Content uploaded by Rauno Parrila
Author content
All content in this area was uploaded by Rauno Parrila
Content may be subject to copyright.
Processing of Derived Forms in High-
Functioning Dyslexics
S. Hélène Deacon
1
, Rauno Parrila
2
, & John R. Kirby
3
1
Dalhousie University, Halifax, Nova Scotia, Canada
2
University of Alberta, Edmonton, Alberta, Canada
&
3
Queen’s University at Kingston, Ontario, Canada
We report on an experiment designed to evaluate processing of derived
forms in high-functioning dyslexics, defined as university students
with a history of reading difficulties who have age-appropriate reading
comprehension skills. We compared high-functioning dyslexics with a
group of normal adult readers in their performance on a lexical deci-
sion task with derived items (such as cloudy and ably) and pseudo-
derived items (such as belly and gravy). Some items contained an
orthographic change (such as able-ably and gravy) and others did
not (such as cloud - cloudy). The results indicated that although con-
trol participants’ response times varied systematically as a function of
morphological complexity, those of high-functioning dyslexics did not.
Further, there was some evidence of a relationship between deriva-
tional processing and reading. It seems that high-functioning dyslex-
ics have persistent difficulties in processing one particular aspect of
morphology; that of derived forms.
Key Words: Derivational processing, high-functioning
dyslexics, morphology, orthographic complexity,
reading difficulties
A long line of research has demonstrated that phonological
awareness is a key and causal factor in literacy development
Annals of Dyslexia, Vol. 56, No. 1, 2006
Copyright ©2006 by The International Dyslexia Association
®
ISSN 0736-9387
103
(Bradley & Bryant, 1983; National Reading Panel, 2000).
Certainly, children and adults with reading difficulties have
marked problems with the manipulation of sounds, or
phonemes, within words (for reviews, see Morais, Carey,
Alegria, & Bertelson, 1979; Snowling, 2000). These problems
persist in adults with a history of reading difficulties long after
they have managed to achieve some level of competence in
their reading (e.g., Pennington, Van Orden, Smith, Green, &
Haith, 1990; Snowling, Nation, Moxham, Gallagher, & Frith,
1997). It is estimated that 22% to 25% of children with a diagno-
sis of dyslexia will recover to the point that their adult reading
is well within the normal range (Lefly & Pennington, 1991). Of
course, there is wide variation in the degree of such compensa-
tion. The study reported here investigates “high-functioning
dyslexics,” individuals with a history of reading difficulties
who have achieved a level of reading comprehension that per-
mits their participation in postsecondary education. A key
question lies in how these individuals have been able to achieve
this relatively high degree of competence in reading compre-
hension despite well-established difficulties in phonological
skills known to be so critical to reading outcomes. One possibil-
ity is that high-functioning dyslexics have relatively strong
morphological processing skills that permit the compensation
for poor phonological processing skills (as suggested by Elbro &
Arnbak, 1996). The evidence to date on this question can be de-
scribed, at best, as mixed. The present study is designed to in-
vestigate the morphological processing of derived forms in
high-functioning dyslexics.
Morphemes are the smallest unit of meaning in language.
The word uncovered, for example, is built out of the prefix un-,
stem cover, and past tense suffix -ed. Several psycholinguistic
models of the lexicon have proposed that affixes are “stripped
off” prior to lexical access (e.g., Taft & Forster, 1975) resulting in
faster response times to real than to pseudo-derived words such
as unlike versus uncle, respectively (Rubin, Becker, & Freeman,
1979; Taft, 1981). As a reaction against this proposal,
Butterworth (1983) proposed a full-listing model in which mor-
phologically complex words are represented in their entirety in
the lexicon (e.g., black, board, and blackboard each have separate
representations). The majority of current models take a compro-
mise position, suggesting that there is a “race” between the two
processing (e.g., Marslen-Wilson, Tyler, Waksler, & Older, 1994;
Taft, 1994). While the precise architecture of morphological or-
ganization remains under investigation, there is good evidence
104
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 105
that morphemes play a role in lexical access in the typical adult
lexicon. Morphological effects in word recognition cannot be ac-
counted for by orthographic or phonological similarities be-
tween the items (e.g., Fowler, Napps, & Feldman, 1985; Napps
& Fowler, 1987; Stoltz & Feldman, 1995) or by semantic overlap
on its own (e.g., Bentin & Feldman, 1990; Feldman, 1992). We
will build on this history of lexical access tasks in the investiga-
tion of the consistency of derivational processing between high-
functioning dyslexics and normal adult readers. To place such a
study within its broader context, we will review morphological
processing as indexed by performance on a wide range of tasks
by poor readers of all ages.
MORPHOLOGICAL PROCESSING
IN YOUNG POOR READERS
We turn first to research that has contrasted the morphological
skills of good and poor reading abilities. Giraudo (2001) com-
pared priming effects from derived words (e.g., pottery - pot) to
those of pseudo-derived words (e.g., potato - pot). Although 8-
year-old normally developing children were significantly faster
at recognizing words primed by morphological relatives, no
such difference emerged in the dyslexic children. Similar evi-
dence of morphological difficulties emerges in a study of oral
and written production tasks. Rubin, Patterson, and Kantor
(1991) found that learning disabled 7-year-olds lagged behind
their peers in both oral and written morphological tasks.
Parallel results emerged in the assessment of the use of mor-
phology in word reading. Carlisle, Stone, and Katz (2001) found
that 10- to 15-year-old readers were less accurate in their nam-
ing and lexical decision responses for words with than without
changes in the phonology of the base form (e.g., natural and cul-
tural, respectively), and these transformations were particularly
problematic for the poor readers. Further evidence of impair-
ments comes from Leong’s (1989b) finding that poor readers in
Grades 4 to 6 were slower than their peers in reading words
presented according to morphological rather than phonological
divisions (e.g., ACTor versus ACtor, respectively) (Taft, 1979),
and in speed of production of base and derived forms to com-
plete sentences (based on Carlisle, 1988; see Leong & Parkinson,
1995 for similar results). Leong (1989a) further noted that out of
a large battery of measures, reaction times in the morphological
production task were the best predictors of reading and spelling
ability within the poor reader group. The generation of base
forms with orthographic changes (e.g., generate begin from be-
ginner) was particularly discriminative. Leong and Parkinson
(1995) reported that similar delays in processing of derived
forms occur in repetition priming tasks. These results suggest
that the morphological processing of poor readers, as assessed
across a wide range of tasks with both accuracy and response
time measures, lags at least behind that of their same age peers.
Certainly, morphology is not the only source of poor read-
ers’ deficits. Fowler and Liberman (1995) found that 7- to 9-
year-old poor readers had particular trouble with the
production of derived forms involving phonological changes
(e.g., heal - health) in comparison to their same age peers.
Similarly, Shankweiler et al. (1995) showed that poor readers
aged 7.5 to 9.5 years were impaired on both phonological and
morphological awareness tasks in comparison to chronological
age controls. It appears that young poor readers are delayed on
morphological and phonological tasks in comparison to their
same aged peers.
A number of researchers have advocated for the value of the
reading level match design in studies of potential causal factors
in dyslexia (e.g., Goswami & Bryant, 1989; Jackson &
Butterfield, 1989; Vellutino & Scanlon, 1989). Bourassa, Treiman,
and Kessler (in press) recently used this approach to show that
dyslexic children were as likely as their younger, normally de-
veloping peers to represent morphological information in their
spelling. Although it is widely agreed that results indicating
that poor readers lag behind reading age matches can be a reli-
able indicator of a potential causal variable in reading, the find-
ing of no differences between poor readers and reading level
matches (as in Bourassa et al., in press, among others) is far less
conclusive. Clearer evidence of difficulties comes from Joanisse,
Manis, Keating, and Seidenberg’s (2000) work, which showed
that dyslexic children were poorer on tasks of productive inflec-
tional morphology than controls with similar word identifica-
tion skills. However, when the groups were equated on both
reading and phonological skills, the performances were similar.
Joanisse et al. took this to reflect general language delays, in-
cluding those in morphology, among the poor readers (see
Casalis, Colé, & Sopo, 2004 for similar findings in French).
In contrast, Carlisle and Stone (2003) showed that poor
readers’ use of morphology in reading was comparable to both
chronological and reading age matches. All three groups
showed greater accuracy on derived than on pseudo-derived
106
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 107
words (e.g., comparing hilly and silly, respectively). Reaction
time results were more ambiguous, perhaps reflecting tradeoff
effects (based on Tyler & Nagy, 1990). Further, Bryant, Nunes,
and Bindman (1997) showed that although poor readers were
worse than age-matched controls in productive morphological
awareness, they had somewhat better morphological awareness
than reading age-matched controls. At the start of the longitudi-
nal study, the poor readers were weaker in phonetic spelling
and stronger in morphologically based spelling than a compari-
son group with similar initial reading abilities (Bryant, Nunes,
& Bindman, 1998). Bryant et al. described these results as indi-
cating that poor readers have a “linguistic strength as well as
linguistic weakness” (1998, p. 509).
MORPHOLOGICAL PROCESSING
IN ADOLESCENT POOR READERS
Shankweiler and colleagues’ research suggested that adolescent
poor readers have greater difficulties with morphological and
phonological tasks in comparison to normal readers. They
showed that a group of 16-year-old learning disabled adolescents
had poorer productive and written morphological skills than 14-
year-old normally developing readers (Shankweiler, Lundquist,
Dreyer, & Dickinson, 1996). A more precise reading ability match
was made by Carlisle (1987) in her comparison of 13-year-old
learning-disabled children with normally developing children
matched on spelling abilities. The learning-disabled children
were more likely to spell only one member of a base-derived pair
correctly (and not consistently the base), suggesting that they
were more likely to represent multimorphemic words as whole
words. Findings of greater difficulties by poor readers than read-
ing age matches indicate that morphological processing is a po-
tential causal variable in reading difficulties.
As in the research with children, there is some suggestion
that morphological processing is, in fact, an area of strength for
adolescent poor readers. Elbro and Arnbak (1996) showed that
dyslexic 15-year-old Danish students were faster at reading the
compound words than the control words (e.g., sunburn versus
window), but that there was no such difference for the 9-year-old
reading comprehension age matched children. Further, the de-
gree of increased speed of response time to the compounds was
correlated with better reading comprehension, suggesting that
morphologically based processing might offer a way for poor
readers to compensate for their phonological difficulties while
reading text. Further evidence of potential compensation comes
from Elbro and Arnbak’s (1996) second study in which dyslexic
adolescents were better than their younger peers in reading text
parsed into morphemes rather than into syllables. These find-
ings, like those of Bryant et al., suggest that morphology is an
area of relative strength that might permit poor readers to sur-
pass some of the hurdles posed by poor phonological skills.
MORPHOLOGICAL PROCESSING
IN ADULT POOR READERS
Rubin et al. (1991) found that adult dyslexics were similar to a
group of 7-year-old children (with whom they were not
matched on reading ability) in performance on morphological
spelling, judgment, and production tests. Fischer, Shankweiler,
and Liberman (1985) included a comparison group that was
matched on reading abilities. They found that in comparison to
good spellers of a similar age, poor adult spellers were insensi-
tive to the morphemic composition of words. In contrast, Bruck
(1993) showed that adult dyslexics had similar use of morphol-
ogy in spelling in comparison to a group of younger readers
matched on reading comprehension, spelling, and word recog-
nition. Leong (1999) found that adult dyslexics were slower
than chronological age controls on timed phonological and
morphological tasks, and they were similar in performance to
reading-ability matched controls. Thus, the vast majority of re-
search on adult readers shows that they are similar to younger
individuals with similar levels of reading in their use of mor-
phology across a wide range of tasks.
THE PRESENT STUDY
The present study is designed to investigate the processing of
derived forms in high-functioning dyslexics, defined here as
university students with a history of reading difficulties who
have age-appropriate reading comprehension skills. Typically,
high-functioning dyslexics demonstrate serious word level and
reading rate difficulties, in addition to phonological processing
difficulties, while achieving age-appropriate levels of reading
comprehension specifically when tested in an untimed manner.
Thus, it is commonly reported that they take longer than nor-
108
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 109
mally achieving students to read material, and the most com-
mon accommodation offered to them is longer time to complete
assignments and tests.
The question of the most appropriate comparison group for
high-functioning dyslexics is a difficult one. One approach in-
volves finding younger individuals with similar word level or
timed reading comprehension abilities, as in the reading level
match design that is ideal for controlling reading experience
toward the establishment of causal factors (e.g., Goswami &
Bryant, 1989). Although this would achieve similarity on some
reading ability dimensions, the groups would be substantially
different in many other respects. For example, dyslexics in uni-
versity and much younger readers would likely show a wide
gap in reading, educational, and world experience. Differences
in reading experience would be particularly problematic, as this
is the key control that the reading level match design is meant
to accomplish. Further, even if it is possible to find groups with
similar levels of print exposure in terms of the amount of text
read, a vital assumption for the reading-level match designs, it
is unlikely that the complexity of the texts read would be com-
parable given that the high-functioning dyslexic group has
made it to postsecondary education, and the control groups
(from elementary or secondary schools) have not yet done so.
Clearly, results from such comparisons could be misleading and
they would not help in identifying strengths, which is a key
goal of research with adult dyslexic samples. Another approach
is to contrast high-functioning dyslexics with a group of normal
adult readers with similar levels of untimed reading compre-
hension. While such a comparison results in dyslexic and con-
trol groups differing in word reading skills (and possibly in
timed reading comprehension skills), it ensures a more similar
reading (at least in terms of text complexity) and world experi-
ence than using a younger word reading level control group.
The present study adopted the second approach.
In addition, we focused specifically on processing of de-
rived forms as assessed with a lexical decision task (Taft, 1979).
In this task, participants decide if a letter string is a word and
respond accordingly with a button press, obviating the need for
a spoken response. Rubin et al. (1979) found that participants
responded faster to prefixed words than to pseudo-prefixed.
These results parallel Carlisle and Stone’s (2003) findings of
greater accuracy in reading derived than pseudo-derived
words. Such comparisons offer a control for possible word
length effects. To date, few studies have included equally com-
plex control words (Elbro & Arnbak, 1996; Giraudo, 2001; Rubin
et al., 1991, are noteworthy exceptions), and poor readers might
have difficulties with longer words in general rather than with
morphologically complex words in particular (see e.g., Ben-
Dror, Pollatsek, & Scarpati, 1991; De Luca, Borrelli, Judica,
Spinelli, & Zoccolotti, 2002; Nation, Marshall, & Snowling,
2001). Longer words are often less frequent and require sublexi-
cal processing, which might increase task demands that may af-
fect dyslexic readers more than nondyslexic readers (e.g.,
Parrila & Georgiou, 2005; van der Leij & van Daal, 1999; see also
Zoccolotti et al., 2005, for findings of word length effects for
both words and pseudowords). Lexical decision tasks also per-
mit the use of a reaction time measure (the sensitivity of which
has been advocated by many; e.g., Seidenderg, 1985; and em-
ployed by even more; e.g., Leong, 1999), which may be particu-
larly sensitive. We will use this comparison to address the status
of processing of derived forms in high-functioning dyslexics.
Based on Taft (1979) and others, we expected faster process-
ing of derived than pseudo-derived forms by normal readers
when the base is both phonologically and orthographically
transparent. If morphological processing offers a compensatory
reading avenue into high-functioning dyslexics (as suggested
by Elbro & Arnbak’s 1996 study of compound word reading),
then we would expect an even greater effect of derivational
structure with the dyslexics than with the controls. Such a find-
ing would suggest that derivational processing skills might per-
mit an avenue of compensation in reading comprehension to
surpass phonological difficulties. A finding of a difference be-
tween processing of derived and control words in the normals,
but not in the high-functioning dyslexics would suggest that,
just like phonological processing, derivational processing is an
area of weakness. The finding of similarity in derivational pro-
cessing in the two groups would be ambiguous.
We were also interested in the role of orthographic complex-
ity in derivational processing. It is clear that morphological pro-
cessing operates independently of orthographic and
phonological overlap (Fowler et al., 1985); specifically, even
with words that share the same initial letters and sounds, those
for which this initial section is a morpheme, offer greater prim-
ing than those that do not (e.g., marking and market). However,
in words in which there is a change in the orthographic repre-
sentation (such as those investigated by Leong, 1989a) between
the base and the derived forms, morphological structure might,
in fact, slow processing. Tyler and Nagy (1990) suggested that
110
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 111
the reaction time gains accrued from the initial identification of
morphological units might be offset by increased time to pro-
cess these units. Orthographic changes between the derived and
base form are likely to bring about such tradeoffs (as have been
found with phonological changes; e.g., Carlisle & Stone, 2003,
2005). We will test whether this is the case in normal readers
and compare their performance to that of high-functioning
dyslexics. The status of orthographic knowledge is a particu-
larly interesting one in relation to dyslexic samples as it has
been argued that it can be particularly strong in this group
(Siegel, Share, & Geva, 1995). Further, as the results of Leong
(1989a) suggested, derived forms involving orthographic
changes might be particularly indicative of reading difficulties.
The measure of derived processing will include items with or-
thographic changes to permit such an investigation.
Finally, we were interested in whether processing of derived
forms is related to reading ability in general. Elbro and Arnbak
(1996) found that morphological processing was correlated with
reading comprehension in their poor readers. They took this as
evidence of morphological processing as a compensatory strat-
egy in reading. We used the same approach to offer a second
way to determine the role that processing of derived forms
plays in the relative level of reading achievement in high-
functioning adult dyslexics.
METHOD
PARTICIPANTS
The high-functioning dyslexic group consisted of 27 partici-
pants (nine males and 18 females) who reported a significant
history of reading difficulties and whose performance on the el-
ementary education section of the modified Adult Reading
History Questionnaire (Parrila, Corkett, Kirby, & Hein, 2003) in-
dicated reading acquisition difficulties in childhood. Nine of
them had recent diagnoses and had received or were receiving
services for learning-disabled students. The average age of the
high-functioning dyslexic group was 30.08 (SD = 9.30; range 18
to 52). The participants in the high-functioning dyslexic group
were all either current university students or recent graduates
(less than six months at the time of initial testing) and were re-
cruited through letters sent by the university’s Student Support
Services, announcements in undergraduate classes, and posters
displayed throughout the university’s campus.
The control group consisted of 28 participants (nine males
and 19 females) who reported no history of reading problems.
The average age of the ND participants was 24.93 (SD = 6.31;
range 18 to 49). The five-year difference in the mean ages of
the groups resulted mainly from the dyslexic participants tak-
ing a longer time after high school before enrolling in the uni-
versity. The control group was recruited through
announcements in undergraduate classes and through posters
displayed throughout the university’s campus. All were cur-
rent university students.
All participants reported English as their spoken language
of preference and normal or corrected to normal vision. As part
of the testing, the participants completed a battery of reading
and phonological processing tests. The reading tests consisted
of computerized versions of Word Identification and Word
Attack subtests from the Woodcock Reading Mastery
Test–Revised (Woodcock, 1987) and of the Comprehension sub-
test from the Nelson-Denny Reading Test (Brown, Fishco, &
Hanna, 1993). The latter test provided two indices: reading rate
and reading comprehension with a 20-minute time limit.
Because of the time limit, many dyslexic participants did not
answer all the questions. To obtain an estimate of untimed read-
ing comprehension ability, we also calculated the percentage
correct of all attempted questions. The phonological processing
tests were the Rosner Auditory Analysis Test (Rosner & Simon,
1971), pseudohomophone choice (the participants had to choose
between two nonwords the one that sounded like a real word),
and rapid automatized naming (RAN) of digits (six digits—2, 7,
4, 5, 3, and 8—presented five times each in a semirandom order
in a 4 x 9 matrix for the total of 36 stimuli) adapted from the
Comprehensive Test of Phonological Processing (Wagner,
Torgesen, & Rashotte, 1999).
First, the performance of the high-functioning dyslexic par-
ticipants with a recent diagnosis were compared to those with-
out a recent diagnosis. The two groups did not differ in any of
the measures (all ps > .171). Next, MANOVAs with the raw
scores of the five reading variables, Wilks’ =.522, F(5, 49) =
8.99, p < .001, or the four phonological processing variables,
Wilks’ =.459, F(4, 50) = 14.71, p < .001, indicated that high-
functioning dyslexic group was significantly different from the
control in their reading and phonological processing skills.
Table I reports the means, standard deviations, and F-values
from subsequent ANOVAs comparing the groups on reading
and phonological processing tests. The two groups were signif-
112
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 113
icantly different in both Word Identification and Word Attack;
the high-functioning dyslexic group’s means reported in table I
correspond roughly to grade equivalencies of 11 for Word
Identification and seven for Word Attack. We should note,
however, that these tests were not presented in the standard-
ized format.
The comprehension subtest of the Nelson-Denny Reading
Test was presented following the standard procedures and ac-
cording to the norms, a reading rate scale score of 197 for the
high-functioning dyslexic group represent an average level of
performance in Grade 9 whereas their comprehension scale
score of 221 represents an average level of performance dur-
ing the second year of university. In terms of end of the fourth
Table I. Means (and Standard Deviations) of Performance on the Reading
and Phonological Processing Tests. F-values Report the Main Effect of
Group.
Group
Control High-functioning Dyslexics
(n = 28) (n = 27)
Tests M M F
Reading
WRMT Word
Identification 102.18 (2.86) 94.85 (6.22) 31.91***
WRMT Word
Attack 38.61 (2.55) 32.96 (6.64) 17.53***
ND Rate 289.61 (94.80) 213.20 (98.62) 8.26**
ND Timed
Comprehension 65.36 (8.42) 57.46 (12.58) 7.45**
ND Untimed
Comprehension 89.68 (8.97) 86.00 (10.51) 1.95
Phonological Processing
AAT 36.54 (2.06) 30.48 (8.29) 14.06***
PC Accuracy 18.14 (1.76) 16.63 (3.56) 4.03*
PC Response
Time 2.16 (0.56) 4.11 (1.49) 41.82***
RAN Digits 13.15 (3.14) 15.96 13.21**
Note. WRMT = Woodcock Reading Mastery Test; ND = Nelson-Denny;
AAT = Rosner Auditory Analysis Test; PC = Pseudohomophone Choice; RAN
= Rapid Automatized Naming.
* p < .05; ** p < .01; *** p < .001
year university norms, the control group’s reading rate corre-
sponds to the 73rd percentile and comprehension to the 53rd
percentile, whereas the respective percentiles for the high-
functioning dyslexic group are 27th and 29th. F values re-
ported in table I show that the two groups were significantly
different in both reading rate and comprehension, but not in
the proportion correct measure that was calculated to esti-
mate untimed reading comprehension and to adjust for the
fact that many dyslexic participants did not finish the task in
20 minutes. These results also suggest that our dyslexic par-
ticipants continued to read text at a slower speed than
nondyslexic university students, but they can understand text
at an age appropriate level, particularly if given adequate
time. Further, it is important to note that timed comprehen-
sion at the second year university level is age-appropriate by
itself. In general, these results are similar to those of earlier
studies showing that university students with reading dis-
abilities can perform similarly to normal readers on reading
comprehension tasks when the time constraints are removed
(e.g., Mosberg & Johns, 1994), but they continue to struggle
on decoding, word reading, reading speed, and phonological
processing tasks (e.g., Aaron, 1989; Gallagher, Laxon,
Armstrong, & Frith, 1996; Hatcher, Snowling, & Griffiths,
2002; Rack, 1997).
MATERIALS
The derived items were made up of a base and a suffix (e.g.,
read + er = reader). The pseudo-derived items were made up of a
smaller word followed by a final letter-sequence that repre-
sented a plausible morpheme (e.g., off + er = offer). In the
pseudo-derived items, neither component was morphologically
related. For the No Change derived items, there was no change
in the sound, stress pattern, or orthographic representation be-
tween the stem and the derived form (e.g., read - reader).
Similarly, with the No Change pseudo-derived items, there
were no orthographic or phonological changes between the
smaller segment and the one-morpheme word (e.g., off - offer).
For the Orthographic Change items, the smaller initial section
of the words was spelled differently in the smaller and longer
words. This smaller word was the base morpheme for the de-
rived items, but it was not for the pseudo-derived items (e.g.,
able - ably and grave - gravy, respectively). As in Carlisle (1987,
1988), these items included both changes in consonant doubling
and in final letter spelling (e.g., fog - foggy and able - ably,
114
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 115
respectively)
1
. There were eight items in each condition and
these are listed in the Appendix.
The frequency of the items and of the smaller word within
the items (e.g., read in reader and off in offer) was balanced for
Kucera-Francis written frequency, obtained through the MRC
psycholinguistic database (ps > .76) (Coltheart, 1981). Words are
also similar in the number of letters and syllables (ps > .22). The
items are listed in the Appendix.
PROCEDURE
The target items were embedded among 134 words and 104
nonwords (88 pronounceable pseudowords and 16 letter
strings) in a standard lexical decision task in which the partici-
pants were asked to press a button on a keyboard (Right
Control for YES and Left Control for NO; buttons labeled ac-
cordingly) to indicate if a letter-string was a word or not. The
words and nonwords were matched on length, which varied be-
tween four and 12 letters. They were presented in large black
type (Times New Roman 42) in random order on the center of a
computer screen against a white background. The participants
first saw a fixation marker (a cross-hair) in the center of the
screen, followed by a word after 1000ms. The experiment gener-
ation program (DirectRT v2004; Empirisoft, 1999-2004) recorded
both the reaction time and the accuracy of the response. The
participants were offered a short break after the first 125 items.
RESULTS
ANALYSIS OF LEXICAL DECISION TASK
Given very high accuracies and the ceiling effect in one cell (as
shown in table II), accuracy data were not analysed further.
Mean accuracy for any one item did not drop below .52, with
the majority above .90. Reliability analyses of the eight items
within each condition show reasonable reliability (alphas > .7).
Response times to correctly responded items in each condi-
tion by each participant group are shown in table II. Data were
cleaned so that response times faster than 200 ms or slower than
1. There were two Orthographic Change items that included slight phonologi-
cal changes (center - central; able - ably). These were included because of the dif-
ficulty in locating items that met all of the criteria for balancing. The high
levels of reliability (reported in the Results section) suggests that there were
similarities in processing across the set of Orthographic Change items.
3000 ms were removed, as were outliers (defined as response
times more than two standard deviations from the mean re-
sponse time for a given individual in that condition). Data were
analysed with a repeated-measures ANOVA with Change Type
(No Change or Orthographic Change), Word Type (Derived or
Pseudo-derived), and Group (Normal or High-functioning
Dyslexic) as factors in the analyses by subjects. Results are re-
ported for analyses by subjects (F1) and by items (F2).
According to Raaijmakers, Schrijnemakers, and Gremmen
(1999), in experiments with highly selected items such as this
one, only significance in the analyses by subjects is required to
reject the null hypothesis.
Analyses suggest that there are differences between the two
groups of participants. There was a trend for faster response
times for the control group than for the dyslexic group, F1(1, 53)
= 3.417, p = .070; F2(1, 28) = 32.826, p < .001.
Most critically for the questions at hand, it appears that
there are differences between the groups in their derivational
116
DEACON, PARRILA, AND KIRBY
Table II. Mean Response Time and Standard Deviation (in parentheses) to
Respond to the Lexical Decision Task.
High-functioning
Word Type Control Dyslexics
Response time to correct
No Change
Derived 718 (171) 838 (218)
Pseudo-derived 793 (177) 851 (261)
Derivational Sensitivity .900 (.198) .985 (.187)
Orthographic Change
Derived 899 (261) 927 (291)
Pseudo-derived 705 (128) 885 (284)
Derivational Sensitivity 1.28 (.296) 1.047 (.197)
Accuracy
No Change
Derived 1.00 (.00) .83 (.12)
Pseudo-derived .98 (.04) .97 (.08)
Orthographic Change
Derived .92 (.10) .98 (.06)
Pseudo-derived .99 (.03) .94 (.09)
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 117
processing. A main effect of Change Type (F1[1, 53] = 8.307, p =
.006; F2[1, 28] = 2.283, p = .142), and significant interactions be-
tween Word Type and Change Type (F1[1, 53] = 18.011, p < .001;
F2[1, 28] = 5.716, p = .024) and between Word Type, Change
Type, and Group (F1[1, 53] = 7.947, p = .007; F2[1, 28] = 6.592,
p = .016).
Analyses of the two-way interaction showed that responses
were faster for the derived than for the pseudo-derived items in
the No Change condition (t1[54] = 2.453, p = .017; F2[1, 14] =
0.829, p = .378) at a level that was slightly above the Bonferroni
corrected p-value (p < .0125). In contrast, response times were
slower for the derived than for the pseudo-derived items in the
Orthographic Change condition (t1[54] = 2.989, p = .004; F2[1,
14] = 4.705, p = .048). These results indicate that the effect of
morphological structure on the processing of derived forms
changed as a function of the type of change between the base
and derived form: speeding processing of items without a
change in the orthographic representation of the base and slow-
ing processing of items with this change. The effects of ortho-
graphic structure also varied across derived and
pseudo-derived conditions, showing a significant slowing of re-
action times for the derived items (t1[54] = 3.743, p < .001; F2[1,
14] = 4.165, p = .061), but not for the pseudo-derived items
(t1[54] = 1.261, n.s.; F2[1, 14] = 0.967, n.s.).
Analyses of the three-way interaction showed that for the
control group, there were main effects of both Word Type (F1[1,
27] = 4.832, p = .037; F2[1, 28] = 2.229, p = .147) and Change Type
(F1[1, 27] = 4.344, p = .047; F2[1, 28] = 1.642, p = .221), indicating
that response times were slower for Orthographic than for No
Change items, and for pseudo-derived than for derived items.
These main effects were qualified by an interaction between
Word Type and Change Type (F1[1, 27] = 26.728, p < .001; F2[1,
28] = 9.382, p = .005). Follow-up paired-sample t-tests of the in-
teraction focused on the effects of derivational structure within
the No Change and Orthographic Change items, respectively, as
this was the primary focus of the experiment. Control partici-
pants were faster with the derived than with the pseudo-
derived items (t1[27] = 2.579, p = .016; F2[1, 14] = 2.090, p = .170)
for the No change items. In contrast, responses were faster to
the pseudo-derived than to the derived items for the
Orthographic Change items (t1[27] = 3.992, p < .001; F2[1, 14] =
7.359, p = .017). If the interaction is examined according to the
effects of orthographic structure, we see that there are effects of
orthographic change far more robust for the processing of
derived (t1[27] = 4.535, p < .001; F2[1, 14] = 6.337, p = .025) than
for the pseudo-derived items (t1[27] = 2.233, p = .034; F2[1, 14] =
1.712, p = .212). The critical finding here is that the control
group’s response times were clearly affected by the derivational
structure of the presented items.
In contrast, for the high-functioning dyslexic group, only
one effect approached significance; response times tended to be
shorter for the No Change items than for the Orthographic
Change items (F1[1, 26] = 4.118, p = .053; F2[1, 28] = 2.482, p =
.126). This effect also appeared for the control group, indicating
that both groups found the words involving orthographic
changes difficult to process (no interaction between Change
Type and Group, p > .25). However, unlike the control group,
the lack of an interaction with Word Type indicates that this ef-
fect held for both the derived and pseudo-derived words for the
high-functioning dyslexic group.
In sum, the control group participants benefited from mor-
phological complexity when the relationship between the base
and derived form was transparent as in No Change condition.
When the relationship was less transparent, morphological
complexity resulted in slower response times. High-function-
ing dyslexics, in contrast, showed no effects of morphological
complexity.
DERIVATIONAL PROCESSING AND READING
To assess the relationship between derivational processing and
reading, we first calculated two indices of derivational sensitiv-
ity for each participant (based on Elbro & Arnbak, 1996). These
indices were calculated with the reaction time to correctly iden-
tify items. We divided this reaction time for the items in the de-
rived condition by that for the pseudo-derived condition items
for each of the No Change and Orthographic Change condi-
tions, respectively (resulting in two indices reported in table II).
These offer an index of how much the individual was speeded
(or slowed, as the case may be) by added morphological com-
plexity. The calculation of these indices for each individual al-
lows the evaluation of correlations across the two sample
groups combined together, alleviating concerns regarding re-
duced sample sizes and power associated with the splitting of
groups.
We calculated correlations between these two measures of
derivational sensitivity and the reading measures (reported in
table III). If derivational processing is linked to reading, then
there should be relationships between the derivational sensitiv-
118
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 119
ity measures and reading. Positive correlations indicate that bet-
ter readers read derived words faster than pseudo-derived
items. We found that derivational sensitivity to No Change
forms was significantly related to word identification and
derivational sensitivity to Orthographic Change forms was re-
lated to timed reading comprehension at a significant level and
to word identification at a level approaching significance (p =
.061). The relationships between processing of the No Change
forms and reading variables are negative, reflecting the fact that
processing is speeded in the derived as compared to the
pseudo-derived condition, while those between Orthographic
Change forms and reading are positive, reflecting the slowing
of processing in the derived condition.
We also calculated these correlations separately for both
samples in the study (reported in table IV). These within-group
correlations show that only the relationship between deriva-
tional sensitivity to Orthographic Change forms and timed
reading comprehension remained at a level approaching signifi-
cance in the control group (p = .057). There were no such trends
in the high-functioning dyslexics.
DISCUSSION
The current experiment was designed to investigate deriva-
tional processing in high-functioning dyslexic adults. We com-
pared individuals with a history of reading difficulties who
Table III. Correlations between Indices of Derivational Sensitivity (DS) for
the No Change and Orthographic Change Conditions and Reading
Measures.
1. 2. 3. 4. 5.
1. D.S. No change
2. D.S. Orth. Change -.149
3. Reading rate .026 .185
4. Untimed Reading
Comp. .007 .248 .139
5. Timed Reading
Comp. -.076 .339* .421** .676***
6. Word
Identification -.295* .254† .238† .194 .472***
Note. † indicates p < .10, * p < .05, ** p < .01, *** p < .001.
were participating in postsecondary education with normal
readers. High-functioning dyslexics were poorer than the con-
trol group on a range of phonological processing and reading
tasks, although they had comparable levels of untimed reading
comprehension. We found striking differences in response time
patterns to morphologically complex words between normal
readers and high-functioning dyslexics. While the control
group’s response times varied according to the morphological
complexity of the words that they encountered, those of the
high-functioning dyslexics did not.
The results added to the existing evidence for a morphologi-
cally mediated route for the identification of derived words in
the lexicon in normal readers as suggested in several models of
lexical access. It seems that for the normal readers, morphologi-
cal structure speeds the recognition of derived words with no
change in orthographic structure from base to derived form.
This finding replicates previous lexical decision results (e.g.,
Taft, 1981). In contrast, when the words involved orthographic
changes (such as between able and ably), responses were slower
to derived than to pseudo-derived words. The effects of the
change in orthographic structure were most robust for the de-
rived items, both in the analysis of two groups together and in
120
DEACON, PARRILA, AND KIRBY
Table IV. Correlations between Indices of Derivational Sensitivity (DS) for
the No Change and Orthographic Change Conditions and Reading
Measures for the Control Group Below the Diagonal and for the High-
Functioning Dyslexics Above.
1. 2. 3. 4. 5. 6.
1. D.S. No change -.118 .290 .028 .053 -.236
2. D.S. Orth.
Change -.025 .174 .047 .143 -.049
3. Reading
rate -.027 -.020 .113 .341† -.093
4. Untimed
Reading
Comp. .087 .317 .075 .609** .134
5. Timed
Reading
Comp. -.049 .364† . 349† .767*** .358†
6. Word
Identification -.133 .125 .241 .037 .326†
Note. † indicates p < .10, * p < .05, ** p < .01, *** p < .001.
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 121
the analysis of the normal readers’ data on its own, suggesting
that the slowing effects of orthographic changes are relatively
specific to derivational access processes. This type of detrimen-
tal effect of morphological complexity has been found in other
investigations, as in the time-consuming recombination of
words with high frequency bases (Taft, 2004) and the rejection
of nonwords with morphologically plausible components
(Caramazza, Laudanna, & Romani, 1992). To our knowledge,
the “costs” associated with recognizing multimorphemic words
with complex orthographic structures have not been previously
identified. This finding is consistent with Tyler and Nagy’s
(1990) proposal of a timing tradeoff between access to and pro-
cessing of morphemes within some morphologically complex
words, as well as Taft and Forster’s (1975) suggestion that the
processing of certain types of morphologically complex words
can be effortful and time-consuming.
In contrast to this sensitivity of normal readers to deriva-
tional structure, the high-functioning dyslexics in this study did
not demonstrate any such appreciation. Response times did not
differ between the pseudo-derived and derived words in either
the No Change or the Orthographic Change condition. This null
result resonates with findings from Giraudo (2001), suggesting
that individuals with reading difficulties do not demonstrate
sensitivity to morphological structure in time-sensitive tasks.
Findings from this experiment extend these results to a new
group, that of high-functioning dyslexics, at least within the do-
main of derivational processing. It may be that this group had
early difficulties in accessing the morphemic structure of de-
rived forms and that remnants of these difficulties remain, as do
their phonological difficulties. At first glance, the finding that
response times for this group were not slowed by the ortho-
graphic complexity of the base forms of the derived words with
orthographic changes might appear to be evidence of a relative
morphological or orthographic strength. However, the
Orthographic Change derivational sensitivity index was posi-
tively related to reading abilities in the analyses across all par-
ticipants and in the normal adult reading group alone, showing
that this slowing of response time by orthographic complexity
was related to efficient decoding and comprehension of text.
These results do not support the existence of the orthographic
strength suggested by Siegel et al. (1995). Further, this link to
reading is reminiscent of Leong (1989a), who found that items
with orthographic changes were particularly indicative of read-
ing and spelling difficulties. Although these findings clearly
need to be replicated with a wider range of items, the differ-
ences in the processing of derived words uncovered in this ex-
periment have implications for the interpretation of previous
studies indicating that high-functioning dyslexics read words in
a qualitatively similar manner to that of normally developing
readers (e.g., Ben-Dror et al., 1991; Watson & Brown, 1992). Our
results suggest that, when faced with derived words, high-
functioning dyslexics may approach words differently than nor-
mal readers (see also Leong, 1999).
The finding that sensitivity to the structure of derived
words, as assessed in the lexical decision task, is related to read-
ing performance is worth some consideration. In the analyses
with all participants, the derivational sensitivity index in the No
Change condition was significantly related to word identifica-
tion, and there was a trend toward such a relationship with the
Orthographic Change derivational sensitivity index. Sensitivity
to derived forms with orthographic changes was also related to
timed reading comprehension and there was an appropriate
trend in the analysis of the normal sample on its own. These re-
sults resonate with those of Leong (1989a), and they provide
some evidence that derivational processing is a component of
accurate decoding of words and efficient comprehension of text
in normal adult readers. Clearly, there was no such relationship
in the high-functioning dyslexics’ data, suggesting that a poten-
tial normal relationship between derivational processing and
reading was not intact in this group. Although lexical decision
tasks have been used extensively to investigate a wide range of
morphological processing in adults, we are not aware of any
published reports that have attempted to relate such measures
to reading ability in adults. While this study provides some first
evidence for a relationship, it requires further investigation, in-
cluding a more thorough coverage of morphological processing
and reading tasks. Future studies could investigate possible in-
teractions between morphological processing and other factors
(e.g., phonological complexity and word length) that have
shown to be important.
One of the hypotheses that we wanted to explore was
whether morphology offered a compensatory mechanism
through which high-functioning dyslexics might have achieved
some degree of reading comprehension, albeit slow. This was a
plausible hypothesis, given the results of prior research (e.g.,
Elbro & Arnbak, 1996). Our results indicate that unlike normal
adult readers, high-functioning dyslexics do not demonstrate
sensitivity to derivational morphology within this timed word
122
DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 123
recognition task. These findings emerged in a comparison of
two groups with similar levels of untimed reading comprehen-
sion and these results clearly need to be replicated with a group
with similar word reading skills. Nevertheless, the present re-
sults make it unlikely that this derivational processing is an area
through which these individuals might have compensated for
reading difficulties, at least not in a manner that permits them
to increase their reading speed. The majority of earlier research
employed untimed production tasks, either oral or written. It is
possible that children and adults with reading difficulties can
use morphology to generate words and spellings in tasks for
which there is no time limit, but they cannot do so in an effi-
cient manner. Response times of the high-functioning dyslexics
were slower than those of controls across all conditions in the
lexical decision task, suggesting that this task is more cogni-
tively demanding for this group than for control participants.
Increased task demands might have interrupted the normal re-
lationship between derivational processing and reading, a link
for which we found some emerging evidence. The comparison
of timed and untimed tasks in new studies would permit the in-
vestigation of this possibility. It is also possible that it is specifi-
cally the semantic aspect of morphology that permits poor
readers to use this strategy. Compound words, such as those in-
vestigated in Elbro and Arnbak (1996), have a high level of se-
mantic transparency, and poor readers might be able to use a
compensatory morphological strategy with forms with salient
meanings (as suggested by Elbro, 1990). Further research needs
to explore the extent of morphological abilities in tasks with a
wide range of morphemes, as well with varying response de-
mands and semantic salience.
Further research should also investigate the performance of
high-functioning dyslexics in comparison to different types of
control groups. In this experiment, we contrasted two groups of
adult readers with similar levels of untimed reading compre-
hension. This is a potentially useful approach in the identifica-
tion of strengths of dyslexic readers. However, there was a
range of differences between these two groups, including per-
formance on phonological and word reading measures. With
adult dyslexics, it may be impossible to find a true control
group, one that differs in only one feature from the target
group. The only pragmatic solution may be to include different
types of comparisons in separate studies to control for various
critical features in separate studies. For example, high-functioning
dyslexics could be compared to much younger individuals with
similar levels of phonological and single-word reading abilities.
Although these groups might not have comparable reading ex-
perience levels (given the wide difference in age of the partici-
pants), they would be similar in underlying processing skills.
This experiment set out to evaluate derivational processing
in high-functioning dyslexics, a question that has not been ad-
dressed to date in experimental research. We did not find evi-
dence of sensitivity to derivational morphology in a timed
lexical decision task within this specific adult sample; the nor-
mal readers’ response times to derived and pseudo-derived
words varied systematically as a function of morphological
complexity, but those of high-functioning dyslexics did not.
Further, there was evidence of a relationship between deriva-
tional processing and reading, a finding that requires further ex-
ploration. There was some indication that this link existed
primarily for the normal readers. In summary, high-functioning
dyslexics may have persistent difficulties in derivational pro-
cessing, especially in timed tasks. Future research needs to ex-
plore the interaction between linguistic processing abilities in a
range of domains, including morphological and phonological,
in this group.
Address correspondence to: S. Hélène Deacon, Department
of Psychology, LSC, Dalhousie University, 1355 Oxford Street,
Halifax, Nova Scotia, Canada, B3H 4J1. Phone: (902) 494-2538;
Fax: (902) 494-6585; E-mail: sdeacon@dal.ca
References
Aaron, P. G. (1989). Qualitative and quantitative differences among dyslexic, normal,
and nondyslexic poor readers. Reading & Writing: An Interdisciplinary Journal, 1,
291–308.
Ben-Dror, I., Pollatsek, A., & Scarpati, S. (1991). Word identification in isolation and in
context by college dyslexic students. Brain and Language, 40, 471–490.
Bentin, S., & Feldman, L. B. (1990). The contribution of morphological and semantic re-
latedness to repetition priming at short and long lags: Evidence from Hebrew.
Quarterly Journal of Experimental Psychology, 42, 693–711.
Bourassa, D., Treiman, R., & Kessler, B. (in press). Use of morphology in spelling by
children with dyslexia and typically developing children. Memory and Cognition.
Bradley, L., & Bryant, P. E. (1983). Categorising sounds and learning to read—a causal
connection. Nature, 301, 419–420.
Brown, J. I., Fishco, V. V., & Hanna, G. (1993). The Nelson-Denny reading test. Chicago:
Riverside Publishing Company.
Bruck, M. (1993). Component spelling skills of college students with a childhood diag-
nosis of dyslexia. Learning Disabilities Quarterly, 16, 171–184.
124 DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 125
Bryant, P., Nunes, T., & Bindman, M. (1997). Backward readers’ awareness of language:
Strengths and weaknesses. European Journal of Psychology of Education, XII,
357–372.
Bryant, P., Nunes, T., & Bindman, M. (1998). Awareness of language in children who
have reading difficulties: Historical comparisons in a longitudinal study. Journal
of Child Psychology and Psychiatry, 39, 501–510.
Butterworth, B. (1983). Lexical representation. In B. Butterworth (Ed.), Language produc-
tion, Vol. 2 (pp. 257–294). London: Academic Press.
Caramazza, A., Laudanna, A., & Romani, C. (1992). Lexical access and inflectional mor-
phology. Cognition, 28, 297–332.
Carlisle, J. F. (1987). The use of morphological knowledge in spelling derived forms by
learning disabled and normal students. Annals of Dyslexia, 37, 90–108.
Carlisle, J. F. (1988). Knowledge of derivational morphology and spelling ability in
fourth, sixth and eighth grades. Applied Psycholinguistics, 9, 247–266.
Carlisle, J. F., & Stone, C. A. (2003). The effects of morphological structure on children’s
reading of derived words. In E. Assink & D. Santa (Eds.), Reading complex words:
Cross-language studies (pp. 27–52). Dordrecht: Kluwer Academic Publishers.
Carlisle, J. F., & Stone, C. A. (2005). The role of morphemes in word reading. Reading
Research Quarterly, 40, 428–449.
Carlisle, J. F., Stone, C. A., & Katz, L. A. (2001). The effect of phonological transparency
on reading derived words. Annals of Dyslexia, 51, 249–274.
Casalis, S., Colé, P., & Sopo, D. (2004). Morphological awareness in developmental
dyslexia. Annals of Dyslexia, 54, 114–138.
Coltheart, M. (1981). The MRC psycholinguistic database. Quarterly Journal of
Experimental Psychology, 33A, 497–505.
De Luca, M., Borrelli, M., Judica, A., Spinelli, D., & Zoccolotti, P. (2002). Reading words
and pseudowords: An eye movement study of developmental dyslexia. Brain and
Language, 80, 617–626.
Elbro, C. (1990). Differences in dyslexia. A study of reading strategies and deficits in a linguis-
tic perspective. Copenhagen: Munksgaard.
Elbro, C., & Arnbak, E. (1996). The role of morpheme recognition and morphological
awareness in dyslexia. Annals of Dyslexia, 46, 209–240.
Feldman, L. B. (1992). Morphological relationships revealed through the repetition
priming task. In P. Downing, S. D. Lima, & M. Noonan (Eds.), Linguistics and lit-
eracy (pp. 239–254). Amsterdam/Philadelphia: John Benjamins.
Fischer, F. W., Shankweiler, D., & Liberman, I. Y. (1985). Spelling proficiency and sensi-
tivity to word structure. Journal of Memory and Language, 24, 423–441.
Fowler, A., & Liberman, I. (1995). The role of phonology and orthography in morpho-
logical awareness. In L. B. Feldman (Ed.), Morphological aspects of language process-
ing (pp. 157–188). Hillsdale, NJ: Lawrence Erlbaum.
Fowler, A., Napps, S. E., & Feldman, L. B. (1985). Relations among regular and irregular
morphological related words in the lexicon as revealed by repetition priming.
Memory and Cognition, 13, 241–255.
Gallagher, A. M., Laxon, V., Armstrong, E., & Frith, U. (1996). Phonological difficulties
in high-functioning dyslexics. Reading and Writing: An Interdisciplinary Journal, 8,
499–509.
Giraudo, H. (2001). Rôle et représentation de l’information morphologique chez l’ap-
prenti lecteur et l’enfant dyslexique (The role and representation of morphologi-
cal information in the beginning reader and in dyslexic children). Annales Fyssen,
16, 81–90.
Goswami, U., & Bryant, P. (1989). The interpretation of studies using the reading level
design. Journal of Reading Behavior, 21, 413–424.
Hatcher, J., Snowling, M. J., & Griffiths, Y. M. (2002). Cognitive assessment of dyslexic
students in higher education. British Journal of Educational Psychology, 72, 119–133.
Jackson, N. E. & Butterfield, E. C. (1989). Reading-level-match designs: Myths and reali-
ties. Journal of Reading Behavior, 21, 387–412.
Joanisse, M. F., Manis, F. R., Keating, P., & Seidenberg, M. S. (2000). Language deficits in
dyslexic children: Speech perception, phonology and morphology. Journal of
Experimental Child Psychology, 77, 30–60.
Lefly, D. L., & Pennington, B. F. (1991). Spelling errors and reading fluency in compen-
sated dyslexics. Annals of Dyslexia, 41, 143–162.
Leong, C. K. (1989a). Productive knowledge of derivational rules in poor readers.
Annals of Dyslexia, 39, 94–115.
Leong, C. K. (1989b). The effects of morphological structure on reading proficiency: A
developmental study. Reading and Writing: An Interdisciplinary Journal, 1, 357–379.
Leong, C. K. (1999). Phonological and morphological processing in adult students with
learning/reading disabilities. Journal of Learning Disabilities, 32, 224–238.
Leong, C. K., & Parkinson, M. E. (1995). Processing of English morphological structure
by poor readers. In C. K. Leong & R. M. Joshi (Eds.), Developmental and acquired
dyslexia: Neuropsychological and neurolinguistic perspectives (pp. 237–261).
Dordrecht: Kluwer Academic Publishers.
Marslen-Wilson, W., Tyler, L., Waksler, R., & Older, L. (1994). Morphology and mean-
ing in the mental lexicon. Psychological Review, 10, 3–33.
Morais, J., Carey, L., Alegria, J., & Bertelson, P. (1979). Does awareness of speech as a se-
quence of phones arise spontaneously? Cognition, 7, 323–331.
Mosberg, L., & Johns, D. (1994). Reading and listening comprehension in college stu-
dents with developmental dyslexia. Learning Disabilities Research & Practice, 9(3),
130–135.
Napps, S. E., & Fowler, C. A. (1987). Formal relationships among words and the organi-
zation of the mental lexicon. Journal of Psycholinguistic Research, 15, 257–272.
Nation, K., Marshall, C. M., & Snowling, M. J. (2001). Phonological and semantic contri-
butions to children’s picture naming skill: Evidence from children with develop-
mental reading disorders. Language & Cognitive Processes, 16, 241–259.
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of
the scientific literature on reading and its implications for reading instruction. Bethesda
MD: National Institute of Child Health and Human Development.
Parrila, R., Corkett, J., Kirby, J. R., & Hein, S. (2003). Adult reading history question-
naire–revised. Unpublished questionnaire. University of Alberta.
Parrila, R., & Georgiou, G. (2005, June). Persistent naming speed problems in high-func-
tioning adult dyslexics: Wherein lies the problem? Poster presented at the annual
meeting of the Society for the Scientific Study of Reading, Toronto, Canada.
Pennington, B. F., Van Orden, G. C., Smith, S. D., Green, P. A., & Haith, M. (1990).
Phonological processing skills and deficits in adult dyslexics. Child Development,
61, 1753–1778.
Raaijmakers, J. G. W., Schrijnemakers, J. M. C., & Gremmen, F. (1999). How to deal with
“the language as a fixed effect fallacy:” Common misconceptions and alternative
solutions. Journal of Memory and Language, 41, 416–426.
Rack, J. (1997). Issues in the assessment of developmental dyslexia in adults: Theoretical
and applied perspectives. Journal of Research in Reading, 20, 66–76.
Rosner, J., & Simon, D. P. (1971). The auditory analysis test: An initial report. Journal of
Learning Disabilities, 4, 384–392.
Rubin, G. S., Becker, C. A., & Freeman, R. H. (1979). Morphological structure and its ef-
fect on visual word recognition. Journal of Verbal Learning and Verbal Behaviour, 18,
757–767.
126 DEACON, PARRILA, AND KIRBY
DERIVED FORM PROCESSING IN HIGH-FUNCTIONING DYSLEXICS 127
Rubin, H., Patterson, P. A., & Kantor, M. (1991). Morphological development and writ-
ing ability in children and adults. Language, Speech, and Hearing Services in Schools,
22, 228–235.
Seidenderg, M. S. (1985). The lexicon in a model of language production. New York: Garland
Publishing.
Shankweiler, D., Crain, S., Katz, L., Fowler, A. E., Liberman, A. E., Brady, S. A.,
Thornton, R., Lundquist, E., Breyer, L., Fletcher, J. M., Stuebing, K. K., Shaywitz,
S. E., & Shaywitz, B. A. (1995). Cognitive profiles of reading-disabled children:
Comparison of language skills in phonology, morphology and syntax.
Psychological Science, 6, 149–156.
Shankweiler, D., Lundquist, E., Dreyer, L. G., & Dickinson, C. C. (1996). Reading and
spelling difficulties in high school students: Causes and consequences. Reading
and Writing: An Interdisciplinary Journal, 8, 267–294.
Siegel, L. S., Share, D., & Geva, E. (1995). Evidence for superior orthographic skills in
dyslexics. Psychological Science, 6, 250–254.
Snowling, M. (2000). Dyslexia (2nd ed.). Oxford, UK: Blackwell Publishers.
Snowling, M., Nation, K., Moxham, P., Gallagher, A., & Frith, U. (1997). Phonological
processing skills of dyslexic students in higher education: A preliminary report.
Journal of Research in Reading, 20, 31–41.
Stoltz, J. A., & Feldman, L. B. (1995). The role of orthographic and semantic transparency of
the base morpheme in morphological processing. In L. B. Feldman (Ed.), Morphological
aspects of language processing (pp. 109–129). Hillsdale, NJ: Lawrence Erlbaum.
Taft, M. (1979). Recognition of affixed words and the frequency effect. Memory and
Cognition, 7, 263–272.
Taft. M. (1981). Prefix stripping revisited. Journal of Verbal Learning and Verbal Behaviour,
20, 289–297.
Taft, M. (1994). Interactive-activation as a framework for understanding morphological
processing. Language and Cognitive Processes, 9, 271–294.
Taft, M. (2004). Morphological decomposition and the reverse base frequency effect.
Quarterly Journal of Experimental Psychology: Human Experimental Psychology,
57A(4), 745–765.
Taft, M., & Forster, K. I. (1975). Lexical storage and retrieval of prefixed words. Journal of
Verbal Learning and Verbal Behaviour, 14, 630–647.
Tyler, A., & Nagy, W. (1990). Use of derivational morphology during reading. Cognition,
36, 17–34.
van der Leij, A., & van Daal, V. H. P. (1999). Automatization aspects of dyslexia: Speed
limitations in word identification, sensitivity to increasing task demands, and or-
thographic compensation. Journal of Learning Disabilities, 32, 417–428.
Vellutino, F. R., & Scanlon, D. M. (1989). Some prerequisites for interpreting results
from reading level matched designs. Journal of Reading Behavior, 21, 361–385.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1999). CTOPP: Comprehensive test of
phonological processing. Austin, TX: PRO-ED.
Watson, F. L., & Brown, G. D. (1992). Single-word reading in college dyslexics. Applied
Cognitive Psychology, 6, 263–272.
Woodcock, R. (1987). Woodcock reading mastery tests–revised. Circle Pines, MN: American
Guidance Services.
Zoccolotti, P., De Luca, M., Di Pace, E., Gasperini, F., Judica, A., & Spinelli, D. (2005).
Word length effect in early reading and in developmental dyslexia. Brain and
Language, 93, 369–373.
Manuscript received November 15, 2005.
Final version accepted March 22, 2006.
APPENDIX
FREQUENCIES OF THE TARGET WORDS AND
FIRST SEGMENTS OF THE TARGET WORDS
Derived No Change Words Pseudo-derived Control Words
ship 83 shipment 2 bread 41 breadth 7
dirt 43 dirty 36 apart 57 apartment 81
dust 70 dusty 16 top 204 topic 9
music 216 musical 85 bell 18 belly 23
farm 125 farmer 23 met 132 metal 61
read 173 reader 43 wit 20 witness 28
paint 37 painter 21 pan 16 panic 22
cloud 28 cloudy 2 ten 165 tenant 5
Mean 96.88 28.50 96.88 28.50
Derived Orthographic Pseudo-derived
Change Words Control Words
center 224 central 164 ham 19 hammer 9
fog 25 foggy 5 war 464 warrant 20
fun 44 funny 41 cop 15 copper 13
begin 84 beginner 34 gut 1 gutter 1
able 216 ably 2 grave 33 gravy 4
bag 42 baggy 4 lad 6 ladder 19
cool 62 coolly 5 ban 7 banner 8
chop 3 choppy 3 let 384 letter 145
Mean 87.50 32.25 116.13 27.38
128 DEACON, PARRILA, AND KIRBY
