INT. J. LANG. COMM. DIS.

, APRIL –JUNE 2004,

VOL. 39, NO. 2, 245–256

Literacy outcomes for students with speech

impairment: long-term follow-up

Suze Leitão and Janet Fletcher

School of Psychology, University of Western Australia, Crawley, Australia
(Received October 2002; accepted July 2003)

Abstract
Background: Theoretical and empirical support now exists for the finding that
many children with expressive phonological impairment experience problems in
acquiring phonological awareness and early literacy skills. Few studies, however,
have examined the long-term academic and literacy outcomes for this
population, in particular as the students leave the final stages of primary school.
Aims: The reported study forms the final stage of a longitudinal research
project that tracked the phonological processing and literacy skills of a group of
children with specific speech impairment from their first year at school (aged
5–6 years). The earlier data provided evidence of a relationship between speech
impairment characterized by the presence of non-developmental error
processes and weaker phonological awareness and literacy skills in the first
2–3 years at school. It was hypothesized that the effect of this relationship
would continue to be apparent as the students completed the final stages of
primary school.
Methods & Procedures: Fourteen of the original set of 36 students were available
for reassessment of their phonological processing, reading and spelling skills at
age 12–13 years.
Outcomes & Results: Those children with an original classification of non-
developmental speech errors performed significantly more poorly than those
with an original classification of developmental errors on phonological awareness
and reading comprehension measures. Reading accuracy and spelling scores
also showing a similar trend.
Conclusions: These findings provide further evidence for the long-term impact of
speech impairment. The follow-up data demonstrated ongoing difficulties for
students who entered school with expressive speech impairment, particularly
those whose speech errors were characterized by non-developmental error
processes. The impact was apparent on tasks measuring phonological awareness,
reading accuracy and spelling (skills that depend on good phonological pro-
cessing skills and clear underlying phonological representations). Weaknesses in

Address correspondence to: Suze Leitão, Child Study Centre, School of Psychology, University of
Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia; e-mail: suzel@iinet.net.au

International Journal of Language & Communication Disorders

ISSN 1368-2822 print/ISSN 1460-6984 online # 2004 Royal College of Speech & Language Therapists
http://www.tandf.co.uk/journals
DOI: 10.1080/13682820310001619478
246 Suze Leitão and Janet Fletcher

reading comprehension were also found. These findings have implications for
the early identification of those at risk. In addition, intervention approaches for
young children with expressive speech difficulties demonstrating these patterns
of error should address weak underlying phonological representations and
develop phonological awareness skills.

Keywords: literacy outcomes, speech impairment, long-term follow up.

Introduction
Theoretical and empirical support exists for the finding that many children with
expressive phonological impairment experience problems in acquiring phonological
awareness (Bird and Bishop 1992, Leitão et al. 1997, Stackhouse 2000, Hesketh
2001). In addition, there is evidence that they are also at risk for difficulties
acquiring literacy skills in the early stages, especially spelling (Bird et al. 1995, Catts
and Kamhi 1999, Leitão et al. 2000). Evidence is accumulating that speech disorders
associated with difficulties at the level of phonological representation have a more
detrimental impact on literacy development than those that are more peripheral or
articulatory in nature (Snowling 2000).
Training phonological awareness skills has recently been introduced as an
intervention approach for young children with speech impairment (Gillon 2000,
Hesketh et al. 2000). The rationale behind this type of intervention is that an early
focus on phonological awareness skills will lead to improvements in speech and
also work to prevent later literacy difficulties (Dodd and Gillon 2001). The small
number of studies in the area have provided limited evidence to support such an
approach, but a great deal more evidence is required before phonological awareness
programmes should be considered a standard procedure for all children with
speech impairment.
Not all children with speech impairment have limited phonological awareness
skills nor will go on to develop literacy difficulties (Stackhouse 1996, Hesketh et al.
2000). Clinicians need to identify at an early age those speech-impaired children
who will develop literacy difficulties in order to plan appropriate intervention.
One way to begin identifying those preschool children at risk of later literacy
difficulties is through longitudinal research. Whilst some follow-up studies exist
(Bishop and Adams 1990, Webster et al. 1997, Lewis et al. 2000), few have
examined the long-term implications for this population as the students leave the
final stages of primary school and enter high school (Stothard et al. 1998, Bernhardt
2002).
Stothard et al. (1998) found the majority of adolescents in their study with a
history of speech–language impairment (SLI) were experiencing academic dif-
ficulties and almost half were receiving special education. Bishop and Adams (1990)
in their earlier study had proposed a critical age hypothesis, whereby a child is less
at risk for literacy difficulties provided their expressive problems have resolved by
the age at which they first learn to read and write. Among the children followed up
by Stothard et al. (1998), it was those who had ongoing SLI on school entry (at
aged 5;6 years) who were at high risk of developing literacy difficulties.
Despite clinical observation of the problems that highly unintelligible children
experience in the area of emergent literacy (Hodson et al. 1989), data regarding the
Literacy outcomes and speech impairment 247

possible relationship of specific speech impairment to literacy acquisition are

conflicting. Comprehensive reviews of the literature considering possible links
between speech impairment, phonological processing skills and literacy outcomes
conclude the results are equivocal (Stackhouse 1996, Leitão et al. 1997, Leitão
1998). The inconsistency in the results across the studies may be explained by a
number of potential confounding factors. These include the specificity and severity
of the speech impairment, the age of the subjects and the pattern of speech errors
(Leitão 1998).
The study reported here is the final phase of a longitudinal research project that
followed two groups of specifically speech-impaired children from age 5–6 years
(on entry to primary school) through their first 2–3 years of primary school (Leitão
1998, Leitão et al. 1997, 2000). A number were retested at the end of their primary
school years, at age 12–13 years. The selection criteria used by Leitão et al. (1997)
ensured the subjects presented with a moderate–severe speech impairment, in the
absence of language impairment, that was ongoing on entry to formal schooling at
5–6 years of age.

Summary of previous studies

Children entering the first year of school aged 5–6 years in 1994 and 1995 formed
two cohorts of specifically speech-impaired children whose progress was monitored
in their second and third year at school, respectively. They were selected based on
intelligence, oral language and speech output.
Subjects were nominated by speech and language therapists and selected from
community clinics, hospitals and language units. All the children were aged 5–
6 years, attending their first year of formal learning at school and in the early
stages of formal reading instruction. The performance component of an ability test,
the Wechsler Preschool and Primary Scale of Intelligence—Revised (WPPSI-R)
(Wechsler 1989) was administered to ensure non-verbal skills were within the
normal range for all subject groups. Language development was assessed using the
Clinical Evaluation of Language Fundamentals—Revised (CELF-R) (Semel et al.
1994). Speech was assessed using the Goldman–Fristoe Test of Articulation
(Goldman and Fristoe 1986). A cut-off point of 1 SD below the mean was used for
selection for all scores except the total language score on the CELF-R for which
the cut-off point was 21.5 SD. The CELF-R has two expressive subtests (word
structure and sentence recall) that rely on intact morphology and hence provide
unreliable measures in children with such poor intelligibility.
Subjects were therefore selected if they presented with cognitive and language
skills within the normal range, but expressive speech scores fell below the normal
range. This led to the selection of a group of children who entered school with a
specific speech impairment. Subject selection criteria are described in more detail in
Leitão et al. (1997, 2000) and Leitão (1998).

Table 1. Subject selection scores

WPPSI-R CELF-R CELF-R GFTA

Subject performance Receptive Language Total Language percentile
group IQ standard score standard score score
Speech impaired w85 w85 w77 ¡16
248 Suze Leitão and Janet Fletcher

The term ‘speech impaired’ encompasses a heterogeneous population. Beyond a

basic severity rating and broad classification, speech error patterns can be described
in detail. This is achieved by carrying out a detailed analysis of speech errors and
phonological processes based on a sample of single words and connected speech.
Commonly used normal developmental phonological processes include such processes
as stopping (a target fricative is realized as a plosive with the same place of
articulation, e.g. ‘sea’ is produced as ‘tea’) or cluster reduction (the deletion of one
or more consonants from a cluster, e.g. ‘spoon’ is produced as ‘poon’). Examples
of non-developmental phonological errors include backing (target front consonants,
usually alveolar, are realized as velar, e.g. ‘tar’ is produced as ‘car’), use of non-
native language phonemes (e.g. use of a bilabial fricative to mark all consonant
clusters) or systematic sound preference (e.g. all consonants realized as ‘s’)
(Grunwell 1985, Dodd et al. 1989). These non-developmental processes do not
usually occur in the speech of a younger normally developing child.
Phonological awareness data collected from the first cohort at a mean age of
6;1 years (range 5;8–6;7 years) indicated the presence of two subgroups. As
previously reported, no relationship was found between performance on the
phonological awareness tasks and measures of language skill or speech severity
(Leitão et al. 1997). Post-hoc analysis demonstrated that performance was linked to
the pattern of speech errors (Leitão et al. 1997). Subjects were classified into two
subgroups based on patterns of error (developmental/non-developmental
processes) using criteria adapted from Dodd (1995). Intra- and interrater reliability
measures on this classification were made and 90% agreement was reached. When
the subgroups of speech-impaired children were compared, i.e. those with
Developmental (DEV) versus those with Non-developmental (NONDEV) speech
errors, some patterns emerged in the data. In general, the DEV subgroups per-
formed significantly better on phonological awareness measures than NONDEV.
In the follow-up study, at mean age 7;1 years (range 6;7–7;8 years), the DEV
subgroup performed significantly better on tasks measuring word identification,
word attack skills and spelling of both real and non-words. While tests of reading
accuracy and comprehension did not show a significant difference, there was a
trend to higher scores among the DEV subgroup (Leitão et al. 1997).
Apart from the work of Dodd et al. (1989), research has tended to include a
composite group of speech-impaired subjects whose group performance may mask
that of existing subgroups. The presence of subgroups may go some way towards
explaining the inconsistent findings in the literature regarding the relationship
between speech impairment and literacy outcomes (Hall and Tomblin 1978, Levi
et al. 1982, Snowling and Stackhouse 1983, Hodson et al. 1989, Bishop and Adams
1990, Catts 1993, Larrivee and Catts 1999).
A second cohort was therefore selected and classified prospectively. Subject
selection followed the same procedure as described above. Additional speech data
were collected to allow classification into subgroups based on patterns of
developmental/non-developmental process use. A cassette recording was made of
spontaneous conversation and talk related to the tasks during the data collection for
initial subject selection. The transcriber listened to the tapes and selected periods
where the child was most talkative (Dodd 1995). A minimum of 30 utterances was
transcribed using broad phonetic transcription.
A phonological process analysis was performed on the 30 utterances.
Phonological process analysis is primarily concerned with identifying and classifying
Literacy outcomes and speech impairment 249

the systematic patterns in the data and comparing these patterns with those found
in normal speech in order to evaluate the communicative implications (Grunwell
1990).
The analysis used the following steps:
(1) Error forms were noted and compared with the target (e.g. ‘shoe’ said as
‘too’).
(2) Phonological rules used by the child were described (e.g. stopping of ‘sh’
to ‘t’).
(3) Phonological process used was identified (e.g. process: stopping).
(4) Processes identified were classified as developmental or non-develop-
mental (e.g. ‘stopping’ is developmental).
The information gained from this sample was then used to classify each subject’s
overall speech pattern as ‘developmental’ or ‘non-developmental’ based on their
process use. The technique applied followed the work of Dodd (1995). Dodd is not
specific about the proportion of non-developmental processes a child must use to
be classified as such. In addition, in order for a child to be classified as exhibiting
non-developmental phonological development, not every error needs to be non-
developmental. From clinical experience and a review of the literature (Ingram
1976), it was concluded that most children who exhibit non-developmental process
use will also use a variety of developmental processes. It is the identification of
non-developmental processes that is crucial, since this indicates that the acquisition
of the phonological system has gone awry (Dodd et al. 1989). The subgroupings are
thus based on major trends rather than all-or-none error patterns.
The operational definition established in this research was that a classification
of ‘non-developmental’ (NONDEV) was given to subjects whose speech samples
contained more than 10% of non-developmental processes, in addition to the pre-
sence of any developmental processes. Subjects were classified as ‘developmental’
(DEV) if their speech sample contained less than 10% of non-developmental
processes. A check by two experienced speech pathologists using 25% of the taped
samples indicated the presence of 10% or more of non-developmental process use
consistently resulted in a child’s speech presenting as ‘clinically disordered’.
A second cohort of 21 specifically speech-impaired children (mean age 5;7
years; range 5;4–6;2 years) was selected in this way, 10 falling into the DEV group
and 11 into the NONDEV group. The results of the phonological awareness data
collected in the first year at school and the follow up speech and literacy data
collected in the third year of school (mean age 7;10 years; range 7;7–8;5 years) have
been reported by Leitão et al. (2000). In summary, the data showed the speech-
impaired group as a whole to have difficulties with the phonological awareness and
spelling tasks, with a marked degree of variation in performance on the reading
tasks. The DEV group performed significantly better than the NONDEV group on
tasks measuring letter name/sound knowledge, phoneme oddity, invented spelling,
real and non-word reading, real and non-word spelling. Other measures of
phonological awareness and reading showed non-significant differences but the data
showed a similar trend.
It was hypothesized that the presence of non-developmental speech errors in
the NONDEV subgroup could be seen as a marker of an underlying difficulty
abstracting and using knowledge about phonology (Dodd 1995, Leitão et al. 1997).
Both non-developmental speech errors and difficulties acquiring literacy skills
250 Suze Leitão and Janet Fletcher

(especially spelling) were felt to reflect faulty processing of phonological infor-

mation and weak or ‘fuzzy’ underlying phonological representations (Stackhouse
and Wells 1997a, Leitão 1998).
In the present study, those students from the original groups who could be
traced were followed up at age 12–13 years, in the final year of primary school or
the first year of high school. Phonological processing, reading and spelling skills
were assessed. The earlier data had provided evidence of a relationship between
speech impairment characterized by the presence of non-developmental error
processes, and weaker phonological awareness and literacy skills in the first 2–
3 years at school. It was hypothesized that this relationship would continue to be
present as the students completed the final stages of primary school.

Method
Information sheets and consent forms were sent to the last contact address for 36
students from the original cohorts. Fifteen families replied and 14 students were
available to be tested. Four subjects came from the original 1994 cohort and were
now 13 years old and in the last year of primary school or the first year of high
school; the remaining 10 subjects came from the second 1995 cohort, were aged
12 years and completing their final year in primary school. Location of testing was
negotiated with each family and took place at either the home or school.
A series of independent sample t-tests was carried out to compare those
students who were followed up with those from the original cohorts who were not
followed up. No significant differences were found between the two groups for any
of the original subject selection measures. The follow up cohort was considered to
be representative of the original group as a whole in terms of intelligence, language
and severity of expressive speech skills.
The standardized test battery aimed to use tests commonly used in the
Australian school system and included the following:
. Neale Analysis of Reading Ability—Revised (NARA-R) (Neale 1988): a
standardized reading test, administered individually, that provides measures
of text reading accuracy and comprehension. The child is asked to read
aloud a series of short passages and then answer a set of open-ended
questions after each one. The test is suitable for children aged 6–12 years
and takes about 20 minutes to give. Raw scores were converted to age-
equivalent scores (months) as no standardized scores are available for this
version of the test.
. South Australian Spelling Test (SAST) (Westwood 1999): a test of real word
spelling that has been standardized across the age range 6–15 years. The raw
score of number of words spelled correctly is calculated and can be com-
pared with the normal range. The test also yields a critically low score (a cut-
off of the bottom 10%) and an age equivalent score, though standard scores
are not available.
. Test of Word Reading Efficiency (TOWRE) (Torgesen et al. 1997): this
timed test quickly and accurately measures word recognition fluency by
asking children to read as many words as possible in 45 seconds, placing an
emphasis on sight word reading. The ability to recognize words quickly and
accurately (also called lexical processing) is a clear indication of skilled
Literacy outcomes and speech impairment 251

reading. Familiar words (lexical items) are stored in the lexicon, a kind of
mental store cupboard for known words. The second subtest requires the
child to read a list of non-words under the same conditions, measuring
phonemic decoding fluency. Results are presented as standard scores with a
mean of 100 and a standard deviation of 15.
. Comprehensive Test of Phonological Processing (CTOPP) (Wagner and
Torgeson 1997): the second version of this test (for ages 7–24 years) was
given. It covers three areas of phonological processing: phonological
awareness, phonological memory and rapid naming, i.e. phonological
retrieval. Each area is comprised of several subtests. Phonological awareness
subtests include elision (deletion), blending words, blending non-words,
segmenting words, segmenting non-words and phoneme reversal. Phono-
logical memory subtests include memory for digits and non-word repetition.
Rapid naming subtests include rapid letter and digit naming. Each area yields
a composite (standard) score with a mean of 100 and a standard deviation
of 15.

Results
The 14 follow-up subjects fell into two subgroups based on their original
classification (DEV and NONDEV), each containing seven children. The relative
performance of the two groups was compared in a series of t-tests.
The results for the NARA-R are shown in table 2. The data are presented as
age-equivalent scores (months) as no standardized scores are available for this
version of the test. The difference between the reading accuracy/reading com-
prehension scores and chronological age is presented as difference scores (months).
These difference scores were used in the analyses.
Turning first to the NARA-R Reading Accuracy scores, the difference between
the two subgroup means was non-significant, although the trend was in the pre-
dicted direction. Original classification in the NONDEV subgroup was associated
with weaker reading accuracy skills. The mean score was 16.57 (SD~23.68)
months below chronological age for the DEV group and 26.14 (16.85) months
below chronological age for the NONDEV group.
The NARA-R Reading Comprehension mean age equivalent score for the
DEV group was 11.71 (SD~14.68) months below chronological age and for the
NONDEV group was 29.14 (13.14) months below chronological age. This dis-
crepancy between the groups was significant (t~22.34, d.f.~12, p~0.037).
The results for the SAST were presented in the same way as no standard scores
are available. The difference between the two subgroup means for SAST was non-
significant, though the trend was again in the predicted direction. The mean spelling
age equivalent score for the DEV group was 16 months below chronological age
(SD~19.15) while it was 31 (18.36) months below chronological age for the
NONDEV group.
The SAST results were profiled according to those subjects who fell above/
below the Critical Low score. This score is said to represent a cut-off point at the
10th percentile. Five of the six subjects falling below the Critical Low score (i.e.
within the bottom 10% of the population) came from the NONDEV group.
The CTOPP phonological awareness mean standard score for the DEV group
was 97.86 (SD~5.11), while that for the NONDEV group was 83.29 (14.78).
252 Suze Leitão and Janet Fletcher

Table 2. Descriptive data on the NARA-R and SAST for the speech-impaired subjects

NARA-R NARA-R NARA-R NARA-R SAST SAST

No. Reading RA/CA Reading RC/CA AE AE/CA
and CA Accuracy difference Comprehension difference score difference
group (months) score (months) (months)1 score (months) (months)1 (months) (months)1
1 D 156 97 259 115 241 117 239
2 D 159 125 234 142 217 134 225
3 D 149 126 223 152 3 131 218
4 D 145 150 5 140 25 141 24
5 D 146 150 4 137 29 166 20
6 D 153 144 29 140 213 126 227
7 D 150 150 0 150 0 129 221
8 N 157 115 242 112 255 120 237
9 N 163 125 238 115 248 117 246
10 N 146 149 3 134 212 141 25
11 N 146 113 233 126 220 108 238
12 N 148 114 234 126 222 99 249
13 N 146 138 28 118 228 141 25
14 N 144 113 231 115 229 107 237
1
The difference between each subject’s chronological age and age equivalent score is presented in
months.

Allowing for non-equivalent variances, the difference between the two groups was
still significant (t~2.46, d.f.~7.42, p~0.04).
t-Tests demonstrated no significant difference between the mean scores of the
DEV and NONDEV subgroups for the tasks measuring phonological memory and
phonological retrieval on the CTOPP, and phonemic decoding efficiency or sight
word reading efficiency using the TOWRE.

Table 3. Standard scores on the TOWRE and CTOPP for the speech-impaired subjects

TOWRE TOWRE CTOPP CTOPP CTOPP

No. Sight Phonemic Phonological Phonological Phonological
and Words Decoding Awareness Memory Retrieval
group SS SS Composite SS Composite SS Composite SS
1 D 84 77 88 121 76
2 D 81 84 97 79 97
3 D 88 84 97 94 85
4 D 107 98 100 79 118
5 D 88 90 103 97 79
6 D 91 82 103 94 94
7 D 90 79 97 91 88
8 N 91 82 70 73 97
9 N 88 82 94 85 88
10 N 92 95 106 118 88
11 N 94 82 67 67 97
12 N 86 72 73 94 100
13 N 83 86 79 94 97
14 N 86 84 94 97 91
Literacy outcomes and speech impairment 253

Discussion
When the children were originally tested early in primary school, the data showed
difficulties with phonological awareness tasks at age 5–6 years, and low average
reading skills and weak spelling skills at age 7 years (Leitão et al. 1997, 1998, 2000).
Furthermore, those children with a speech impairment characterized by non-
developmental error processes were significantly more likely to have weak phono-
logical awareness skills and go on to struggle with literacy acquisition, especially
non-word reading and spelling (real and non-words).
The follow-up data at aged 12–13 years demonstrated significant differences
between the two subgroups for phonological awareness and reading comprehension
scores with the Developmental subgroup performance being stronger than that of
the Non-developmental subgroup. Reading accuracy and spelling scores also
showed a strong trend in the same direction.
Ongoing difficulties with reading accuracy, phonemic decoding and spelling
reflect weaknesses in areas that depend on good skills in processing phonological
information and clear underlying phonological representations. Weak decoding
skills now appeared to be influencing reading comprehension. These students were
originally selected as demonstrating cognitive and oral language skills within the
normal range. Whilst follow-up language testing did not take place, the observed
ongoing difficulties with reading comprehension are interpreted as more likely to be
a consequence of decoding difficulties rather than due to a more general language
comprehension deficit (Cornoldi and Oakhill 1996, Perfetti et al. 1996).
As hypothesized by Leitão (1998) and Leitão et al. (1997, 2000), those children
who enter school with a moderate–severe speech disorder characterized by non-
developmental speech errors appear to have a subtype of speech impairment where
the errors can be considered a surface indicator of an underlying breakdown in the
speech processing system.

Implications of non-developmental process use

Non-developmental process use can be considered a symptom of a breakdown in
the speech processing system at the level of processing phonological information, a
cognitive–linguistic impairment. This will lead to difficulties in segmentation and in
identifying and classifying phonemes. As such, a pattern of non-developmental process
use can be linked to impairments in abstracting and using knowledge about
phonology to generate both spoken and written output (Dodd 1995).
Bird and Bishop (1992) concluded that although the phonologically impaired
subjects in their study evidenced perceptual difficulties, these were not significant
enough to explain their production deficits. They proposed that the difficulty lay
not so much in discrimination of speech sounds as in learning how to categorize
speech input (Bird and Bishop 1992).
This line of thinking is consistent with research on normal young language
learners that suggests that speech errors may result from a child forming an
incomplete phonological representation of a word that includes information about
stress and syllabic structure and some additional information about the acoustic
features associated with particular syllables rather than a full specification of the
signal as a sequence of phonemes (Echols 1993). In the phase model described by
Stackhouse and Wells (1997a), children’s first words are stored and retrieved as a
254 Suze Leitão and Janet Fletcher

holistic pattern of interacting elements. During the period of rapid vocabulary

expansion, holistic representation of lexical items becomes increasingly inefficient,
many words begin to overlap in acoustic properties and pressure builds to
implement representations that would facilitate more efficient storage (Walley
1993). Over time, a refining of the unit of underlying representation takes
place, narrowing to that of intrasyllabic and phonemic units. Early vocabulary
may therefore originally be represented at a more holistic level, with lexical
representations becoming increasingly segmental between 1 and 8 years of age
(Fowler 1991).
Children with persisting speech difficulties may therefore perform poorly on
phonological awareness tasks and be at risk for literacy problems because their
speech processing system is underdeveloped (Stackhouse and Wells 1997b, 2001).
Difficulties can begin at the whole word phase of the Stackhouse and Wells model,
with children progressing through to the systematic simplification phase without
having resolved the way their phonological representations are encoded and stored.
Their speech simplifications or productions may therefore become more systematic
and consistent, but the error processes will be disordered, based on unsegmented
or poorly segmented analyses and therefore on weak or incorrect phonological
representations.
In addition to segmentation difficulties, when children with disordered expres-
sive phonology do segment at the level of the phoneme, they may focus on features
that are not distinctive, selecting the wrong parameters of the speech signal as
salient in their native phonology and marking the differences in an unusual way.
The child therefore develops an aberrant phonological system in which inappro-
priate criteria are used to establish phoneme identity (Dodd et al. 1995).
Poor segmentation and phoneme identification skills can lead not only to
unusual speech errors, but also to literacy difficulties. Children who come to the
task of learning to read and spell with well-specified representations are well
placed to establish links between the letters of printed words and the sounds of
spoken words. This helps them crack a critical key to successful literacy acquisition:
the ‘alphabetic code’ (Westwood 2001). On the other hand, children may struggle
if they are trying to develop orthographic–phonological mappings based on
coarse grained, indistinct and even inaccurate underlying representations (Elbro
1998).

Conclusions
The long-term follow up of a small group of subjects has provided further evidence
that speech disorders characterized by non-developmental error processes and
associated with difficulties at the level of phonological representation place a child
at greater risk of difficulties with acquiring literacy skills. This information should
help in the quest for the early identification of those at risk and those who should
benefit the most from intervention approaches which include phonological aware-
ness skills training. Intervention should be preceded by an assessment of the
child’s abilities, profiling speech, phonological awareness and phonological pro-
cessing skills using a framework such as that of Stackhouse and Wells (1997a),
which allows principled goal setting to be linked to underlying processing strengths
and weaknesses.
Literacy outcomes and speech impairment 255

Acknowledgements
The follow-up study described here was partly supported by a clinician’s grant
awarded to S. L. in 2001 by Speech Pathology Australia. Thanks are also due to Dr
John Hogben, School of Psychology, University of Western Australia, for help at
various stages throughout the research. The authors also thank the children and
families who took part in the research.

References
BERNHARDT, B., 2002, Long-term outcomes in language, metaphonology and literacy for children in
British Columbia with a history of phonological disorders or cleft palate. Paper presented at the
2002 Speech Pathology Australia National Conference, Alice Springs, Australia.
BIRD, J. and BISHOP, D., 1992, Perception and awareness of phonemes in phonologically impaired
children. European Journal of Disorders of Communication, 27, 289–311.
BIRD, J., BISHOP, D. and FREEMAN, N., 1995, Phonological awareness and literacy development in
children with expressive phonological impairments. Journal of Speech and Hearing Research, 38,
446–462.
BISHOP, D. and ADAMS, C., 1990, A prospective study of the relationship between specific language
impairment, phonological disorders and reading retardation. Journal of Child Psychology and
Psychiatry, 31, 1027–1050.
CATTS, H., 1993, The relationship between speech–language impairments and reading disabilities.
Journal of Speech and Hearing Research, 36, 948–958.
CATTS, H. and KAMHI, A., 1999, Language and Reading Disabilities (Boston: Allyn & Bacon).
CORNOLDI, C. and OAKHILL, J., 1996, Reading Comprehension Difficulties: Processes and Intervention (Hillsdale:
Lawrence Erlbaum).
DODD, B., 1995, The Differential Diagnosis and Treatment of Children with Speech Disorder (London: Whurr).
DODD, B. and GILLON, G., 2001, Exploring the relationship between phonological awareness, speech
impairment, and literacy. Advances in Speech–Language Pathology, 3, 139–147.
DODD, B., GILLON, G., OERLEMANS, M., RUSSELL, T., SYRMIS, M. and WILSON, H., 1995, Phonological
disorder and the acquisition of literacy. In B. Dodd (ed.), Differential Diagnosis and Treatment of
Children with Speech Disorder (London: Whurr), pp. 125–146.
DODD, B., LEAHY, J. and HAMBLY, G., 1989, Phonological disorders in children: underlying cognitive
deficits. British Journal of Developmental Psychology, 7, 55–71.
ECHOLS, C., 1993, A perceptually-based model of children’s earliest productions. Cognition, 46, 245–296.
ELBRO, C., 1998, When reading is ‘readn’ or somethn. Distinctness of phonological representations of
lexical items in normal and disabled readers. Scandinavian Journal of Psychology, 39, 149–153.
FOWLER, A., 1991, How early phonological development might set the stage for phoneme awareness.
In S. Brady and D. Shankweiler (eds), Phonological Processes in Literacy: A Tribute to Isabelle Y.
Liberman (Hillsdale: Lawrence Erlbaum), pp. 97–117.
GILLON, G., 2000, The efficacy of phonological awareness intervention for children with spoken
language impairment. Language, Speech and Hearing Services in Schools, 31, 126–141.
GOLDMAN, R. and FRISTOE, M., 1986, Goldman–Fristoe Test of Articulation (Circle Pines: American
Guidance Service).
GRUNWELL, P., 1985, PACS: Phonological Assessment of Child Speech (Windsor: NFER-Nelson).
GRUNWELL, P., 1990, Introduction. In P. Grunwell (ed.), Developmental Speech Disorders (Edinburgh:
Churchill Livingstone), pp. 3–14.
HALL, K. and TOMBLIN, J., 1978, A follow-up study of children with articulation and language disorders.
Journal of Speech and Hearing Disorders, 43, 227–241.
HESKETH, A., 2001, Phonological awareness, speech impairment, and literacy: the missing link. Advances
in Speech–Language Pathology, 3, 155–158.
HESKETH, A., ADAMS, C., NIGHTINGALE, C. and HALL, R., 2000, Phonological awareness therapy and
articulatory training approaches for children with phonological disorders: a comparative
outcome study. International Journal of Language and Communication Disorders, 35, 337–354.
HODSON, B., NONOMURA, C. and ZAPPIA, M., 1989, Phonological disorders: impact on academic
performance? Seminars in Speech and Language, 10, 252–259.
INGRAM, D., 1976, Phonological Disability in Children, vol. 2 (London: Edward Arnold).
256 Literacy outcomes and speech impairment

LARRIVEE, L. and CATTS, H., 1999, Early reading achievement in children with expressive phonological
disorders. American Journal of Speech–Language Pathology, 8, 137–148.
LEITÃo, S., 1998, Speech Impairment, Phonological Processing Skills and Literacy Outcomes. Unpublished PhD,
University of Western Australia.
LEITÃo, S., FLETCHER, J. and HOGBEN, J., 2000, Speech impairment and literacy difficulties: underlying
links. Australian Educational and Developmental Psychologist, 17, 63–75.
LEITÃo, S., HOGBEN, J. and FLETCHER, J., 1997, Phonological processing skills in speech and language
impaired children. European Journal of Disorders of Communication, 32, 91–111.
LEITÃo, S., HOGBEN, J. and FLETCHER, J., 1998, Does speech impairment mean later literacy difficulties?
Proceedings of the 1998 Speech Pathology Australia National Conference, pp. 221–229.
LEVI, G., CAPOZZI, F., FABRIZI, A. and SECHI, E., 1982, Language disorders and prognosis for reading
disabilities in developmental age. Perceptual and Motor Skills, 54, 1119–1122.
LEWIS, B., FREEBAIRN, L. and TAYLOR, H., 2000, Follow-up of children with early expressive phonology
disorders. Journal of Learning Disabilities, 33, 433–444.
NEALE, M., 1988, Neale Analysis of Reading Ability—Revised (Hawthorn: Australian Council for
Educational Research).
PERFETTI, C., MARRON, M. and FOLTZ, P., 1996, Sources of comprehension failure: theoretical
perspectives and case studies. In C. Cornoldi and J. Oakhill (eds), Reading Comprehension
Difficulties (Hillsdale: Lawrence Erlbaum), pp. 137–165.
SEMEL, E., WIIG, E. and SECORD, W., 1994, Clinical Evaluation of Language Fundamentals—3 (Cleveland:
Psychological Corporation/Harcourt Brace Jovanovich).
SNOWLING, M., 2000, Dyslexia (Oxford: Blackwell).
SNOWLING, M. and STACKHOUSE, J., 1983, Spelling performance of children with developmental verbal
dyspraxia. Developmental Medicine and Child Neurology, 25, 430–437.
STACKHOUSE, J., 1996, Speech, spelling and reading: who is at risk and why? In M. Snowling and J.
Stackhouse (eds), Dyslexia, Speech and Language (London: Whurr), pp. 12–30.
STACKHOUSE, J., 2000, Barriers to literacy development in children with speech and language difficulties.
In D. V. M. Bishop and L. B. Leonard (eds), Speech and Language Impairments in Children: Causes,
Characteristics, Intervention and Outcome (Hove: Psychology Press), pp. 73–97.
STACKHOUSE, J. and WELLS, B., 1997a, Children’s Speech and Literacy Difficulties (London: Whurr).
STACKHOUSE, J. and WELLS, B., 1997b, How do speech and language problems affect literacy
development? In C. Hulme and M. Snowling (eds), Dyslexia: Biology, Cognition and Intervention
(London: Whurr), pp. 182–211.
STACKHOUSE, J. and WELLS, B., 2001, Children’s Speech and Literacy Difficulties 2: Identification and Intervention
(London: Whurr).
STOTHARD, S., SNOWLING, M., BISHOP, D., CHIPCHASE, B. and KAPLAN, C., 1998, Language impaired
preschoolers: a follow-up into adolescence. Journal of Speech, Language and Hearing Research, 41,
407–418.
TORGESEN, J., WAGNER, R. and RASHOTTE, C., 1997, Test of Word and Nonword Reading Efficiency (Austin:
Pro-Ed).
WAGNER, R. and TORGESON, J., 1997, The Comprehensive Test of Phonological Processes in Reading (Austin:
Pro-Ed).
WALLEY, A., 1993, The role of vocabulary development in children’s spoken word recognition and
segementation ability. Developmental Review, 13, 286–350.
WEBSTER, P., PLANTE, A. and COUVILLION, L., 1997, Phonological impairment and pre-reading: Update on
a longitudinal study. Journal of Learning Disabilities, 30, 365–375.
WECHSLER, D., 1989, Wechsler Preschool and Primary Scale of Intelligence – Revised (Cleveland: Psychological
Corporation).
WESTWOOD, P., 1999, Spelling: Approaches to Teaching and Assessment (Camberwell: Australian Council for
Educational Research).
WESTWOOD, P., 2001, Reading and Learning Difficulties: Approaches to Teaching and Assessment (Camberwell:
Australian Council for Educational Research).