kagan & Kogan, 1970; Kogan, 1980),
successfully or unsuccessfully.
Cognitive/learning styles have been viewed in three main ways, as a structure
(content), as a process or as both. The implications of these differing views are that if
cognitive style is viewed as a structure, then the focus is on its stability over time¡ as
such, style is a ’given’ in a training or an educational setting. Once the style in the
setting is identified, the training material can be adapted or ‘matched’ to the indivi-
dual’s cognitive style. However, if cognitive style is viewed as a process, then the focus
 Cognitic›e Styles 195

is on how it changes¡ as such, trainers may even try to foster that change. Style areas
can be built upon and can be used to compensate for or strengthen weaknesses. Style is
seen as dynamic, not ‘frozen forever’. For others, cognitive style is viewed as both
process and structure. It may be relatively stable, not changeable like liquid with no
form of its own, yet at the same time always in 8ux. In such a view, style structure is
continually modified as new events in8uence it directly or indirectly.

Gurry’s Onion Model

Curry (1983) proposed that all cognitive/learning style measures may be grouped into
three main types or ’strata resembling layers of an onion’. By organising learning style
measures like this, “learning behaviour is fundamentally controlled by the central
personality dimension, translated through middle strata information processing dimen-
sions and, given a final twist by interaction with environmental factors encountered in
the other strata”.
Downloaded by [University of Chicago] at 11:00 05 July 2012

According to Curry, the outermost layer of the onion and the most observable style
is what she labels as 'instructional preference’. Examples of measures of this style are
the ’Learning Preference Inventory’ (Rezler & Rezmovic, 1981); Grasha Reichmann
Smdents Learning Styles Scales (Reichmann & Grasha, 1974). Instructional prefer-
ence refers to the individual’s choice of environment in which to learn. Since this is the
layer that is most exposed to the learner’s environment, learner expectations, teacher
expectations and other external feamres, Curry believes instructional prefer-
ence to be the least stable, the most easily inBuenced level of measurement in the
learning styles area.
The second layer of the onion model is referred to as the ‘information processing
style’. This is considered as the individual’s intellectual approach to assimilating
information and, because this processing does not directly involve the environment,
Curry believes that measures of this style are more stable than instructional prefer-
ence, but are still modifiable by learning strategies. Learning style measures such as
the ‘Learning Style Inventory’ (Kolb, 1976); ’Cognitive Preference Inventory’ (Tamir
& Cohen, 1980) and ’Inventory of Learning Processes’ (Schmeck et af., 1977) are all
thought to be dealing with information processing style concepts applicable at the
intersection between fundamental personality level individual differences and environ-
mentally offered learning format choices.
The third and the innermost layer of the hypothetical learning style onion is
cognitive personality style. The cognitive personality style is defined as the individual’s
approach to adapting and assimilating information, which does not interact directly
with the environment, but is an underlying and relatively permanent personality
dimension that is expressed indirectly and is apparent only when an individual’s
behaviour is observed across many learning instances.
Examples of measures which Curry suggested assess this third type of learning style
are the ‘Embedded Figures Test’ (Witkin, 1962)¡ ‘Myers-Briggs Type Indicator’
(Myers, 1962) and ‘Matching Familiar Figures Test’ (Kagan, 1965).

Styles versus Strafegies

‘Cognitive style’ needs to be distinguished from 'learning strategy’. In the present
context, a style is considered to be a fairly fiaed characteristic of an individual, while
strategies are the ways that may be used to cope with situations and tasks. Strategies
 196 R. Riding ‹Sr I. Gheema

may vary from time to time, and may be learned and developed. Styles, by contrast are
static and are relatively in-built features of the individual.

The Relationship Between Cognitive Styles

Cognitive Syle Fainilies
Like Riding & Buckle (1990), Coan (1974), Fowler (1980) and Brumby (1982), Miller
(1987) argued that a number of cognitive style labels are, but different conceptions of
the same dimension. With respect to what will be termed the Wholist-Analytic cognitive
style, this view is also shared by Nickerson et at. (1985, p. 50) who stated,
. . the view that there are two qualitatively different types of thinking is
widely shared. Among the terms used to describe one type are analytic,
deductive, rigourous, constrained, convergent, formal and critical. Represen-
Downloaded by [University of Chicago] at 11:00 05 July 2012

tative of the terms used to describe the other type are synthetic, inductive,
expansive, unconstrained, divergent, informal, diffuse and creative. No doubt
the partitioning of thinning into two types involves something of an oversim-
plification, but possibly a useful one.
Messick (1984) identified 19 cognitive styles. However, a literature search by the
authors, revealed over 30 labels referred to as cognitive/learning styles. Analysis of
these suggests that they may be grouped into two principal cognitive styles and a
number of learning strategies as follows.

Wholist-Analytic Co iriee Style Family

Field dependence-independence (Witkin, 1962); impulsivity—reBectivity (Kagan,
1965); holist-serialist (Pask, 1972)¡ leveller-sharpener (Holzman & Klein, 1954);
simultaneous-successive (Das, 1988b); diverging—converging (Hudson, 1966); toler-
ant-intolerant (Gardner et al., 1959); Bexible control—constricted automatisation c.
restructuring (Gardner ec al., 1959)i compartmentalization; concepmal articulation.
Other terms used include: conceptual integration; cognitive simplicity e. cognitive
complexity; scanning (Broad r. Narrow)¡ risit taking v. cautiousness; conceptual
differentiation; global-analytic; broad—narrow categorisation; surface—deep¡ formarticu-
lation—element articulation¡ conceptualising styles-relational-analytic.

Verbalir er-I mager Cogniiit e Style Family

—Sensory modality preferences (Bartlett, 1932)
—Verbalizer—imager (Riding & Taylor, 1976)
—Verbalizer—visualiser (Richardson, 1977).

Learning Strategies
—Myers-Briggs Type Indicator (Myers, 1962)
—Learning Styles Inventory (Dunn er al., 1975)
—Divergers-assimilator—converger—accommodator (Kolb, 1977)
—Activist—reBective—theorist—pragmatic
—Surface-deep (Biggs, 1987).
 Whilst each entry had been an object of an empirical research effort, these efforts
have varied from a handful of studies on one extreme, to a proliferation of studies at
the other. Furthermore, of the 19 styles listed by Messicit, the large majority have
attracted little subsequent attention and, in most cases, they were the author’s own
words.
The two fundamental cognitive style families will be brieBy reviewed.

Wholist-Analytic Style
Various terms have been used to describe this style dimension. What appear to be the
principal ones of these will be reviewed and the evidence for the relationship between
them examined. The criteria used for selecting the cognitive style measures in this
dimension were that, (a) only those cognitive style measures would be examined which
were ’established’ as judged by the amount of research worit carried out using it since
Downloaded by [University of Chicago] at 11:00 05 July 2012

they were first observed, (b) it should be applicable for cognitive tasks and classroom
teaching, and (c) it shared similar characteristics to the most prominent style label,
field-dependence—independence.

1. Field-Dependence-Independence
Description
The concept of the field dependence—independence cognitive style emerged as a result
of the work of Titkin (Within, 1962, 1978) and derived its theoretical basis from
Werner's (1948) organismic theory of development.
Witkin & Goodenough (1981) described how the conception of the field-dependence
—independence construct has evolved over time. That is, from the narrow 'perception
of the upright’, to ‘perceptual-analytic ability’. Later with the awareness that compe-
tence at disembedding in perceptual tests is strongly associated with competence at
disembedding in non-perceptual problem-solving tasks, it was broadened to encompass
both perceptual and intellectual activities, and referred to as the ’global-articulated
dimension’. However, with additional evidence on self-consistency, extending to the
areas of body concept, sense of self, and controls and defences, the construct became
more comprehensive and was labelled as ’differentiation’.

Assessment
Measurement of field dependence—independence tendencies has been assessed by the
Body Adjustment Test, the ’Rod and Frame’ Test and the Embedded Figures Test
(EFT and GEFT). All of these measures involve the disembedding of a shape from its
surrounding field. Since these are generally well known they will not be described here.

Effect
Cognitive. Regarding teaching methods, Within, Moore, Goodenough & Cox (1977)
reported that field independent teachers prefer more formal approaches, whilst field
dependent teachers prefer frequent interaction with learners. It further appears that
field independent teachers use questions as instructional tools, whilst field dependent
 198 R. Riding T' I. Gheema

teachers use questions as a check on what has been learnt (Moore, 1973). Field
independent teachers have a tendency to emphasise their own standards and to
formulate principles themselves when explaining subject matter to learners, but field
dependent teachers tend to involve learners more in organising the content and
sequences of the teaching-learning process, and encourage learners to formulate
principles themselves (Within ci ef., 1977; Gordon & Gross, 1978). Field independent
teachers are more inclined to correct the learner and explain why they are incorrect,
whilst field dependent teachers are reluctant to express critical feedbaclt. Field
dependent teachers are mainly interested in creating and maintaining positive attitudes,
and encouraging a good classroom ecosystem with not much emphasis on subject
content, whilst field independent teachers have an opposite set of priorities. Moreover,
field independent teachers are perceived by learners as encouraging them to apply
principles, in contrast to field dependent teachers who are seen as simply providing
facts (Witkin er al., 1977). In view of the range of teacher and pupil variables affected
Downloaded by [University of Chicago] at 11:00 05 July 2012

by field dependence/field, independence cognitive style, it is not surprising then that

this cognitive style dimension is associated with learning outcomes.
Witkin & Goodenough (1981) claimed that field independent individuals rely on an
internal frame of reference, while field dependent individuals rely on an external frame
of reference. These differences are rejected in the cognitive restructuring skills
employed by field independent and field dependent learners. In addition, they have
identified three separate, but related skills within cognitive restructuring (i) providing
structure for an ambiguous stimulus complex, (ii) breaking up an organised field into
its basic elements, and (iii) providing a different organisation to a field than that which
is suggested by the inherent structure of the stimulus complex. Field independent
individuals are perceived as being more capable in all three areas.

Social. Witkin et al. (1977) have shown that the field dependence/field—independence
cognitive style continuum affects preference for, and response to, different kinds of
learning/teaching methods. Whilst field dependent individuals have a preference to
learn in groups and to interact frequently with one another as well as the teacher, field
independent learners may respond better to more independent and more individualised
approaches. Also, field independent learners are more likely to have self-defined goals
and to respond to intrinsic reinforcement, whilst field dependent learners require more
extrinsic reinforcement and more structured worit by the teacher. Whereas the field
independent learners prefer to structure their own learning, and like to develop their
own learning strategies, field dependent learners may need more assistance in problem-
solving strategies or more exact definitions of performance outcomes (Within ct of.,
1977).
Field independent individuals are more able to deal with situations requiring
impersonal analysis whilst field dependent individuals are better equipped to deal with
situations requiring social perceptiveness and interpersonal skills.

Style Implications
Widiger ei al. (1980) have argued that field dependence does not exist as a style
distinct from abilities such as those involved in measures of spatial intelligence. It
turns out that measures of field-independence correlate as highly with these intelli-
gence measures as the latter correlate with each other. This point will be considered in
the concluding section of the paper.
 GoJriiiiee Syles 199

2. Impul«ivity-Reneetivity

Another cognitive style dimension derived from perceptual tasks is that itnown as
impulsivity-re8ectivity, or concepmal tempo. This dimension was originally intro-
duced by Kagan ct al. (1964) to describe the individual differences in the speed with
which decisions are made under conditions of uncertainty.

Assessment
Such differences are most obvious and are ’measured’ in matching-to-sample exercises,
e.g. Matching Familiar Figures Test (MFFT), where the individual is shown a line
drawing of an object or figure and has to indicate which of a number of other drawings
is exactly the same as the standard provided. Using such tasks, Kagan noted learners
Downloaded by [University of Chicago] at 11:00 05 July 2012

could be grouped into two main types, those that responded by inspecting the
alternatives brieBy and then quickly selecting the one they considered to be correct
(cognitively impulsives), and those who deliberate before responding, carefully consi-
dering each alternative (cognitively re8ective).

Effect
Overall, it seems that impulsivity—re8ectivity holds for tasks that are taught in school,
as well as those mastered naturally. For academic tasks, impulsiveness may play a
particularly important role, given the fact that impulsiveness in the early stages of
learning especially can have cumulative negative effects, whereas the effects of
overcautiousness on latency in the early stages may be largely overcome through
practice.
The relationship between the impulsivity-re8ectivity dimension and its effect on
cognitive tasks has been investigated by Kagan et at. (1966) who studied the
relationship between conceptual tempo to children's skill in inductive reasoning tests.
Using 6-year-olds as subjects, their smdy showed that there was a significant relation-
ship between re8ectivity and inferential skills.
More recently, Zelniker & Jeffrey (1979) drew attention to a parallel asymmetry in
their style research. They found that reBectives consistently performed better than
impulsives on tasks requiring detail processing, but that impulsives usually performed
only as well as re8ectives on tasits thought to require global processing.

Relationship to Other Labels

Banta (1970) reported a highly significant correlation between field dependence—inde-
pendence and impulsivity-reBectivity in a sample of lower class children aged 3-6
years. Massari & Massari (1973) using disadvantaged preschoolers obtained significant
correlations between MFFT and EFT scores. Schleifer & Douglas (1973) using the
EFT and impulsivity-re8ectivity measures on middle class and disadvantaged pre-
schoolers, found identical significant correlations for both samples.
Messer (1976) in reviewing impulsivity-re8ectivity style, quoted additional studies
evidencing a moderate relationship between field dependence—independence and
impulsivity—re8ectivity. Re8ectives are significantly more field independent than
 200 2t. Riding T• I. Gheema

impulsives (Campbell & Douglas, l972¡ Keogh & Donlon, 1972, Massari, 1975;
Neimark, 1975).

3. Convergent-Divergent Thinking
Description
The term convergent-divergent thinking was proposed by Guilford in the early 1950’s,
when he introduced his ’model of the intellect’. Guilford described the convergent
thinker as one who can be distinguished by his ability in dealing with problems
requiring one conventional correct answer clearly obtainable from the information
provided. Examples of such questions are those seen in intelligence tests, multiple
choice tests and even many of the psychological tests, e.g. EFT; MFFT, where each
question is provided with several choices, but only one is correct.
Downloaded by [University of Chicago] at 11:00 05 July 2012

The divergent thinner, however, is identified as one being highly adept in problems
requiring the generation of several equally acceptable answers where the emphasis is
on the quantity, variety and originality of responses. Whilst the convergingZdiverging
processes are not exclusive as certain convergent problems also require diverging, in
assessment, the test items in convergent and divergent thinking tests do encourage
different approaches.

Assessment
Assessment of 'convergent thinking’ ability has usually been inferred from success on
tests which may be based on ’figural’ input (such as the EFT¡ MFFT) or may be based
on verbal or numerical input (such as the items on intelligence tests). However, in
assessing ’divergent thinking’ the emphasis is not on finding the correct solution (as
there is not one), but on the ability to ’generate’ answers. Examples of such ’open-
ended’ test items again can be non-verbal/figural (such as the ’circles’, the ’squares’
and the barrelled lines tests) or verbal (such as the Uses of Objects test, Consequences
test and the ’S’ test).

Effect
Co gnititie. The educational implications stemming from the convergent—divergent
thinking style research are far reaching. Certain teaching strategies commonly found in
mathematics, science and technology are characterised by logical, structured presenta-
tions, and consequently, encourage mostly convergent thinning and discourage
divergent thinking. In contrast, certain teaching strategies in arts, which provide
students with an area of interest and ask them to generate a project based upon their
study of the area encourage mainly divergent thinking. Furthermore, such biases were
found by Hudson (1966) to exist even among children who had not yet begun to
specialise, thus suggesting the need to look beyond subject specialism. In response to
this, Haddon & Lytton (1968) investigated the effects of differing primary school
teaching strategies on divergent thinking abilities. Their results showed that pupils
from the ‘informal’ schools were significantly better in divergent thinking, compared
with ’formal’ schools.
Further support for the view that certain approaches foster convergent thinking and
 Co gzzitiwe Skies 201

others divergent thinking comes from studies such as Barber-Lunn (1970), Crutchfield
(1965) and Covington et al. (1974).
Hudson also investigated the effects of matched and mismatched teaching strategies.
He identified students at a teaching hospital as either convergent or divergent thinkers.
Then, their teachers were identified as convergers or divergers by Hudson, the
teachers’ students and fellow teachers (there was unanimous agreement). The results
indicated that the convergent students learned best from the convergent teachers,
while the divergent students did best with the divergent teachers.
In general, Hudson has found that convergers prefer formal problems and tasks that
are better structured and demand greater logical ability than the more open-ended
problems favoured by divergers. Convergers apparently are more emotionally inhibited
than divergers, and appear to keep the different aspects of their lives ’compartmental-
ised’ (Austin, 1971). One explanation is that convergers prefer to structure their
experience at all levels, more than divergers do and are more capable of utilising any
structure present.
Downloaded by [University of Chicago] at 11:00 05 July 2012

Social. Overall, evidence seems to suggest that whilst divergent thinking is an

invaluable cognitive quality, socially, it is considered as irritating, disruptive and even
threatening by teachers. Indeed, Getzels & Jackson (1962) found that teachers
preferred learners who were low in divergent thinking (i.e. conformist and orderly) to
those higher in divergent thinking, even though all the learners were of similar
intelligence, and even though the divergent thinkers produced more imaginative and
original responses. Since schools are inherently rule-bound and conservative institu-
tions this comes as no surprise. This inevitably means that much of the divergent
thinking (creativity) is likely to contrast or even con8ict with what is routine, familiar,
expected and ’correct’.

Evidence of the Relationship Beiu›een Labels

Although there is no hard evidence to connect the convergent—divergent thinking style
with other styles (e.g. field dependence—independence, impulsivity-reflectivity,
leveller-sharpener), however, from what has already been said about cognitive styles
(Kogan, 1976) i.e. that, “they employ operational indices that reBect level of accuracy
of performance”, this suggests that they are related. Noppe & Gallagher (1977)
reported a correlation of 0.44 between Mednick & Mednick’s (1967) Remote Associ-
ates Test (a measure of creativity) and the GEFT for college undergraduates.
Field independent subjects have often been regarded as more creative than the field-
dependent subjects. Studies indicate that although the all subjects who score high on
field independence are not necessarily creative, those scoring high on divergent
thinking tend to score higher on field independence (Bloomberg, 1971; 1976; Noppe &
Gallagher, 1977; Spotts & Mackler, 1967). Furthermore, learners studying ‘more
creative’ disciplines have been found to be more field independent than learners
studying ‘less creative’ subjects. For example, Bergum (Bergum, 1977; Morris and
Bergum, 1978) found that students of architecture regard themselves as more creative
than business students and are also more field independent than business students.
 202 fi. Riding ‹Sr I. Gheema

4. 6evelter-Sbarpener
Description
Holzman & Klein (1954), first used the term ’levelling and sharpening’ to describe
another dimension of cognitive style. The distinction between ’levellers’ and ’sharpen-
ers’ is again based on how the visual task is perceived. Holzman and Klein noted that
whilst some individuals resulted in over simplifying their perceptions (’levelling’),
others had a tendency to perceive the task in a complex and differentiated fashion;
showing little assimilation (’sharpening'). Indeed, whilst ’levellers’ tend to assimilate
new events with previously stored ones, ’sharpeners’ in contrast tend to accentuate the
perceived events and treat them more discretely from those already stored. How well
the perceived task was assimilated led to the dimension of assimilation, with ’sharpen-
ers’ at one end, showing very little assimilation to ‘levellers’ at the other end, showing
high levels of assimilation.
Gardner et al. (1960, p. 122) described this dimension as “the characteristic degree
Downloaded by [University of Chicago] at 11:00 05 July 2012

to which current precepts and relevant memory traces interact or assimilate in the
course of registration of the current precepts and memories”.

Assessment
’Levelling and sharpening’ has generally been measured by the Schematising Test.
This entails in the subjects first becoming dark adapted and then judging the sizes of
squares of light of increasing size. Overall, there is a tendency to underestimate the
sizes of squares as new squares of increasing size are added. Subjects who make greater
underestimates are ’levellers’ whilst those that make less errors are ’sharpeners’.

From the few studies that have been carried out examining the effects of leveller-
sharpener cognitive style on cognitive tasks (which invariably have consisted of
estimating sizes of squares or rebelling stories) alarming effects have been observed.
Indeed, it has been noted that this cognitive style has two main forms of manifestation,
either as a tendency to ‘gloss over’ inconsistencies; the condensing of information and
simplifying, on the one hand to the caricaturising on the other.

Evidence of the Relationship BeMeen Labels

Overall, the leveller—sharpener dimension resulted in relatively few studies, (and most
of these were carried out by the authors on adults), this means that out of these,
studies examining the relationship between it and other cognitive styles are few.
However, Riding & Dyer (1983) did a factor analytic study of styles and found that
leveller-sharpener loaded on the same factor as field dependence—independence.

5. Holists-Serialists
Description
Another dimension of cognitive style is that labelled by Pask (1972) as the ‘holist-seri-
alist’. According to Pask, these two ‘competencies’ are clearly demonstrated when
 Co gnitit›e Styles 203

learners are provided the ’free-learning situation’. In such circumstances, learners who
‘showed intention to search for specific data’ are labelled as ’serialists’, whilst those
who ’test a large predicate or relational hypothesis’ are considered to be ’holists’.
Throughout his work, Past observed that whereas serialists have a tendency to
examine less data and use a step-by-step approach to confirm or disconfirm their
hypothesis¡ holists in contrast, prefer to scan large amounts of data, searching for
patterns and relationships. Pask comments that although in the end both groups of
learners show similar understanding, their ‘routes’ to attaining that understanding are
different. Indeed, whereas the serialists tend to put much more emphasis on the
separate topics and the logical sequences connecting them only late in the process, the
holists perceive the learning task in an overall content from the start.

dssessmenr

The holists-serialist dimension was initially identified by Pask & Scott (1972) by
Downloaded by [University of Chicago] at 11:00 05 July 2012

giving the learners a problem-solving task (i.e. the keyboard task) in a 'free-learning
situation’. Overall, this resulted in two broad types of individual competence—holist
and serialist. In order to establish that these indiYidual differences were not just
restricted to this type of task, but indeed would manifest in other simations, Pask &
Scott devised a series of esperiments again requiring a deep level of understanding.
The students in this case had to establish for themselves the principles of classification
underlying the division of two imaginary species of Martian animals—the Clobbits and
the Gandlemullers—into a series of sub-species. Information about Clobbits was
provided in the form of 50 cards. These were placed face down in ten columns (each
column representing a separate sub-species). The five rows contained separate categ-
ories of information about the 10 sub-species (e.g. habitat, physical characteristics,
drawings of animals, etc.). In addition, students could also write their own information
cards if they found it helpful.
Students were instructed to turn over the cards to obtain the information they
required. However, they were to turn the cards over one at a time and to give a reason
for ‘choosing’ that card. Each reason amounted to a hypothesis about the nature of the
classification system which the information on the card was expected to test. A record
was kept of the order in which the cards were used and also of the hypothesis given at
each step. Finally, smdents were to ‘teach-back’ to the experimenter what they had
learned about the Martian animals.
Pask noted that some students concentrated on a step-by-step strategy in which they
used simple hypotheses to assess one property (of the animals) at a time, e.g. ’Do
Gandlemullers have sprongs?’ Whilst others used more complex hypotheses which
combined several properties simultaneously, e.g. ’Are there more kinds of Gandlers
with mounds (dorsal or cranial) than Plongers?’
Students who used the step-by-step approach, indicated a logical linear progression
from one hypothesis to the next. The focus was narrow, with the student concentrating
on each step of the argument, cautiously and critically (serialists). In contrast, students
adopting the holist approach, had a tendency to use global strategies. Holists who in
addition to having broad perspectives, personalise their learning to the extent that they
almost ‘create’ information, have been labelled by Pask as ‘redundant holists’. Past
argued that whilst the redundant holists succeeds in attaining the same level of
understanding as both the holist and the serialist, unlike these, they had to rely
on personal (redundant) elaboration to aid that understanding. Indeed, in certain
 204 R. Riding T• I. Cheema

cases, there was even incorrect information being reinterpreted from the presented
information.
Subsequently, Pask (1976) developed a ’coversational theory’ of learning, which
emphasised the need for the learner to ’teach-back’ all that she/he has learnt. Past
argued that the holist and serialist strategies are manifestations of important underly-
ing stable differences in the way individuals perceive problems. He claimed that
whereas some learners are disposed to act ‘lilte holists’ whenever they are given the
opportunity, others behave ’like serialists’.
In observing students adopting extreme styles, Past noted that there appear to be
two main types of ’pathologies’ from which such learners suffer from. Those that rely
more on serialist-like style and are unable to take a global view and are likely to suffer
from ‘improvidence’ pathology; similarly, those adopting the holist-like style have
a tendency for ’globe-trotting’, that is to mike hasty decisions from insufficient
evidence.
Downloaded by [University of Chicago] at 11:00 05 July 2012

Effect
Pask maintained that whilst the holists and serialists use different learning strategies,
the level of understanding achieved by each is the same. Pask (1976) on assessing
students on standardised tests for various cognitiYe traits (logical reasoning, embedded
figures, analogy completion, perceptual discrimination and a test for divergence),
found that the only differences in the mean score between holist and serialists are on
the analogies test and the divergence test, that is holists score higher than serialists.
Pask et al. researched the effects of matching and mis-matching learning materials
with students’ learning strategies. Programmed learning materials were designed
according to serialist and holist principles, and then allocated to students according to
their learning styles. Students were matched and mis-matched with the materials. The
matched students (serialist teaching programme and holist to holist programme)
performed much better when tested on their knowledge of the programme than did the
mis-matched students. Furthermore, the matched students showed a significantly
greater ability to generalise from their knowledge.

E c›idence of the Relationship Betu›een Labels

In comparison with most of the other cognitive styles already considered, it should be
noted at the outset, that the holist—serialist dimension has two main wearinesses. First,
it is based only on a relatively small sample of students, and secondly, all the students
were above the age of 15 years. Despite this, the outcomes of Past’s research is
suggestive of individual differences that could be very important in the educational
system; consequently, this has led to the widely recognition of the holist-serialist as a
cognitive style (Bernini, 1986¡ Entwistle, 1981). Whilst the style has many educational
implications, it is unfortunate that as far as the authors are aware, as yet, it has not
been empirically correlated with other styles of learning. That is, no empirical evidence
exists 1inlting the holist-serialist dimension with other styles of the wholist-analytic
family.

The Relationships Between the Labels

The above five labels and their measures have been discussed, and from their
 descriptions, correlations, methods of assessment and effect on behaviour, it can be
seen that they are liite1y to be correlates of the same single cognitive style, which may
be termed the wholist-analytic dimension.
Overall, it appears that the labels may be categorised as in Table 1.

TnBLE I.
WhOliStS

Field dependence Field independence

Impulsive Reflective
Levellers Sharpeners
Divergers Convergers
Holists Serialists
Downloaded by [University of Chicago] at 11:00 05 July 2012

Verbaliser-imager Style
Int:roduction
How individuals vary in their modes and methods of thought has been of intriguing
interest and dates back experimentally at least to Galton’s (1883) ’breakfast table’
questionnaire, which inquired into the quality of images elicited by individuals when
visualising certain scenes (e.g. the breakfast they ate that morning).
A number of imagery questionnaires have been developed since then, (see Richard-
son, 1969). Currently used self-report measures of imaginal abilities include the Betts
Inventory (Betts, 1909), the shortened version of Betts scale developed by Sheehan
(1967), the Marks Vividness of Visual Imagery Questionnaire (Marks, 1973), and the
Gordon’s Scale of Imagery Control (Gordon, 1949). All of these scales require self-
ratings of images on dimensions of quality. These consequently have proven difficult
to validate and often fail to show correlations with objective performance on tasks
where good images would be helpful (Ernest, 1977). Furthermore, these scales lack
internal controls for response sets such as acquiescence and social desirability. In order
to overcome these shortcomings, and to broaden the basis for measurement of both
imaginal and verbal traits, various questionnaires have been constructed, e.g. the
Individual Difference Questionnaire (IDQ).
All these measures, are developed from the perspective of dual coding theory
(Paivio, 1971) and, consequently, endeavour to tap verbal abilities, habitats and
preferences as well as tapping imaginal habits, preferences and abilities. A distinctive
feature of the IDQ is the inclusion of both positively and negatively worded versions
of most items in order to control for individual differences in the response set of
acquiescence.
Whilst several measures eaist for assessing imagery, only three main measures exist
for assessing the verbaliser—visualiser cognitive dimension (i.e. Paivio, 1971; Richard-
son, 1977; Riding & Taylor, 1976; Riding & Calvey, 1981; Riding and Bucitle, 1990),
since Kirby et al. (1988) multidimensional verbal—visual learning style questionnaire
still awaits further investigation.
Although it is 1iite1y that most individuals can and do switch strategies according to
the nature of the task, examination of individuals attempting given tasks has shown
that there are some who consistently appear to be heavily dependent upon one or other
 206 R. Riding ‹Sr I. Cheema

of the two main types of coding. Indeed, it is possible to argue that perhaps preference
is governed by ability in that if one finds a particular mode more easily comprehensible
then they are going to resort to that mode regardless of whether it is appropriate for
the set task tErnest and Paivio, 1971).

Differences in Approach
Richardson’s Verbaliser-Visualises Questionnaire (VVQ) is a subset of Paivio’s Indivi-
dual Differences Questionnaire, in fact, Sheehan ci al. (1983) described the VVQ as an
‘abbreviated form of the IDQ’. This inevitably means that they are highly correlated.
Riding utilises a different approach in that the subjects are given an opportunity to
display the preference of their mode of thinking. That is assessment of the learner’s
preference to use a particular habimal mode is dependent on the ratio of the responses.
Downloaded by [University of Chicago] at 11:00 05 July 2012

Indeed, whereas the other two measures are very much introspective questionnaires
that require subjects to say whether they think they can or cannot easily generate
images, Riding’s test of verbal-imagery allows the individual to respond in their
habitual manner of behaving. The three styles will now be brieBy examined.

1. Individual Differences Questionnaire

Description and Assessment
Paivio’s (1971) Individual Difference Questionnaire, is an 86—item Ways of Thinking,
true-false questionnaire which is scored on a single dimension such that subjects with
visual tendencies obtain high scores, those with verbal tendencies low scores.

Effect
Visualisers have been shown to be better than verbalisers in the recall of high imagery
material, and to have more regular breathing patterns (Richardson, 1977), have
produced less uniform eye movement responses (Richardson, 1978) and report more
spontaneous visual eatrasensory experiences (Irwin, 1979).

Style Evidence
Paivio (1986) examined that the dual coding approach incorpoated in the Individual
Difference Questionnaire allows these distinctions to be made between habitual
verbalisers and habitual imagers. Indeed, when Paivio and Harshman (1983) carried
out a detailed factor-analysis based on two separate samples of more than 300 subjects
each, their study showed a highly stable factor structure. First, a two-factor solution
provided a reasonable description of the factor structure, one factor being identified by
items that refer to a preference for imagery and the other by items referring to verbal
thinking. Interestingly, each factor correlated more than 0.9, with the respective total
imagery and verbal scores based on all of the original items, thus validating the
inmitive construction of items based on dual coding theory.
 Go$nitic›e Styles 207

2. Verbaliser-Visualiser Questionnaire
Description and Assessmenr
The Verbaliser—Visualiser Questionnaire (VVQ) measures individual differences on a
verbaliser—visualiser dimension of cognitive style. It is a 15-item true-false test, is a
subset of Paivio’s (1971) Ways of Thinking questionnaire, and is scored on a single
dimension such that subjects with visual tendencies obtain high scores and those with
verbal tendencies, low scores.

Effect
Since its construction, the Visualiser—Verbaliser Questionnaire (like the Individual
Differences Questionnaire), has been used little in instructional research. Indeed, the
majority of the VVQ studies have been concerned with the instrument’s reliability.
Downloaded by [University of Chicago] at 11:00 05 July 2012

For example, Alesandrini (1981) reported that visualisation tendencies were in-
versely, but not highly, related to both science learning in a verbal context
(r = —0.24) and verbal-analytic ability (r —0.12). Furthermore, there has been
little research examining the relationship of the VVQ to mental abilities. Indeed,
Alesandrini’s study showed no relationship between spatial ability (a figure-completion
test) and visual-verbal preference (r = 0.04), and only a weak one (r = —0.12,
P<0.05) with verbal—analytic ability (WAIS similarities).

3. Verbal-imagery Cognitive Style

Description
Riding & Buckle (1990) described imagers to be similar to those which Paivio (1971),
Richardson (1977) and other researchers in the field have called 'visualisers’. That is,
individuals 'who think’ in terms of mental picmres as opposed to ’verbalisers’ who are
considered to ’think’ in terms of words.
Furthermore, Riding like others, e.g. Walter (1953), believes that whilst most
individuals are capable of using either mode of representation, i.e. visual or verbal,
there are some individuals who have a tendency to use just one mode only, either the
visual or the verbal.

Assessment
In order to examine the major mode in which a person is functioning, Riding & Taylor
(1976) initially developed an imagery code test. The test basically required the 7-year-
old subjects to listen to 10 short prose paragraphs, each of which was followed by a
question. The question required answers that were not provided in the paragraphs, but
which could be ’read off’ by the learner if an image had been generated by the
information that had been read. For example, for the statements, ‘Finally, Giacco saw a
little house under a mulberry tree. He knocked at the door’, the question was 'What
colour was the door?’ The response time from the end of the question to the beginning
of the answer was noted. Riding and Taylor made the assumption that children who
formed an image of the information during reception would answer quickly by
supplying the answer to the question from the image, whereas those who did not form
a clear image, or any image, would take longer to respond as they would have to search
 208 2t. Riding T• I. Gheema

their lexical memory for suitable door colours. On their mean response times, Riding
and Taylor divided the children into three groups—fast, moderate and slow respon-
ders. All three groups, next listened to a concrete and an abstract prose passage, each
followed by an immediate test of questioned recall. A significant interaction was found
between response speed and passage type such that fast responders performed best on
concrete passage, whilst the slow responders excelled on the abstract passage, thus
suggesting that fast responders are imagers and slow responders verbalisers. However,
this approach did not directly assess verbal coding, but only inferred it from slow
imagery responses.
Riding & Calvey (1981) extended the Riding and Taylor test by including questions
to assess verbal coding as well as imaginal coding. Furthermore, by obtaining the ratio
of the response times to questions assessing the verbal and imagery performance, it
was considered that an individuals verbal or imaginal coding preference could be
detected. However, whilst the verbal-imagery code test was effective in demonstrating
the existence of the verbal-imagery style (see Riding & Calvey, 1981; Riding &
Downloaded by [University of Chicago] at 11:00 05 July 2012

Ashmore, 1980; Riding & Dyer, 1980), it was difficult to administer and time
consuming to score.
To overcome difficulties in administration Riding et al. (1989) devised a computer
presented test of verbaliser-imager cognitive style. Basically, the test consists of the
computer presentation of pairs of words. The individual has to decide whether the
relationship between the first word presented and the second word belongs to the same
category. The two categories are: (1) same group; (2) same colour. There are 24 pairs
of words in each category, half of which are true, whilst the others are false in each
caregory. It is anticipated that imagers respond more quicitly to the same colour
category since they can readily generate images for the objects in question; whilst
verbalisers will have a shorter response time for the same group category because of
its verbal associations and which would be difficult to visualise. A ratio of response
times gives an indication of an individual’s position on the verbaliser-imager style
continuum.

Go,giiiiiee tasks. With respect to the mode of presentation of information and learning
performance, generally imagers learn best from pictorial presentation, while verbalisers
are superior from text (Riding & Ashmore, 1980; Riding er al., 1989; Riding & Buckle,
1990).
In terms of the type of content, imagers find concrete and readily visualised
information easier than semantically and acoustically complex details, with verbalisers
the reverse (Riding & Taylor, l976¡ Riding & Calvey, 1981; Riding & Dyer, 1980).
This is also rejected in the initial learning to read performance in 7-year-olds where
verbalisers are superior to imagers (Riding & Anstey, 1982).
In summary, it appears that verbalisers prefer and perform best on verbal tasks,
whilst imagers on concrete, descriptive, imaginal ones. When there is a mismatch
between learning style and material or mode of presentation, attainment is usually
relatively poor.

Social (c›erbaliser-imager/extraoersion-introzersion). Eysenck (1967) described intro-

verts as individuals which show greater cortical arousal than their counterparts—the
extraverts. Whilst most of the evidence is not based directly on brain activity, but on
 Gogriifior Styles 209

various behaviours, overall they show that introverts condition more easily, can
concentrate on routine tasks for longer periods and matte better use of their long-term
memory than extraverts. Whilst the parallels between these findings and brain
activities are strong, the explanations are still speculative.
Indeed, of particular interest is the present context, is the interpretation of the
introversion-extraversion scale as being related to the cognitive dimension of imager-
verbaliser. Support for this postulation comes from various smdies, e.g. Costello
(1957), Huckabee (1974), Richardson (1969), and Riding & Dyer (1980).
Costello (1957) described a classification of imagery types which coincides with
Eysenck’s dimensions of introversion—extraversion and neuroticism. Huckabee (1974)
examined this by administering the Eysenck Personality Inventory and a mental
imagery questionnaire, which had 16 concrete and 16 abstract nouns arranged in a
random order. For each noun, subjects had to rate the ease with which it evoked a
’mental picture’. Regarding the introversion-extraversion scale, the scores were related
to both forms of imagery, and high imagery scores are more likely to be obtained by
Downloaded by [University of Chicago] at 11:00 05 July 2012

introverts. Furthermore, this tendency is much stronger for concrete nouns than for
abstract ones.
Riding & Dyer (1980), investigated the relationship between extraversion and the
verbal imagery learning style in 12-year-old children. They found that extraversion
was significantly and negatively correlated with the verbal-imagery code ratio ( —0.67
boys; —0.76 girls). Overall, the results showed a clear relationship between extraver-
sion and verbal-imagery performance implying that verbalisers are extraverts and
imagers are introverts. They suggested that the cognitive verbal-imagery continuum
parallels the social extraversion-introversion dimension.
Studies on voluntary imagery have also found a relationship between the vividness
and ease of imagery control with extraversion. For example, Gale ct al. (1972), and
Morris and Gale (1974) reported significant correlations between extraversion and the
vividness of actively evoked images on the Betts test. Similarly, Euse and Haney
(1975) found positive correlations between extraversion, and image controllability and
clarity.
Riding & Dyer (1980) suggested that in terms of the personality dimension of
introversion—extraversion, introverts experience a large amount of spontaneous imag-
ery when they think about information, but these mental pictures are difficult for the
individual to control, are unstable and likely to be replaced by fresh images as thinking
progresses. In contrast, extraverts experience little spontaneous (involuntary) imagery,
but they are better able voluntarily to generate and control an image in response to a
decision to do so. The situation is said to reverse for verbal associations, with the
extraverts experiencing greater 8uency, but lesser control, whilst introverts experience
less 8uency and more control.

Conclusion
Factor Analysis
Evidence for the relationship between the measures has been examined within the two
broad categories of the wholist—analytic and the verbal imagery style dimensions.
There have been few overall comparisons of these various styles showing how they are
related or unrelated to each other.
Riding & Dyer (1983) examined the relationship between several learning styles and
 210 2t. Riding T' I. Gheema

personality tests. Twelve-year-old children were given the Junior Eysenck Personality
Inventory (jEPI) which gives a measure of introversion—extraversion, and neuroti-
cism-stability, the GEFT, the Schematising Test (leveller—sharpener), the verbal—
imagery Code Test and the MFFT. Using only subjects who had a complete set of
results and a lie score of less than three on the JEPI (75 boys and 75 girls), Riding and
Dyer used a factor analysis which indicated four factors, which were subjectively each
assigned a name and will be described in descending order of size.

Factor I (differentiation). On this Factor, field independence and leveller-sharpener

loaded highly, suggesting field independents to be sharpeners and dependents to be
levellers, and reBectivity loaded moderately.

Factor II (representation). Both extraversion and the verbal-imagery code ratio loaded
on this Factor. The Betts Test of Imagery Control also loaded on this Factor, but less
highly than on Factor IV (neuroticism). This factor provides support for the view
Downloaded by [University of Chicago] at 11:00 05 July 2012

(Riding & Dyer, 1980) that extraYerts tend to be verbalisers and introverts imagers.

Factor III (response eJcien‹yj. Response efficiency loaded highly whilst reflectivity
produced a modest loading.

Factor IV (conool-neuroticism). Neuroticism loaded quite highly on this Factor whilst

Betts Imagery Control and ReBectivity produced modest loadings.

Two Fundamental Cognitive Styles and their Relacionship

Research, then, suggests that there are two basic dimensions of cognitive style: (1) the
wholist-analytic style of whether an individual tends to process information in wholes
or parts; and (2) the verbal imagery style of whether an individual is inclined to
represent information during thinking verbally or in images. These two styles may be
thought of as independent of one another, that is the position of an individual on one
dimension of cognitive style does not affect their position on the other. For instance a
person may be an imager and a wholist, and another an imager and an analytic, or
another may be a verbaliser and a wholist, while someone else may be a verbaliser and
analytic (Fig. 1).

Assessment
A fundamental weakness of most measures is that they assess positively only one end
of a dimension. For exañiple, on the GEFT a high score indicates field independence.
An individual obtaining a low score is assumed to be field dependent, although the
score could also be due to other factors such as low motivation, inability to follow the
instructions or a visual defect. This deficiency may have been responsible for causing
field dependence-independence to be thought to be related to intelligence, and may
certainly explain why field independents usually do best on many types of task.
These two dimensions may be assessed using the Cognitive Styles Analysis (Riding,
1991) which is a computer presented test developed from the Verbal-Imagery Code
Test, with the addition of sub-tests to assess the wholist-analytic dimension. It
positively assesses both ends of the wholist-analytic and verbal—imagery dimensions.
It contains three sub-tests. The first assesses the verbal-imagery dimension by
 Co gnitic e Styles 211

Whol i st - an0l y tic

dimension

who I ist

Verbo I — i maqery
dimension

verba liser imager

Downloaded by [University of Chicago] at 11:00 05 July 2012

anal yti c

Flo. 1.

presenting statements one at a time to be judged true or false. Half of the statements
contained information about conceptual categories while the rest described the appear-
ance of items. Half of the statements of each tfpe were true and the rest false. It was
assumed that imagers would respond more quickly to the appearance statements,
because the objects could be readily repesented as mental pictures and the information
for the comparison could be obtained directly and rapidly from these images. In the
case of the concepmal category items, it was assumed that verbalisers would have a
shorter response time because the semantic conceptual category membership is ver-
bally abstract in nature and cannot be represented in visual form. The computer
records the response time to each statement and calculates the Verbal-imagery Ratio. A
ratio of less than one corresponds to a verbaliser and a ratio of more than one to an
imager. It may be noted that in this approach individuals have to read both the verbal
and the imagery items so that it is not a test of reading ability or of reading speed.
The second two subtests assess the wholist—analytic dimension. The first of these
presents items containing pairs of complex geometrical figures which the subject is
required to judge either the same or different. Since this task involves judgements
about the overall similarity of the two figures, it was assumed that a relatively fast
response to this task would be possible by wholists. The second presents items each
comprising a simple geometrical shape (e.g. a square or a triangle) and a complex
geometrical figure, and the subject is asked to indicate whether or not the simple shape
is contained in the complex one by pressing one of the two marked response keys. This
task requires a degree of disembedding of the simple shape within the complex
geometrical figure in order to establish that it is the same as the stimulus simple
shape displayed, and it was assumed that analytics would be relatively quiciter at this.
The computer records the latency of the responses, and from these calculates the
 212 R. Riding T• I. Gheema

wholist—analytic ratio. A ratio of less than one corresponds to a wholist and a ratio of
more than one to an analytic.
This test differs from the traditional method of assessing field dependence-indepen-
dence (e.g. by using the Embedded Figures Test) in three significant ways. First, it
positively measures the wholist tendency and does not simply assume that if a person
does poorly on a disembedding task that they are field dependent. This overcomes a
major objection to the notion of field independence being a learning style raised by
those who have argued that since generally field independents are superior to field
dependents, it is simply a correlate of intelligence or general ability. Secondly, it
compares a person’s relative performance on the two halves of the continuum. Thirdly,
by using computer presentation, it allows more sensitive timing of the taslt.
The Cognitive Styles Analysis provides a simple, quick and convenient means of
assessing an individual’s position on the two fundamental cognitive style dimensions.
Downloaded by [University of Chicago] at 11:00 05 July 2012

Correspondence: Dr Richard Riding, Assessment Research Unit, School of Education,

University of Birmingham, Birmingham, B15 2TT, England.

REFERENCES
ALfsnxnnixi, K.L. (1981) Pictorial-Verbal and Analytic-Holistic Learning Strategies in Science Learning,
Journal of Educoi:tonal Psychology, 73, pp. 358-368.
ALLPORT, G.W. (1937) Personality: A Psychological Interpretation (New York, Holt & Co).
AUSTix, M. (1971) Dream recall and the bias of intellectual ability, Naiuce, 231, p. 59.
BxHTA , T.J. (1970) Tests lot the evaluation of early childhood education: The Cincinnati Autonomy Test
Battery (CATB), in: J. HELLMUTH (Ed.) Cognirieir Studies, Vol. 1 (New Yorl, Brunner-Mazel).
Bnexrn-La, J.C. (1970) Srreami tg in the Primary School (Slough, NFER).
BnRTLETT, F.C. (1932) Remembering. A Study in Experimental and Social Pychology (Cambridge,
Cambridge University Press).
ERGUM B.O. (1977) Undergraduate self-perceptions of creativity and independence, Perceptual and Moi‹n'
Skills, 44, pp. 187-190.
BERTIMI, M. (1986) Some implications of field dependence for education, in: M. BERTlNi, L. Pizzxuiorio &
S. WAPHER (Eds) Field Dependents in Psychological The‹›ry, Research and Applicai:ion. Tnio Syniposia in
ñfem‹my of M. A. lPitâin (Hillsdale, LEA Publisher).
BEms, G.H. (1909) The Distribution and Functions of Menial Imagefy (1few York, Teachers’ College).
BiGGs, J.B. (1987) Student approaches to teeming and studying (Melbourne, Australian Council for
Educational Research).
BLCH3MBERG, M. (1971) Creativity as related to field independence and mobility, ]oumal of Genetic
Psychology, 118, pp. 3-12.
BRUMBY, M.N. (1982) Consistent differences in cognitive styles shown for qualitative biological problem-
solving, British 'Journal of Educai:tonal P rycholoffy, 52, pp. 244—257.
BELL, S.B. & DOUGLAS, V.I. (1972) Cognitive styles and responses to the threat of frustration,
Canadian Journal of Behavioural Science, 4, pp. 30-42.
Coxx, R. (1974) The Optimal Personality: An Empirical and Theoretical Anafys‘u (New York, Columbia
University Press).
COSTELLo, C.G. (1957) The control of visual imagery in mental disorder, Journal o/ñfetitof Science, 102,
pp. 840—849.
Covixozox, M. CRUTCHFIELD, R., Drives, L. & OLTox, R. (1974) The Productive Thinking Pro am: A
Course in Learning to f7iinA (Columbus, Merrill).
CRUTCHmEl.o, R.S. (1965) Creativity thinking in children: its teaching and testing, in: 0. G. Beni, R. S.
CRUTCHmELn & W. H. How x (Eds), Intelligence Perspechaes 1965• Ttie Temian- Otis Memorial
Lectures (New York, Harcourt, Brace & World).
CURRY, L. (1983) An organization of learning styles theory and constructs. ERIC Document 235 185.
Dns, J.P. (l988a) Implications for School Learning, in: R. R. Scituccx (Ed.) Learning Strategies end
Learning Syles (New Work, Plenum Press).
 Go,gnirior Styles 213

Dns, J.P. (1988b) Simultaneous-Successive Processing and Planning, in: R. Scituccii Learning Strategies
end Learning Styles (New York, Plenum Press).
Do, R., Direr, K. & PRICE G.E. (1975, 1979, 1981, 1985) Leeming Syles Inaent‹ny (Lawrence, Price
Systems).
EHTvISTLE N. (1981) 5;yJes of Learning and Teaching (Chichester, Wiley).
EnxzsT, C.H. (1977) Imagery ability and cognition: A critical ceviexr , ]aurnal of Mentol Imagery, 2,
pp. 181—216.
Ear, C.H. & Pnivio, A. (1971) Imagery and verbal associative latencies as a function of imagery ability,
Canadian Journal o/ Psychology, 25, pp. 83—90.
EUSE F.J. & HArrcY, J.N. (1975) Clarity, controlability and emotional intensity of image: Correlations with
introversion, neuroticism and subjective anxiety, Perceptual and ñfotor Skills, 40, pp. 443-447.
EvsExcz, H.J. (1967) The Biological Bias of Personaliy (Springfield, Thomas).
FoWLER, W. (1980) Cognitive differentation and developmental learning, in: H. Rs£s & L. LiPsrrr,
Advances in child Detielopmeni and Behaz'iour, Vol. 15, pp. 163-206 (New York, Academic Press).
Gni.E, A., Monnis, P.E., LuCns, B. & RiCunRDsON, A. (1972) Types of imagery and imagery types: An EEG
study, Brit‘uh Journal of Pychology, 4, pp. 523-531.
GnLTON, F. (1883) Inquii-res inio Human F'aculy and in D evelopment (London, MacMillan & Co).
Downloaded by [University of Chicago] at 11:00 05 July 2012

GnnDNER, R.W., House, P.S., REIN, G.S., Li›rrox, H.B. & SPENCE, D.P. (1959) Cognitive control: a
study of individual consistencies in cognitive behaviour, Ppchologicol Issues, 1, Monograph 4.
GxRDNER R.W., JnCxsox, D.J. & Mzsslcx, S.J. (1960) Field articulation, constricted-flexible control, and
the abilities of flexibility of closure and spatial relations and orientation, Psychological Hesses, 2(4), No. 8,
Part 3.
GETZELS, J.W. & Jncxsox, P.W. (1962) Creativity and Intelligence (New York, Wiley).
Gonnox, R. (1949) An investigation into some of the factors that favour the formation of stereotyped
images, British ]oumal of Psycholoffy, 39, pp. 156—167.
Gonnox, R. & GROSS R. (1978) An exploration of the interconnecting perspective of teaching style and
teacher education, GtirricuJxrri Studies, 10, pp. 151-157.
Hannox, F.A. & LYrrox, H. (1968) Teaching approach and the development of divergent thinking abilities
in primary school, Brit‘ah Journal of Educational Psychology, 38, pp. 171—180.
Hommx, P.S. & KLEIN, G.S. (1954) Cognitivie system-principles of leveling and sharpening: Individual
differences in visual time-error assimilation effects, Jouritol o/ Psychology, 37, pp. 105-122.
HucxnBEE, M.W. (1974) Introversion-extraversion and imagery, Psychological Repons, 34, pp. 453—454.
Hunsox, L. (1966) €otirrat¡y Imaginations (Harmondsworth, Penguin).
IRWiN, H.J. (1979) Coding preferences and the form of spontaneous, extrasensory experiences, Journal of
Porapsychology, 43, pp. 205-220.
Knonx, J. (1965) Impulsive and reflective children: Significance of conceptual tempo, in: ). D. KniJniBocTz
(Ed.) Learning and the Educational Process (Chicago, Rand McNally).
Knoxx, J. (1965) Individual differences in the resolution of response uncertainty, Journal o/ Personally
and Social Psychology, 2, pp. 154-160.
KAGAN, J. & KocxN, N. (1970) Individual variation in cognitive processes, in: P. H. MUSSEL (Ed.)
Carmichael’s Manual of Child P chology, 3rd edn, Vol. 1 (New York, Wiley).
KAGAN, J., PEnRSON, J. & WzLcii, L. (1966) Modifiability of an impulsive tempo, ]oumal o/ Educational
Pychology, 57, pp. 357-365.
KAGnx, J., Rosmx, B., Dxv, D., ALBERT, J. & PHILLIPS W. ( 1964) Information processing and the child:
Significance of analytic and reflective attitudes. P¡ychologicol Afonograpfis, 78 (No. 1, whole number 578).
KEOGH, B.K. & DOHLON, G. (1972) Field dependence, impulsivity and learning disabilities, ]ounial of
Learning Disabilities, 5, pp. 331—336.
Kinky, J.R., Moonc, P.). & SCHOmELD N.J. (1988) Verbal and Visual Learning Styles, Contemporary
Educational P chology, 13, pp. 169-184.
KoGAx, N. (1976) Go itice Styles in Infant:y and Early Childhood (Hillsdale, N.J., Ekbaum).
KOsAx, N. (1980) A style of life, a life of style, Goaiempore›;y P rycholoffy, 25, pp. 595-598.
Kota, D.A. (1976) The Learning Style Int eniory: Technital Manual (Boston, McBer & Company).
Kota, D.A. (1977) Leeming Syles Inventory: A self description of preferred learning modes (Boston, McBer
6t Company).
LEwis, B.N. (1976) Avoidance of aptitude-treatment trivialities, in: S. Marc t (Ed.) Individuality in
Leartiitig (San Francisco, Jossey-Bass).
MnRxS, D.F. (1973) Visual imagery differences in the recall of pictures, British ]ournof of Psychology, 64,
pp. 17—24.
Downloaded by [University of Chicago] at 11:00 05 July 2012
 Cogziirioe Biles 215

styles as an aid to individualised computer-based training, Educational and Training Technology

International, 26, pp. 393-398.
Scimrirce, M. dt DOUGLAS V.I. (1973) Moral judgements, behaviour and cognitive style in young children,
Canadian Journal of Behavioural Science, 5, pp. 133-144.
SCHMECR, R.R., RiBlcH, F.D. & RnunHM, N. (1977) Development of a self-report inventory for assessing
individual differences in learning processes, Applied Pfychological 3feastirernent, 1, pp. 413—431.
SHEEHnN, P.W. (1967) A shortened form of Bett’s questionnaire upon mental imagery, Jouwiaf of Clinical
Psychology, 23, pp. 386-389.
SHEEHAN, P.W., flSHTOx, R. & WHITE, K. ( 1983) Assessment of mental imagery, in: A. A. SHEiRH (Ed.)
Imagery—Current Theory, Research and Application Chichester, Wiley).
SPOTTS, J.V. & MACRLER B. (1967) Relationships of field-dependent and field-independent cognitive test
performance, Perceptual and Motor Skills, 24, pp. 239-268. Monograph Supplement, pp. 2-24.
TnMiR, P. & COHEN, S. (1980) Factors that correlate with cognitive preferences of medical school teachers,
Journal of Educational Research, 74 (2).
WALTER W.G. (1953) 77te Lieing Brain (London, Duckworth).
MERNER H. (1948) Gomparatioe Pfychology of Mental Dn›elopment (Chicago, Collett).
WIDIGER T., KxuosOx, R. & RORER, L. (1980) Convergent and discriminant validity of measures of
Downloaded by [University of Chicago] at 11:00 05 July 2012

cognitive styles and abilities, ]ournat of Personally and Social Pychology, 39, pp. 116-129.
Wrriux, H.A. (1962) Pychologicat Differentotion; Studies of Deoelopmeni (New York, Wiley).
WrfKlx, H.A. (1978) Cognitive Styles in Personal and Cultural Adaptation, ?feiitz Fuzzier Lecture Series,
Vol. 11.
WrrxiN, H.A. & GooDENOUGH, D.R. (1981) Gognitioe byte: Essence and Origins (New York, International
Universities Press).
Wrmix, H.A., Orrwx, P.K., RnSRix, E. & KnRP, S.A. (1971) A Manual f‹rr the Embedded F igures Ttst
(Palo Alto, Consulting Psychologists Press).
WiTRIN H.A., Moons, C.A., GooDENOUGH, D.R. & Cox, P.W. (1977) Field-dependent and Field-
independent cognitive styles and their educational implications, Reoino o/ £:ducotionoI 2teseerch, 47,
pp. 1—64.
ZELNIRER, T. & Jxrrncv, W.E. (1979) Attention and Cognitive Style in Children, in: G. A. He.E &
M. LEWIS (Eds) Aneniion ond Gognitive Development, pp. 275-296 (New York, Plenum).