Neuropsychologia 40 (2002) 2144–2155

Executive functioning in adults and children with developmental dyslexia

Mark Brosnan∗ , James Demetre, Stephen Hamill, Kate Robson, Haidee Shepherd, Gerard Cody
Department of Psychology, School of Social Sciences and Law, University of Greenwich, London SE9 2UG, UK
Received 28 March 2001; received in revised form 18 March 2002; accepted 20 March 2002

Abstract
The performance of developmentally dyslexic children and adults was studied upon a range of tasks that involved executive functioning.
Both adult and child samples of dyslexics were found to under-perform on the group-embedded figures test. This test required the
identification of constituent parts from within complex visual arrays, with good performance necessitating the inhibition of the processing
of the surrounding context. A general deficit on visual–spatial tasks was eliminated as an explanation as dyslexics performed normally
upon a range of other non-verbal assessments. The dyslexics consistently demonstrated a deficit in digit span tasks, a decrement that was
increased with distractors, again suggesting difficulties in inhibiting the processing of the surrounding context. A deficit was also identified
upon a verbal fluency task without a deficit in vocabulary level. Additionally, a specific deficit in the recollection of the temporal order
of the presentation of items was in evidence, without a deficit in the recognition of the items themselves. The findings taken as a whole
suggest that dyslexic individuals show deficiencies in executive functions relating to inhibition of distractors and to sequencing of events,
a set of tasks associated with left prefrontal cortex functioning in the acquired neuropsychology literature.
© 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Group-embedded figures test; Inhibition of distractors; Left prefrontal cortex

Developmental dyslexia (DD) is defined by the World
Federation of Neurology [85] as a disorder manifested by
difficulty in learning to read despite conventional instruc-
tion, adequate intelligence and sociocultural opportunity.
This definition is consistent with the DSM-IV classifica-
tion of specific reading difficulty, for which prevalence is
estimated to be around 5–10% in the school-aged popu-
lation [1], although some estimates are as high as 20%
of schoolchildren [72]. The earliest theorising of DD has
implicated possible abnormalities in cerebral lateraliation,
specifically a failure to acquire left hemisphere speciali-
sation for language [57]. Evidence that hemispheric spe-
cialisation may play a role is suggested by the relatively
high incidence of left-handed or ambidextrous individuals
within the dyslexic population [51]. Geschwind and Behan
[31] hypothesised that left-handedness and DD are related
to a common factor affecting the development of the left
hemisphere in utero, which results in right hemisphere
dominance. Neuroimaging studies consistently reveal that
the dyslexic brain is distinguished from the non-dyslexic
brain in the degree to which language areas in the brain are
asymmetric (see [35]; Table 20.1, p. 624–5 for a summary,
∗ Corresponding author. Tel.: +44-20-8331-9622;
fax: +44-20-8331-8905.
E-mail address: m.j.brosnan@greenwich.ac.uk (M. Brosnan).
see [61] for a meta-analysis). In the most comprehensive
MRI study of brain anatomy in DD, Eliez et al. [17] report
a significant decrease in grey matter on the left side and a
(possibly consequent) non-significant greater-than-normal
rightward symmetry of grey matter.
A rightward symmetry is significant as the right hemi-
sphere primarily subserves the processing of the relatively
global level, whilst the left hemisphere primarily subserves
the processing of the relatively local level [37]. With respect
to reading, initial reading processes focus on the (relatively
global level) word, or ‘logographic’ [25] level (primarily un-
dertaken by the right hemisphere) with a developmental shift
to a focus on the processing of (relatively local level) letters
(and the subsequent formation of grapheme–phoneme corre-
spondences subserved mainly by the left hemisphere) occur-
ring at around 7 years of age. This refocusing of processing
onto local level elements requires the inhibition of global
level processing [9]. A major theory of DD [3–6] suggests
that the abnormal cerebral symmetry underpins a failure to
shift from the word level to the letter level of processing, ar-
resting reading development at the (relatively more global)
logographic level [25]. Unlike normal readers, dyslexic in-
dividuals do not show increased left hemisphere activation
when processing language nor a right ear advantage on di-
chotic listening tasks [6,51,78] highlighting that the DDs’
reading strategies are less phonemically based [50,56].

0028-3932/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 8 - 3 9 3 2 ( 0 2 ) 0 0 0 4 6 - 5
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
Thus, DD can be described as a deficit in the inhibitory
processes required to shift from the logographic level of
reading and as such we would expect the abnormal symme-
try described above to be evident in the prefrontal regions,
which are associated with the control of inhibition. Whilst
most neuropsychological studies focus on more posterior ar-
eas of the brain associated with language, Hynd et al. [36]
found that atypical symmetry of the prefrontal regions cor-
related with deficits in reading skills. Abnormal symmetry
has been found to be evident in the prefrontal cortex in
language impaired children, in the absence of such abnor-
malities in more posterior areas associated with language
[28,39]. This would suggest that the hemispheric symmetry
underpinning the deficit in DD would be manifest in other
cognitive assessments not directly related to reading, such
as tasks benefiting from inhibitory control. The main aim of
the four studies reported here is to evaluate this hypothesis
by assessing dyslexic individuals on a series of tasks that
are not specifically associated with assessments of reading.
Given the prefrontal symmetry that has been identified in
DD, we would predict biases in assessments of executive
functioning, which include the capacity for inhibition.
Executive functions refer to a collection of cognitive
abilities such as planning, sequencing, organisation and
inhibition, associated with the functioning of the prefrontal
cortex [27,49]. DDs have been shown to demonstrate poor
strategy formation and a lack of planning and organis-
ing [11,12,41,52,62]. DDs, for example, under-perform on
Tower of Hanoi (ToH) assessments of planning and strategy
formation. Despite the potentially significant role of inhibi-
tion in the development of reading, this executive function
has been largely neglected in the assessment of DD. The
potential significance of inhibitory processes is highlighted
by the consistent deficit that has been identified in verbal
digit span tasks, which assess aspects of working memory.
Increasingly taxing working memory resources results in
increased difficulty in controlling inhibition [60]. Working
memory demand and the demand for inhibition have been
identified as the two dimensions critical for understanding
the breadth of executive functioning tasks. Tasks with a
high demand for either or both of these dimensions are
sensitive to prefrontal function and may be mediated by a
single cognitive mechanism (see [64] for a review).
This is consistent with a large body of research that argues
the capacity for inhibition (or ‘resistance to interference’)
within the perceptual domain is a major factor in cognitive
development that is intimately associated with the operation
of the prefrontal cortex; young children ‘find it difficult
to resist misleading, irrelevant visual and auditory stimuli’
([14], p. 224). This sensitivity to interference declines
throughout the school years through increased effectiveness
in inhibiting or suppressing stimuli or associations that are
not relevant to the task at hand [27,49]. This sensitivity
also increases again during the ageing process, beginning
in the 50s [46]. A task that assesses the effectiveness of
these inhibitory processes and mirrors this developmental
2145
trend is the group-embedded figures test (GEFT; [84]).
The GEFT requires participants to identify a simple shape
(such as a cube) within a complex visual array, designed
to provide distracting context. Effective inhibition of the
distracting context allows for successful completion of the
task. The poor GEFT performance of young children and
elderly individuals is similar to the performance of frontal
lobe patients. Frontal lesions affect the ability to control
interference from external and internal sources, without af-
fecting established IQ [75] and ERP studies have confirmed
the role of the frontal lobes in GEFT performance, specif-
ically the left prefrontal cortex [19,20]. This is particularly
pertinent to reading as most text passages contain abundant
sources of interference (e.g. ambiguities, see [14,15] for a
full account). Some text processing difficulties exhibited
by DDs maybe due to inefficient inhibition as they are
with frontal patients, as ‘s/he will have difficulty remaining
“. . . within the selective system of connections given by
the text” ([49, p. 286]; [14, p. 62]). Thus, in addition to its
impact on reading strategies, inhibition plays a significant
role in learning—particularly in the presence of interference
from distractors [14].
Whilst the prefrontal cortex is not a homogeneous struc-
ture, having several distinct cytoarchitectonic regions, the
capacity for inhibition is an intrinsic property of the pre-
frontal cortices as a whole [65]. The abnormal cerebral sym-
metry in dyslexia discussed above is particularly relevant
to this aspect of executive functioning as the left prefrontal
region inhibits the processing of the right prefrontal region
[24] raising the possibility that a deficit in the control of in-
hibition in the left prefrontal cortex may underpin DD. ERP
studies have identified the left frontal lobe as particularly
important for the selective processing of the local level [86],
a process that requires the inhibition of the global level
processing (as described above). These inhibitory processes
mediate selective attention and the inability to inhibit dis-
tractors may underlie the distractibility frequently found in
frontal patients [24] and is also reported in dyslexics [83].
In normal processing, increased blood flow in the left
prefrontal cortex is associated with efficient planning and
organisation [54]. The verbal fluency task, for example, is
associated with left prefrontal cortex functioning which re-
quires the generation of words that begin with a specified
letter (F, S). Patients with bilateral or left frontal lesions but
not right frontal lesions have been found to generate signif-
icantly fewer words than matched controls [38,53]. These
findings from the acquired literature are comparable to the
study of developmental dyslexics who also display a deficit
on this task [44,47]. Whilst it is probable that the represen-
tations of words are stored in the temporal lobes, the left
prefrontal cortex is involved in generating appropriate re-
sponses and inhibiting inappropriate responses [26].
As Bakker’s [3–6] proposal of abnormal symmetry in DD
extends to the prefrontal cortex, we would predict deficits
in executive functions associated with the left prefrontal
cortex, one of which is a failure to inhibit the processing
2146 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155

of distractors by the right hemisphere. This is consistent Table 1

with Eliez et al. [17, p. 642] who conclude ‘Although the Means (S.D.) for study 1
reading disorder that defines dyslexia may primarily reflect Dyslexic Non-dyslexic P-value
left-sided anatomical and functional abnormalities, the more (N = 9) (N = 9)
subtle involvement of homologous right-sided regions may Age 34 (11.80) 30 (7.25) NS
contribute to additional deficits that are not yet well recog- DS 13.44 (4.28) 18.22 (2.95) <0.01
FDS 7.67 (2.00) 10.11 (1.97) <0.01
nised or understood’. The present study aims to incorporate
BDS 5.78 (2.68) 8.11 (1.27) <0.05
the lateralisation to right or left frontal regions associated GEFT 10.11 (4.76) 14.33 (3.00) <0.05
with executive functions to better understand the cognitive Phoneme fluency 5.78 (3.87) 7.89 (2.37) NS (<0.1)
deficits associated with dyslexia. We predict that a failure WAIS sequencing 15.33 (4.47) 17.78 (2.11) NS
to inhibit the processing of irrelevant context will be re- ToH errors 78.44 (28.26) 78.67 (26.82) NS
ToH initial thinking 20952 (11917) 27239 (18954) NS
flected in a poor GEFT performance in DDs. Additionally
ToH subsequent thinking 3480 (4145) 3204 (2971) NS
we would predict a poor performance on the verbal fluency Free recall 1 15.33 (3.46) 17.56 (3.94) NS (<0.1)
task that has also been associated with left prefrontal func- Free recall 2 20.56 (4.50) 23.11 (3.22) NS (<0.1)
tioning. Planning will be assessed using the ToH task, as this ARC 1 (clustering) 0.60 (0.37) 0.46 (0.24) NS
has been used previously to assess this aspect of executive ARC 2 (clustering) 0.89 (0.13) 0.83 (0.13) NS
functioning. DS, digit span (F—forward, B—backward); ToH, Tower of Hanoi; ARC
The ability to resist the effects of interference also ex- 1, first instance when items were randomly arranged; ARC 2, second
plains the development of recall skills more effectively than instance when items were placed in semantic categories. Items are in
Appendix A.
the development of strategies, such as clustering [15,69].
The organisation of recall will be also be examined as the
reading ability of dyslexic children has been shown to bene- 1.2. Design
fit from explicit training on these skills (such as clustering),
suggesting that a failure to develop processes that generate The executive functions of inhibition, planning, sequenc-
and apply effective reading strategies impacts on the read- ing and organisation were examined in an undergraduate
ing ability of dyslexic children [77,79]. Intervention based sample of ‘compensated’ DDs, who may be able to com-
around organisational strategies for DDs can, therefore, im- pensate for strategies that can be explicitly demonstrated
pact upon their reading abilities. A particular group of DDs, through interventions but not for capacity to inhibit interfer-
who have entered higher education, despite their reading dif- ence, such as cues that are not relevant to the task at hand.
ficulties, will be studied initially. The childhood problems As mentioned above, a deficit in digit span is characteristic
with DD extend into adulthood generally including those of DD and this measure was put in place to ensure that our
who enter university education [10,22,33,74]. These ‘com- adult dyslexics still demonstrated a deficit in digit span.
pensated dyslexics’ still demonstrate underlying difficulties This ensured that the ‘compensated’ dyslexics have a profile
and a profile similar to younger DDs despite compensating consistent with the DD literature. The development of digit
for their reading difficulties and progressing through the edu- span is dependant upon effective inhibitory processing as is
cation system [59]. For example, the deficit in digit span task an effective GEFT performance. These inhibitory processes
is consistently identified in dyslexia across the age span. It is are associated with the left prefrontal cortex, as is an effec-
conceivable, therefore, that the ‘compensation’ occurs in the tive verbal fluency performance. Verbal fluency has been
explicit strategies employed to facilitate reading but not in found to be very sensitive to left prefrontal cortex damage
the capacity for inhibition. The first study will examine the [7] probably as the unusual nature of the task requires par-
executive functioning of planning, sequencing, organising ticipants to devise their own strategy. Parker and Crawford
and inhibition in a population of dyslexic undergraduates. [58] suggest that this is the most widely used formal test of
frontal dysfunction.
Planning and sequencing are also associated with pre-
1. Study 1 frontal cortex functioning. The planning task used previ-
ously is the ToH task. The present version allows for motor
1.1. Participants ability to be factored out (see below). For the sequencing
task, the participant must be sensitive to subtle changes in
Nine dyslexic and nine non-dyslexic male undergraduates. the pictures to enable him/her to order a sequence of events.
The dyslexic sample had all received a formal diagnosis of This kind of comprehension of event structure is associated
dyslexia from an appropriately qualified professional during with prefrontal cortex functioning [27,49]. Finally, the or-
their formal schooling, prior to their arrival at University. ganisation of memory was assessed to allow for an index of
The sample was matched upon academic year and major, free recall and the extent to which the free recall has been
current academic grades and SES and undertook additional clustered, both when superordinate categories are and are
assessments, which are reported elsewhere. The ages are not made salient. This allows us to investigate absolute recall
reported in Table 1. in addition to an index of the clustering strategy employed.
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
1.3. Tasks/measures
1.3.1. Digit span
The digit span procedure from the WAIS-R [82] was used
for both forward and backwards digit span. Participants were
read a series of numbers that they had to repeat back in the
correct order. Two instances of each digit span length were
given. A point was recorded for all correct responses and
testing ceased when an error occurred in both instances of a
digit span length. Forward digit span required the digits to
be repeated in the order they were presented and backward
digit span required the order to be reversed.
1.3.2. Inhibition
The group-embedded figures test [84] was used to assess
the ability to inhibit the processing of irrelevant context. Af-
ter a practice session to ensure the principles are understood,
participants undertook two 5 min testing sessions, each of
which contained nine items (and a possible maximum score
of 18). Participants located a simple figure within a complex
visual array. The purpose of the visual array was to provide
a context that made disembedding the searched for part diffi-
cult. As an example, one of the simple figures resembles the
outline of an upside-down kite. In the complex visual array,
the lines that make up this outline are extended and over-
laid with many parallel and perpendicular lines (of identical
colour and width). The complex visual array, therefore, pro-
vides a series of potential outlines that are similar (though
longer or shorter or disproportionately wider or narrower)
to the simple figure. Where the lines merge to form a geo-
metric figure, for example a triangle, this element is lightly
shaded to ‘reinforce given sub-wholes’ providing additional
distraction that has to be ignored to complete the task suc-
cessfully. There are a range of simple figures and complex
visual arrays to prevent practice effects. This measure has
been found to be both valid and reliable [13,84].
1.3.3. Verbal fluency
The verbal fluency task lasted 1 min during which time
participants has to generate as many words as possible (ex-
cluding plurals and proper nouns) beginning with a given
letter (either F or S). Only one trial was undertaken per par-
ticipant.
1.3.4. Planning
Planning was assessed through a ToH task (called ‘stock-
ings of Cambridge’ and referred to as ‘ToH’ in Table 1)
that required balls to be moved from one position to an-
other to move from a start state to a goal state. There were
two instances of two, three, four and five move problems.
The problems were presented through the Cambridge neu-
ropsychological test automated battery (CANTAB version
2.3; CeNeS Ltd. 1998) system that in addition to noting er-
rors made, recorded the timing of initial thinking time (prior
to the first move) and subsequent thinking time (between
subsequent moves). The system uses a touch screen display.
2147
Participants had to touch the ball they wish to move and then
touch the stocking in which they wanted to drop the ball.
This procedure was practised to ensure that participants were
familiar with the touch screen technology and the task. After
participants attempted the problems they were then asked to
simply follow the display as it moves. In fact these moves
were the moves that the participant made whilst attempting
the problems. The subtraction of this motor element allows
for the calculation of thinking times. Initial thinking time
was the time prior to the first move and subsequent thinking
time was the time between the first move and completion.
Each task had a minimum number of moves for completion,
additional moves were counted as errors.
1.3.5. Sequencing
The WAIS-R [82] contains a picture arrangement subtest
that required participants to arrange a series of pictures in a
sequence. Participants were presented with a series of cards
that had to be rearranged into a specific sequence for the
pictures to form a coherent story. The tasks were timed and
had a maximum score of 20.
1.3.6. Organisation
Organisation of memory was used to assess recall under
two conditions, namely when superordinate categories were
not made salient and then again when this potential organis-
ing structure was made salient. Participants underwent two
recall tasks of the same 30 items to investigate categorical
clustering in long-term memory. The to-be-recalled (TBR)
list comprised 30 items from three superordinate categories.
Previous research has shown that each of these items had
produced at least 90% agreement on name [73]. Each item
was represented as a black line drawing on laminated white
card. Appendix A provides a list of the TBR items. A mixed
measures design was employed, with independent groups on
group (dyslexic and non-dyslexic) and repeated measures on
trial (randomised and categorised TBR list). One TBR list
was used for both trial conditions in order to eliminate po-
tential inter-list differences contaminating any obtained ef-
fects involving trial. Counterbalancing of two different lists
was viewed as a risky design strategy, given the small sam-
ple sizes anticipated in this study. A fixed order was used for
the presentation of the two trials, with the randomised list
preceding the categorised list. A point was recorded for each
item correctly recalled (maximum of 30). In Table 1, recall
1 is the number of items recalled from the randomised list
and ARC 1 the adjusted ratio of clustering (ARC) for this
initial list. Similarly recall 2 and ARC 2 relate to the same
measures for the second (categorised) instance of the list.
1.4. Procedure
For the first trial, the TBR cards were shuffled prior to the
participant’s arrival. Once the participant had been seated,
E explained that a number of pictures of common objects
would be presented on the table before him/her, and that
2148 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
he/she will be asked to name each item, and would then be
given 1 min to study these, and subsequently, remember as
many as possible in any order. The cards were arrayed in
three rows of ten, and the participant was asked to name
each item. The participant was then given 1 min to study the
items, but was not allowed to manipulate or move the cards.
Once the minute had elapsed, E collected the cards, and then
asked the participant to undertake a brief repetition task of
repeating back each line of a poem that was read out by the
experimenter. This filler task took approximately 1 min, and
was intended to reduce the size of the recency effect, and
hence, result in recall that was dependent upon the control
processes that govern the long-term memory component [2].
Participants were given 2 min to recall items, their recall
being recorded on audiocassette tape.
The second trial commenced approximately 40 min after
the first trial, with the other tasks above being interposed
between the two trials. The same procedure was followed as
for the first trial, with the exception that each of the three
rows comprised items from one superordinate category. The
recordings were transcribed, and two measures obtained:
total correct recall, and an index of categorised recall, the
ARC [66]. ARC indexes the extent to which the frequency
of category adjacencies in a given recall protocol departs
from what would be expected by chance, given the number
of items recalled and the number of instances recalled from
each category. Thus, ARC is independent of the absolute
level of accurate recall. ARC has a possible maximum of +1
(indicating perfect categorical clustering of recalled items),
with a score of 0 indicating a chance-level of clustering.
ARC can also assume a negative value (below-chance levels
of clustering).
In between these tasks the GEFT, verbal fluency, the
stockings of Cambridge, the WAIS-R digit span and se-
quencing tasks were undertaken in a random order. Partic-
ipants were tested individually in a quiet room away from
any distractions.
1.5. Results
Table 1 shows the group means and standard deviations
for the dyslexic and non-dyslexic groups. As expected the
dyslexic group had a significantly reduced digit span com-
pared with the non-dyslexic group (t = 2.76, d.f. = 16,
P < 0.01). This was the case for both forward (t = 2.62,
d.f. = 16, P < 0.01) and backward (t = 2.36, d.f. = 16,
P < 0.05) digit spans when they were separated out. The
GEFT, too, is consistent with the hypothesis that the dyslexic
group would underachieve upon this task (t = 2.25, d.f. =
16, P < 0.05). For all the assessments of planning, sequenc-
ing and organisation of memory, there were no significant
differences between the groups. The improvement in recall
and ARC scores on the memory task when the superordinate
categories were made salient was also comparable across
groups. The only trends (P < 0.1, one-tailed) were for the
dyslexic group to under-perform upon the verbal fluency
task (by 27% compared with controls) and the two recall
tasks (by around 2.5 TBR items compared with controls).
1.6. Discussion
The difference in digit span confirmed that the student
dyslexic sample retained the typical dyslexic pattern of re-
duced verbal working memory, despite undoubtedly having
to compensate for their deficit [55]. The deficit was found
in both forward and backward digit spans. There were no
group differences for any of the executive functioning mea-
sures of planning, sequencing or organisation of memory.
As hypothesised, however, there was a significantly poorer
performance upon the GEFT, suggesting an inability to in-
hibit irrelevant context in the dyslexic sample.
The results of the planning, sequencing and organising
tasks are not consistent with previous research, some of
which used the same measures of executive functioning as
the present study. The use of adult ‘compensated’ dyslexics
in the present study suggests that the compensation that
occurs may be related to certain executive functions. Sim-
ilarly, despite a trend to a lower recall score, DDs did not
demonstrate a deficit in the use of categorical clustering as a
strategy (ARC scores). This is consistent with other research
that found that certain executive function deficits diminish
as DDs progress through the education system [12]. The
executive measures that did show a difference between the
groups were those requiring inhibition and associated with
the left prefrontal cortex. The deficit in GEFT performance
is consistent with a description of a failure to inhibit the
processing of contextual information to facilitate the extrap-
olation of constituent parts. The trend to a deficit in verbal
fluency is also consistent with a description of a deficit in
left prefrontal cortex activity. Benton’s [7] original study
identified damage to the left prefrontal cortex resulting in
a 30% decrease in word generation. This is comparable to
study 1, where the dyslexic sample demonstrated a 27%
decrement on the verbal fluency task.
It is conceivable that the stockings of Cambridge and the
WAIS-R sequencing tasks did not target one frontal lobe
specifically. Deficits in the WAIS-R sequencing task do not
appear to be specific to either the right or left frontal lobe
and indeed greater impairment occurs with more posterior
damage [80]. Additionally, whilst Shallice [70] originally re-
ported impairment upon the ToH task following left frontal
damage, a subsequent study failed to replicate the finding
[71, p. 347]. This suggests that the compensation that occurs
in DD may be restricted to executive functions that are not
associated with the left prefrontal cortex. The frontal sym-
metry described in the introduction may underpin the lack of
ability to compensate for the executive functions associated
with the left prefrontal cortex.
The poor GEFT and digit span performances suggest
that DDs under-perform when the demands are high for
the critical dimensions of executive functioning [60]. The
deficit in digit span is a common finding in studies of DD
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
Table 2
Means (S.D.) for study 2
Dyslexic (N = 30) Non-dyslexic (N = 30) P-value
Age 14.0 (0.8) 13.8 (0.6) NS
GEFT 8.2 (5.7) 13.3 (2.4) <0.05
but identification of a poor GEFT performance is a novel
finding. Having identified a deficit in the GEFT in a small
number of adult dyslexics in study 1, study 2 sought to
replicate the finding with a larger and younger sample.
2. Study 2
2.1. Participants
Thirty dyslexic males from a school for dyslexic children
and 30 non-dyslexic males matched for age, demographic
variables, SES and academic performance from a nearby
school. The sample was selected to be as young as possible
for the reliability of the measure. The youngest participant
was 13 years of age. For ages much below this, the children’s
embedded figures test (CEFT) is recommended [84]. The
ages are presented in Table 2.
2.2. Design, results and discussion
All participants were given the GEFT as before. As pre-
dicted, study 2 replicates the findings from study 1. Table 2
shows that dyslexics have greater difficulty inhibiting the
processing of the visual context to identify the parts of stim-
uli than matched controls (t = 4.40, P < 0.001). This
finding from study 1 is, therefore, extended to a larger and
younger dyslexic sample.
3. Study 3
The results from studies 1 and 2 are consistent with de-
scriptions of DD as a deficit in inhibiting irrelevant visual
cues, associated with a deficit in left prefrontal cortical
processing. This has been demonstrated using the GEFT,
an assessment that is visual–spatial in nature. An alterna-
tive hypothesis might contend that just as DDs have poor
reading ability, they also have poor visual–spatial ability
(generally). Normally, whilst phonological aspects of work-
ing memory are primarily subserved by the left hemisphere,
visual–spatial aspects of working memory are normally
primarily sub served by the right hemisphere [29]. In DD,
however, it is conceivable that the over-involvement of
the right hemisphere in reading compromises non-verbal
abilities [42,51]. Indeed Fein et al. [21] report that dyslex-
ics show working memory deficits for both verbal and
non-verbal material. This is consistent with research that ar-
gues the maintenance of serial order shares common features
2149
across memory tasks presented in different modalities [40].
Other research, however, reports no differences in the
visual–spatial abilities of dyslexics compared with matched
controls [32,34]. The sensitivity to interference described
above has also been argued to facilitate success within cer-
tain domains [83]. The visual–spatial abilities of dyslexic
students were, therefore, further examined in the third study.
3.1. Participants
Fifteen dyslexic university students and 15 matched con-
trols, diagnosed and matched as in study 1, aged 18–29 years
with eight females and seven males in each sample. None
of these participants took part in study 1.
3.2. Design
The digit span assessment from the WAIS-R was admin-
istered as before. Additionally, the CANTAB touch screen
system described above was used to administer four tests of
visual spatial cognition. These were as follows.
• Spatial span: A pattern of white squares was shown on
the screen. Some of the squares change in colour, one by
one, in a variable sequence. At the end of the sequence,
a tone indicates that the subject should touch each of the
boxes coloured by the computer in the same order as they
were originally presented. Two instances are presented at
each sequence level that increases in length by one item
until the participant fails both instances of the same level
(mirroring exactly the verbal digit span procedure).
• Spatial recognition: The participant was presented with
a white square that moves in sequence to five different
places on the screen. The participant then sees a series of
five pairs of squares, one of which they have already seen
in the presentation phase and one of which was not seen
in the presentation phase. The participant has to touch the
previously seen square. Locations are tested in the reverse
order of presentation order. This procedure is repeated
three times, each time with five new locations.
• Delayed matching to sample: The participant was shown
a complex visual pattern and then after a brief delay, four
choice patterns. The participant has to point to the pre-
viously seen pattern. These patterns are designed so they
cannot easily be given verbal labels. Each pattern is made
up of four sub-elements each of a different colour. To dis-
courage encoding strategies based upon encoding single
quadrants, all four choices have one quadrant in common.
The choices are presented either simultaneously or with
0, 4 or 12 s delays. Ten examples of each delay were used,
giving a total of 40 presentations.
• Pattern recognition: The participant was presented with a
series of 12 visual patterns (as above) in the centre of the
screen. In the recognition phase, participants are presented
with a previously seen pattern and a novel pattern. The
participant has to touch the previously seen square. The
2150 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155

Table 3 working memory is part of the working memory model but

Means (S.D.) for study 3 would appear not to suffer the effects of high demand as
Dyslexic Non-dyslexic P-value verbal working memory does.
(N = 15) (N = 15) The results of study 3 are consistent with the proposal that
Age 23 (3.6) 24 (4.1) NS the deficit evidenced in the GEFT is not due to a general
Digit span 14.13 (4.2) 17.73 (4.3) <0.001 deficit in visual–spatial abilities. The pattern of results of the
Spatial span 7.13 (1.3) 7.53 (0.9) NS first three studies is consistent with the proposal that DD can
Spatial recognition 83.3 (46) 80.7 (38) NS
Pattern recognition 93.1 (42) 88.1 (43) NS (<0.1)
be characterised as a specific deficit in control of inhibition,
Delayed matching 90.2 (40) 88.0 (37) NS assessed using the GEFT. The relative weakness of executive
functioning associated with the left prefrontal cortex was
also implicated in study 1 and was further examined in the
final study.
recognition phase is presented in the reverse order of the
presentation order.
4. Study 4
The scores for the latter three assessments are presented
as percentages and did not deviate significantly from a
normal distribution. The tests were presented in a random Study 4 aimed to examine executive functioning in a larger
order. and younger sample with tasks specifically associated with
the left prefrontal cortex. In addition to study 1 containing a
relatively small sample size, controlling for some sort of lit-
3.3. Results eracy standard is recommended for clinical assessment [48]
and was incorporated into the final study. It is clear that a
The difference in digit span was again in evidence, reaf- neuropsychologically informed understanding of executive
firming that dyslexic students again demonstrate a deficit functioning is required to characterise the deficits associated
in this task. Table 3 highlights, however, that there were no with DD. From the study 1, we can infer that there will be
significant differences upon the visual–spatial measures. specific deficits that are associated with left frontal func-
tioning. This was confirmed in study 2 for the GEFT in a
3.4. Discussion younger sample and the final study examined verbal fluency
in a younger sample. Additionally, a specific aspect of se-
Once again the deficit in digit span provides support for quencing tasks that is associated with lateralised executive
the contention that dyslexics at university still retain verbal functioning was assessed.
working memory deficits. There were no visual spatial dif- Temporal order judgements can be seen as a specific
ferences, however, suggesting neither enhanced nor inhibited example of sequencing and have been found to be deficient
visual spatial abilities (associated with the right hemisphere) in dyslexic children and adults [21,44]. Patients with pre-
in DD. Examining the means highlights that dyslexics frontal lesions (bilateral, left) have also been found to suffer
averaged higher scores (albeit non-significantly) than the selective impairment in the recall for order of presentation
non-dyslexics on the visual–spatial measures. The pattern with the absence of recognition deficits for the same images
recognition difference shows a trend at P < 0.08, two [43,68]. Aspects of retrieval other than the temporal order,
tailed, for example. Overall, however, dyslexic individuals such as recognition, have not been associated with the pre-
perform comparably to non-dyslexic individuals on visual frontal cortex [68,76]. Executive functions associated with
spatial tasks associated with right hemisphere speciali- the left prefrontal cortex would selectively impair the judge-
sation. ment of the order of items to be recognised without necessar-
It would appear that the abnormal symmetry does not ily impacting upon the recognition itself. The introduction
enhance cognitive abilities associated with the right hemi- to the present series of studies also provided evidence to
sphere. The aim of these assessments was to confirm that support the contention that the inability to inhibit interfer-
DDs do not have a general visual–spatial deficit. Only a ence extends beyond the visual plane to other modalities.
small range of visual–spatial tasks were used and it may be A specific hypothesis is drawn from the acquired literature
that other assessments could yield an advantage. The failure that a deficit associated with the left prefrontal cortex will
to identify any advantage on these tasks may be due to the become manifest in an inability to inhibit the processing
normal right hemisphere specialisation for non-verbal tasks of distracting verbal material [55]. The susceptibility of
being compromised by the verbal activity that the hemi- DDs to verbal distraction has been observed in the class-
sphere undertakes. room [62].
Previous findings of impaired spatial span in DD were not Therefore, specific aspects of planning (verbal fluency),
replicated. As with study 1, this may be due to the ability sequencing (temporal order) and inhibition (of distraction)
of ‘compensated dyslexics’ to compensate for deficits not associated with left prefrontal functioning were examined
directly associated with left hemisphere processing. Spatial in study 4. Younger children were selected to minimise any
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155 2151

effects of strategy tuition that may occur through the educa- Table 4
tion process. Means (S.D.) for study 4
Dyslexic Non-dyslexic P-value
4.1. Participants (N = 16) (N = 16)
Age 9.9 (0.9) 9.9 (0.9) NS
Sixteen dyslexics aged 8–10 years (6 girls and 10 Verbal fluency 7.0 (3.7) 10.5 (3.2) <0.01
Vocabulary 98.7 (11.8) 93.9 (12.1) NS
boys) had all received a diagnosis from an educational
Digit span silent 5.0 (1.5) 5.8 (1.3) <0.05
psychologist with no history or evidence of sensory im- Digit span noisy 4.4 (1.0) 5.8 (1.2) <0.001
pairments, hyperactivity or emotional problems. The DS difference 0.6 (1.1) 0.0 (1.0) NS
dyslexics attended a mainstream school and 16 age- and Temporal order 3.6 (1.0) 4.2 (0.8) <0.05
gender-matched controls were selected as before from class Recognition 9.2 (1.9) 9.9 (0.5) NS
peers.
As the measures produce parametric data and we are
4.2. Design and procedure
predicting that the dyslexic group will under-perform upon
As before, the verbal fluency task lasted 1 min during specific tasks, parametric one-tail tests were performed.
which time participants has to generate as many words as
possible (excluding plurals and proper nouns) beginning 4.3. Results
with a given letter (either F or S). Only one trial was un-
dertaken per participant. To ensure that all participants had As predicted the dyslexic group produced significantly
a similar level of vocabulary, the british picture vocabulary fewer items on the verbal fluency task than the non-dyslexic
scale (BPVS [16]) was also administered. This requires par- group (t = 2.90, d.f. = 30, P < 0.01) with no significant
ticipants to point to one of four pictures that correspond to difference on the BPVS scores (t = 1.14, d.f. = 30, NS).
the word spoken by the experimenter. Indeed the dyslexic group scored insignificantly higher upon
Forward digit span was assessed using the standard this latter task. The dyslexic group also obtained a lower
WISC-III procedure [81]. An increasing series of digits digit span in the silent condition (t = 1.65, d.f. = 30,
is read to the participant who recalls the digits immedi- P = 0.05). This difference was enhanced in the background
ately. Each sequence length occurs twice and when both noise condition (t = 3.51, d.f. = 30, P < 0.001) and the
sequences of that length are recalled erroneously, testing decrement for the dyslexic group showed a trend towards a
ceases. Digit span is the highest sequence length cor- greater decrement than the non-dyslexic group (t = 1.48,
rectly recalled. This occurred under the traditional con- d.f. = 30, P = 0.075). This can be described as a trend for
dition of silence and under an additional condition with dyslexics to be distracted, especially as the distraction had
background voices. A different set of sequences was read no impact upon the performance of non-dyslexics. Finally
out as above. To the right of the participant 0.5 m away and again in line with prediction the dyslexic group scored
a tape recorder played a tape of continuous speech (no significantly lower on the recall of temporal order (t = 1.77,
numbers). The tape recorder was set at 30 db. At the d.f. = 30, P < 0.05) with no significant difference for the
end of each session the children were asked if they had recognition scores (t = 1.42, d.f. = 30, NS). All the means
heard the digits clearly. All the children reported that they are reported in Table 4.
had.
To assess long-term memory processes children were 4.4. Discussion
shown ten cards with simple line drawings from Snodgrass
and Vanderwart (as study 1). Each card was presented for Once again the deficit in digit span was evident in the
3 s and then removed untill all ten cards had been presented. younger sample. This difference increased with verbal dis-
A brief filler task was undertaken whilst the ten cards were tractors, which is consistent with the left prefrontal cortex
shuffled (as study 1) and paired with a previously unseen failing to inhibit analysis of irrelevant background verbal
card. The participant was then presented with a pair of material [55] and observations of dyslexic children’s dis-
cards and asked to point to the card that has been previously tractibility in the classroom [62]. The detrimental impact of
seen. The order was random as was whether the previously additional interference suggests that a sensitivity to distrac-
seen card was on the right or left. After this assessment tion may be related to the diminished digit span capacity in
of recognition the previously seen cards were presented in DD.
pairs (1st and 10th, 2nd and 9th, 3rd and 8th, 4th and 7th, The results are also consistent with the proposal that the
5th and 6th) and participants were asked to point to the verbal fluency and temporal order tasks associated with the
picture that had been shown first in the original presentation left prefrontal cortex are impaired independent of the vocab-
(identified as temporal order in Table 4). The ordering of ulary and recognition tasks, which are not associated with
the pairs was random and a point recorded for each correct prefrontal functioning. The decrement in verbal fluency was
response (maximum of 5). 33%, again similar to the 30% decrement identified in left
2152 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
frontal patients [7]. These three findings are all consistent
with the proposal that DD can be characterised as a deficit in
the executive functioning associated with the left prefrontal
cortex.
5. General discussion
Eliez et al. [17] suggested that an understanding of right
and left hemisphere processing is required to fully appreciate
the deficits associated with DD. A rightward symmetry in the
frontal lobes has been identified in DD, even in the absence
of more posterior symmetry, and consequently the aim of
this paper was to examine executive functioning in child and
adult dyslexics. Studies 1 and 2 demonstrated a deficit upon
the GEFT in dyslexic children and adults. This is a novel
finding, not previously identified. The GEFT requires the
identification of a specific element within a more complex
display. The task requires the inhibition of the processing of
irrelevant context to facilitate the identification of the salient
part.
This perspective is consistent with a large body of liter-
ature that has identified the right hemisphere as processing
a relatively global level of stimuli and the left hemisphere
as processing a relatively local level. The rightward sym-
metry, therefore, favours a processing of the word (logo-
graphic) level for reading in preference to the more efficient
letter level. The present study suggests that this account of
dyslexia is underpinned by a failure to inhibit the process-
ing of the logographic level that is required to establish the
normal reading process via grapheme–phoneme correspon-
dences [25]. The development of these reading processes is
facilitated by inhibiting the processing of distractors at the
more global level. Recent neuroimaging research has iden-
tified that the normal processing of distractors, which indi-
cated the presence of potentially relevant information at the
unattended level, resulted in delayed lateralization of pro-
cessing [18]. Distractors delayed the asymmetric processing
of the local level to the left hemisphere and the global level
to the right hemisphere. This is consistent with the neuropsy-
chological evidence that DDs fail to specialise language to
the left hemisphere. Thus, the shaky grapheme–phoneme
representations associated with dyslexia (see [74] for an ac-
count) can be viewed as a specific manifestation of a more
general bias towards delayed asymmetry of processing—
a process delayed by a failure to inhibit the processing of
interference.
The contention that these deficits in DD identified in the
present study can be characterised as an inability to inhibit
the processing of to-be-ignored information is supported
by the poor performance of dyslexics upon the Stroop task
[45]. This performance upon the Stroop task also mirrors
the performance of left frontal lobe patients [8]. A deficit in
left prefrontal cortex processing in DD is also suggested by
the comparable deficit in the dyslexic samples (27 and 33%)
to the left frontal lobe patients (30%) on verbal fluency
tasks. Similarly the deficits in temporal order judgements
and recall are consistent between dyslexics and frontal lobe
patients, in isolation from any deficits in recognition tasks or
general sequencing tasks [68]. The left and right prefrontal
cortices are not independent from each other and one aspect
of the functioning of the left prefrontal cortex is to inhibit
aspects of the processing of the right prefrontal cortex.
Thus, a single neuropsychological deficit may underpin the
cognitive performances of a bias towards failing to inhibit
the processing of context (and poor GEFT performance,
distractibility) and the decrement in verbal fluency and tem-
poral order judgements. This concurs with assumptions that
the resistance to interference is a fundamental feature of the
cognitive system (i.e. a psychological primitive), that cannot
be explained in terms of other cognitive processes [15].
The impact of irrelevant background verbal material of
digit span in study 4 confirms this bias extends to the au-
ditory modality. DDs were less able to inhibit the process-
ing of the auditory context to maximise performance upon
the digit span task. DDs, however, did not demonstrate a
deficit on the spatial span task. Whilst the control of these
two working memory span systems resides within executive
functioning [2], the processing during span tasks activates
more posteriors areas in the left and right hemispheres, re-
spectively [29]. There are many links from the prefrontal
cortices to posterior areas that are reciprocal. It is possible
that the neuropathology resides outside of the prefrontal cor-
tex, in a closely connected brain system. From the range of
tasks described above, however, it seems more plausible that
the left prefrontal cortex itself is responsible for a deficit in
inhibiting interference.
Working memory processes are involved in the inhibi-
tion of inappropriate activity [64] and the extent to which
dyslexia could be described as a working memory deficit
is unclear. As mentioned, DDs demonstrate a pattern sim-
ilar to frontal lobe patients with deficits on recall (study 1)
and temporal order (study 4) tasks without deficits upon
recognition tasks [68]. Pennington et al. [60] argue that
temporal order and recall tasks utilise a memory system
that allows an individual to access, organise and manipulate
memories, therefore, placing strong demands on working
memory. The authors argue that this can be distinguished
from recognition memory, which relies on effective stor-
age and consolidation and place relatively few demands
on working memory. The concept of working memory is
not distinct from inhibitory processing, however, as in-
creasing demands on working memory results in increasing
difficulty in controlling inhibition and a high demand for
either or both of these dimensions are sensitive to prefrontal
function. A smaller digit span will, by definition, approach
capacity earlier (than a normal digit span) and demand in-
hibitory resources earlier. Additionally deficits in inhibitory
control will result in a poorer digit span performance,
highlighting how these two critical dimensions of execu-
tive functioning are inter-dependant and pertinent to under-
standing DD.
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155 2153

A potential contradiction resides in the failure to find a Appendix A. TBR items

difference in clustering in study one. Both groups recall
improved by around five items (from a list of 30) on the
Animals Objects Fruits/vegetables
second recall condition with no differences between groups
Frog Broom Lemon
or conditions for the clustering (ARC) scores. The acquired
Horse Anchor Pumpkin
literature would suggest that a left prefrontal deficit would
Cat Pencil Strawberry
manifest itself in a deficit in spontaneously applying a
Elephant Toothbrush Pear
clustering strategy and a consequently disproportionate in-
Owl Hammer Grapes
creased in free recall when the strategy was made explicit
Kangaroo Axe Onion
[67]. In the present study, the words were presented or-
Rabbit Candle Pineapple
thographically (rather than verbally), participants were not
Dog Belt Banana
explicitly told to cluster and related word lists were used
Fish Comb Carrot
on both recall trials—variations that reduce the demand
Camel Candle Apple
upon the left prefrontal cortex [23]. Future research could
explicitly manipulate these variables for the dyslexic pop-
ulation. It is interesting that the trend in DDs for a deficit
in free recall in both conditions is independent of cluster- References
ing, suggesting that the inability to inhibit interference may
contribute to recall deficits [30]. [1] American Psychological Association, DSM-IV, 1994.
It has been noted by Evans et al. [18] that it is biologi- [2] Baddeley A. Working memory. Oxford: Oxford University Press,
1986.
cally adaptive to be susceptible to a degree of environmental
[3] Bakker D. Hemispheric differences and reading strategies: two
distractibility, if the organism is to be alerted to potential dyslexias? Bulletin of the Orton Society 1979;29:84–100.
dangers outside of attentional focus. There could also be ad- [4] Bakker D. Neuropsychological treatment of dyslexia. New York:
vantages on assessments that require sensitivity to the pro- Oxford University Press, 1990.
cessing of context. This may be the advantage provided by a [5] Bakker D, Bouma A, Gardien C. Hemisphere-specific treatment of
dyslexia subtypes: a field experiment. Journal of Learning Disabilities
deficit in inhibition of right hemisphere processing. The pat-
1990;23:433–8.
tern recognition task in study 3 may be an example of this. [6] Bakker D, Smink T, Reitsma P. Ear dominance and reading ability.
There was a trend for DDs to perform better than controls Cortex 1973;9:301–12.
on this task, which was designed to be non-verbal and to [7] Benton A. Differential behavioural effects in frontal lobe disease.
discourage part-based encoding strategies. It is perhaps im- Neuropsychologia 1968;6:53–60.
portant to also reiterate that dyslexics did not show deficits [8] Benson D, Stuss D, Naeser M, Weir W, Kaplan E, Levine H. The
long-term effects of prefrontal leukotomy. Archives of Neurology
upon a large array of tasks (planning, sequencing, recall 1981;38:165–9.
and clustering in memory, spatial span, spatial recognition, [9] Briand K. Selective attention to global and local structure objects:
pattern recognition, delayed matching to sample and over- alternative measures of nontarget processing. Perception and
all level of vocabulary). Interestingly, the opposite findings Psychophysics 1994;55:562–74.
to those described above, have been identified for another [10] Bruck M. Word-recognition skills of adults with childhood diagnosis
of dyslexia. Developmental Psychology 1990;26:439–54.
developmental disorder. Individuals with Autism have been
[11] Catts H. Speech production deficits in developmental dyslexia.
characterised as extremely insensitive to context and per- Journal of Speech and Hearing Disorders 1989;54:422–8.
form exceptionally well upon the GEFT. This cognitive pro- [12] Condor A, Anderson V, Saling M. Do reading disabled children have
file has been characterised as a ‘right hemisphere weakness’ panning problems? Developmental Neuropsychology 1995;11:485–
[63, p. 777]. 502.
This series of studies has characterised dyslexia as a spe- [13] Day D, McRae L, Young J. The group-embedded figures test: a
factor analytic study. Perceptual and Motor Skills 1990;10:835–9.
cific range of cognitive deficits (digit span, verbal fluency, [14] Dempster FN. The rise and fall of the inhibitory mechanism: toward
temporal order, inhibition of distractors), that are linked by a unified theory of cognitive development and ageing. Developmental
neuropsychological theory to deficits associated with left Review 1992;12:45–75.
prefrontal cortical processing. Further, it may be that it is on [15] Dempster FN, Corkhill A. Neo-interference research and the
executive functions not specifically associated with the left development of intelligence. In: Anderson M, editor. The
development of intelligence. Hillsdale, NJ: Psychology Press, 1999.
prefrontal cortex, that compensation can take place for the
[16] Dunn L, Whetton C, Burley J. British picture vocabulary scale. 2nd
dyslexic student. ed. Nelson, Windsor: NFER, 1997.
[17] Eliez S, Rumsey J, Giedd J, Schmitt J, Patwardhan A, Reiss A.
Morphological alteration of temporal lobe gray matter in dyslexia:
Acknowledgement an MRI study. Journal of Child Psychology and Psychiatry 2000;41:
637–44.
[18] Evans M, Shedden J, Hevenor S, Hahn M. The effect of variability
The authors would like to express their thanks to the of unattended information on global and local processing: evidence
two anonymous referees for their comments on the original for lateralization at early stages of processing. Neuropsychologia
version of this paper. 2000;38:225–39.
2154 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
[19] Federico PA. ERP correlates of cognitive styles, abilities and
aptitudes. Personality and Individual Differences 1984;5:575–85.
[20] Federico PA, Froning JN, Calder M. Validation of brain ERPs
as indicators of cognitive styles, abilities and aptitudes. US Navy
Personnel Research and Development Technical Report 83-11, 1983.
[21] Fein G, Davenport L, Yingling CD, Galin D. Verbal and non-verbal
memory deficits in pure dyslexia. Developmental Neuropsychology
1988;4:181–97.
[22] Felton R, Naylor C, Wood F. Neuropsychological profile of adult
dyslexics. Brain and Language 1990;39:485–97.
[23] Fletcher PC, Shallice T, Dolan RJ. The functional roles of prefrontal
cortex in episodic memory. I. Encoding. Brain 1998;121:1239–48.
[24] Foster JK, Eskes GA, Stuss DT. The cognitive neuropsychology
of attention: a frontal lobe perspective. Cognitive Neuropsychology
1994;11:133–47.
[25] Frith U. Cognitive processes in spelling. London: Academic Press,
1985.
[26] Frith CD, Friston KJ, Liddle PF, Frackowiak RSJ. A pet study of
word finding. Neuropsychologia 1991;29:1137–48.
[27] Fuster J. The prefrontal cortex. 2nd ed. New York: Raven, 1989.
[28] Galaburda A. Developmental dyslexia: a review of biological
interaction. Annals of Dyslexia 1985;35:21–33.
[29] Gathercole S. Cognitive approaches to the development of short term
memory. Trends in Cognitive Science 1999;3:410–8.
[30] Gershberg FB, Shimamura AP. Impaired use of organisational
strategies in free recall following frontal lobe damage. Neuro-
psychologia 1995;33:1305–33.
[31] Geschwind N, Behan P. Left handedness: association with immune
disease, migraine and developmental learning disorders. Proceedings
of the National Academy of Sciences 1982;79:5097–100.
[32] Gillies M, Light P. Spatial cognition in language impaired children.
European Journal of Disorders of Communication 1997;32:159–66.
[33] Hanley R. Reading and spelling impairments in undergraduate
students with developmental dyslexia. Journal of Research in Reading
1997;20:22–30.
[34] Hicks C. The ITPA visual sequential memory task: an alternative
explanation and the implications for good and poor readers. British
Journal of Educational Psychology 1980;50:16–25.
[35] Hynd G, Marshall R, Hall J, Edmonds J. Learning disabilities:
neuroanatomic asymmetries. In: Davidson R, Hugdahl K, editors.
Brain asymmetry. Bradford: MIT Press, 1995.
[36] Hynd G, Semrud-Clikman M, Lorys A, Novey E, Eliopulos D.
Brain morphology in developmental dyslexia and ADHD. Archives
of Neurology 1990;47:919–26.
[37] Ivry R, Robertson L. The two sides of perception. Bradford: MIT
Press, 1997.
[38] Janowski JS, Shimamura AP, Kritchevsky M, Squire LR. Cognitive
impairment following frontal lobe damage and its relevance to human
amnesia. Behavioural Neuroscience 1989;103:548–60.
[39] Jernigan TL, Hesselink JR, Sowell E, Tallal PA. Cerebral structure
on magnetic resonance imaging in language- and learning-impaired
children. Archives of Neurology 1991;48:539–45.
[40] Jones D, Farrand P, Stuart G, Morris N. Functional equivalence of
verbal and spatial information in serial short term memory. Journal
of Experimental Psychology: Learning Memory and Cognition
1995;21:1008–18.
[41] Kappers E. Outpatient treatment of dyslexia through stimulation
of cerebral hemispheres. Journal of Learning Disabilities 1997;30:
100–25.
[42] Kershner J, Morton L. Directed attention and dichotic listening in
reading disabled children: A test of four models of maladaptive
lateralization. Neuropsychologia 1990;28:899–902.
[43] Kesner RP, Hopkins RO, Fineman B. Item and order dissociation
in humans with prefrontal cortex damage. Neuropsychologia
1994;32:881–91.
[44] Kinsbourne M, Rufo DT, Gamzu E, Palmer RL, Berliner AK.
Neuropsychological deficits in adults with dyslexia. Developmental
Medicine and Child Neurology 1991;33:763–75.
[45] Lazarus P, Ludwig R, Aberson B. Stroop color word test: a screening
measure of selective attention to differentiate LD from non LD
children. Psychology in the Schools 1984;21:53–60.
[46] Lee J, Pollack R. The effect of age on perceptual problem solving
strategies. Experimental Ageing Research 1978;4:37–54.
[47] Levine BE. Organizational deficits in dyslexia: possible frontal
lobe dysfunction. Developmental Neuropsychology 1990;6:95–
110.
[48] Lezak M. Neuropsychological assessment. Oxford: Oxford University
Press, 1983.
[49] Luria AR. Higher cortical functions in man. London: Tavistock, 1966.
[50] Malatesha Joshi R, Aaron P. Specific spelling disability: factual or
artifactual? Reading and Writing 1990;2:107–25.
[51] Masutto C, Bravar L, Fabbro F. Neurolinguistic differentiation of
children with subtypes of dyslexia. Journal of Learning Disabilities
1994;27:520–6.
[52] Mati-Zissi H, Zafiropoulou M, Bonoti F. Drawing performance in
children with special learning difficulties. Perceptual and Motor Skills
1998;87:487–97.
[53] Milner B. Some effects of frontal lobectomy in man. In: Warren JM,
Akert G, editors. The frontal granular cortex and behaviour. New
York: McGraw-Hill, 1964.
[54] Morris RG, Ahmed S, Syed GM, Toone BK. Neural correlates of
planning ability: frontal lobe activation during the Tower of London
test. Neuropsychologia 1993;31:1367–78.
[55] Nathaniel-James DA, Fletcher P, Frith CD. The functional anatomy
of verbal initiation and suppression using the hayling test.
Neuropsychologia 1997;35:559–66.
[56] Newman S, Fields H, Wright S. A developmental study of
specific spelling disability. British Journal of Educational Psychology
1993;63:287–96.
[57] Orton S. Reading, writing and speech problems in children. New
York: Norton, 1937.
[58] Parker D, Crawford J. Frontal lobe dysfunction. In: Crawford J,
Parker D, McKinlay W, editors. A handbook of neuropsychological
assessment. Hove, East Sussex: LEA, 1992.
[59] Paulesu E, Frith U, Snowling M, Gallagher A, Morton J,
Frackowiak R, Frith C. Is developmental dyslexia a disconnection
syndrome? Evidence from PET scanning. Brain 1996;119:
143–57.
[60] Pennington B, Bennet L, McAteer O, Roberts R. Executive functions
and working memory: theoretical and measurement issues. In: lyon G,
Krasnegar NA, editors. Attention, memory and executive functions.
Baltimore, MD, Paul H. Brookes Publishing Co., 1996.
[61] Peterson B. Neuroimaging in child and adolescent neuropsychiatric
disorders. Journal of the American Academy of Child and Adolescent
Psychiatry 1995;34:1560–76.
[62] Pollock J, Waller E. Day-to-day dyslexia in the classroom. London:
Routledge, 1994.
[63] Rinehart N, Bradshaw J, Moss S, Brereton A, Tonge B. Atypical
interference of local detail of global processing in high functioning
autism and asperger’s disorder. Journal of Child Psychology and
Psychiatry 2000;41:769–78.
[64] Roberts R, Huger L, Heron C. Prefrontal cognitive processes: working
memory and inhibition of the antisacade task. Journal of Experimental
Psychology: General 1994;123:374–93.
[65] Roberts A, Robbins T, Weiskrantz L. Discussions and conclusions.
In: Roberts A, Robbins T, Weiskrantz L, editors. The prefrontal
cortex. Oxford: Oxford University Press, 1998.
[66] Roenker D, Thompson C, Brown S. Comparison of measures for
the estimation of clustering in free recall. Psychological Bulletin
1971;76:45–8.
[67] Savage C, Deckersbach T, Heckers S, et al. Prefrontal regions
supporting spontaneous and directed application of verbal learning
strategies: evidence from PET. Brain 2001;124:219–31.
[68] Schacter D. Memory, amnesia and frontal lobe dysfunction.
Psychobiology 1987;15:21–36.
 M. Brosnan et al. / Neuropsychologia 40 (2002) 2144–2155
[69] Schneider W, Pressley M. Memory development between 2 and 20.
New York: Springer, 1989.
[70] Shallice T. Specific impairments of planning. Philosophical
Transactions of the Royal Society of London 1982;298B:199–209.
[71] Shallice T. From neuropsychology to mental structure. Cambridge:
Cambridge University Press, 1988.
[72] Shaywitz S. Dyslexia. Scientific American 1996;275:98–104.
[73] Snodgrass J, Vanderwart M. A standardised set of 260 pictures:
norms for name agreement, form and visual complexity. Journal
of Experimental Psychology: Human Learning and Memory
1980;6:174–215.
[74] Snowling M, Nation K, Moxham P, Gallagher A, Frith U.
Phonological processing skills of dyslexic students in higher
education: a preliminary report. Journal of Research in Reading
1997;20:31–41.
[75] Stuss D, Benson D. The frontal lobes. New York: Raven, 1986.
[76] Swick D, Knight RT. Contributions of prefrontal cortex to
recognition memory: electrophysiological and behavioural evidence.
Neuropsychology 1999;13:155–70.
[77] Tarver SG, Hallahan DP, Kauffman JM, Ball DW. Verbal rehearsal
and selective attention in children with learning disabilities: a
developmental lag. Journal of Experimental Child Psychology
1976;22:375–85.
2155
[78] Thompson M. A comparison of laterality effects in dyslexia and
controls using verbal dichotic listening tasks. Neuropsychologia
1976;14:243–6.
[79] Torgesen J, Goldman T. Verbal rehearsal and short-term memory in
reading-disabled children. Child Development 1977;48:56–60.
[80] Warrington E, James M, Macjiewski C. The WAIS as a lateralising
and localising diagnostic instrument: a study of 656 patients with
unilateral cerebral lesions. Neuropsychologia 1986;24:223–39.
[81] Wechsler D. Wechsler Intelligence Scale for children. 3rd ed. Sidcup,
Kent: Psychological Corporation, 1976.
[82] Wechsler D. Wechsler Adult Intelligence Scale—revised. Sidcup,
Kent: Psychological Corporation, 1981.
[83] West T. In the mind’s eye—visual thinkers, gifted people with
dyslexia and other learning difficulties, computer images and the
irony of creativity. New York: Prometheus Books, 1997.
[84] Witkin H, Oltman P, Raskin E, Karps S. A manual for the group-
embedded figures test. Paulo Alto, CA: Consulting Psychologists
Press, 1971.
[85] World Federation of Neurology report of research group on dyslexia
and illiteracy. Dallas: WFN, 1968.
[86] Han S, Fan S, Chan L, Zhuo Y. Modulation of brain activities by
hierarchical processing: a high density ERP study. Brain Topography
1999;11:171–83.