TUGAS : PHONOLOGICAL THERAPY WITHIN A PSYCHOLINGUISTIC FRAMEWORK: PROMOTING CHANGE IN A CHILD WITH PERSISTING SPEECH DIFFICULTIES

 

PHONOLOGICAL THERAPY WITHIN A PSYCHOLINGUISTIC FRAMEWORK: PROMOTING CHANGE IN A CHILD WITH PERSISTING SPEECH DIFFICULTIES

Created by :

Name               : Anni Alimatul Azizah

St. Number      : 1501070224

Class                : B

 

A.    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 speech-disordered 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).

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 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.

B.     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 , which returns to the specific questions posed earlier.

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,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?

Questions addressed by the intervention programme

Area	Question	Answers
Single Word Speech	Will phase I intervention result in an increased count of final consonants in the treated word lists (A and B)?	Yes
	After intervention phases I and II, will Katy’s final consonant count (FCC) for list C (untreated controls) also improve beyond chance level?	Yes
	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)?	No
Connected Speech	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?	No
	Phase III specifically addresses connected speech. Will it result in significantly increased FCCs in connected speech productions in the treatment lists A and B?	Yes
	After intervention phase III, will Katy’s FCC for list C items (untreated controls) in connected speech also improve beyond chance level?	Yes
Spelling	Will Katy’s ability to indicate final segments in spelling improve following three phases of intervention for speech?	Yes
Auditory Discrimination	Will improved speech production result in improved ability to discriminate between treatment stimuli and phonetically similar words?	Yes
	Will increased experience with production of final consonants result in improved discrimination of novel words that differ in terms of final segments?	No

 

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 ).

Figure: The relationship between the phases of speech and literacy development: Developmental model for speech and literacy (from Stackhouse and Wells 1997; Frith 1985).

 

 

 

 

 

 

 

 

 

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 (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.

a)      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 long-term, 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.

b)      Speech

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 spel-ling 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.

c)      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 re-assessments. 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 top-down 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 top-down processing route more effectively.

d)     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 phono-logical 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.

C.    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.

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 practice should be acted upon. Studies such as the one presented here contribute to that body of evidence.

D.    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.

BISHOP, D., 1989, Test for Reception of Grammar (2nd edition) (University of Manchester: Age and Cognitive Performance Research Centre).

MUTER, V., HULME, C. and SNOWLING, M., 1997, Phonological Abilities Test (PAT) (London: The Psychological Corporation).

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.

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.