See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/11380532

The Prevalence of Dyslexia Among Art Students

Article  in  Dyslexia · January 2002

DOI: 10.1002/dys.211 · Source: PubMed

CITATIONS READS

99 3,225

2 authors, including:

Ulrika Wolff
University of Gothenburg
14 PUBLICATIONS   327 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

School for Quantitative Research Methods in Education https://qrm.gu.se/english View project

All content following this page was uploaded by Ulrika Wolff on 04 December 2017.

The user has requested enhancement of the downloaded file.

& The Prevalence of Dyslexia
Among Art Students
Ulrika Wolff and Ingvar Lundberg*
Department of Psychology, G.oteborg University, Box 500, SE-405 30 G.oteborg, Sweden

It is widely held opinion that dyslexia is associated with remarkably

artistic creativity. Speculations on different brain structures and
brain functions have been proposed as an explanation. Very few
objective studies have been reported that confirm the conjectures on
the relationship between dyslexia and artistic creativity. Two studies
are reported on the prevalence of dyslexia among university
students}one group of art students and one group of students
from non-art disciplines. The admission to the art schools were
extremely demanding, possibly implying that the students were
genuinely talented, and that their choice of training did not reflect a
compensation for failure in conventional academic fields. Art
academy students reported significantly more signs of dyslexia than
non-art university students. Objective testing showed that art
students had significantly poorer phonological skills than non-art
students. Thus, according to self-reports combined with objective
testing, the incidence of dyslexia was far higher among art
students. Copyright # 2002 John Wiley & Sons, Ltd.
Keywords: dyslexia; art students; self-report; word recognition

INTRODUCTION

T
here is a common belief that dyslexia is often connected with creativity or
artistic talents. The basis for this notion is primarily anecdotic and
informal observations. Case studies of unusually talented people with
dyslexia have also been reported (West, 1997). However, there is a surprising
scarcity of more empirical studies. (For a review see Winner et al., 2001.) In
explaining the nature of the hypothetical association between dyslexia and
creativity several possibilities can be conceived.
First, the association is assumed to be genuine and based on a specific
neurological wiring resulting in both original information processing and
dyslexic difficulties. Geschwind and Galaburda (1987) speculated along these
lines referring to the size and symmetry of the planum temporale among

*Correspondence to: Department of Psychology, Go. teborg University, Box 500, SE-405 30
Go. teborg , Sweden.

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
DOI: 10.1002/dys.211
Dyslexia among art students 35

dyslexics. The evolutionary resistance of the dyslexic genes might then be

explained by the compensating extraordinary talents in other fields.
Secondly, there might be a more general ‘co-morbidity’ where an unknown
general factor has caused both dyslexia and creativity without any direct causal
connection between the two conditions.
Thirdly, the association between dyslexia and creativity might reflect
compensation for early failure in highly valued skills in school. Thus, dyslexic
children tend to look for opportunities to succeed in other areas.
Fourthly, the problems with reading and writing and the resulting frustration
may activate original and unconventional coping strategies and modes of thinking.
Fifthly, the association might be an illusion based on the conspicuous
discrepancy between reading achievement and artistic talent. Such talents might
be equally distributed among non-dyslexics and dyslexics, but they are just more
visible among dyslexics. Non-dyslexics have their doors wide open to any field
whereas talented dyslexics are restricted to non-verbal domains, such as art,
design and crafts.
As already mentioned, systematic empirical studies are few in this area.
Everatt et al. (1999) compared a group of dyslexic and non-dyslexic adults and
children concerning creativity. They found that dyslexic adults showed signs of
greater creativity and more innovative styles of thinking in tasks requiring
novelty as compared to the non-dyslexic adults. As to visual-spatial skills no
clear difference was observed. Neither were any differences in these respects
observed in the younger groups.
Winner et al. (2000) assessed dyslexic and non-dyslexic young adults with a
number of perceptual and spatial tasks. No significant differences in the expected
direction were observed. On some tasks the non-dyslexic subjects rather
outperformed the dyslexics. They actually obtained one remarkable difference
to the advantage of the dyslexics. The dyslexic subjects detected impossible
figures with a higher speed than non-dyslexic subjects. Winner et al. (2000)
suggested that there might be a subtype of dyslexia not easily detected in a
heterogeneous group that may possess special talents.
The examination of the relationship between dyslexia and creativity in the
cited studies has taken its departure from comparisons between dyslexic and
non-dyslexic groups. However, the comparability between the groups is a
problematic issue. The groups may differ in other respects than being dyslexic vs
non-dyslexic. Therefore, the matching procedure must be done with great care.
Furthermore, the tasks measuring the creativity and artistic talents have often
unknown validity.
In the studies to be presented here, we examined the prevalence of dyslexia in
a population of individuals assumed to be highly artistically talented, namely art
academy students. The strict admission policy of the prestigious art schools at the
university level should be a sufficient guarantee of a generally high level of
artistic talents. These talents are assumed to be genuine or even constitutional in
nature. Thus, the art students’ selection of study program probably does not
reflect an attempt to escape the literacy demands in more traditional academic
fields but rather a genuine choice based on a very early discovery of
extraordinary talents.
The prevalence of dyslexia among the art students was compared with
the corresponding prevalence among a matched group of students

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
36 I. Lundberg and U. Wolff

following more traditional university programs. The approach to examine a

population with known artistic talents has, to our knowledge, not been reported
previously.
A critical issue in this context is the determination of dyslexia. The present
paper takes its departure from a conception of dyslexia as a basic phonological
deficit with word decoding problems as its major manifestation. Thus, the
current consensus definition proposed by the International Dyslexia Association
with the de-emphasis of the discrepancy criterion has provided the guidelines for
our approach.
The screening for dyslexia in our studies was based on self-reported signs of
dyslexia, word recognition, and (in Study 2) phonological skills. Self-reports of
dyslexia among adults have proven to be surprisingly reliable and valid
procedures (Lefly and Pennington, 2000). Our questionnaire is in some respects
similar to the revised version of Finnucci’s Reading History Questionnaire
examined by Lefly and Pennington (2000). However, our factor analysis indicated
two subscales, one on dyslexia and one on reading interests.

STUDY 1

Method
Participants: Participants were 74 students (34 females, 40 males) from two art
schools at Go. teborg university, Sweden (Department of fine arts Valand and
Department of photography) and 80 students (43 females, 37 males) from the
School of economics and commercial law at the same university.. The mean age of
the art students was 27.5 years and of the non-art students 24.5 years. The age
difference between the groups was significant t(145) = 3.19; p50.01. All invited
students at the participating schools agreed to join the study.
Instruments: The screening for dyslexia was based on a self-report and a word
recognition test. Reading habits or exposure to print were assessed by an author
recognition task. The instruments were administrated in group settings in the
following order: word recognition, author recognition and self-report.
The questionnaire included 14 items covering practical problems with reading in
daily life, practical problems with writing, motivation for literacy activities and
items concerned directly with the history of reading and writing difficulties. Two
items were only indirectly related to literacy but covered typical symptoms such
as problems with the tables and spoken words of high phonological complexity.
Each item consisted of a 4-point scale on which the degree of a problem was
rated. The direction of the scale was reversed and randomized across the 14
items. A low total score on the 14 items indicated that the respondent had
problems with literacy. Six of the items were regarded as typical dyslexia
indicators also including non-reading aspects. A low sum score on these six items
was a strong dyslexia index.
The word recognition test consisted of chains of words written together without
inter-word spaces. Each chain was built up by three common words. The task
was simply to separate the words with pencil marks. A large number of chains
were available (40 on each page). The performance was expressed as the number
of correctly marked chains within a period of 3 min. Successful performance

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
Dyslexia among art students 37

requires a high level of automaticity and orthographic processing. This test has
been widely used as a valid and reliable procedure for assessing word
recognition skill in Sweden (Jacobson, 1993). An English version is now also
available (Milller Guron, 1999).
The author recognition test was modelled after West et al. (1993) and was used to
obtain an indicator of reading habits or print exposure. It consisted of a list of 40
names of which 20 were names of reasonably well-known Swedish and
international authors and 20 were foil names (non-authors). The score was
expressed as the number of correctly marked names minus the number of
incorrectly marked names. This indirect technique is assumed to circumvent the
compliance normally accompanied conventional questionnaires on reading habits.

RESULTS

First we report results separately for each instrument where art and non-art
students are compared. Then, our estimate of the prevalence dyslexia in the two
groups is reported.
Self-reports: The mean total score on the questionnaire for the art students was
43.8 (maximum 56 points). The standard deviation was 7.8. The students of
economics had a mean of 48.0 (S.D. = 3.9). The difference was significant
t(1 5 2) = 4.24; p50.001). The variation of scores was obviously higher among art
students, to some extent possibly due to ceiling effect.
The mean score on the six dyslexia items was 18.4 (S.D. = 4.6) for the art
students and 20.8 (S.D. = 2.5) for the students of economics (t(152) = 5.57;
p50.001). Also on this subset of items the variation was higher among art
students.
Word recognition: The mean number of correctly marked chains on the
wordchains test was 71.3 (S.D. = 12.2) for the art students and 81.3 (S.D. = 9.4)
for the students of economics. Also this difference was highly significant
(t(152)=5.68; p50.001).
Author recognition test: The mean score for the art students was 14.5 (S.D. = 4.1)
and for the students of economics 11.3 (S.D. = 3.7). This time the art students
significantly outperformed the students of economics (t(152) = 3.00; p50.01).
Thus, despite the poorer word recognition skill and more reported dyslexia signs
the art students seemed to show a wider readership probably reflecting a
stronger cultural orientation.
The incidence of dyslexia: The primary purpose of Study 1 was to determine the
incidence of dyslexia in the two groups of students. The screening criteria for
dyslexia were one standard deviation below the mean for the self-reported
dyslexia indices on the questionnaire (16 points or below) and 1 S.D. below the
mean on the wordchains test (65 points or below). Since the questionnaire was
designed such as the expected result for an individual without dyslexic problems
would approach the maximum score on the six items, a sum score of 16 points
would indicate some significant signs of dyslexia. Among the 74 art students, 11
individuals (3 females, 8 males) or 15% met these criteria; among the 80 students
of economics only one individual (female), 51.3% could be regarded as dyslexic
according to our criteria (z = 3.14; p501). Even if we put the cut-off point at 1.5
S.D. below the mean the incidence of dyslexia would be far higher among art

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
38 I. Lundberg and U. Wolff

students; 7% vs 0%. Thus, our expectation of a higher incidence of dyslexia

among artistically gifted individuals was empirically confirmed. The gender
difference often observed was also confirmed in this study.

STUDY 2

Study 2 was designed with the aim of replicating Study 1 and to modify and
extend the promising questionnaire for self-reports of dyslexic problems. The
phonological nature of dyslexia was also more explicitly recognized in Study 2
with a special instrument for assessing phonological skill as part of the screening
criteria. We also wanted a larger number and a broader spectrum of students
from various university programs.

Method
Participants: A total of 194 students of art (148 females, 46 males) participated
representing the following art programs at Go. teborg university: Fine arts (24),
photography (14), design and crafts (1 1 5), and architecture (41). The expanded
range of art programs in Study 2 included design and crafts with a clear gender
bias in favour of females. This explains most of the gender imbalance in the total
sample of art students. A total of 202 non-art students (98 females, 104 males)
participated. The following programs were represented: political science (48),
economics (43), civil engineering (62), and psychology (49). The average age was
about 26 years in both groups (no significant difference). Of all 387 students
available only one refused participation.

Instruments
The questionnaire now included 29 items. The additional items served the purpose
of increasing the power of the self-report instrument. A principal component
analysis indicated that two factors could well explain the total variance. The first
factor was based on 18 items and the second factor was based on 6 items. The first
factor was interpreted as a dyslexia-indicator factor, and the second factor was
regarded as an interest or motivation factor. Typical dyslexia items concerned
reading and writing skills as well as more basic verbal and information
processing skills such as difficulties with remembering new words, tables, pin
codes, names, alphabetic order, and foreign language learning. In addition some
of the dyslexia items were concerned with earlier diagnosis of reading disability
or dyslexia, remedial help in school, and family incidence of dyslexia. The
interest factor was indicated by items concerning reading habits and preferences,
library use, and the enjoyment of reading.
The validity of the dyslexia subscale of the questionnaire was assessed by
relating the score on this scale to a composite measure of phonological skills
(latency of non-word reading, spoonerism, reversed spoonerism, non-word
spelling) individually assessed with a subset of subjects participating in a third
study (N = 78) which will be reported in a separate publication. The correlation
was as high as 0.77 which is probably almost as high as the reliability permits.

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
Dyslexia among art students 39

Table 1. Mean and standard deviation of the assessments1

Assessment All Art Non-art

Self-report of dyslexia 0 (1.0) 0.25 (1.08) 0.24 (0.86) t(394)=5.05; p50.001

signs (factor scores)
Self-report of reading 0 (1.0) 0.13 (0.90) 0.12 (1.07) t(394)=2.51; p50.01
interest (factor scores)
Wordchains 80 (12.6) 79 (13.0) 81 (12.0) t(394)=0.94; n.s.
Phonological choice 33 (8.9) 32 (8.8) 34 (8.7) t(394)=2.80; p50.05
1
Standard deviations are presented within parenthesis.

This correlation is based on a combined sample of dyslexic and non-dyslexic

students. With only the dyslexic group included (N = 38) a strong restriction-of-
range effect attenuates the correlation to 0.40; p50.01.
Word recognition: The same test was used as in Study 1 (wordchains).
Phonological choice: This task was a modification and a Swedish adaptation of
the silent phonological choice task designed by Olson et al. (1994) and consisted
of a large number of triplets of non-words. In each triplet, one of the words
would sound like a real word if pronounced (gan–pab–kat). In order to mark the
correct alternative the subject must be able to use a phonological strategy. The
total working time was 2 min and the performance was expressed in the number
of correctly marked words within this working period.

Results
Table 1 presents a comparison of the two main student categories (art vs non-art
students) on the assessments.
On the subscale of the self-report indicating dyslexic signs there was a clear
difference between the student categories. The art students had a significantly
lower mean demonstrating more dyslexia signs in this group. In the non-art
group there was a significant gender difference (t(200) = 1.98; p50.05) in favour of
males, whereas there was no significant gender difference among art students.
On the subscale concerning reading interest the pattern of differences between
art and non-art students was the opposite. Despite higher dyslexia load the art
students showed a stronger preference for literacy activities reflecting a stronger
cultural orientation. There was still no significant gender difference among art
students. However, females showed significantly stronger reading interest than
males among the non-art students (t(200) = 5.90; p50.001).
As for actual reading performance indicated by the wordchains test there was
no significant difference between the art and non-art students. A gender
difference was only found among art students. Females significantly out-
performed males (t(192) = 2.47; p50.05). The performance on the phonological
test was significantly lower for the art students. None of the groups showed a
significant gender difference.
The results were broken down on the various study programs in which the
student of both main categories were involved. For each of the assessments a one
way ANOVA was performed with study program as the independent variable. A

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
40 I. Lundberg and U. Wolff

Table 2. Estimates of the incidence of dyslexia according to different criteria

Art Non-art

Self-report (dyslexia signs) 44 (22.7%) 17 (8.4%) z=3.97; p50.001

1 S.D. below the mean

Self-report (dyslexia signs) 25 (13%) 10 (5%) z=2.86; p50.01

1.5 S.D. below the mean

Self-report and phonological choice 17 (8.8%) 5 (2.5%) z=2.75; p50.01

1 S.D. below the mean

Self-report, phonological choice and wordchains 9 (4%) 3 (1.5%)

1 S.D. below the mean

majority of the comparisons did not demonstrate differences across the main
categories (art–non-art). However, a few exceptions to this pattern were
observed. Post hoc analyses indicated that students of architecture had
significantly fewer self reported dyslexic signs than students of crafts. Further
more the students of civil engineering showed lower reading interest than
students of political science and psychology. These minor exceptions from the
major pattern justified the collapsing of data across study programs within the
main categories.

The incidence of dyslexia

The first straightforward and simple criterion for identifying the cases of
dyslexia was completely based on the self report dyslexia factor. The cut-off
limit was set to 1 S.D. below the mean. Table 2 shows that 44 art students and
17 non-art students met this criterion. This estimate of the prevalence of
dyslexia is probably too high for a population of university students. With
a cut-off of 1.5 S.D. below the mean a more reasonable estimate was obtained
(13 vs 4%).
The next step was to supplement the self report data with the objective
assessment of phonological skill (the phonological choice task). Thus, the
criterion specified results below 1 S.D. the mean of the self report factor scores
and 1 S.D. below the mean on the phonological choice task. With this criterion, as
indicated in Table 2 the number of dyslexics in the art category was 17, whereas
only 5 non-art students met this stricter criteria. The third step set up dyslexia
criterion requiring scores of 1 S.D. below the mean on three assessments (self
report, phonological choice, wordchains). Very few students met this criterion (9
art students and 2 non-art students).
In summary, Table 2 demonstrates that the incidence of dyslexia is far higher
among art students regardless of the strictness of the criterion. The observed
pattern is the same when even stricter criteria are applied by taking cut-off points
of 1.5 S.D., although the proportion of cases will be lower. There were no
significant gender differences within or between the groups.

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
Dyslexia among art students 41

DISCUSSION

The two studies have demonstrated that dyslexia signs are more frequently
occurring among art students as compared to students enrolled in non-art
programs at the university. In the second study, however, the gap between the
two student groups was smaller in this respect. One reason for the difference
between the studies might be related to the fact that a broader spectrum of study
programs was included in the second study. For example, in the art group of the
second study, students of architecture, design and crafts also participated. Some
of these students might, in fact, not have been exceptionally talented in arts.
Among the non-art students we included students taking less demanding
courses with a less stringent admission program (for example some of the
psychology courses). The instruments used in the second study were also
different with a more extended questionnaire and an assessment more directly
focussed on phonological skills. Thus, in contrast to Study 1 the selection criteria
also involved phonological skill in Study 2.
A current consensus definition of dyslexia emphasizes the phonological
weakness as a marker sign of the disability (Hoien and Lundberg, 2000). By
disregarding word recognition the possibility of compensation among older
students at the university level was acknowledged. The dyslexia criteria applied
in the present studies were primarily based on self-reported signs of dyslexia.
This procedure was justified by reference to earlier studies on the validity of self-
reports among adults (Lefly and Pennington, 2000) and our own validation of the
questionnaire. By supplementing the self-report data with objective measures of
phonological skills and word recognition we might have come close to the
current consensus view of dyslexia. The cut-off limits set-up were of course
statistically arbitrary; as the criteria were made stricter the frequency of dyslexia
decreased. However, the relative difference in prevalence between the two
groups remained.
The higher rate of dyslexia observed among art students cannot be interpreted
in causal terms. The study is correlational in nature and thus gives no basis for
causal conclusions. However, we have good reasons to assume that the students’
selection of art studies did not reflect avoidance behaviour or an attempt to
escape the literacy demands in more traditional academic fields. It was rather
interpreted as a genuine choice based on a very early discovery of extraordinary
talents. In this sense, then, the relationship between artistic talent and dyslexia
might be a real relationship. This possibility will be further enlightened in a
forthcoming study including careful interviews and detailed individual testing of
the dyslexic cases.
A deeper understanding of the nature of the connection between dyslexia and
creativity might be obtained by more neuro-biologically oriented studies. Strong
visual skills and creative talents might represent a compensation contributing to
the evolutionary resistance of the dyslexic genes. The neuro-biological correlates
observed in dyslexia are primarily an unusual symmetry of planum temporale
(Larsen et al., 1990) with numerous cortical ectopias (Galaburda et al., 1989).
Dyslexic individuals confronted with phonological tasks also show different
activity patterns in their brains as compared to normal individuals (Pugh et al.,
2000). The differences observed might be indications of a different distribution of
talents.

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)
 42 I. Lundberg and U. Wolff

These speculations are given as an alternative to the more simplistic

interpretation of the relationship between creativity and dyslexia where the
early failure in school forces some dyslexic children to look for other fields in
which they might find opportunities for compensatory success. Although this
might be true in some cases, the extraordinary artistic talents of many dyslexics
have been revealed even before they started to learn how to read and write.

References
Everatt, J., Steffert, B. and Smythe, I. (1999) An eye for the unusual: creative thinking in
dyslexics. Dyslexia, 5, 28–46.
Galaburda, A., Rosen, G. and Sherman, G. (1989) The neural origin of developmental
dyslexia: Implications for medicine, neurology and cognition. In A. Galaburda (Ed.), From
Reading to Neurons, pp. 377–388. MIT Press: Cambridge, MA.
Geschwind, N. and Galaburda, A.M. (1987) Cerebral Lateralization: Biological Mechanisms,
Associations, and Pathology MIT Press: Cambridge, MA.
Hoien, T. and Lundberg, I. (2000) Dyslexia: FromTheory to Intervention. Kluwer Academic
Publisher: Dordrecht NL.
Jacobson, C. (1993) Manual till Ordkedjor. Psykologifo. rlaget: Stockholm.
Larsen, J.P., Hoien, T., Lundberg, I. and Odegaard (1990) MRI evaluation of the size and
symmetry of the planum temporale in adolescents with developmental dyslexia. Brain and
Language, 39 289–301.
Lefly, D.L. and Pennington, B.F. (2000) Reliability and validity of the adult reading history
questionnaire. Journal of Learning Disabilities, 33, 286–296.
Miller Guron, L. (1999) Wordchains. A Word Reading Test for all Ages. NFER-Nerlson:
Windsor.
Olson, R., Forsberg, H., Wise, B. and Rack, J. (1994) Measurement of word recognition,
orthographic, and phonological skills. In G.R. Lyon (Ed.), Frames of References for the
Assessment of Learning Disabilities. New Views on Measurement Issues, pp. 243–277. Paul H.
Brookes: Baltimore, MD.
Pugh, K.R., Mencl, E.W., Shaywitz, B.A., et al. (2000) The angular gyrus in developmental
dyslexia: Task-specific differences in functional connectivity within posterior cortex.
Psychological Science, 11, 51–56.
West, T.G. (1997) In the Mind’s Eye: Visual Thinkers, Gifted People with Dyslexia and other
Learning Difficulties, Computer Images, and the Ironies of Creativity. Prometeus: Amherst, NY.
West, R.F., Stanovich, K.E. and Mitchell, H.R. (1993) Reading in the real world and its
correlates. Reading Research Quarterly, 28, 34–50.
Winner, E., von K!arolyi, C. and Malinsky, D. (2000) Dyslexia and visual-spatial talent: no
clear link. Perspectives The International Dyslexia Association, 26, 26–30.
Winner, E., von Karolyi, C., Malinsky, D., French, L., Seliger, C., Ross, E. and Weber, C.
(2001) Dyslexia and visual-spatial talents: compensation vs deficit model. Brain and
Language, 76, 81–110.

Copyright # 2002 John Wiley & Sons, Ltd. DYSLEXIA 8: 34–42 (2002)

View publication stats