INT. J. LANG. COMM. DIS., APRIL–JUNE
VOL.
40,
NO.
2005,
2, 189–220
Research Report
Phonological therapy within a psycholinguistic
framework: Promoting change in a child with
persisting speech difficulties
Michelle Pascoe, Joy Stackhouse and Bill Wells
Department of Human Communication Sciences, University of Sheffield,
Sheffield S10 2TA, UK
(Received 20 August 2003; accepted 14 July 2004)
Abstract
Background: Single case studies are a valuable means of providing information
about the outcomes of speech and language intervention. Many previous studies
have used phonological analysis as a theoretical basis, while others have used
psycholinguistic models. The present study combines these approaches to
assessment, intervention and evaluation of outcomes.
Aims: The aim of the research was to determine if intensive psycholinguistically
based intervention could result in (a) specific and (b) generalized improvements
in the speech production of a child with severe and persisting phonological
difficulties.
Methods and Procedures: A single subject research design was used with pre- and
post-intervention assessment carried out. Assessment took place at two levels:
the macro level focused on global change in the child’s speech processing
system using psycholinguistic speech profiling and phonological analyses;
the micro assessment focused on specific, treated (and matched control)
stimuli. There were three phases of intervention with a total of 30 hours of
therapy.
Outcomes and Results: Micro evaluation showed significant changes in Katy’s
single word and connected speech production – as well as in other areas such as
spelling and auditory discrimination of closely related real word pairs. Macro
evaluation revealed significant improvement in speech severity indices (PCC,
PPC), and gains in her ability to discriminate between closely related real word
pairs. These changes were maintained at follow-up seven months after
intervention had ceased.
Conclusions: Specific and intensive intervention brought about significant
improvements in this child’s speech, spelling and auditory discrimination at
Address correspondence to: Michelle Pascoe, Department of Human Communication Sciences,
University of Sheffield, Sheffield S10 2TA, UK. e-mail: hcpøø[email protected]
International Journal of Language & Communication Disorders
ISSN 1368-2822 print/ISSN 1460-6984 online # 2005 Royal College of Speech & Language Therapists
http://www.tandf.co.uk/journals
DOI: 10.1080/13682820412331290979
190
M. Pascoe et al.
the micro level, and in speech and auditory discrimination at a macro level. The
macro level assessments reveal persisting core deficits requiring further focused
intervention before changes can be observed more widely. When carefully
defined and evaluated, speech and language intervention can have positive
outcomes for children with severe and persisting speech difficulties.
Keywords: phonological disorder, psycholinguistics, intervention, outcomes.
Introduction
There is an ongoing call for intervention studies in speech and language therapy
literature (Sommers, Logsdon and Wright 1992, Enderby and Emerson 1995,
Frattali 1998, Crosbie and Dodd 2001, Gibbon, McNeill, Wood and Watson 2003).
Intervention is of benefit not only to the individual child but also in contributing to
evidence-based practice. Approaches to providing this evidence base differ: some
have emphasized the need for randomized controlled trials (RCTs), the ‘gold
standard’ of efficacy research (e.g. see Fahey, Griffiths and Peters 1995, Glogowska,
Roulstone, Enderby and Peters 2000). Others have carried out effectiveness studies
in real life clinical settings focusing on a single child, or small numbers of individual
children (e.g. Broom and Doctor 1995, Holm and Dodd 1999, Crosbie and Dodd
2001, Spooner 2002). Each of these approaches has value: RCTs are a powerful
means for testing experimental hypotheses using methods designed to reduce bias.
However, such studies may fail to provide information on how intervention works,
the duration or intensity of intervention that is needed to bring about change, or the
specific details of treatment that may be effective in the treatment of certain
individuals (Frattali 1998). Single subject designs are able to answer some of these
questions, and solve the problems associated with subject homogeneity in that
subjects serve as their own control. Single case studies afford the opportunity to ‘get
inside therapy’ as urged by authors such as Clark and Elliot (1992) and Byng, Van
der Gaag and Parr (1998).
Intervention studies have value not just in their contribution to the efficacy
database, but also in terms of increasing our theoretical knowledge (Bishop 1998,
Baker, Croot, McLeod and Paul 2001). Psycholinguistic models can be used to
develop and test theories about the processes that underlie speech production in
individual children. When the intervention is carried out in a controlled way, the
outcomes of the programme allow one to return to the theoretical starting point,
and reconsider the nature of the speech and language processing system.
Psycholinguistic approaches to the assessment and remediation of speechdisordered children are well established (e.g. Stackhouse and Wells 1997, 2001; Chiat
2000). Such approaches are valuable in giving clinicians a theoretical base from
which to drive the intervention process. Psycholinguistic assessment relies on
theoretical models of speech processing from which hypotheses about the level of
breakdown leading to the speech difficulties can be generated and systematically
tested (Stackhouse and Wells 1997). The aim is to find out where on the model the
child’s speech processing skills are breaking down, and to consider how one might
remediate these difficulties. Examples of model-based interventions include the
single case studies reported by Bryan and Howard (1992), Broom and Doctor
(1995), Waters, Hawkes and Burnett (1998), Norbury and Chiat (2000), Crosbie and
Dodd (2001), Spooner (2002) and Stiegler and Hoffman (2001).
Phonological therapy within a psycholinguistic framework
191
Psycholinguistic approaches have developed partly in response to dissatisfaction
with more traditional medical diagnostic categories. The traditional approach to the
classification of speech and language disorders does not focus on each person as an
individual with a unique deficit in his or her processing system. The focus is on
grouping people with broadly similar aetiologies or symptoms, by implication
suggesting that the same treatment might be applicable to all members of the group.
However, studies have shown that children with superficially similar speech
difficulties may have very different patterns of underlying processing deficit (e.g.
Ruscello 1995, Stackhouse, Nathan, Goulandris and Snowling 2002, Chiat 2000,
Dodd and Bradford 2000). The psycholinguistic approach is concerned with
investigating underlying processing skills. If intervention is carefully targeted at an
individual’s specific point of breakdown, and carried out with an awareness of the
strengths and weaknesses that underlie the individual’s speech processing system,
then it seems more likely that (a) intervention will be successful in bringing about
change in the speech processing system, and (b) if intervention is not successful then
it is possible to isolate the level of the speech processing system that therapy tasks
were tapping, and make appropriate revisions.
However, speech and language processing models have inherent limitations, and
even if further refined, it is doubtful if they could ever shape the clinical process in
isolation. Not all single case studies use psycholinguistic models as their theoretical
springboard. Many single case studies have relied mainly on linguistic theory and
phonological analyses in planning and evaluating interventions (e.g. Weiner 1981,
Monahan 1986, Saben and Ingham 1991, Bernhardt 1992, Barlow 2001). This paper
aims to show how a psycholinguistic approach can be combined with a linguistic
approach in intervention with a school age child with a phonological disorder. The
intervention study draws on knowledge from two key areas: developmental
psycholinguistics and child phonology. It is suggested that the psycholinguistic
approach is useful in answering the question: ‘How?’ - How is intervention going to
work, i.e. how is change to be brought about in the individual’s speech processing
system? Knowledge from linguistics – in this case phonology - enables us to answer
the more specific ‘what?’ question, i.e. what is the content of intervention? e.g. what
are the stimuli that will be used in the activities?
There are few model-based intervention case studies that have attempted to
couch phonological intervention within an explicit psycholinguistic framework.
Bryan and Howard (1992) described intervention for a five-year-old child with
severe phonological difficulties. The child’s speech processing difficulties were
investigated through a series of psycholinguistically-motivated tasks and interpreted
in the light of current models of speech and language processing. In addition a
phonological analysis of the child’s surface speech errors took place, with both sets
of data used to inform intervention planning. Waters et al. (1998) more explicitly
emphasized the need to integrate psycholinguistic information with phonological
information in their report of intervention with a five-year-old boy with
unintelligible speech. They suggested that while phonological analysis and
psycholinguistic assessment are essential for a principled approach to intervention,
they may not always be sufficient: children’s attitudes, behaviours and preferred
learning styles also need to be taken into account. More recently Ebbels (2000)
investigated the speech and language processing skills of a 10-year-old child with a
hearing impairment. Specific points of breakdown for individual phonological
contrasts were identified, with detailed input and output phonological analyses
192
M. Pascoe et al.
interpreted within a broader psycholinguistic framework. The results of the
investigation showed that for some children there is not a single level of breakdown,
but rather there may be multiple levels of difficulty with specific phonological
contrasts implicated at particular levels.
Each of these papers approached assessment and intervention in different ways.
However, they share a common concern with the nature of their participant’s
underlying phonological representations. Edwards, Fourakis, Beckman and Fox
(1999) outline the evolution of representation-based approaches to understanding
children’s phonology, and suggest that characterizing children’s phonological
competence in terms of representations and the constraints acting on them allows
for a richer conceptualisation of phonological development, than traditional
derivational and ‘normalizing’ approaches. A further reason for careful investigation
of underlying phonological representations and phonological processing ability is
because of the close relationship between these skills and reading and spelling
abilities. The association between phonological processing difficulties and reading
and spelling problems has been shown in a number of single case studies (e.g.
Campbell and Butterworth 1985, Snowling, Stackhouse and Rack 1986) and
experimental investigations comparing dyslexic children with normally developing
readers (e.g. Wagner and Torgeson 1987). For school-age children with persisting
speech problems, understanding of the child’s underlying difficulties can have
important implications for speech, language and literacy support.
In this study, we describe the assessment, intervention planning, intervention
and evaluation that took place with a girl aged 6;5 at the start of the study. The child,
who we call Katy, had severe and persisting speech difficulties. The aim of the
research was to determine if psycholinguistically-based intervention could result in
(a) specific and (b) generalized improvements in the speech production of a child
with severe and persisting difficulties.
Methods and Procedures
Participant
The participant, Katy, was aged 6;5 years at the start of the project. Katy has no
hearing difficulties and comes from a monolingual English home where she lives
with both parents and one older sibling. Katy attends a mainstream school where
she has a statement of special educational needs. This provides for a learning
support assistant who works with her in the classroom as well as taking her out of
the classroom for additional one-to-one support. This includes work on her speech
that is carried out under the guidance of a National Health Service speech and
language therapist. IQ results (WASI, Wechsler, 1999) indicated a verbal IQ of 83
(low average), a performance IQ of 78 (borderline) and a full scale IQ of 78. Katy’s
class teacher described her as being in the ‘weaker ability group’ for all her subjects.
She is described as a popular girl with many friends, who enjoys school despite her
difficulties.
Katy has a history of receptive and expressive language delays, although her
expressive delay outweighs her receptive delay. From approximately age 2;0 she
relied heavily on gestures and pointing in order to communicate. At the start of the
intervention project, Katy was in her second year of formal schooling and remained
highly unintelligible to unfamiliar listeners. Further details of Katy’s speech
Phonological therapy within a psycholinguistic framework
193
difficulties and her speech and language intervention history are presented in the
following section.
Intervention history
Katy has been known to the local speech and language therapy service from the time
that she was 2;0 years, and has had regular periods of therapy. These are outlined in
table 1 which shows that intervention has focused mainly on general communication
and early language development, with some specific sessions focusing on
articulation, and more recently phonological therapy with incorporation of
phonological awareness. Progress made has been limited and Katy’s attention has
often been described as limited for her age, making both therapy and follow-up
work at home challenging. During the intervention project, Katy received no
additional direct speech and language therapy. In addition to the normal classroom
instruction she attended a ‘speech and listening group’ twice a week for half an hour.
Five children were in the group run by a learning support assistant. This group
involved general language stimulation activities and phonological awareness work,
e.g. clapping of syllables, identification of initial and final sounds.
Speech difficulties
The Phonological Assessment of Child Speech (PACS) (Grunwell 1985) and
Edinburgh Articulation Test (EAT) (Anthony, Bogle, Ingram and McIsaac 1971)
were used to assess Katy’s speech. Table 2 outlines the findings from these
assessments.
The severity of Katy’s speech difficulties was estimated by using PCC
(percentage of consonants correct), PVC (percentage of vowels correct) and PPC
(percentage of phonemes correct) indices following guidelines from Shriberg,
Austin, Lewis, McSweeny and Wilson (1997) and Dodd (1995), and a set of 100
randomly selected single words. Katy has difficulties with both accurate vowel and
consonant production. Her speech difficulties were severe with only 22% of
consonants accurately realized in single words. Her syllable structure is typically
open (CV, CVCV). Katy’s phonetic inventory shows a marked mismatch between
the contrasts available to her word-initially and medially, and those she is able to
indicate word finally. She has all plosives in her word initial and medial repertoire,
and other consonants emerging in these word positions. In word final position her
repertoire evidences no consonants. She was stimulable for a wide range of sounds
with the exception of /v/, /ð/, /#/, / /, /t#/.
Guidelines from PACS were used to yield percentage usage of each phonological
process by counting the number of instances of use of a particular process in
relation to the potential opportunities for use of each process. The predominant
phonological processes found in Katy’s speech were (1) cluster reduction (100%) in
which clusters are typically reduced to one element that is realised as a substitution
(e.g. /sp/ and /st/ typically produced as [b] and [d] respectively), (2) final consonant
deletion (96%), (3) pre-vocalic voicing (40%), (4) stopping of fricatives and affricates
(21%) and (5) gliding (21%). Vowel distortions were the main example of deviant
processes in Katy’s speech and these are likely due to her motor difficulties.
194
M. Pascoe et al.
Table 1. Summary of Katy’s speech and language intervention history
Katy’s
chronological age
(years; months)
2;3
3;0
3;6
4;2
Description
General advice on
language development
given in home
Non-directive,
multidisciplinary therapy
at child development centre
Therapy block focusing
on use of Rebus symbols
as well as oral-motor
exercises ‘to address
articulation problems’
Instrumentation to provide
increased feedback used
in university clinic
Duration
Comments
66weekly
visits
Concerns about
delayed speech
Unknown
Focus on signing,
Makaton and
neuro-developmental
therapy
No comments
made regarding
progress. Katy
thought to be
more confident
Some limited
improvement
noted in CV
production.
CVCV remains
more challenging
Increased used
of gesture and
pointing; increasingly
sociable
Increased awareness
of word final
consonants; within
word consonants
beginning to emerge
66weekly
106weekly
sessions
Nursery 4;4
Observation and advice
given to nursery staff
One visit
Reception/
Year 1 5;1
Syllable clapping to
improve awareness of
length and timing; expand
repertoire of consonants
to include /p, m, b, d,
n, c/ in word final position
Year 2 6;1
Continuation of
programme from Year 1
Two to three
termly visits
from speech
and language
therapist. Speech
and language
therapist’s
programme carried
out on daily basis
by learning support
assistant in 1:1
sessions or
small groups
As for Year 1
Concerns emerging
re influence of
speech on spelling;
improved production
of final consonants
at single word level
but limited
generalization to
sentence level; increased
awareness of listener’s
needs; intelligibility
remains poor
Phonological therapy within a psycholinguistic framework
195
Table 2. Katy’s speech data at pre-intervention assessment (T1)
Assessment
Severity indices
Phonetic inventory
Stimulability
Phonological processes
analysis (per cent use)
Single-word
speech sample
Connected
speech sample
Comments
Percentage consonants correct (PCC): 22%
Percentage vowels correct (PVC): 74.1%
Percentage phonemes correct (PPC): 41.9%
Word initial position: [m, n, p, b, d, t, k, c, f, w, j, r, ]
Word medial position: [m, b, d, t, k, c, w]
Word final position:
All consonants except /v/, /ð/, /#/, / /, /t#/
Developmental processes: cluster reduction (100%); final
consonant deletion (96%); prevocalic voicing (40%); stopping
of fricatives and affricates (21%); gliding (21%)
Non-developmental processes: vowel distortion (25%)
[bæ] for BAG
[æ"b ] for APPLE
[we] for WEB
["gæwI] for GARAGE
[vI] for FISH
["vebe] for VEGETABLES
[gI"me] for CHRISTMAS
[dI] for SINK
[bæ] for PRAM
[jaI] for LIGHT
[e] for EGGS
[ki:] for QUEEN
[bI] for BEES
[gæ] for CLASS
[k: le: w "ab ua ] for HER LEG WAS BROKEN
[a l l taI "a g u] for A LONG LONG TIME AGO
[I "bet na] for ITS BETTER NOW
[aaI aæ "peItap i: a "wedI] for I HAD PLAYTIME ALREADY
[I "pw .bI "dIa taI na] for ITS PROBABLY DINNER TIME NOW
In terms of connected speech, Katy’s PCC was 25%, similar to that of her single
word speech. Her speech rate is typically slow and gives the impression of being
syllable timed rather than stress timed.
Design
A single subject research design was used with pre- and post-intervention
assessment carried out. The research design is shown in figure 1. Three phases of
Figure 1. Research design.
196
M. Pascoe et al.
intervention took place with each phase involving a different treatment. Baseline
testing was carried out at two levels: initially a ‘macro assessment’ was carried out to
obtain a broad picture of Katy’s speech processing skills (T1a). A range of speech
and language tests was administered. Data from these tests assisted in intervention
planning, and resulted in the development of a specific set of stimuli to be used in
intervention. Baseline measures (T1b) were then obtained to provide information on
Katy’s processing abilities directly related to the stimuli set to be used in therapy, in
what is referred to in this paper as the ‘micro assessment.’ The difference between
Katy’s speech severity indices (PCC, PVC and PPC) at T1a and T1b was not a
significant one (t(99)52.575, n.s) indicating a stable pre-intervention baseline. Reassessment at the micro level was carried out on completion of each intervention
phase (T2, T3, T4a) to identify changes over the course of intervention. The study is
completed with a return to the macro assessment (T4b) to allow for a more global
evaluation of any changes that have occurred. Finally longer-term follow-up took
place 7 months after the completion of intervention (T5), to evaluate maintenance
of any progress that had occurred. This follow-up assessment involved both macro
and micro assessment. No speech and language therapy took place between T4
and T5.
Macro assessment and intervention planning
(a) The psycholinguistic perspective
The speech processing profile and model developed by Stackhouse and Wells (1997)
were used for organizing and understanding the macro assessment data. The speech
processing profile is a clinical tool which allows clinicians to consider potential levels
of breakdown in the speech processing system by asking questions such as: ‘Can the
child access accurate motor programs?’ and ‘Can the child discriminate between
real words?’ At each level of the profile, tests were carried out in order to
provide answers to these questions. Katy’s speech processing profile is presented in
figure 2.
In terms of input processing Katy had specific difficulties in discriminating
between non-words (level B) as well as between real words (level D). On the
output side she shows a range of difficulties at each of the successive levels.
She experienced difficulties in picture naming tasks (level G), phonological
manipulation (level H, e.g. spoonerism tasks), real and non-word repetition tasks
(levels I and J) and some oro-motor tasks at level K. Results of the receptive
language and literacy assessments not incorporated on the speech processing profile
are shown in Table 3.
The speech processing profile is based on a speech processing model – a box
and arrow model used by Stackhouse and Wells (1997) to conceptualize speech
processing and production. Task analysis and comparison led to the development of
hypotheses about the loci of Katy’s speech processing deficits. Figure 3 illustrates
where Katy’s main areas of difficulty are on this model.
Katy’s difficulties included phonological recognition on the input side, and
retrieval of stored motor programmes, creation of online motor programmes and
motor planning on the output side. These levels of processing are discussed in
further detail below.
Phonological therapy within a psycholinguistic framework
197
Figure 2. Katy’s Speech Processing Profile at age 6;5 (from Stackhouse and Wells 1997) !5age
appropriate performance X51 s.d below the expected mean for her age; XX52 s.d below
the expected mean for her age.
198
M. Pascoe et al.
Table 3.
Summary of Katy’s receptive language and literacy assessment at CA 6;5
Assessment
Receptive language
Test for the
reception of
grammar (TROG;
Bishop 1989)
British Picture
Vocabulary Scale
(BPVS; Dunn et al.
1997)
Literacy measures
Schonell Graded
Reading Test
(Subtest of Aston
Index; Newton and
Thompson 1982).
Schonell Spelling
Test (Subtest of
Aston Index,
Newton and
Thompson 1982)
Area tapped
Standard score
Centile
Age equivalent
Receptive
grammar
95
40
6;0
Receptive
vocabulary
80
10
4;6
Reading
single words
reading age56;8 years
Writing single
words from
dictation
spelling age55;9
& Input processing
Katy experienced many difficulties with the auditory discrimination of real and nonwords, suggesting that she was not able to recognize fine phonological differences,
i.e. phonological recognition (in Figure 3). Her phonological representations were
however, found to be relatively accurate. It is difficult to specifically tap into
phonological representations without processing being affected by weaknesses in
auditory discrimination. Katy performed in an age-appropriate way when asked to
identify rhyme and alliteration from pictures (level F, figure 2). Specially devised
‘posting tasks’ were carried out to provide information about Katy’s phonological
representations and the sound contrasts involved. The format of these tasks was to
introduce Katy to two post-boxes, each one used for posting of a different sound,
e.g. pictures of /b/ words (e.g. BAT, BEE) into a green box, and pictures of /s/ words
(e.g. SAT, SEE) into a red box. Katy was required to pick up picture cards and put
them into the relevant box. The therapist did not name the items for the child. Katy
performed well on these tasks: obtaining or approximating 100% for all tasks with
both word initial and word final contrasts. Most challenging for her was the voicing
contrast of /p/ and /b/ where she favoured the /b/ box and scored correctly on
only 75% of items. Katy did have voicing errors in her speech and typically voiced
plosives pre-vocalically. It was concluded that for Katy phonological representations
are a relative strength.
Retrieval of stored motor programmes and creation of online motor
programmes
Picture naming will involve a child’s access of semantic information and a linked
motor programme. Real word repetition tasks do not necessarily involve access of
&
Phonological therapy within a psycholinguistic framework
199
Figure 3. Speech processing model (from Stackhouse and Wells 1997) with circled areas indicating
Katy’s main areas of difficulty.
lexical representations, and non-word repetition tasks cannot by definition involve
the lexical representation. Non-word repetition tasks tap auditory input processing,
phonological representations (in order to check that there is no phonological
representation) and the mapping of this input form by means of the online motor
programming device onto an output form. Unlike the children described in other
studies who were found to have notable differences between naming and repetition
skills (Ebbels 2000) and real and non-word production (Bryan and Howard 1992),
Katy was found to have equally low scores for each of these tasks. Because nonword repetition and picture naming involve distinct processing routes, it was
hypothesized that both the stored motor programmes (central to picture naming)
and the online motor programming (central to non-word repetition) were affected.
The tasks carried out at level H (figure 2) of the speech processing profile, the
spoonerism task and rhyme generation task, suggested that Katy may have some
difficulty with the internal processing required to manipulate these phonemes and
create new motor programmes. However, because her speech difficulties are severe
it was difficult to accurately judge her ability to carry out these tasks.
200
M. Pascoe et al.
& Motor planning
Motor planning is considered to involve phonetic aspects of speech production,
moving beyond the abstract linguistic knowledge of the previous stage. It is at this
level that co-articulation comes into play. While motor programming is conceived as
being a single word level, motor planning involves the connection of words into
strings of speech. Katy’s speech production at a connected speech level had a similar
degree of accuracy to her single word speech. Many of the words in her connected
speech were produced as single words with primary stress and many pauses between
items that would normally be elided or assimilated in the connected speech of
normally developing children (Newton and Wells 2002). It was considered that once
motor programmes had been revised, motor planning might need to be more
specifically addressed.
Katy had difficulties with both input and output processing. The first important
intervention-planning question was whether to address input (auditory discrimination) or speech production. This is a question that has been frequently debated in
the literature (e.g. Locke 1980, Evershed Martin 1991, Bird and Bishop 1992,
Rvachew 1994, Rvachew, Rafaat and Martin 1999, Waters 2001). Bishop, Brown and
Robson (1990) found that children with severe speech output difficulties - like Katy
- typically have difficulties in discriminating between similar items, particularly nonwords. They suggest this is because such children have had limited opportunities to
experiment and play with related sound strings in the way that normally developing
children do. This theory would suggest that Katy’s discrimination difficulties are
secondary to her speech problems, and thus intervention targeted at the speech
problems may have the effect of improving her discrimination abilities.
Focusing on Katy’s range of output difficulties, which level of processing should
intervention target? Katy has been diagnosed with ataxic cerebral palsy and has
some muscular weakness (level K of the speech processing profile, figure 2).
However, detailed assessment at this level revealed that Katy’s sound production
skills are adequate for speech: she was able to produce all speech sounds in isolation
and had no specific difficulties in terms of the strength, range of movement or coordination exhibited during speech (and non-speech tasks). Katy’s difficulties begin
at the single word level of speech, and her difficulties are concentrated on the output
side of the profile.
It is important to consider children’s strengths and weaknesses in intervention
planning. Katy’s relative strengths included her orthographic knowledge, her
semantic knowledge and her phonological representations. She enjoyed written
language activities and showed age-appropriate letter knowledge. Intervention
would ideally involve activities that give Katy explicit opportunities to use her
strengths. Thus activities could include reading (drawing on her orthographic
awareness), meaningful minimal pair work (drawing on her semantic knowledge)
and tasks involving picture naming which give Katy the opportunity to
access her own (relatively accurate) phonological representation and relate it to
the (inaccurate) stored motor programme, giving her the opportunity to revise the
latter.
(b) The phonological perspective
The Phonological Assessment of Child Speech (PACS) (Grunwell 1985) and
Edinburgh Articulation Test (EAT) (Anthony et al. 1971) were used for the
phonological analyses, to aid in intervention planning and stimuli selection, i.e. what
201
Phonological therapy within a psycholinguistic framework
stimuli would be used in the activities outlined in the previous section? Table 2
outlined the findings from these assessments.
‘Whole word phonology’ is a widely accepted way of conceptualizing children’s
early phonological patterns (e.g. Ferguson and Farwell 1975, Macken 1979, Velleman
and Vihman 2002). In recent years this theory has been applied to clinical settings.
Velleman (2002) describes ‘phonotactic therapy’ as an intervention for unintelligible
children which focuses on building up accurate phonotactic frames, before focusing
on ‘filling up’ the phoneme slots within the frames with accurate phonetic
realizations. Along similar lines, Stackhouse and Wells’ psycholinguistic framework
includes a developmental phase model of speech and literacy in which the ‘whole
word’ phase is one of the earliest phases of speech development, preceding the
‘systematic simplification’ phase in which systematic substitutions dominate
(Stackhouse and Wells, 1997: 197, 2001:410). Katy’s speech output shows
characteristics of the whole word phase: she relies heavily on CV syllable structure.
Expanding her potential syllable structures to include CVC, would enable her to
make a much greater range of lexical contrasts. Although she has some patterns (e.g.
stopping) characteristic of the systematic simplification phase, it was developmentally appropriate to focus intervention on the earlier whole word phase.
The intervention programme aimed to encourage Katy to produce exemplars of
the CVC frame. However, children do not talk in single words, and the ultimate aim
of the intervention was for Katy to use final consonants in CVC words embedded in
sentences. In order to reach these goals, three phases of therapy were devised as
follows:
– Phase I: therapy on a specific set of single words
– Phase II: therapy on a wider range of single words
– Phase III: therapy on connected speech
Three lists of stimuli (lists A, B and C) were devised (table 4). Lists A and B were
designated as treatment lists. In phase I of intervention, lists A and B would be
Table 4.
Matched lists of stimuli
Item number
List A
List B
List C
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
NOTE
PLANE
HEART
NAIL
CAGE
SLIDE
WHEEL
RAKE
STORK
LEAF
SAUCE
ICE
SOAP
PIPE
BARN
ROAD
BOAT
RAIN
CART
WHALE
AGE
HIDE
SEAL
CAKE
FORK
HALF
FLEECE
SLICE
ROPE
PEEP
DAWN
SWORD
GOAT
TRAIN
PART
HAIL
PAGE
LIED
KNEEL
STEAK
WALK
HOOF
PURSE
DICE
GRAPE
SHEEP
LINE
TOAD
202
M. Pascoe et al.
treated in different ways. Words in list A were used in speech production tasks that
included explicit consideration of spelling. This involved making links between
written and spoken word forms targeted in the intervention, i.e. drawing Katy’s
attention specifically to final consonants in written forms. List B was used in speech
production tasks that did not involve written forms. In subsequent phases of
intervention, list A and B were not differentiated. List C was randomly selected as a
non-treatment control list.
The selection criteria for stimuli are listed below.
(a) Each list consisted of 16 monosyllabic words.
(b) CVC words were favoured, although in some instances CCVC words had
to be used since all requirements for matched items could not be met.
Where CCVC words were selected they were matched across the three lists.
(c) Items were matched in terms of rhyme across the three lists. Where rhyme
could not be matched, items were matched by coda segment.
(d) Items were matched across the lists for age of acquisition and spoken
language frequency using the MRC psycholinguistic database (http://
www.psy.uwa.edu.au/mrcdatabase).
(e) Irregular orthographic forms were weighted in terms of their complexity
and balanced across the lists.
(f ) Items were chosen to highlight the functional importance of final
consonants and, thus were words that, with the final consonant removed,
made another real-word minimal pair (e.g. BOAT/BOW). The vowel nucleus
was therefore either a long vowel or a diphthong.
Details of stimuli for the single word phases of intervention have been provided.
Stimuli for the connected speech phase of treatment (phase III) were also chosen
according to phonological criteria. A graded hierarchy of sentences was devised
around each of the target single words shown in table 4, moving from a facilitatory
context to a more demanding one. For example, in the case of the target word ROPE
the facilitatory sentence used as a starting point was: THIS ROPE PULLED THE CAR,
where the onset consonant of the following word PULLED is the same as the coda
consonant of the target word ROPE. Given her phonological abilities at the beginning
of the study, it was thought that Katy should be able to produce the initial [p] in
PULLED even if she omitted the final [p] in ROPE. In order to achieve an acceptable
realization of this final consonant, she would merely have to lengthen the closure
phase for the (single) consonant articulation. At the next level, Katy would be
required to produce a sentence such as THERE’S ROPE ON THE ROAD with the target
ROPE being followed by a vowel. Most challenging was a sentence such as THIS ROPE
GOT FRAYED where she was required to change her place of articulation (and voicing)
between the final [p] in ROPE and the following consonant [c].
Summary
Macro assessment and intervention planning focused on providing answers to two
questions: (a) How will intervention be carried out, i.e. types of activities to be
carried out (b) What would the content of these activities be, i.e. what phonological
stimuli would be used in the activities? Psycholinguistic profiling addressed (a): it
was suggested that intervention should give Katy opportunities to revise and update
Phonological therapy within a psycholinguistic framework
203
her stored motor programmes. Her relative strengths such as reading, making
phonological judgments (that do not rely on auditory input) and using semantic
knowledge to distinguish between minimal pairs in meaningful contexts were
considered important ways that could be used to bring about change in motor
programming. The linguistic approach addressed (b): the specific pattern of motor
programme to be addressed was CVC words since this fundamental template was
lacking from Katy’s motor programme store.
Micro assessment and baseline measures
Pre-intervention baseline assessment was carried out. For each of the items in lists
A, B and C (Table 4), Katy was required to
(a) name pictures of each of the target single words
(b) produce the target single word in a short phrase (‘the TARGET in the
picture’)
(c) write the target single words
(d) discriminate between the target single word and a closely related item that
differed only in terms of the final segment in spoken production (e.g. are
these word pairs the same or different: ROPE and ROW? ROPE and ROTE?).
This set of tasks constituted the micro assessment. It was carried out as a preintervention baseline and after each phase of intervention.
Research questions
This project aimed to determine whether Katy could make significant progress in
speech production when given intensive phonological therapy structured within a
framework of psycholinguistic assessment. More specific questions were formulated
regarding her progress in a range of areas including speech at a single word and
connected level, spelling, and auditory discrimination. Table 5 summarizes these questions.
Intervention
Intervention was carried out in Katy’s school on a twice-weekly basis. Sessions lasted
for one hour each time, with ten hours of intervention being given in each phase.
Therapy took place in a quiet room with only the child and therapist present. All
intervention sessions were audio taped using a SONY MZ-R30 portable minidisc
recorder, with the child and her parents’ consent. Most of the sessions involved
tabletop activities and games, e.g. minimal pair card games, barrier games, board
games, making pictures and reading stories. Katy was 6;8 at the start of the
intervention itself and was 7;5 on completion of the final phase of intervention.
Further details of the intervention aims and activities for each phase are provided
below.
Phase I
Aim: To increase Katy’s awareness of final consonants and encourage her to
produce CVC stimuli items from the treatment lists A and B as listed in table 4.
204
M. Pascoe et al.
Table 5. Questions about the intervention programme
Area
Question
Single word
speech
Connected
speech
Spelling
Auditory
discrimination
&
&
&
Will phase I intervention result in an increased count of final consonants
in the treated word lists A and B?
Will exposure to orthographic forms promote faster learning,
i.e. after phase I will list A (speech and spelling treatment) show more
improvement than list B (speech only treatment)?
After intervention phases I and II, will Katy’s final consonant count
(FCC) for list C (untreated controls) also improve beyond chance level?
Will Katy’s FCC for target words used in a short carrier phrase improve
in phase I and II, as Katy’s speech processing system is modified?
Phase III specifically addresses connected speech. Will it result in significantly
increased FCC’s in connected speech productions in the treatment lists A and B?
After intervention phase III, will Katy’s FCC for list C items
(untreated controls) in connected speech also improve beyond chance level?
Will Katy’s ability to indicate final segments in spelling improve following three
phases of intervention for speech?
Will improved speech production result in improved ability to discriminate
between treatment stimuli and phonetically similar words?
Will increased experience with production of final consonants result in
improved discrimination of novel words that differ in terms of final segments?
List A items were worked on with explicit use of written forms
List B items were given a speech-only treatment without the associated
literacy activities.
In order to avoid confusion between the two types of intervention, oddnumbered sessions addressed list A stimuli with the speech-spelling
treatment, and even-numbered sessions worked on stimuli from list B.
Phase II
Aim: To encourage generalization of CVC production to a broader range of single
words.
&
&
&
Intervention in this phase was guided by broad themes (e.g. animals,
numbers, household objects).
Sessions gave Katy the opportunity to produce CVC words in a wider and
more natural range of contexts.
Written forms of the words were used to remind Katy about her final
consonant production in some instances, together with silent posting tasks
and ‘meaningful minimal pair’ activities (following Weiner 1981) which also
incorporated exposure to written forms.
Phase III
Aim: To facilitate production of CVC items from lists A and B in sentences graded
in terms of phonetic difficulty.
&
&
Intervention in this phase revolved heavily around literacy.
Katy was required to read the stimuli sentences written below illustrations of
the items, as well as matching sentences with appropriate pictures.
Phonological therapy within a psycholinguistic framework
205
Results
Micro evaluation
Results from the micro evaluation are considered in terms of each of the main
areas measured: single word speech, connected speech, spelling and auditory
discrimination.
Single word speech
The results for Katy’s single word speech production over the three phases of
therapy and at each of the assessment points are presented in figure 4.
Overall, a statistically significant main effect for intervention was found using
two-way mixed ANOVA (F(2, 44)538.310, p,.001). Katy’s single word production
was assessed by means of a picture-naming task that formed part of the baseline
micro assessment, and was carried out following each successive phase of
intervention. The focus of the assessment was on Katy’s final segment production:
she was awarded two points for final consonants that were accurately realized, one
point for using an inaccurate final consonant, and no points for omission of a final
consonant. Raw scores were converted to percentages. Two-tailed paired samples
t-tests showed significant gains for each of the stimuli lists from baseline (T1) to
post phase I (T2) (list A treated words, t(15)524.392, p5.001; list B treated words,
t(15)527.652, p,.001; list C untreated words, t(15)24.038, p5.001). Further
significant gains were made with treated items in list A and the untreated items in list
C from T2 to T3 (t(15)523.955, p5.001), following phase II, with the treated items
now approaching ceiling at 94%. However, at T4 reassessment, following
intervention phase III that targeted connected speech, there was a significant
decrease in performance on final consonant production in CVC words for each of
the three lists (list A treated words, t(15)53.416, p,.005; list B treated words,
t(15)54.392, p5.001; list C untreated words, t(15)53.101, p,.05). At long-term
follow-up (T5) following a period with no intervention, significant gains were made
Figure 4. Katy’s single word speech production.
206
M. Pascoe et al.
from T4 (t(31)522.47, p,.05) for the treated words. For the matched untreated
controls there was no significant change after intervention ceased from T4 to T5.
Connected speech
Katy’s ability to produce CVC words in connected speech was assessed by asking her
to repeat the stimuli items in a short carrier phrase. As for the single word speech
assessment, the focus was on Katy’s final segment production of the stimuli words:
she was awarded two points for final consonants that were accurately realized, one
point for using an inaccurate final consonant, and no points for omission of a final
consonant. Raw scores were converted to percentages. Results are shown in figure 5.
A statistically significant main effect for intervention was found using two-way
mixed ANOVA [F(2, 44)5108.477, p,.001]. Initially, Katy found this a very
challenging task and did not use any final consonants in connected speech at the T1
pre-intervention assessment, or T2 or T3 assessments. However, following
intervention phase III which specifically targeted connected speech at the T4
assessment, there was a significant change [F(2, 45)567.623, p,.001] for each of the
three stimuli lists. Her treated word performance for connected speech did not
change significantly after intervention ceased between T4 and T5. For the matched
untreated controls (list C), further significant gains were made from T4 to T5
(t(15)523.093, p,.01) for this set of stimuli.
Spelling
Katy’s written output of the CVC targets was investigated using a single word
dictation task. As for the other assessments, the focus was on Katy’s final segment
production of the stimuli words: she was awarded two points for final consonants
that were correctly written, one point for the use of any final consonant, and no
points for omission of a final consonant. Raw scores were converted to percentages.
Results are shown in figure 6.
Katy’s written representations of the targets were significantly more accurate
than her spoken representations pre-intervention (t(47)55.657, p,.001) and at
Figure 5. Katy’s CVC production in connected speech.
Phonological therapy within a psycholinguistic framework
207
Figure 6. Katy’s written production of CVC stimuli.
long-term follow-up (T5) (t(47)53.483, p5.001). However, the effect size for single
word speech (eta squared5.635) was greater than that for spelling (eta
squared5.390). Overall, a statistically significant main effect for intervention was
found using two-way mixed ANOVA [F(2,44)514.083, p,.001].
For the list A and B treated items, steady but not significant increases were noted
at each assessment when compared to the previous evaluation. The change overall
from T1 to T5 was a significant one (t(31)522.6, p,.05). For the untreated items
in list C, a significant increase was noted when comparing scores from T3 with those
at T4 (t(15)522.739, p,.05). Untreated items improved significantly after the final
phase of intervention and these gains were maintained at the long-term follow-up
at T5.
Auditory discrimination
Katy’s auditory discrimination skills were investigated by asking her to make same/
different judgments about pairs of closely related words (e.g. ROPE/ROTE). Results of
Katy’s performance on this task are presented in figure 7.
At T1 assessment, Katy found the task relatively easy with scores between 70
and 90% for the wordlists A, B and C. Overall there was not a significant main effect
of time. However, some significant changes were noted on the list A items: when
comparing Katy’s performance at T1 with T2 (t(15)522.782, p,.05) and when
comparing her overall discrimination performance on the list A words from T1 with
T5 (t(15)522.150, p,.05). No significant changes were noted for the untreated
non-word items.
Macro evaluation
The macro assessment procedures carried out at the start of the project were readministered on completion of the three intervention phases at T4 in order to
208
M. Pascoe et al.
Figure 7. Katy’s auditory discrimination judgments of closely related CVC word pairs.
evaluate the intervention from a global perspective, and again at T5 some seven
months later to evaluate maintenance of progress following a period with no
intervention taking place.
(a) Psycholinguistic perspective
Tests used to build up the original speech processing profile at CA 6;5 (Stackhouse
and Wells 1997, see figure 2) were re-administered. It was found that Katy’s profile at
CA 7;7 (and CA 8;2) was largely unchanged from the initial profile (figure 8). Only
one of the levels revealed changes: that of level D, which poses the question: ‘Can
the child discriminate between real words?’ Katy showed improvement on three
assessments carried out in order to tap this level (Wepman and Reynolds 1987,
Newton and Thompson 1982, Bridgeman and Snowling 1988). She now performed
at an age appropriate level for each of these tasks. However, at level B (‘Can the
child distinguish between non-words?’) no improvement was noted. On the output
side of the profile no significant gains were made at any level.
Reporting standardized test results for children with severe speech results can
mask true change. Although comparisons with chronologically-matched peer groups
can provide evidence of functional gains, these tests are not sensitive in their
evaluation of children with severe difficulties. The following section provides further
qualitative information about changes in Katy’s speech from a macro perspective.
(b) Phonological perspective
Phonological analyses using the PACS (Grunwell 1985) and EAT (Anthony et al.
1971) were carried out in order to evaluate changes in Katy’s phonology after the
intervention. Results are summarized in Table 6.
Katy made significant gains with her PPC (t(99)524.662, p,.001) when
comparing results from T1 with those at T5. Significant gains were made for PCC
(percentage consonants correct) from pre- to post-intervention (t(99)526.051,
p,.001). Her PVC (percentage vowels correct) did not change significantly from
pre- to post-intervention. Her word final phonetic inventory expanded although she
is not yet able to use all phonemes to realize appropriate contrasts. She is now able
to produce all plosives and nasals - although not always appropriately - in
comparison to her complete lack of consonants available in this position prior to
Phonological therapy within a psycholinguistic framework
209
Figure 8. Katy’s Speech Processing Profile at age 7;6 (from Stackhouse and Wells 1997) with areas
of changed performance highlighted. !5age appropriate performance; X51 s.d below the
expected mean for her age; XX52 s.d below the expected mean for her age.
210
Table 6.
M. Pascoe et al.
Comparison of Katy’s speech data at CA 6;5 (pre-intervention) with CA 8;2 (postintervention)
Assessment
CA 6;5 pre-intervention
Severity
indices
PCC 22%
PVC 74.1%
PPC 41.9%
Word initial position:
[m, n, p, b, d, t, k,
c, f, w, j, r, ]
Word medial position:
[m, b, d, t, k, c, w]
Phonetic
inventory
Word final position:
Stimulability
Phonological
processes
analysis (% use)
Single word
speech sample
Connected
speech sample
All consonants except
[v, ð, #, , t#]
Developmental
processes: cluster
reduction (100%);
final consonant
deletion (96%);
prevocalic voicing
(40%); stopping of
fricatives and affricates
(21%); gliding (21%)
Non-developmental
processes: vowel
distortion (25%)
[bæ] for BAG
[we] for WEB
[vI] for FISH
[gI"me] for CHRISTMAS
[bæ] for PRAM
[e] for EGGS
[bI] for BEES
[k: le: "ab ua ] for
HER LEG WAS
BROKEN
[a l l taI "a g u]
for A LONG LONG
TIME AGO
[I "bet na] for
ITS
BETTER NOW
[aaI aæ "peItapi:
a "wedI] for I HAD
PLAYTIME ALREADY
[I "pw .bi "dIa
taI na] for ITS
PROBABLY DINNER
TIME NOW
CA 8;2 post-intervention
PCC 49%
PVC 73.2%
PPC 58.2%
Word initial position:
[m, n, p, b, d, t, k,
c, f, s, w, j, r, l, ]
Word medial position:
[m, n, p, b, d, t,
k, c, f, w, l]
Word final position:
[m n, , p, b, t, d, k, g]
All consonants except
[v, ð, #, , t#]
Developmental processes:
cluster reduction (70%);
final consonant deletion
(54%); prevocalic voicing
(12%); stopping of fricatives
and affricates (20%);
gliding (21%)
Non-developmental
processes: vowel
distortion (25%)
[bæg] for BAG
[web] for WEB
[vI] for FISH
[gI"me] for CHRISTMAS
[pæm] for PRAM
[e] for EGGS
[bI] for BEES
[daI n d nu: du: f : ak:]
for TRY ON THE NEW
SHOE FOR HER
[ai:
bi: "dI.t a aa li:d]
for
HE BE SITTING
ON A LEAD
[ai: "raIaI a :t] for
HE’S RIDING A HORSE
[aI "g aI "a v fe:] for
HE’S GOING OVER
THE FENCE
[dI f :n dãu de ] for
SHE FELL DOWN
THE STAIRS
intervention. Word initially [s] and [l] now appeared in her repertoire. She still has
difficulties in producing certain consonants in isolation.
The incidence of phonological simplifications in her speech also shows changes,
with final consonant deletion, the process specifically addressed in intervention,
211
Phonological therapy within a psycholinguistic framework
reduced from 96% to 54%. Cluster reduction was reduced from 100% to 70%, with
Katy now frequently attempting to produce both elements of the cluster (e.g. [bwId]
for BRIDGE and [pweIn] for PLANE). /s/ + stop clusters remained challenging for her,
for example [dep] produced for STEP. Pre-vocalic voicing was decreased from 40%
to 12% whereas Katy’s stopping of fricatives and affricates had decreased minimally.
Discussion
This paper aimed to determine if psycholinguistically-based phonological therapy
could lead to (a) specific and (b) generalized improvements in the speech production
of Katy, a child with severe and persisting difficulties. Results from the micro
evaluation showed significant improvements in Katy’s speech – as well as other areas
such as spelling and auditory discrimination between real words. Katy’s response to
the intervention programme is summarized in table 7, which returns to the specific
questions posed earlier.
Single word speech
In terms of speech at the single word level, the following questions were asked: Will
phase I intervention result in an increased count of final consonants in the treated
word lists (A and B)? Will exposure to orthographic forms promote faster learning,
Table 7. Questions addressed by the intervention programme
Area
Single word
speech
Connected
speech
Spelling
Auditory
discrimination
Question
Will phase I intervention result in an increased count of
final consonants in the treated word lists (A and B)?
After intervention phases I and II, will Katy’s final
consonant count (FCC) for list C (untreated controls)
also improve beyond chance level?
Will exposure to orthographic forms promote faster learning,
i.e. after phase I will list A (speech and spelling treatment)
show more improvement than list B (speech only treatment)?
Will Katy’s FCC for target words used in a short carrier
phrase improve in phase I and II, as Katy’s speech processing
system is modified?
Phase III specifically addresses connected speech. Will it result
in significantly increased FCCs in connected speech
productions in the treatment lists A and B?
After intervention phase III, will Katy’s FCC for list C items
(untreated controls) in connected speech also improve
beyond chance level?
Will Katy’s ability to indicate final segments in spelling improve
following three phases of intervention for speech?
Will improved speech production result in improved ability
to discriminate between treatment stimuli and
phonetically similar words?
Will increased experience with production of final consonants
result in improved discrimination of novel words that differ
in terms of final segments?
Answers
yes
yes
no
no
yes
yes
yes
yes
no
212
M. Pascoe et al.
i.e. after phase I will list A (speech and spelling treatment) show more improvement
than list B (speech only treatment)? After intervention phases I and II, will Katy’s
final consonant count (FCC) for list C (untreated controls) also improve beyond
chance level?
After the first phase of intervention, Katy’s use of final consonants in single
words increased significantly not only for the treatment lists but also for the
untreated, matched control set. This suggests that generalized change had been
brought about, rather than being limited to the specific items that she had been
introduced to in the intervention. Intervention phase I offered different treatments
for the stimuli lists A and B, with A items being given a treatment that explicitly
utilised written forms to promote speech, and B items focusing solely on speech.
There was no significant difference in the outcomes from these two different
treatments. This result is surprising given the theoretical argument that spelling
ought to promote speech (e.g. Foorman, Francis, Novy and Liberman 1991, Gillon
2002), and also in terms of the initial identification of Katy’s strengths that included
awareness of written forms. It has been noted that as a non-word reading task was
not administered, Katy may have been utilizing a whole word reading strategy
or visual memorization to recognize the words rather than decoding per se.
Her apparent orthographic knowledge may not be phonologically-based, i.e. she
may be at a logographic stage of reading and lacking the phonological base that
would support speech development. Stackhouse and Wells (1997) link their
developmental phase model with Frith’s (1985) model of literacy development
(figure 9).
It has been suggested that Katy was in the whole-word phase of speech
development at the start of the intervention. This is consistent with her literacy
development being at the logographic stage. Katy progressed to the systematic
simplification phase of speech development, but has yet to master the skills of the
assembly and metaphonological phases of speech development before she will have
the skills associated with Frith’s alphabetic literacy stage.
After the second phase of intervention, further significant gains were made in
single word speech production. Again, this change was not limited to the treatment
lists but also for the untreated, matched control sets suggesting that generalized
change had been brought about. Following the third phase of intervention, which
focused on connected speech, Katy showed a decrease in her production of CVC
stimuli in single word naming tasks. This decrease may be attributable to the focus
of the intervention in each of the phases: phase I and II focused on single word
production and had an effect at this level for each of the three matched stimuli lists.
Phase III involved work on connected speech only: no work was done directly on
single word production. The single word task may have been perceived by Katy as
less important than tasks involving connected speech. On the other hand, a recency
effect may have been acting so that at each post-intervention assessment Katy
performed well on whatever had been addressed most recently in the preceding
intervention, but these gains were not maintained in the longer term. However, data
from T5 (long-term follow-up) suggests that a recency effect was not operating:
Gains in connected speech were maintained after intervention ceased, and the
decline in her single word speech production did not continue, a slight increase in
performance being noted at T5.
It is clear that Katy made significant gains in her speech production at the micro
level. Results from the macro evaluation were less clear-cut. The speech assessments
Phonological therapy within a psycholinguistic framework
213
Figure 9. The relationship between the phases of speech and literacy development: Developmental
model for speech and literacy (from Stackhouse and Wells 1997; Frith 1985).
(e.g. Word-finding Vocabulary Test, Renfrew 1995, Edinburgh Articulation Test,
Anthony et al. 1971) used at level G of the speech processing profile, showed that
Katy had not improved in relation to her peers. However, more sensitive measures
such as speech severity indices revealed that significant improvements had been
made in Katy’s PCC and PPC. Her final consonant deletion had decreased and her
phonetic inventory for the word-final position had increased.
The lack of improvement at the macro level suggests that her motor
programming difficulties, targeted in intervention, are a core deficit in her speech
processing system. The notions of whole-word phonology and phonotactic therapy
(Velleman and Vihman 2002, Velleman 2002) were central to this intervention. They
may account to some extent for the specific speech improvements noted at a micro
level, and not at the macro level. Intervention aimed to establish a new phonotactic
frame in Katy’s motor programming system – and this succeeded; but intervention
to date has not focused on the full and accurate specification of the final consonant
within the template. In terms of the developmental phase model (Stackhouse and
Wells 1997, 2001) Katy’s speech is now more characteristic of the systematic
simplification phase, having been helped to progress from the earlier whole word
phase. If systemic simplifications (notably stopping) can be addressed in future
intervention, more global changes in her speech may be observed.
214
M. Pascoe et al.
Connected speech
Relating the findings from the new speech processing profile (figure 8) to the
speech-processing model (Stackhouse and Wells 1997) enables us to reconsider the
theory underlying Katy’s intervention. It was asked: Will Katy’s final consonant
count for target words used in a short carrier phrase improve in phase I and II, as
Katy’s speech processing system is modified? Katy was not able to make
improvement in her connected speech until phase III when connected speech
was specifically addressed. In terms of the speech processing model, the first two
phases of intervention focused on motor programmes, while the third phase was
targeting motor planning. For Katy, and possibly for other similar children,
generalization of single words into connected speech may be dependent on the
specific targeting of motor planning. The intervention in phase III was very
successful in getting her to use the CVC stimuli in sentences, something which she
had been completely unable to achieve before. It is likely that Katy’s attention had
shifted to the production of larger units of speech which involve motor planning,
rather than being focused at the single word level.
Again, this change was not limited to the treatment lists but also extended to the
untreated, matched control lists suggesting that generalized change had been
brought about. Gains made with connected speech were maintained in the longterm, after a period of no intervention, when T4 and T5 performances were
compared. It seems that improvement in connected speech was only brought about
by specifically addressing connected speech in a carefully structured way. An
important question to consider in future research is whether the single word
intervention phases were necessary prior to the connected speech phase, or whether
intervention might have started with the connected speech work.
Spelling
It was asked if Katy’s ability to indicate final segments in spelling would improve
following three phases of intervention for speech. Katy’s spelling did improve
significantly, even though the specific incorporation of literacy in Phase I did not
seem to have particular benefits for her speech. It may be that in therapy, the effects
of working on speech and spelling are uni-directional: working on speech improves
spelling, but working on spelling does not necessarily improve speech. Alternatively,
there may have been no difference between the progress made on the two stimuli
lists because the inclusion of spelling – albeit only for one of the wordlists – was
sufficient to highlight phonological awareness for Katy and have resulting influence
on the other list.
Given that Katy was being exposed to new words and literacy teaching over
the course of the intervention project, it may not seem surprising that her spelling skills improved over the intervention period. Pre-intervention assessment of
Katy’s spelling using the Schonell spelling test from the Aston Index (Newton
and Thompson 1982) revealed a spelling age of 5;9 years. One year later on
completion of intervention, Katy’s spelling age was 6;10 years. This suggests that her
spelling skills had not increased more than one might expect given the amount of
time that had elapsed. Therefore her improvements in spelling of the micro
assessment stimuli can not be regarded as specific and due to the effects of the
intervention.
Phonological therapy within a psycholinguistic framework
215
Auditory discrimination
For auditory discrimination it was asked: Will improved speech production result in
improved ability to discriminate between treatment stimuli and phonetically similar
words? Will increased experience with production of final consonants result in
improved discrimination of novel words that differ in terms of final segments? Katy
improved at both the micro and macro level in her ability to discriminate between
real words.
How does this improvement relate to the theory of Bishop et al. (1990)
suggesting that children with motor output problems are likely to experience
auditory discrimination difficulties secondary to these output difficulties? Katy may
have made sufficient improvement in her speech to bring about changes in her
auditory discrimination. However, the fact that she improved in her real word
discrimination and not in her non-word discrimination suggests that this is not the
case and that there may be other mechanisms at play. Real word discrimination was
not directly addressed in the intervention, but Katy’s real word discrimination ability
was re-assessed several times between the intervention phases. She may have
improved in this area due to the exposure and practice afforded by the reassessments. The fact that she improved in her real word discrimination but not in
her ability to discriminate between non-words (at even a micro level) suggests that
these are distinct abilities using different processing routes. Phonological
representations are tapped by real word auditory discrimination tasks, where topdown knowledge can be used. Phonological recognition is tapped into in non-word
auditory discrimination tasks which rely on bottom-up processing. The fact that
Katy improved in her real word auditory discrimination and not in her non-word
discrimination suggests that her ability to map from phonological recognition to her
phonological representations had improved so that she was now able to use this topdown processing route more effectively.
Clinical implications
Speech and language therapy often focuses on children’s production of specific
speech sounds or production of single words (e.g. Forrest, Elbert and Dinnsen 2000,
Williams 2000, Barlow and Gierut 2002). Some children are able to apply what
they have learnt at a segmental or whole-word level to conversational speech
(Wright, Shelton and Arndt 1969, Elbert, Dinnsen, Swartzlander and Chin 1990,
Almost and Rosenbaum 1998). However, this is not always the case, and there is
little research addressing the relationship between connected speech and single
word speech production in intervention. Connected speech has important
implications from a functional point of view and in terms of intervention efficiency.
It is important to identify which children will generalize automatically to connected
speech, and which children will not. This may depend on severity, age, the
underlying nature of the speech processing system or the particular phonological processes involved. One child, Zoe, described by Stackhouse and Wells
(1993) had severe speech difficulties but was able to generalize from single
words to connected speech. This child did not have final consonant deletion,
and it may be that this process makes children vulnerable in terms of limited
generalization.
216
M. Pascoe et al.
Conclusion
Katy’s intervention was effective in improving her speech production at the single
word and connected speech level by reducing the frequency of final consonant
deletion and increasing her percentage of consonants correct. However, there are
two important caveats to note. Firstly, the study is limited to one child and results
cannot be generalized to other similar children. The findings from this study need to
be viewed in conjunction with related case study interventions such as those carried
out by Bryan and Howard (1992) and Ebbels (2000), and the expanding body of
intervention research carried out using psycholinguistic frameworks such as
Stackhouse and Wells’ (e.g. Constable et al. 1997, Waters et al. 1998, Dent 2001).
Secondly, one needs to consider the different levels of change that were brought
about in Katy’s speech processing system. The micro evaluation considered specific
changes in the treated and untreated (but carefully matched) stimuli. The macro
evaluation aimed to provide evidence of any global changes in her speech processing
system. The results at each of these levels are summarised in table 8.
Table 8. Changes in Katy’s speech processing system using two levels of evaluation
Speech
Spelling
Auditory discrimination
Real words
Non-words
Micro evaluation*
Macro evaluation**
!
!
!
X
!
X
!
X
!, Significant (p,0.05) change noted; X, no change noted.
*Stimuli shown in table 4 and the appendix
**Based on severity measures for speech (table 6); standardized test results for spelling (table 3) and
auditory discrimination (figure 8).
One needs to consider both macro and micro levels of change in order to gain a
full understanding of the outcomes achieved. We need to consider the two levels of
change as closely interlinked. Ongoing and intensive intervention brings about
micro changes that may eventually result in macro changes. The present study aimed
to evaluate the effectiveness of a particular intervention. The design of the
intervention does not allow for comparisons with other approaches to intervention:
we do not know if other approaches to intervention would have been more or less
effective. It is for this reason that further detailed intervention studies including
wide-ranging outcomes measures are required. However, this present approach
offers an explicit framework for understanding intervention studies and interpreting
the results, drawing on psycholinguistic and phonological theory.
Recent randomized control studies (Glogowska et al. 2000) have seemed to show
that interventions for children with speech and language impairments do not work.
However, in evaluating such studies we need to consider the dosage and nature of
therapy that is given. Clearly, in the case of children with severe, specific and
persisting speech difficulties, intervention can be successful when the intervention is
specific and intensive. The case presented here provides evidence of the value of
direct and specific intervention for a child with severely disordered speech. Law and
Conti-Ramsden (2000) urge practitioners and managers to offer a more flexible
package of interventions, suggesting that the results of a body of evidence-based
Phonological therapy within a psycholinguistic framework
217
practice should be acted upon. Studies such as the one presented here contribute to
that body of evidence.
Acknowledgements
The authors thank Katy, her family and teachers for participating in the study.
Thanks to the local Speech and Language Therapy Agency for support with this
work. This case study formed part of the first author’s PhD research, which was
funded by the Overseas Research Students Award Scheme, and studentships from
the University of Sheffield, UK and University of Cape Town, South Africa.
References
ALMOST, D. and ROSENBAUM, P., 1998, Effectiveness of speech intervention for phonological disorders: a
randomized controlled trial. Developmental Medicine and Child Neurology, 40, 319–352.
ANTHONY, A., BOGLE, D., INGRAM, T. and MCISAAC, M., 1971, Edinburgh articulation test (Edinburgh:
Churchill Livingstone).
BAKER, E., CROOT, K., MCLEOD, S. and PAUL, R., 2001, Psycholinguistic models of speech development
and their application to clinical practice. Journal of Speech, Language and Hearing Research, 44, 685–702.
BARLOW, J., 2001, The structure of /s/-sequences: evidence from a disordered system. Journal of Child
Language, 28, 291–324.
BARLOW, J. and GIERUT, J., 2002, Minimal Pair Approaches to Phonological Remediation. Seminars in Speech
and Language, 23, 57–68.
BERNHARDT, B., 1992, The application of non-linear phonological theory to intervention with one
phonologically disordered child. Clinical linguistics and phonetics, 6, 283–316.
BIRD, J. and BISHOP, D., 1992, Perception and awareness of phonemes in phonologically impaired
children. European Journal of Disorders of Communication, 27, 289–311.
BISHOP, D., 1989, Test for Reception of Grammar (2nd edition) (University of Manchester: Age and Cognitive
Performance Research Centre).
BISHOP, D., 1998, Uncommon Understanding (London: Psychology Press).
BISHOP, D., BROWN, B. and ROBSON, J., 1990, The relationship between phoneme discrimination, speech
production, and language comprehension in cerebral-palsied individuals. Journal of Speech and
Hearing Research, 33, 210–219.
BRIDGEMAN, E. and SNOWLING, M., 1988, The perception of phoneme sequence: a comparison of
dyspraxic and normal children. The British Journal of Disorders of Communication, 23, 245–252.
BROOM, Y. and DOCTOR, E., 1995, Developmental phonological dyslexia: a case study of the efficacy of a
remediation programme. Cognitive Neuropsychology, 12, 725–766.
BRYAN, A. and HOWARD, D., 1992, Frozen phonology thawed: the analysis and remediation of a
developmental disorder of real word phonology. European Journal of Disorders of Communication, 27,
343–365.
BYNG, S., VAN DER GAAG, A. and PARR, S., 1998, International initiatives in outcomes measurement: A
perspective from the United Kingdom. In C. Frattali (ed.), Measuring Outcomes in Speech-language
Pathology (New York: Thieme), pp. 558–578.
CAMPBELL, R. and BUTTERWORTH, B., 1985, Phonological dyslexia and dysgraphia in a highly literate
subject: A developmental case with associated deficits of phonemic processing and awareness.
Quarterly journal of experimental psychology, 37, 435–475.
CHIAT, S., 2000, Understanding children with language problems (Cambridge: Cambridge University Press).
CLARK, N. and ELLIOT, S., 1992, Clinical Audit in the Nursing and Therapy Professions (Maidstone
Hospital, Maidstone Priority Care Group).
CONNERY, V., 1994, The Nuffield Centre Dyspraxia Programme (London: The Nuffield Hearing and Speech
Centre).
CONSTABLE, A., STACKHOUSE, J. and WELLS, B., 1997, Developmental word-finding difficulties and
phonological processing: the case of the missing handcuffs. Applied Psycholinguistics, 18, 507–536.
CROSBIE, S. and DODD, B., 2001, Training auditory discrimination: A single case study. Child Language
Teaching and Therapy, 17, 173–194.
218
M. Pascoe et al.
DENT, H., 2001, Electropalatography: A tool for psycholinguistic therapy. In J. Stackhouse and B. Wells
(eds), Children’s Speech and Literacy Difficulties: Identification and Intervention (London: Whurr
Publishers), pp. 204–248.
DODD, B., 1995, Differential diagnosis and treatment of children with speech disorder (London: Whurr Publishers).
DODD, B. and MCCORMACK, P., 1995, A model of speech processing for differential diagnosis of
phonological disorder. In B. Dodd (ed.), Differential diagnosis and treatment of children of children with
speech disorder (London: Whurr Publishers), pp. 65–90.
DODD, B. and BRADFORD, A., 2000, A comparison of three therapy methods for children with different
types of developmental phonological disorder. International Journal of Language and Communication
Disorders, 35, 189–209.
DUNN, L., DUNN, L., WHETTON, C. and BURLEY, J., 1997, British Picture Vocabulary Scale (2nd ed) (Windsor:
NFER-Nelson).
EBBELS, S., 2000, Psycholinguistic profiling of a hearing-impaired child. Child Language Teaching and
Therapy, 16, 3–22.
EDWARDS, J., FOURAKIS, M., BECKMAN, M. and FOX, R., 1999, Characterizing Knowledge Deficits in
Phonological Disorders. Journal of Speech, Language and Hearing Research, 42, 169–186.
ELBERT, M., DINNSEN, D., SWARTZLANDER, P. and CHIN, S., 1990, Generalization to conversational speech.
J Speech Hear Disord, 55, 694–699.
ENDERBY, P. and EMERSON, J., 1995, Does speech and language therapy work? (London: Whurr Publishers).
EVERSHED MARTIN, S., 1991, Input training in phonological disorder: A case discussion. In M. Yavas (ed.),
Phonological Disorders in Children (London: Routledge), pp. 152–172.
FAHEY, T., GRIFFITHS, S. and PETERS, T., 1995, Evidence based purchasing: understanding results of clinical
trials and systematic reviews. British Medical Journal, 311, 1056–1060.
FERGUSON, C. and FARWELL, C., 1975, Words and sounds in early language acquisition. Language, 51,
419–439.
FOORMAN, B., FRANCIS, D., NOVY, D. and LIBERMAN, D., 1991, How letter-sound instruction mediates
progress in first grade reading and spelling. Journal of Educational Psychology, 84, 456–469.
FORREST, K., ELBERT, M. and DINNSEN, D., 2000, The Effect of substitution patterns on phonological
treatment outcomes. Clinical Linguistics and Phonetics, 14, 519–531.
FRATTALI, C., 1998, Outcomes measurement: definitions, dimensions and perspectives. In C. Frattali
(ed.), Measuring Outcomes in Speech-Language Pathology (New York: Thieme), pp. 1–27.
FREDERIKSON, N., FRITH, U. and REASON, R., 1997, Phonological Assessment Battery (PhAB) (Windsor: NFERNelson).
FRITH, U., 1985, Beneath the surface of developmental dyslexia. In K. Patterson, J. Marshall and M.
Coltheart (eds), Surface Dyslexia (London: Routledge and Keagan Paul).
GIBBON, F., MCNEILL, A., WOOD, S. and WATSON, J., 2003, Changes in linguapalatal contact patterns during
therapy for velar fronting in a 10-year-old with Down’s syndrome. International Journal of Language
and Communication Disorders, 38, 47–64.
GILLON, G., 2002, Follow-up study investigating the benefits of phonological awareness intervention for
children with spoken language impairment. International Journal of Language & Communication
Disorders, 37, 381–400.
GLOGOWSKA, M., ROULSTONE, S., ENDERBY, P. and PETERS, T., 2000, Randomised controlled trial of
community based speech and language therapy in preschool children. British Medical Journal, 321,
923–926.
GRUNWELL, P., 1985, PACS - Phonological Assessment of Child Speech (Windsor: NFER-Nelson).
HOLM, A. and DODD, B., 1999, An intervention case study of a bilingual child with phonological
disorder. Child Language Teaching and Therapy, 15, 139–158.
LAW, J. and CONTI-RAMSDEN, G., 2000, Treating children with speech and language impairments. British
Medical Journal, 321, 908–909.
LOCKE, J., 1980, The inference of speech perception in the phonologically disordered child. Part I: A
rationale, some criteria, the conventional tests. Journal of Speech and Hearing Disorders, 45, 445–468.
MACkEN, M., 1979, Developmental reorganization of phonology: a hierarchy of base units of
acquisition. Lingua, 49, 11–49.
MONAHAN, D., 1986, Remediation of common phonological processes: four case studies. Language speech
and hearing services in schools, 17, 199–206.
MRC Psycholinguistic database: http://www.psy.uwa.edu.au/mrcdatabase
MUTER, V., HULME, C. and SNOWLING, M., 1997, Phonological Abilities Test (PAT) (London: The
Psychological Corporation).
Phonological therapy within a psycholinguistic framework
219
NEWTON, M. and THOMPSON, M., 1982, Aston Index (Wisbech: LDA).
NEWTON, C. and WELLS, B., 1999, The development of between-word processes in the connected speech
of children aged between three and seven. In N. Maassen and P. Groenen (eds), Pathologies of
speech and language: Advances in clinical phonetics and linguistics (London: Whurr Publishers), pp. 67–75.
NORBURY, C. and CHIAT, S., 2000, Semantic intervention to support word recognition: a single-case study.
Child Language Teaching and Therapy, 16, 141–163.
RENFREW, C., 1969, The Bus Story (Oxon: Winslow Press).
RENFREW, C., 1995, Word Finding Vocabulary Test (Oxon: Winslow Press).
RUSCELLO, D., 1995, Visual feedback in treatment of residual phonological disorders. Journal of
Communication Disorders, 28, 279–302.
RVACHEW, S., 1994, Speech perception training can facilitate sound production learning. Journal of Speech
and Hearing Research, 37, 347–357.
RVACHEW, S., RAFAAT, S. and MARTIN, M., 1999, Stimulability, speech perception skills and the treatment of
phonological disorders. American Journal of Speech-Language Pathology, 8, 33–43.
SABEN, C. and INGHAM, J., 1991, The effects of minimal pairs treatment on the speech-sound production
of two children with phonologic disorders. J Speech Hear Res, 34, 1023–1040.
SHRIBERG, L., AUSTIN, D., LEWIS, B., MCSWEENY, J. and WILSON, D., 1997, The Percentage of Consonants
Correct (PCC) Metric: Extensions and Reliability Data. Journal of Speech, Language and Hearing
Research, 40, 708–722.
SNOWLING, M., STACKHOUSE, J. and RACK, J., 1986, Phonological dyslexia and dysgraphia: A developmental
analysis. Cognitive Neuropsychology, 3, 309–339.
SOMMERS, R., LOGSDON, B. and WRIGHT, J., 1992, A review and critical analysis of treatment research related
to articulation and phonological disorders. Journal of Communication Disorders, 25, 3–22.
SPOONER, L., 2002, Addressing expressive language disorder in children who also have severe receptive
language disorder: A psycholinguistic approach. Child Language Teaching and Therapy, 18, 289–313.
STACKHOUSE, J., NATHAN, L., GOULANDRIS, N. and SNOWLING, M., 2002, The relationship between speech
disorders and literacy problems: Identification of the at risk child. Report on a 4 year
longitudinal study. Department of Human Communication Sciences, University of Sheffield.
STACKHOUSE, J. and WELLS, B., 1993, Psycholinguistic assessment of developmental speech disorders.
European Journal of Disorders of Communication, 28, 331–348.
STACKHOUSE, J. and WELLS, B., 1997, Children’s Speech and Literacy Difficulties I: A psycholinguistic framework
(London: Whurr Publishers).
STACKHOUSE, J. and WELLS, B., 2001, Children’s Speech and Literacy Difficulties 2: Identification and intervention
(London: Whurr Publishers).
STIEGLER, L. and HOFFMAN, P., 2001, Discourse-based intervention for word-finding in children. Journal of
Communication Disorders, 34, 277–304.
VANCE, M., STACKHOUSE, J. and WELLS, B., 1994, ‘‘Sock the Wock the Pit-Pat-Pock’’ - children’s responses to
measures of rhyming ability. Work in Progress, National Hospitals College of Speech Sciences, 4, 171–185.
VELLEMAN, S. and VIHMAN, M., 2002, Whole-word Phonology and Templates: Trap, Bootstrap or some of
each? Language Speech and Hearing Services in Schools, 33, 9–23.
VELLEMAN, S., 2002, Phonotactic Therapy. Seminars in Speech and Language, 23, 1–18.
WAGNER, R. and TORGESON, J., 1987, The nature of phonological processing and its causal role in the
acquisition of reading skills. Psychological Bulletin, 101, 192–212.
WATERS, D., 2001, Using input processing strengths to overcome speech output difficulties. In J.
Stackhouse and B. Wells (eds), Children’s speech and literacy difficulties: Identification and intervention
(London: Whurr Publishers), pp. 164–203.
WATERS, D., HAWKES, C. and BURNETT, E., 1998, Targeting speech processing strengths to facilitate
pronunciation change. International Journal of Language and Communication Disorders, 33, 469–474.
WECHSLER, D., 1991, Wechsler Abbreviated Scale of Intelligence (WASI) (Psychological Corporation: UK).
WEINER, F., 1981, Treatment of phonological disability using the method of meaningful contrast: two
case studies. Journal of Speech and Hearing Disorders, 46, 97–103.
WEPMAN, J. and REYNOLDS, W., 1987, Wepman’s Auditory Discrimination Test, (2nd edition) (Western
Psychological Services: Los Angeles).
WILLIAMS, A., 2000, Multiple oppositions: Case studies of variables in phonological intervention.
American Journal of Speech-Language Pathology, 9, 289–299.
WRIGHT, V., SHELTON, R. and ARNDT, W., 1969, A task for evaluation of articulation change III: Imitative
task scores compared with scores for more spontaneous tasks. Journal of Speech and Hearing
Research, 12, 875–884.
220
Phonological therapy within a psycholinguistic framework
Appendix
Qualitative changes in Katy’s single-word speech production
List A: treated words
List B: treated words
List C: untreated
words
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Stimulus
T1
Pre-intervention
NOTE
PLANE
HEART
NAIL
CAGE
SLIDE
WHEEL
RAKE
STORK
LEAF
SAUCE
ICE
SOAP
PIPE
BARN
ROAD
BOAT
RAIN
CART
WHALE
AGE
HIDE
SEAL
CAKE
FORK
HALF
FLEECE
SLICE
ROPE
PEEP
DAWN
SWORD
GOAT
TRAIN
PART
HAIL
PAGE
LIED
KNEEL
STEAK
WALK
HOOF
PURSE
DICE
GRAPE
SHEEP
LINE
TOAD
nu
peI
a
neI.
keI
daI
wi.
reI
d .g
jif
d
aI
d
paI
ba
w
b
weI
ka
weI.
eI
aI
si
keI
f
a
fi
daI
w
pi
d
d
gu
teI
ba
e I.
peI
laI
ni
deI
w
u
bk
daI
keI
i
jaIn
tu
T2
T3
T4
n ut
peIn
ha
neI.j
keI
taIt
wi.
jeIk
tk
lI
d
aI
d up
baIp
ba
w
b ud
weIt
kat
weI.j
eI
a It
dI.
jeIk
f
a
fI t
tslaIt
w up
p Ip
d
dd
k ut
teI
pa
eI.j
bed
jaI
nij
teI
wg
uf
pk
daI
geI
ti
jaI
tu
n ut
pen
hat
ni.j
keIt
tsaId
wi.
weIk
tk
liv
d
a It
d up
paIp
ban
w uv
b ud
reIn
kat
w I. l
e Id
e Id
si.
keIk
fk
af
fit
saI
w up
p Ik
dn
dd
g ut
teI
pat
e I. Il
peId
laId
ni.
steIk
wk
huf
pk
daI
geIp
sip
laIn
t ud
n ut
pweI
at
neI.
keIt
taIt
wi.
weIk
dk
lif
d
aI
du
paI
ban
wu
b ut
weIn
ka
weI.
e Id
aI
di.
keIk
f
a
fwi
daI
w up
p Ip
du
d
g ut
t#eI
ba
e I.
beI
laIf
ni.
deIk
wk
uf
pkt
daI.
geIp
#ip
laI
du
T5
Follow-up
n ut
plem
hat
neI.j
keIk
daId
wi.j
weI
dk
ji
d
aIt
d up
paIp
ban
w ud
b ut
weIn
ka
weI
eI
aI
si.j
keIk
fk
af
f It
daIt
w up
b Ip
dn
sd
gu
teIn
ba
eI.j
beIt
jaId
ni.
deIk
wk
uf
pkd
daIt
geIp
s Ip
jaIn
du