UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigating sentence shaper : a processing prosthesis Albright, Erin 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-0354.pdf [ 9.1MB ]
Metadata
JSON: 831-1.0092659.json
JSON-LD: 831-1.0092659-ld.json
RDF/XML (Pretty): 831-1.0092659-rdf.xml
RDF/JSON: 831-1.0092659-rdf.json
Turtle: 831-1.0092659-turtle.txt
N-Triples: 831-1.0092659-rdf-ntriples.txt
Original Record: 831-1.0092659-source.json
Full Text
831-1.0092659-fulltext.txt
Citation
831-1.0092659.ris

Full Text

INVESTIGATING SENTENCE SHAPER: A PROCESSING  PROSTHESIS  by ERIN ALBRIGHT B A . , The University o f British Columbia, 2004  A THESIS S U B M I T T E D IN P A R T I A L F U L L F I L L M E N T OF THE R E Q U I R E M E N T S F O R THE D E G R E E OF  M A S T E R OF SCIENCE  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Audiology and Speech Sciences)  T H E U N I V E R S I T Y OF BRITISH C O L U M B I A October 2006 © E r i n Albright, 2006  ABSTRACT This study explores the possible uses o f Sentence Shaper (Linebarger, M c C a l l , & Berndt, 2004), a software program designed as a language remediation tool for individuals with nonfluent aphasia. In particular, Sentence Shaper allows users to practice message production by facilitating sentence construction. Previous studies indicate that after extended use o f this program, individuals with nonfluent aphasia may be able to produce more morphosyntactically complex narratives. A s Sentence Shaper allows users to record and save messages, there is the potential that this program could be used to augment communication. The goals o f this study were, first, to partially replicate the Linebarger et al. (2004) study and, second, to explore ways in which Sentence Shaper could be used to augment communication in everyday life. These goals were investigated in a four-month case study with a woman with nonfluent aphasia and her mother. Models o f social approaches to aphasia intervention informed the design o f the study, which included a treatment component and an ethnographic component. For the treatment component, the participant with aphasia practiced producing messages using Sentence Shaper for 12 weeks. Treatment effects were measured by comparing pre- and post-treatment unaided narratives, as well as post-treatment aided narratives. The ethnographic component involved regular meetings with the participants. Integration o f qualitative data from fieldnotes, recorded conversations, brief interviews, and questionnaires revealed the uses o f Sentence Shaper and the ways in which the participants' life situation affected its use. Treatment results demonstrated an increase in the morphosyntactic complexity o f the participant's narratives, while measures o f informativeness and narrative structure  ii  remained relatively unchanged. Given these conflicting findings, a judgment task was also conducted with two groups o f listeners (speech-language pathologists and peers). Findings from the ethnographic component revealed that, although the participant with aphasia used Sentence Shaper messages for e-mail and in conversation, neither participant readily accepted the use o f the program to augment communication in everyday life. Reasons for this lack o f acceptance are explored. Finally, ways in which a social approach contributed to a deeper understanding o f the findings, as well as implications for future research and clinical practice, are discussed.  iii  T A B L E OF CONTENTS ABSTRACT  ii  T A B L E OF C O N T E N T S  iv  LIST OF T A B L E S  viii  LIST OF FIGURES  ix  LIST OF E X C E R P T S  x  ACKNOWLEDGEMENTS  xi  C H A P T E R ONE: R E V I E W OF T H E L I T E R A T U R E 1.1 Introduction 1.2 ICF as an Organizing Framework 1.2.1 Description 1.2.2 Application o f the I C F to Aphasia 1.3 The Nature o f Nonfluent Aphasia 1.3.1 The Impairments o f Nonfluent Aphasia 1.3.2 Effects o f Nonfluent Aphasia at the Activity/Participation Level 1.3.3 The Social Impact o f Aphasia 1.4 Treatment o f Aphasia 1.4.1 Treatment within an ICF Framework 1.4.2 Treatment Addressing the Impairment 1.4.3 Treatment Focusing on the Activity Level 1.4.4 Targeting Impairment and Activity Limitations through a Processing Prosthesis 1.5 A A C and Aphasia: Focusing on Participation 1.5.1 Conversation and the Impact o f A A C 1.5.2 Recent Advances i n A A C Technology 1.5.3 Sentence Shaper: Further Possibilities 1.6 Research Objectives CHAPTER TWO: M E T H O D O L O G Y  1 1 5 5 7 10 10 14 18 21 21 21 25 27 33 38 40 42 43 46  2.1 Introduction 2.2 The Case Study 2.3 Overview 2.4 Participants 2.4.1 Initial Participant Guidelines 2.4.2 Ethical Considerations 2.4.3 Communication Profile  iv  46 46 47 48 48 50 51  2.5 Component One: Procedures 2.5.1 Sentence Shaper Training 2.5.2 Data Collection 2.5.2.1 Narrative samples 2.5.3 Data Analysis 2.5.3.1 Morphosyntactic analysis...., 2.5.3.2 Informativeness and efficiency analysis 2.5.3.3. Story structure analysis 2.6 Component Two: Procedures 2.6.1 Data Collection 2.6.1.1 Participant observation 2.6.1.2 Interviews 2.6.1.3 Sentence Shaper log 2.6.1.4 Quality o f Communication Life Scale and Communicative Effectiveness Index 2.6.2 Data Analysis 2.6.2.1 Analysis o f fieldnotes and interviews 2.6.2.2 Transcription o f conversations 2.6.2.3 Analysis o f Q C L and C E T I 2.7 Relating Findings to the Research Objectives C H A P T E R THREE: E M E R G E N T FINDINGS 3.1 Introduction 3.2 Study Participants 3.2.1 Communication Profile Results 3.2.2 Nicole as a Communicator 3.3 Outline o f Study 3.4 Data Set 3.4.1 Narrative Samples 3.4.2 Interviews, Conversations, and Sentence Shaper Messages 3.5 Challenges o f Communicating with Nicole CHAPTER FOUR: RESULTS  52 52 53 53 54 54 54 55 56 56 56 57 57 58 58 59 60 62 62 63 63 63 64 65 67 69 69 70 71 75  4.1 Use o f Sentence Shaper: Overview 4.2 Component One 4.2.1 Quantitative Production Analysis 4.2.1.1 Preparation o f the data 4.2.1.2 Q P A analysis o f unaided narratives 4.2.1.3 Q P A analysis o f aided narratives 4.2.2 Correct Information Unit Analysis 4.2.2.1 C I U analysis o f unaided narratives 4.2.2.2 C I U analysis o f aided narratives 4.2.3 Analysis o f Story Schema and Informative Content i n The Cowboy Story Unaided and A i d e d Narratives  v  75 75 76 76 77 80 81 81 83 83  4.2.4 Listeners' Perceptions 4.2.4.1 Perceptions o f pre- and post-treatment narratives 4.2.4.1 Perceptions o f post-treatment unaided and aided narratives 4.2 Component T w o 4.2.1 Participation Opportunities: Uses o f Sentence Shaper in Everyday Life 4.2.1.1 The training phase 4.2.1.2 Nicole's strategies for creating Sentence Shaper messages 4.2.1.3 Use o f Sentence Shaper with e-mail 4.2.1.4 Use o f Sentence Shaper messages i n conversation 4.2.1.5 Sentence Shaper as language therapy 4.2.2 Contextual Factors 4.2.2.1 Personal factors: Nicole as a communicator 4.2.2.2 Environmental factors: The social context 4.2.3 Themes and Patterns: Interpretive Support 4.2.3.1 Language remediation as meaningful activity 4.2.3.2 "Sentence Shaper is not real life" 4.2.3.3 Seeking perfection 4.2.3.4 Nicole as a dependent communicator 4.2.3.5 Nicole as an independent user o f technology 4.2.4 Outcomes o f Sentence Shaper 4.2.4.1 Narrative versus conversational speech 4.2.4.2 Final participant responses to Sentence Shaper 4.2.4.3 The participants' ideas for future use o f Sentence Shaper 4.2.4.4 Q C L and C E T I outcomes CHAPTER FIVE: DISCUSSION  85 86 88 90 91 92 93 94 95 96 97 98 102 105 105 107 112 114 117 118 119 120 121 123 127  5.1 Introduction 5.2 The Researcher's Perspective 5.2.1 Comparison o f N i c o l e ' s Performance to Linebarger et al.'s (2004) Participants 5.2.2 Possible Explanations for these Differences 5.2.3 Implications for Processing versus Linguistic Approaches to Intervention of Nonfluent Aphasia 5.2.4 Generalization to Conversational Speech 5.2.5 The Integration o f Sentence Shaper into Conversation: Lessons from Nicole 5.3 The Participants' Perspectives 5.3.1 Changes in Communication Ability 5.3.2 Acceptance o f Sentence Shaper as an A A C Device 5.4 A n Integration o f Perspectives 5.5 H o w D o the Principles o f a Social M o d e l A p p l y to this Study?  vi  127 128 128 131 134 137 138 139 139 141 143 144  5.6 5.7 5.8 5.9  Implications for Clinical Practice Limitations o f the Study Directions for Future Research Conclusion  146 147 147 148  REFERENCES  150  APPENDICES  159  Appendix A : Consent Forms Consent Form for Person with Aphasia Consent Form for the Communication Partner Aphasia Friendly Consent Form Appendix B : Certificate of Approval for Ethics Appendix C : Interview Guide Appendix D : K e y to Transcription Appendix E : Schedule of Data Collection Appendix F: List of Sentence Shaper Messages Appendix G : Transcribed Narratives Appendix H : Q C L and C E T I Scores The Quality of Communication Life Scale The Communicative Effectiveness Index  vn  160 160 163 165 178 179 181 182 185 187 192 192 193  List o f Tables Table 3.1  Nicole's assessment scores  65  Table 3.2  Summary o f Transcribed Conversations  70  Table 4.1  Excerpts from The Three Little Pigs pre and post narratives. Underlined portions are those words included i n the Q P A  78  Q P A analysis results for Nicole's unaided narrative samples pre/post use o f Sentence Shaper  79  Q P A analysis results for aided narrative samples following use of Sentence Shaper  80  Table 4.4  Sentence Shaper Narratives  80  Table 4.5  C I U analysis results for Nicole's unaided narrative samples pre/post use o f Sentence Shaper. Results presented Linebarger et al. (2004) are provided for comparison  82  C I U analysis results for Nicole's aided narrative samples. The time required to create the narratives is also included  83  Table 4.2  Table 4.3  Table 4.6  Table 4.7  The percentage o f core propositions found in the setting, complication, and resolution o f The Cowboy Story unaided and aided narratives 84  Table 4.8  Listeners' perceptions o f The Three Little Pigs and The Cowboy Story unaided narratives  87  Listeners' perceptions o f The Three Little Pigs and The Cowboy Story post-treatment unaided narrative versus the aided narrative  89  Table 4.9  Table 4.10 Sample o f Sentence Shaper message and corresponding message unaided  viii  109  List o f Figures Figure 1.1  The Sentence Shaper work area (Linebarger, et al., 2004)  ix  29  List o f Excerpts Excerpt 3.1  ^1  Excerpt 3.2  73  Excerpt 4.1  93  Excerpt 4.2  9  4  Excerpt 4.3  9  9  Excerpt 4.4  1  0  4  Excerpt 4.5  1  1  0  Excerpt 4.6  1  1  0  Excerpt 4.7  1  1  2  Excerpt 4.8  1  1  2  Excerpt 4.9  1  1  3  Excerpt 4.10  1  1  9  Excerpt 4.11  1  2  2  Excerpt 4.12  1  2  3  ACKNOWLEDGEMENTS I would like to express my deep gratitude and thanks to D r . Barbara Purves. Y o u r enthusiasm for this project, endless support, guidance, patience, constructive input, and editing have been greatly appreciated throughout this process. Our numerous discussions over the course o f this project have increased my excitement and passion for the profession o f speech-language pathology and for working with people who have aphasia. For this I am truly grateful. Thank you to Dr. Jeff Small and Dr. Judith Johnston for agreeing to be members of the thesis committee. Y o u r suggestions, contributions, and support have been valuable and much appreciated. I wish to express my profound thanks to Nicole and Eileen who took part in this study over the course o f four months. Their perspectives contributed greatly to my understanding o f aphasia and the use o f technology for communication. I am extremely appreciative o f the time and effort Nicole and Eileen committed to this project. I am also grateful to the eight people who listened to Nicole's narratives and gave their perceptions. Their insights added an interesting aspect to this study. Thank you to the British Columbia Medical Services Foundation for the financial support which allowed me to pursue this project. I would also like to thank Dr. Marcia Linebarger for providing me with a copy o f Sentence Shaper to use i n my project. I am extremely thankful to my family, especially my M o m and Dad, who have provided me with the much needed emotional and financial support throughout my university career. It is this support which has given me the confidence to believe that I can achieve whatever I put my mind to. Thank you to my friends for taking an interest in my study and providing encouragement throughout this process. A special thanks goes to Caroline Sniatynski for editing the many drafts o f this thesis. Finally, I wish to thank Markus Tschopp for being my rock throughout this entire process. I could not have done this without your unwavering support and encouragement.  xi  Chapter One: Review o f the Literature  1.1 Introduction Aphasia is an acquired language disorder resulting from brain damage which can cause impairments in verbal comprehension and production, reading, and writing (Chapey & Hallowed, 2001). Each o f these language modalities can be impacted to varying degrees in individuals with aphasia. There are various presentations o f aphasia, and many treatment methods have been developed for each type o f aphasia. O f particular interest in this study is nonfluent aphasia, which is characterized by slow, effortful speech with relatively short utterances. A computer software program, previously referred to as a communication system ( C S ; Linebarger, Schwartz, Romania, K o h n , & Stephens, 2000, Linebarger, Schwartz, & Kohn, 2001, Linebarger, M c C a l l , & Berndt, 2004) and now called Sentence Shaper, has been developed as a language remediation tool for this type o f aphasia. This software works as a processing prosthesis to support individuals with nonfluent aphasia in sentence production and allows the user to record his or her voice to produce longer utterances and to build these single utterances into longer narratives. Previous studies have shown that following an extended use o f this program, some individuals with nonfluent aphasia demonstrated increased structural and grammatical complexity in narrative production. The impact o f aphasia can have far-reaching effects. The ways in which language impairments affect a person with aphasia i n his or her everyday life is a complex issue and is dependent on each person's life situation. Aphasia can have devastating effects on an individual's ability to participate in work and social activities (Hinckley, 2002).  1  Relationships and social roles are often affected as well as the individual's perceived life satisfaction, confidence, and self-identity (Shadden, 2005). Clearly, treatment o f aphasia involves more than language remediation, but also includes supporting the individual, so that communicating in everyday life with his or her current language ability is possible. Sometimes the natural speech o f a person with aphasia is not sufficient in adequately expressing his or her message. In these instances, individuals with aphasia may rely on alternative and augmentative communication ( A A C ) to supplement or replace their existing natural speech. Because Sentence Shaper allows users to record and save messages they have created, there is the potential that the program could be used for more than language remediation. O f interest in this study is how Sentence Shaper could be incorporated into the communication strategies o f an individual with nonfluent aphasia to facilitate everyday communication. Traditionally, treatment o f aphasia focused primarily on the remediation o f language impairments. This medical model approach to healthcare concentrates on fixing the problem, but, with a chronic impairment such as aphasia, a complete recovery is often not possible. In the past two decades, there has been a movement in healthcare to adopt a more holistic, social approach to assessment and treatment o f health conditions. With this change, the emphasis shifts from merely "fixing" the impairment to an overall assessment o f how a person is able to function in society. Treatment o f aphasia is now frequently conducted under the framework devised by the W o r l d Health Organization's ( W H O ) International Classification o f Functioning, Disability and Health (ICF) which incorporates ideas from both the medical and social models ( W H O , 2001). In the revised ICF, the focus has shifted from the "consequences o f disease" to an emphasis on  2  "components o f health" (p. 4). The I C F framework takes into account all variables contributing to one's ability to live with one's health condition. The first chapter w i l l discuss the various components o f the I C F and their application to aphasia. The revised I C F is clearly an important change i n the way we view health and disability. While the I C F is not a social model, it considers health in a social context which was lacking in previous healthcare frameworks. The social aspect o f the I C F encourages clinicians to address issues such as quality o f life and social functioning. However, Hewitt and B y n g (2003) caution that our therapy goals for people with aphasia should not be entirely focused only on the functional aspects o f living, but rather should help people with aphasia become 'engaged' in life. To them engagement has many aspects, but overall it means "being involved in something which has real purpose, feels personally meaningful and valuable" and provides "a sense o f real connection to other people with whom you can identify" (p. 63). The I C F and the social model o f healthcare suggest that people with aphasia should not be viewed in isolation, but rather as members of a community, and therapy should therefore be focused on improving participation and engagement within that community. To explore the impact o f Sentence Shaper on language remediation and on communication in everyday life, several areas o f literature must be reviewed in order to fully understand this topic. A thorough explanation o f the I C F framework is necessary to describe the different factors which contribute to an individual's health and level o f functioning. In addition, the principles underlying a social approach to healthcare w i l l be reviewed. Following this, a description o f the impairments, activity and participation limitations and treatment approaches associated with nonfluent aphasia is provided.  3  Theoretical accounts o f nonfluent aphasia are discussed in the context o f the treatment approaches described. Subsequently, a discussion o f A A C is presented with regard to acceptability and usability, as well as recent advances in A A C technology for individuals with aphasia. These topics w i l l be covered in the first chapter. The purpose o f this study is to investigate the ways in which Sentence Shaper can be used. The first objective o f the study is to partially replicate and extend previous language treatment studies done with this program. The second objective is to further explore the possibility that Sentence Shaper could be used i n everyday life to aid i n communication. Accordingly, acceptance and usability issues w i l l also be explored.  4  1.2 ICF as an Organizing 1.2.1  Framework  Description The World Health Organization ( W H O ) is a department within the United  Nations with the mandate to ensure that all people are able to obtain the highest level o f health possible ( W H O , 2001). The W H O ' s Constitution defines health as including not only the physical aspects o f health, but mental and social well-being as well. The International Classification o f Functioning, Disability and Health (ICF) is a classification system devised by the W H O which serves many purposes ( W H O , 2001). It is extremely relevant to researchers within health disciplines as it provides a scientific framework for studying health. A l s o , the I C F aims to improve communication among health professionals by establishing a universal language to use when discussing and describing health and health-related states. The revised I C F is composed o f two parts: functioning/disability and contextual factors. A person's functioning and disability is conceptualized as a dynamic interaction between his or her health condition, the environment, and personal factors. Within the I C F , functioning is defined as " a l l body functions, activities, and participation" ( W H O , 2001, p. 8) and disability encompasses "impairments, activity limitations, or participation restrictions" (p. 8). Body functions are defined as "the physiological functions o f body systems" and impairments are problems with these functions (p. 10). In relation to language, an example o f a body function would be the "the specific mental functions o f recognizing and using signs, symbols and other components o f a language" (p. 58). Activities are specific tasks or actions done by an individual and activity limitations are difficulties executing those actions. Finally, participation is defined as an individual's  5  "involvement in a life situation" (p. 10), and participation limitations occur when an individual has difficulty taking part in normal life situations. A n example o f an activity/participation aspect o f language is conversation occurring with one or many people. The I C F identifies two components o f contextual factors: environmental factors and personal factors. A n individual's environment includes physical, social, and attitudinal aspects which impact daily life. Personal factors are defined as an individual's background, including factors such as gender, race, and age. The I C F regards a person's functioning as "a dynamic interaction between health conditions (diseases, disorders, injuries, traumas, etc.) and contextual factors" (p. 8). Because a person's health and functioning is conceptualized within the ICF framework as an interaction between individual and environmental factors, a person's disability is not regarded as residing solely within the individual. In this classification system, each component contains a list of domains which can be quantified in terms of their level o f functioning or disability. Thus, the I C F provides a framework for the assessment and treatment o f health conditions in which each level (i.e. impairment, and activity and participation) can be targeted. The I C F regards activity and participation as one component as it is often difficult to separate the two aspects. However, in Ross and Wertz's (2005) discussion o f the ICF and its application to aphasia, they separate activity and participation into two components which can be targeted in the assessment o f and intervention for aphasia. They imply that activities focus on only the individual whereas participation is the individual's life situation and, thus, is not isolated to the individual, but includes other  6  people as well. For the purpose o f this study, activity and participation w i l l be discussed as two separate components following Ross and Wertz's distinction.  1.2.2. Application  of the ICF to Aphasia  Traditionally, assessment and treatment o f aphasia have been based on a medical model approach to disability and functioning focusing purely on the individual with aphasia. Those who adopt a medical model view a decrease in functioning as the result o f an individual's health condition, and therapy typically focuses on relieving that impairment ( W H O , 2001). However, a complete recovery is often not possible (Simmons-Mackie, 2001). With its emphasis on the impairment level o f disability, the medical model is limited in its application to the treatment o f aphasia. This approach does not take into consideration how the person with aphasia actually functions i n the real world with the help o f compensatory strategies such as gesture and communication aids, and o f conversation partners. Furthermore, the collaborative aspect o f conversation is ignored when an individual's language ability is viewed in isolation. In the past decade, a social approach to aphasia has been emphasized. The I C F moves beyond a medical model and is defined as a "biopsychosocial" model o f health (p. 20). A definition of aphasia which is consistent with the ICF framework comes from Simmons-Mackie (2001); here, aphasia is defined as "an impairment due to brain damage in the formulation and reception o f language, often associated with diminished participation i n life events and reduced fulfillment o f desired social roles" ( p . 248). Ross and Wertz (2005) propose that assessing and treating individuals with aphasia in the context o f the ICF framework w i l l allow for "socially valid treatment" (p.  7  861). However, they state that the measures currently available may not lead to socially valid treatment as they were developed prior to the ICF. Some assessment tools may not measure a level discretely, but may also include information about other levels o f the ICF, which according to Ross and Wertz, is problematic. They suggest that assessment tools addressing each level o f the I C F independently as well as measures o f quality o f life for individuals with aphasia need to be developed. In response to Ross and Wertz's application o f the I C F to aphasia, Penn (2005) suggests that developing standard measurements o f the I C F levels w i l l i n fact limit our understanding o f the intricacies o f an individual's language impairment. She argues that the constructs o f the I C F cannot be measured independently as the very nature o f the framework includes interaction between each level. In Perm's view, it is not possible "to fragment and to separately delineate some o f these domains" (p. 876) especially for an area as complex as language. Furthermore, Ross and Wertz fail to mention longestablished qualitative methods such as ethnography and conversation analysis which have been used successfully to investigate contextual factors. In closing, Penn asserts that new measures are not required but, rather, that we need to "move to understanding and explaining things in their context, using valid and established philosophies o f assessment ... that are attuned to the complexity o f the human condition" (p. 878). A s previously discussed, the I C F is not a social model, but rather, an integration of medical and social models, which considers the contributions o f an individual's social environment to his or her level o f functioning. The I C F is broad enough that different models of healthcare can be accommodated within the framework; therefore, adopting the framework o f the I C F does not mean that assessment and treatment w i l l necessarily  8  be conducted with a social approach. In the literature, the term social model or approach is often used, although rarely defined. B y n g and Duchan (2005) recognized this problem and proposed five principles underlying the social model philosophy. The first principle is "equalizing the social relations o f service delivery" (p. 908) which proposes that users of healthcare services should have more control over the services provided. This principle highlights an idea which is at the forefront o f the social model: people with disabilities are "experts in their own conditions" (p. 907), and therefore, service providers should learn from these experts and deliver the services people with disabilities feel are necessary. B y n g and Duchan's second principle states that service users should be involved authentically in their own care. In particular, they should be able to make decisions about the services provided in their situation, as well as healthcare in general. The third principle o f the social model is that clinicians should provide therapy which allows for engaging experiences for the service user. Engaging experiences are considered those which are valued by the service user and others and are also associated with activities found in daily life which are personally meaningful. The fourth principle proposes that service users should have control o f organizations which provide service to them. In order for authentic control to be achieved, people with disabilities should be involved in all levels o f the organization and decision making processes. The fifth and final principle states that healthcare service providers should be accountable to the service users, which "requires not only the provision o f high-quality direct services but also that organizations involve, listen to, and inform users in understandable ways about various facets o f the organization" (p. 909).  9  It is evident that the social model extends beyond the framework o f the I C F to include principles which suggest ways o f developing relationships between healthcare service providers and users and providing service to people with disabilities. Therefore, Ross and Wertz's (2005) claim that appraisal o f aphasia within the I C F framework w i l l allow for socially valid treatment is not entirely accurate. While the I C F takes into account social aspects relating to a person's health and functioning, there are features o f the social model, as described by B y n g and Duchan, which are not specified in the I C F . A s a framework, the I C F describes what areas o f health and functioning require consideration; however, the framework, which encompasses both social and medical approaches, does not explain how healthcare should be conducted.  1.3 The Nature of Nonfluent  Aphasia  Thus far, the I C F framework and the principles o f a social model o f healthcare have been described with attention given to the application o f this framework and model to aphasia. A t this point it is necessary to describe the effects o f nonfluent aphasia in terms o f the impairment and activity/participation limitations, as these factors have important implications for intervention.  1.3.1 The Impairments of Nonfluent  Aphasia  Individuals with aphasia may present very different impairments depending on the extent to which various aspects o f language, oral production, auditory comprehension, reading, and writing have been impaired (Goodglass, Kaplan, & Barresi, 2001). In terms o f production, aphasia can be categorized into two subtypes: fluent and  10  nonfluent aphasia. Nonfluent aphasia is characterized by effortful, slow speech output with relatively short utterances. In contrast, individuals with fluent aphasia speak with normal intonation patterns, articulation, and utterance length, but their productions often suffer from inappropriate word and sound substitutions called paraphasias. Comprehension deficits are associated with both these types o f aphasia in varying ways. Within the category o f nonfluent aphasia, further differentiations have been suggested with regard to an individual's tendency to produce or omit grammatical morphemes. Saffran, Berndt, and Schwartz (1989) attempted to describe the nature o f nonfluent production and to investigate the differences thought to clinically distinguish two types o f nonfluent aphasia. Agrammatic speech is characterized by the frequent omission o f "some bound and free-standing grammatical morphemes" (p. 446). In contrast, individuals with nonfluent, nonagrammatic aphasia do not demonstrate this omission o f grammatical morphemes in their nonfluent speech. Saffran and her colleagues developed a quantitative production analysis for aphasic speech, to determine whether the procedure could reliably differentiate the productions o f individuals clinically judged to have agrammatism versus nonfluent, nonagrammatic speech. In this production analysis, the individual's telling o f a classic fairytale was analyzed in terms o f various morphological and structural characteristics. The quantitative production analysis was found to be very reliable in distinguishing the participants with agrammatic speech from those with normal speech as well as from those with nonfluent nonagrammatic speech. The two groups with nonfluent aphasia were not found to differ in terms o f speech rate, with both demonstrating a reduction in rate relative to normal speakers. When the productions were measured for structural elaboration o f subject noun phrases  11  and verb phrases, both the agrammatic and nonagrammatic groups produced sentences which were much simpler than the normal control participants. Furthermore, no difference was seen in the frequency of embedding with both the agrammatic and nonagrammatic groups showing a reduction in embedding. However, Saffran et al. were able to find several measures which differentiated the two groups. In terms o f morphological measures, the participants with agrammatic speech demonstrated a significant reduction in the use o f closed class words, obligatory determiners, pronoun usage, verb inflection, verb usage, and elaboration o f the verb. In contrast, the participants with nonfluent, nonagrammatic speech were not found to differ significantly from the normal control participants on these measures. The participants with agrammatic speech were found to have much more difficulty producing propositional utterances, while the nonfluent, nonagrammatic participants were less impaired in this measure and were not significantly different from the normal control participants. W i t h regard to the proportion o f sentences that were well-formed, all three groups performed differently. The participants with agrammatic speech had significantly more ill-formed speech than that o f the nonfluent, nonagrammatic participants, who were i n turn much worse than the normal control participants. Interestingly, a difference was noted in the source o f the ill-formedness, with the participants with agrammatic speech failing to produce the required grammatical morphemes and the participants with nonfluent, nonagrammatic speech typically omitting the required arguments. It must be noted, however, that while each as a whole conformed to the clinical judgments made initially when the subjects were separated into groups, inconsistencies were apparent with some of the participants with agrammatic speech showing normal ability for particular  12  measures. Furthermore, although the two groups were statistically different, there was continuity between the groups for both the morphological and structural measures. Saffran et al. conclude with several questions. First, given the apparent continuity between the groups o f agrammatism and nonfluent, nonagrammatism, should nonfluent nonagrammatic speech be characterized as mild agrammatism or categorized separately? They also question whether all nonfluent speech contains some degree o f agrammatism, thus casting further doubt on the argument for distinct categories. Clearly, further research in this area is warranted to determine whether these groups can be qualitatively distinguished in addition to the quantitative differences found by the researchers. The categories o f agrammatism and nonfluent, nonagrammatism were investigated further by B i r d and Franklin (1996) in a study using the Q P A methodology developed by Saffran and her colleagues. Five individuals with aphasia participated i n the study: one was judged to be agrammatic, nonfluent; two were nonagrammatic, nonfluent; and two o f the participants had fluent aphasia. A s in the first Q P A study, participants were asked to tell the story o f Cinderella. The differences on the Q P A measures between the two nonfluent groups reported by Saffran et al. were largely replicated in B i r d and Franklin's study. In addition, the deficits o f the fluent participants were found to be similar to the nonfluent, nonagrammatic participants in several measures (i.e., well-formedness, elaboration, and word-class measures). This replication adds further support to the quantitative differences found to separate the categories o f agrammatism and nonfluent, nonagrammatism, but does not address the question o f whether these categories are distinct.  13  The findings o f these studies raise important issues when analyzing research done in the field o f nonfluent aphasia. First, i f these disorders are qualitatively different, it is possible that the language impairments seen in agrammatism and nonfluent, nonagrammatism should be targeted with different interventions. Secondly, i n evaluating existing research, descriptions o f participants are often unclear in regards to precisely what the language characteristics are. Different assessment tests and various labels (i.e. Broca's aphasia, nonfluent aphasia, agrammatism) are used when assessing and describing the participants in these studies. Thus, it is often difficult to determine what type o f aphasia the results o f the studies were based on, making comparability across studies challenging.  1.3.2. Effects of Nonfluent Aphasia at the Activity/Participation  Level  Clearly the categories o f agrammatism and nonfluent, nonagrammatism within nonfluent aphasia are problematic. However, it is not only at the level o f impairment where variability in the speech characteristics o f nonfluent aphasia can occur, but at the activity and participation levels in different speaking contexts as well. When one examines the discourse o f an individual with aphasia (consistent with the I C F ' s definition of activity as a specific task or action done by an individual), several interesting observations can be made. Typically, samples of agrammatic speech are obtained through picture description, retelling a fairy tale, or a semistandardized interview (Heeschen & Schegloff, 2003). In these contexts, both substitutions and omissions have been found to occur in agrammatic speech. These elicitation techniques have the effect o f maximizing the turns o f the individual with aphasia as the interviewer generally withholds his or her  14  conversation and gives the individual with aphasia a long time to respond. However, this contrasts with normal conversation in which there is a bias for turn minimization. When examining agrammatic speech in natural conversations, it is very different from samples obtained through test-like situations. Here, agrammatic speech is characterized almost entirely by omissions o f grammatical elements with almost no substitutions, resembling what is known as telegraphic speech (Heeschen & Schegloff, 2003). This phenomenon can also be explained by comparing the varying task requirements o f narrative and conversation. A narrative typically necessitates the use o f more formal language than a conversation and also requires the expression o f perspective and tense, both o f which require the use o f specific morphemes and syntax, whereas in a conversation this is not a requirement. Therefore, when individuals with agrammatism are telling a narrative, substitutions may occur when they attempt to express the morphosyntactic elements the story requires. In contrast, those morphosyntactic elements may not play as integral a role in the expression o f meaning in a conversation and are, therefore, typically not included. It is evident from Heeschen and Schegloff s observations that even when picture description techniques are used to assess an individual's verbal output, a complete representation o f an individual's language abilities may not be achieved. Clearly, the task demands o f an interaction play an important role in determining the speech characteristics o f those with agrammatism; thus, assessment o f activity limitations should be done in multiple contexts. Individuals with nonfluent aphasia often develop different strategies to adapt to their language impairment. The context o f conversation is an important area to assess when analyzing the language characteristics o f people with aphasia, given that it presents  15  an opportunity for the individual with aphasia to incorporate conversational partners as a strategy. For example, in conversation, individuals with nonfluent aphasia may have the conversational partner expand their impoverished utterances to facilitate the transfer o f ideas. Some individuals may also use fdlers, such as "um", to hold their place in the conversation while they formulate an utterance. These behaviours are consistent with definitions such as Simmons-Mackie and Damico's (1997), which define compensatory strategies as "a new or expanded communicative behaviour, often spontaneously acquired and systematically employed, to overcome a communication barrier i n an effort to meet both transactional and interactional communicative goals" (p. 770). The transactional function involves the exchange o f information and content, whereas the interactional function serves a social purpose o f establishing and maintaining relationships (Brown and Yule, 1983). This definition was created after Simmons-Mackie and Damico observed the strategies used by two participants with aphasia i n multiple contexts. The new definition they created differs i n several ways from traditional definitions o f compensatory strategies. Firstly, their definition suggests that these strategies serve both transactional and interactional communicative goals. Previously, compensations were thought o f as performing primarily a transactional function in message transmission; however, they found that compensatory strategies were used for interaction at a social level as well. Secondly, this definition puts forth the idea that compensatory strategies are more often natural and spontaneously acquired than those trained in therapy. The results from this study illustrate that the majority o f strategies were either new behaviours spontaneously used i n conversation or normal communicative behaviours which occurred at a higher than normal frequency or were  16  used for different functions. While trained strategies were utilized, they were less frequent and used primarily for a transactional purpose. Perm and Cleary (1988) investigated the compensatory strategies used by six individuals with a closed head injury. Their findings indicate that the participants employed a wide variety o f compensatory strategies; however, not all o f these strategies were deemed effective. These results highlight the fact that some strategies utilized by individuals with aphasia may in fact be maladaptive or inefficient in conversation. When analyzing the speech o f individuals with nonfluent aphasia, it becomes apparent that verbal output is not the only method used to get one's message across. Individuals with aphasia and those with whom they regularly communicate are often extremely creative in the different ways they are able to 'talk'. Compensatory strategies such as gesture, writing, and drawing are often used simultaneously with or in place o f natural speech. Nonfluent aphasia can create activity and participation limitations in speaking, listening, reading and writing. Examples include difficulty having conversations with others, making phone calls, and reading the paper. O f course, the way deficits in these areas affect daily life is dependent on each individual's life situation. Individuals with aphasia may attempt to overcome these deficits by developing compensatory strategies. A s Simmons-Mackie and Damico's study suggests, the strategies individuals with aphasia develop on their own are often used more often and more effectively than those suggested by the speech-language pathologist.  17  1.3.3 The Social Impact of Aphasia A social model o f aphasia requires therapy goals which move beyond focusing on the impairment. Simmons-Mackie (2001) suggests that intervention o f aphasia should "promote membership in a communicating society and participation in personally relevant activities for those affected by aphasia" (p. 246). This statement broadens the view of aphasia therapy to consider the individual within a social environment, with research in this area motivated by the need to understand the social impacts o f aphasia and how best to support people living with those effects. The social consequences for an individual affected by aphasia can be devastating. A person's ability to work and participate in social activities, as well as his or her perceived life satisfaction (Hinckley, 2002), confidence and identity (Shadden, 2005) can all be affected by aphasia. Furthermore, Kagan (1995) claims that feelings o f reduced competency may also result because people with aphasia are often no longer able to demonstrate their social and cognitive competencies through communicating with others. Ross and Wertz (2002) reviewed research which indicates that adults with chronic aphasia score significantly lower on quality o f life ( Q O L ) measures than non-braininjured adults (Ross, 1999, as cited in Ross and Wertz, 2002); however, they report that this decrease in Q O L is not closely associated with language ability. Drawing on work by the W H O Quality o f Life Group, Q O L is defined as people's "perceptions o f their position in life in the context o f the culture and value systems where they live and in relation to their goals, expectations, standards, and concerns" (Ross & Wertz, 2003). The World Health Organization's Quality o f life Instrument ( W H O Q O L - 1 0 0 ) identifies 24 facets contributing to Q O L grouped into six domains: physical, psychological,  18  independence, social, environment, and spiritual. In an effort to determine which facets are most relevant for individuals with aphasia, Ross & Wertz (2003) compared the results on the W H O Q O L - 1 0 0 between groups o f individuals with aphasia and non-brain-injured individuals. They found that the domains o f independence, social relationships, and environment differentiated the two groups. More specifically, within the area o f independence, activities o f daily living, getting around, and working were distinguishing factors between the two groups. In the domain o f social independence, satisfaction with support from friends and with their sex lives were distinguishing aspects o f Q O L , and within the environment domain, accessibility o f information, health services, and transportation separated the aphasia and non-brain-injured groups. However, there is an important caveat to this research finding in that the researchers compared individuals with aphasia to non-brain-injured people. With this comparison, it is not possible to separate the contribution o f aphasia to Q O L versus that o f a brain injury without language impairment. Furthermore, as the W H O Q O L - 1 0 0 was not developed to specifically measure the effects o f communication on Q O L , it is questionable, for instance, whether difficulties "getting around" stem from deficits in communication or from physical impairments. Another study exploring Q O L for people with aphasia used assessments developed specifically to explore the effects o f communication on Q O L . Code (2003) used the Social Network o f Aphasia Profile ( S N A P ) and the Communicative Effectiveness Index (CETI) to investigate the relationship between aphasia and social activities. In this study, Code makes the assumption that the time people spend taking part in different activities outside the house contributes to one's "functional outcome,  19  subjective well-being, and social support" (p. 388). The results from 38 people with chronic aphasia reveal that they spent an average o f 20 hours per week out o f the house, with individual results ranging from 1.5 to 60 hours. In contrast to Ross and Wertz's findings, severity o f aphasia was the most significant factor contributing to these results, with age and physical and motor limitations as additional related factors. Overall, these studies suggest that specific factors which contribute to one's quality o f life can be affected negatively by aphasia. A n important component missing from the previous studies is the voice o f the people with aphasia. Scores on assessment tests and calculations o f time spent in social activities provide a rather limited account o f the quality o f life o f those living with aphasia. What is needed is research which explores the perspectives o f people with aphasia by discussing these topics with them. Although interviewing people with aphasia is not an easy undertaking, L e Dorze and Brassard (1995) investigated perceptions o f the consequences o f aphasia by interviewing eleven pairs o f individuals with aphasia and a relative or friend. The "handicaps" or participation limitations identified by the participants with aphasia included difficulties in communication and interpersonal relationships, loss o f autonomy, restricted activities, and stigmatization. Furthermore, the researchers reported that the majority o f these consequences were related to the participant's language impairments. These results are important as they begin to illuminate what people living with aphasia view as the consequences o f their communication impairment. Clearly, an individual's language impairment can have devastating effects on many aspects o f daily life and on the ability to participate fully in society. Consideration o f the impairments and activity and participation limitations  20  associated with nonfluent aphasia has led to a number o f treatment approaches, which w i l l be discussed in the next section.  1.4 Treatment of Aphasia 1.4.1. Treatment within an ICF  Framework  The ICF proposes that in order to overcome disability, each component o f a person's health condition (i.e., impairment, activity, and participation) must be addressed. Therefore, when developing a treatment plan for someone with aphasia, the aim should be improvement in functioning in all components o f the I C F . However, treatment o f aphasia typically targets recovery o f language function at the impairment and activity levels only. Intervention at these levels is often based on improving verbal output, but may also target reading and writing. What this type o f intervention fails to address is how the individual with aphasia is able to function in his or her life situation. Often a full recovery o f functioning at the impairment and activity levels is not possible and individuals with aphasia are left with a chronic language impairment which can significantly affect their lives. Intervention at the participation level is required to find ways to facilitate the involvement o f an individual with aphasia in his or her home and community life in ways that are important to that person.  1.4.2 Treatment Addressing the Impairment One model that has proven central to many interventions o f agrammatism is Garrett's model o f speech production (e.g., Byng, Nickels, & Black, 1994; Jones, 1986; Lesser, 1990). In this model, Garrett proposes three levels in the speech production  21  process: a conceptual level, a language-specific sentence level, and a motor level (Fromkin & Bernstein Ratner, 1998). The conceptual level is the level at which the message is generated. The sentence level is composed o f two levels: the functional level and the positional level. A t the functional level, lexical selection and thematic role assignment occur. This information is then used in the positional processing stage, where lexical items are mapped onto the morphosyntactic structure o f the sentence and receive phonological representation. Finally, the motor level is responsible for the motor commands necessary to produce the articulatory movements o f speech sounds i n the vocal tract. Researchers and speech-language pathologists have attempted to treat agrammatism at the impairment level in a number o f different ways; some are based on empirical observations o f sources o f difficulties for individuals with nonfluent aphasia while others are based on theoretical accounts o f agrammatism. In a review o f treatment literature, Schwartz, Fink, and Saffran (1995) suggest that the following language areas could be targeted in intervention: "morphosyntactic production; retrieval o f verbs and verb-stated mapping rules; assignment of thematic roles to syntactic constituents (i.e. mapping)" (p. 105). These components can be seen as loosely reflecting the functional and positional levels proposed in Garrett's model o f speech production. Schwartz et al. allude to several different types o f techniques aimed at treating morphosyntactic production. For instance, behavioural or direct-production therapy targets morphology by shaping production o f specific responses. A n example is H e l m ' s Elicited Language Production Program for Syntax Stimulation ( H E L P S S ) which focuses on retrieval o f morphosyntax (Helm-Estabrooks, Fitzpatrick, & Barresi, 1981). It is based  22  on the assumption that retrieval o f specific forms w i l l improve by having individuals with agrammatism hear and say a variety o f sentences o f a particular syntactic form. The program consists o f 11 ranked sentence types o f increasing difficulty (i.e. imperative intransitive, imperative transitive, Wh-interrogative, declarative transitive, declarative intransitive, comparative, passive, yes-no questions, direct and indirect object, embedded sentence, and future). This hierarchy is based on a study done by Gleason, Goodglass, Green, Ackerman, and Hyde (1975) in which 14 sentence types were elicited from a group o f individuals with Broca's aphasia. These results allowed the order o f difficulty for the sentence types to be empirically determined, in contrast to approaches in which the order o f difficulty is established on theoretical grounds (e.g. Friedmann & Grodzinsky, 1997, Hagiwara, 1995, Schwartz et al., 1987). While improvements in sentence production following intervention with H E L P S S are documented (Helm-Estabrooks & Ramsberger, 1986), there is currently no evidence suggesting that these improvements generalize to untrained syntactic forms or to spontaneous speech. In fact, results from a study done by Doyle, Goldstein, and Bourgeois (1987) demonstrate that improvement was limited to the sentence types trained using H E L P S S ; generalization did not occur to untrained sentence types, and the effects on spontaneous speech were not explored. Verb retrieval approaches to intervention are exemplified by Marshall, Pring, and Chiat's (1998) case study which involved a treatment program used to facilitate verb retrieval and sentence production. The participant's speech was characterized as agrammatic with a significant verb deficit. When cued with a verb, the participant was more successful at producing a correct sentence than i f cued with a noun. Therefore, it  23  was hypothesized that therapy aimed at improving access to verbs would also benefit sentence production. In the experiment, 35 verbs, from five semantic categories, were treated. Production o f these verbs and 35 control verbs from the same five categories were tested pre- and post-treatment by a picture description task. Treatment o f the verbs began with comprehension and reading tasks, in which the participant was asked to match a written verb to a picture and choose from a series o f verbs which one was the odd one out. After this was completed, the participant was presented with a noun and asked to say an associated verb and was asked to produce a verb after hearing a scenario. Following 20 hours o f treatment, the participant's verb naming o f the trained verbs increased significantly and a small, though nonsignificant increase was also found for the control verbs. In addition, although never treated directly, sentence production involving both the treated and control verbs increased significantly in the picture description task. Mapping therapy, another treatment approach, is based on the hypothesis that individuals with agrammatism have a deficit in their ability to map thematic roles onto sentence constituents (Thompson, 2001); in other words, there is a breakdown from the functional to positional level in Garrett's model. Treatments based on mapping theory focus explicitly on training the relationships between verbs and their arguments. This deficit has been targeted in different ways in a variety o f research studies. One study by Jones (1986) targeted the mapping deficit through written sentences. The program involved multiple steps in which the participant was asked to identify the verb, the agent, and the object, facilitated by prompting the participant with ' w h o ' or 'what' questions. Once the participant could consistently identify these aspects, verb structures were introduced which could be prompted with the questions 'where', 'when', ' w h y ' , and  24  ' h o w ' , and the complexity o f the sentences was increased. In this program, the participant was not required to produce any o f the targeted structures, and verbal production was discouraged during therapy. Following this therapy, the participant's sentence comprehension improved and spoken language became more structurally complex. These examples o f therapies targeting morphosyntactic production, verb retrieval, and mapping o f thematic roles focus on different aspects o f language affected by agrammatism and are similar in that each focuses on the impairment level. These therapies aim to improve processing and production o f a specific language component, more specifically, retrieval o f morphosyntax, retrieval o f verbs, and mapping o f thematic roles.  1.4.3. Treatment Focusing on the Activity Level With activity level interventions, attention to the task is particularly relevant for individuals with agrammatism in light o f the variability in speech output in different interactional tasks (i.e., narratives versus conversation). Heeschen and Schegloff s observations suggest that conversation supports the use o f telegraphic speech by individuals with agrammatism. In contrast to the impairment level therapy approaches previously discussed, which attempt to increase the structural and grammatical complexity o f speech, other treatment approaches focus on compensating for the impairment rather than reducing it and are thus considered activity level interventions. A technique based on this principle is Reduced Syntax Therapy ( R E S T ; Schlenck, Schlenck, & Springer, 1995, as cited in V a n de Sandt-Koenderman & Bonta, 1998). In R E S T , people with severe agrammatism are supported in combining words to create  25  agrammatic utterances. Emphasis is placed on choosing words which w i l l provide the most information as opposed to trying to produce complete sentences. R E S T is based on K o l k ' s idea o f agrammatism in which he suggests that the "telegraphic speech" o f individuals with agrammatism is a "positive symptom" (Van de Sandt-Koenderman & Bonta, 1998, p. 218). K o l k argues that there is no loss o f syntactic knowledge or ability in agrammatism, but rather a slowing o f the system that handles the computation and retrieving o f grammatical morphemes (Kolk, 1995). He proposes that this slow-down o f syntactic computation can lead to de synchronization during syntactic tree formation. This also leads to an asynchrony between the time when a syntactic slot is produced and the time when the grammatical morpheme is retrieved. Therefore, many errors occur when individuals with agrammatism attempt to produce morphemes that come from a complex part o f the syntactic tree. K o l k hypothesizes that because o f this slowing down, the sentence production system must adapt. This adaptation is thought to occur at the message level o f Garrett's model, which is the "level at which the 'what-isto-be-said' is represented in conceptual terms" (p. 293). Due to these individuals' reduced resource capacities, there is an overload on the system at this level which results in a tendency to select simpler message level verb-argument structures. The agrammatic speech these individuals produce is a strategy adopted as a way to adapt to this difficulty. K o l k (1998) proposes that adaptation to the timing deficit occurs i n two ways: simplification and restarts. Restarts occur when a speaker attempts to say the same thing multiple times, thereby benefiting from the activation remaining from previous attempts and reducing the time needed to reach the activation threshold. K o l k (1995) suggests that telegraphic speech is chosen because it reduces the load on short-term memory. The other  26  option available to individuals with agrammatism is to use very slow, simplified, grammatical speech; however, K o l k argues that this would present too much difficulty for the limited short-term memory capacity. From a treatment perspective, K o l k proposes that therapy should rather stimulate this adaptive strategy instead o f focusing on increasing complexity o f sentence production (Van de Sandt-Koenderman & Bonta, 1998). However, as K o l k ' s account o f agrammatism addressing the impairment o f nonfluent aphasia, consideration o f the task is important when using this intervention approach, in light of Heeschen and Schegloff s (2003) findings.  1.4.4. Targeting Impairment and Activity Limitations through a Processing  Prosthesis  Recently a software program has been developed for individuals with nonfluent aphasia, o f both the agrammatic and non-agrammatic type, to use in language remediation. This program, originally referred to as a communication system (CS) (Linebarger, Schwartz, Romania, Kohn, & Stephens, 2000, Linebarger, Schwartz, & Kohn, 2001, Linebarger, M c C a l l , & Berndt, 2004) and now called Sentence Shaper, was designed to support these individuals in the production o f longer and more grammatically complex utterances. Sentence Shaper is regarded as a 'processing prosthesis' which provides a form o f compensation for the users by decreasing the processing demands o f language production; thus, Sentence Shaper can be considered a therapy tool that addresses activity limitations. However, language improvements have been found to occur in users' unaided speech following extended use o f the program, so in addition to addressing activity limitations, the program also can be considered to target the impairment level.  27  Linebarger et al. (2001) suggest that by having processing support provided, individuals with nonfluent aphasia w i l l be able to access sentence structure unavailable to them in their unaided speech. This idea builds on K o l k ' s hypothesis that individuals with aphasia have reduced resource capacities which results in their impoverished speech. However, unlike K o l k , who suggests that individuals with nonfluent aphasia should be encouraged to use telegraphic speech, Linebarger et al. propose that providing these individuals with processing support w i l l allow them to produce more complex sentences. These proposals contrast with linguistic approaches to the intervention o f agrammatism, such as mapping therapy and H E L P S S , which suggest that the impaired linguistic structures should be targeted directly in therapy. Sentence Shaper is able to provide processing support by allowing users to "maintain sentence elements long enough to assemble them into larger structures and to monitor and complete these intermediate structures without the time pressure o f normal conversation" (Linebarger, et al., 2000, p. 419). The software program allows users to record their own speech which is then represented on the computer screen by a visual icon. The user can replay the message by clicking on the icon and can also add more to the message, building it into sentences and narratives. Sentence Shaper also has two forms o f lexical support. There is a word finder which allows the user to access a "hierarchical list o f printed words and phrases" (Linebarger and Schwartz, 2004, p. 5). In addition, there are side buttons found on both sides o f the work space which display frequently used prepositions, verbs, pronouns, and question words. For both tools, the user clicks on the button to hear the word and then he or she must record the word to put  28  it in the sentence shaping area. Figure 1.1 displays the Sentence Shaper work area, including the side buttons. Access to the word finder is depicted by a question mark.  In  For  I  Go  P" ' |  Make  1  o  ! To ! | With  j On  From | 1  '•  j  Under  Over  About 1 i  ; Behind  j Give i Want Take 1 Put  j  1  Say  Have  : went  Ask  Had  j Was  At |'  He  She  What [who !  1  You  It  They  Before  When  ; Where  :  r—~~j After  [  i  | Why  Figure 1.1 The Sentence Shaper work area (Linebarger, et al., 2004)  Sentence Shaper is supports language production in a number o f ways. The system provides users with memory support because they are able to replay what has already been recorded. This allows the user unlimited time to construct his or her message and helps to alleviate the processing load. The program allows the individual to avoid typing as it can be a difficult activity for individuals with aphasia who often have some form o f right side paresis; in addition, individuals may also have difficulty with retrieval o f written words. Furthermore, speaking into a microphone is faster than typing which increases the time efficiency o f this program. In order to aid with word finding  29  difficulties, Sentence Shaper also has the preprogrammed lexical support in the form o f the word finder and side buttons (Linebarger & Schwartz, 2004). A significant benefit o f the program is that the recordings are generated by the individual with aphasia and are not dependent on input from family members or caregivers. In the study investigating the A A C device, P R O S E , the fact that the narratives were constructed by family members was found to be limiting for the person using the device (Waller, Dennis, Brodie, & Cairns, 1998). Studies investigating Sentence Shaper have found some promising results. In one study, six individuals with agrammatism used Sentence Shaper for approximately 15 sessions, creating narratives on topics o f personal interest, before being tested on narrative production (Linebarger, et al., 2000). The participants were then asked to watch silent-film videos and produce free narratives and narratives using Sentence Shaper, but without the linguistic support available in the side buttons and word finder. For five out of the six participants, the results indicated that Sentence Shaper significantly facilitated production o f morphosyntactic structure. It was noted that the one participant who did not show significant improvements i n language production performed at chance on a grammaticality judgment task while the other participants performed well above chance. These language improvements indicate that the program may be beneficial for language therapy and suggest that direct intervention o f linguistic structures may not be required. In a second study, two participants took part in a treatment program which involved using Sentence Shaper and another language therapy program incorporating natural language understanding (Linebarger, et al., 2001). The therapy program, i n which locative and directional modifiers were trained, was preceded and followed by  30  approximately 15 hours o f practice in which the participants used Sentence Shaper at home with little instruction from the experimenters. A s in Linebarger et al.'s (2000) study, the word finder and side buttons were disabled so that no linguistic support was provided, and participants were instructed to construct narratives o f personal interest. Both participants showed improvement i n their production, with increases i n median length o f utterance and mean sentence length. Furthermore, the participants demonstrated increases in complex argument structure as well as improvements in verb naming. Linebarger et al. hypothesized several possible reasons for these improvements. Firstly, it is possible that the program was able to teach the strategy o f "chunking" output into small units o f word or phrase size, thus reducing the demands on short-term memory. This strategy also encourages the user to move forward despite word-finding difficulties. Secondly, the language improvements may have resulted from the large amount o f practice these participants had in producing free narratives, rather than simple picture descriptions. Finally, it was speculated that the processing support provided by Sentence Shaper may have allowed the participants to use structures that they had previously been unable to access except in very constrained tasks. A third study involving Sentence Shaper required two individuals with nonfluent aphasia to use the program for 11 weeks independently after a training period (Linebarger, et al., 2004). Each week, participants were asked to watch a different episode o f a television program and then use Sentence Shaper to narrate the episode. In contrast to the preceding studies, participants had access to both the word finder and the side buttons to aid in narrative production when needed. Results o f the study found significant improvements in one participant's production o f unaided spoken narratives  31  involving more structured and informative speech. The second participant, who had a more severe lexical impairment, did not show the same improvements for unaided speech, but with the lexical support provided by Sentence Shaper, he was able to produce more structured narratives. Clearly, with the proper support, the second participant was able to improve his communicative ability. In each o f Linebarger et al.'s studies, narratives were used as the focus o f intervention, with participants instructed to create narratives using Sentence Shaper about personally relevant topics or television episodes. However, it is interesting to note that the effect o f narratives on language remediation was not discussed in these studies. Peach and Wong (2004), drawing on Garrett's model o f speech production, questioned whether addressing the message level in treatment could influence sentence level processes. After a therapy program which involved the retelling o f narratives, the individual with agrammatism was found to have improved expressive syntax, although the results o f their study do not clearly identify the influence o f narratives on intervention. In Peach and Wong's (2004) and Linebarger et al's (2000, 2001, 2004) studies, narratives are targeted in intervention; however, the effect o f the treatment on narrative structure is not investigated, leaving the interaction between sentence processing and supra-sentence level processes (i.e., story schema) unexplored. Linebarger and Schwartz (2005) discuss the possibility o f Sentence Shaper as an A A C device. They suggest that technology developed for people with aphasia can provide two types o f support: direct and indirect. Similar to the Sentence Shaper's word finder and side buttons, direct support provides linguistic information to the user in the form o f words, phrases, sentences, or structural templates, to the user. Indirect support, in  32  contrast, helps the user take advantage o f his or her preserved language production abilities. In addition to providing direct lexical support, Sentence Shaper is also considered a form o f indirect support as the program offers processing support to facilitate production o f longer and more structurally complex utterances. A s Linebarger et al.'s (2000, 2001) studies demonstrate, some users o f technology may only require indirect support to overcome some or all o f their difficulties. However, for other individuals with more severe aphasia, support in the form o f linguistic information may be needed to help compensate for their language impairments. A s was the case with the second participant in Linebarger et al.'s (2004) study, the language improvements made with the use o f technology may not transfer to an individual's unaided speech; thus, in order to maintain this increased language ability, technology may be necessary to augment communication i n everyday life.  1.5 AAC and Aphasia: Focusing on  Participation  With the focus o f language therapy becoming broader to encompass not only work at the impairment level, but at the activity and participation levels as well, alternative and augmentative communication ( A A C ) strategies may help people with aphasia communicate more effectively and facilitate their participation i n social contexts. A A C strategies range on a continuum from low to high tech. Gesture, drawing, communication books, and alphabet boards are examples o f low tech A A C approaches. In contrast, high tech approaches to A A C have incorporated electronic technology (Hux, Manasse, Weiss, & Beukelman, 2001). In some cases, electronic devices (e.g. Dynavox, Lingraphica) have been developed specifically as communication aids. A s computer  33  technology has become prevalent in everyday life, there has been increasing interest in developing software programs or otherwise adapting existing technology to meet the needs o f individuals with aphasia. A A C strategies can replace, supplement, or scaffold remaining natural speech (Hux et al., 2001). When considering A A C options for an individual with aphasia, the person's lifestyle and communication capabilities must be taken into consideration (Garrett & Beukelman, 1998). Moreover, potential A A C users differ i n the severity o f their communication disorder and how the deficits "affect individual abilities to meet current needs and to participate in communication exchanges" (Garrett & Beukelman, 1998, p. 338). A s a group, speech-language pathologists tend to view A A C as a method for replacing natural speech when an individual has failed to regain enough language in order to "convey even the most basic messages" (Hux et al., 2001, p. 677). These individuals with severe disorders would be considered "basic choice communicators", requiring help from their communication partners to make basic choices (Garrett & Beukelman, p. 338). Efficient use o f augmentative and/or alternative communication systems, particularly those involving technology, by an individual with aphasia requires training and practice. Unfortunately, speech-language pathologists tend to introduce A A C systems much too late, often when the client is nearing the end o f his or her therapy, thus leaving insufficient time for the required training and practice (Hux et al., 2001). Historically, due to this insufficient instruction and preparation, there has been little success in A A C replacement therapy. In contrast, research investigating the type o f A A C system used to supplement natural speech rather than replace it has shown greater success (Hux et al., 2001). The  34  type o f individual with aphasia able to use such a system has been labeled the "comprehensive communicator" by Garrett and Beukelman (p. 341). This communicator is defined as someone who needs support in order to communicate effectively as his or her retained speech and language skills are limited and inconsistent. Typically, comprehensive communicators have a variety o f compensatory strategies, such as drawing and gestures, which allow them to get across their meanings to some extent. These communicators may benefit from an A A C program that enables them to talk about their personal interests and also deals with their own individual communication needs and participation patterns. A high-tech A A C device could be a useful addition to the comprehensive communicators' repertoire o f communication strategies. Unfortunately, these communicators are often overlooked by speech-language pathologists as possible A A C users (Hux et al., 2001). It is possible that the reason for this is until recently, hightech A A C devices have been designed primarily as alternative rather than augmentative systems. Despite the promise o f A A C systems, many factors are involved in the acceptance and use o f an A A C device. Lasker and Bedrosian (2001) describe their A A C Acceptance Model as involving three components: milieu, person, and technology. Each area involves several factors which relate to the overall acceptance o f A A C systems. For instance, the area o f milieu contains factors such as the communication partner, the communication environment, and available funding options. Within the person component o f Lasker and Bedrosian's model, factors including the characteristics o f the user's medical condition, attitude, personality, age, skills, needs, and intervention history  35  can all contribute to the possible acceptance o f an A A C system. Finally, various characteristics o f the technology itself play an important role in acceptance. V a n de Sandt-Koenderman (2004) discusses several factors which have been found to contribute to the successful use o f low-tech A A C . A s in Lasker and Bedrosian's model, the type o f aphasia and cognitive abilities o f the user influence potential success with the A A C system. Acceptance o f the A A C system is obviously imperative to its effective use, as is the role the A A C system is expected to serve. The A A C user should be supported in developing realistic expectations for the system as no A A C system can entirely replace natural speech. In addition to these factors, vocabulary which is relevant to the user's needs and appropriate training are essential. Communication partners also play an important role and should be trained to support the individual with aphasia i n using his or her A A C system. V a n de Sandt-Koenderman suggests that these factors also influence the success o f high-tech A A C devices. M a n y interacting factors have been identified as contributing to the potential level o f use o f an A A C system. Clearly, finding an appropriate A A C system for an individual with aphasia is not simply a case o f matching a system to an individual's language abilities. Despite the apparent promise o f these systems to support the communication o f individuals with aphasia, people's attitudes towards A A C devices have been a hindrance to their use. A study by Lasker (1997) as cited in H u x et al. (2001) was conducted i n which three groups o f people - adult peers, speech-language pathologists, and family members - listened to an individual with aphasia tell stories, first using unaided natural speech, then a communication notebook, and finally a digitized speech output device. Overall, the three groups reported that it was easiest to understand the person with  36  aphasia when he/she was using the high-tech speech output device. However, attitudes between the groups differed when asked to rate which system they would prefer i f they were the communication partner. The adult peer group preferred the digitized speech output device, while the SLPs and family members favored natural speech. The preference for natural speech was found to result from the belief that with further therapy and effort, the person with aphasia would be able to communicate without using the A A C system. Therefore, even though the high-tech A A C device produced the highest efficiency and understanding, attitudes o f family members and SLPs towards A A C systems may limit their use. Currently, A A C technologies have been underutilized in aphasia interventions (Mollica, 1999). This is often due to limitations o f the A A C system with regard to available vocabulary, physical design, and complexity o f the system. Furthermore, some individuals are unable to handle the trade-offs that occur with high-tech A A C systems. For instance, increased flexibility o f the A A C system and faster access to messages often require the user to deal with increased mental load. Moreover, there is the issue o f generalization from the therapy environment to daily life. Although A A C interventions have been proven to improve communication in therapy, generalization often fails (Jacobs, Drew, Ogletree, & Pierce, 2004). Clearly, the acceptance and effective use of an A A C system is a very individualized process involving many contributing factors. Researchers and clinicians need to take these issues into account when designing A A C systems and recommending them to their clients with aphasia.  37  1.5.1 Conversation and the Impact of AAC The structure o f conversations is typically quite predictable (Beukelman & Mirenda, 1998). Conversations usually begin with a greeting and small talk and may then proceed to a period o f information sharing, although others may remain at the level o f small talk. Following this, the conversation ends with wrap-up remarks and a farewell. The different stages o f conversation each come with their own demands and requirements for the A A C user. For instance, in order for small talk and social etiquette to be effective, these remarks must be communicated in a prompt and accurate manner (Stuart, Lasker, & Beukelman, 2000). While small talk and social etiquette are generally predictable and generic, information sharing is often unique, which means that the A A C device should be capable o f creating novel messages. The ability to tell narratives adds further requirements in that original stories must be created and stored for future use either by the A A C user or by a facilitator. In order to be successful in conversation, A A C users must be able to adapt to the various stages o f conversation and have systems which are flexible and broad enough to meet the demands each stage puts forth. In addition to contending with the variety of requirements posed by each conversational stage, A A C users must also attempt to satisfy the social norms of conversation with regard to timing and rate. In everyday conversations, people typically speak at a rate o f 150-250 words per minute; however, A A C users are much slower than this, communicating at a rate o f approximately 8 words per minute (Beukelman & Mirenda, 1998). A person's ability to respond quickly and to interject are important factors in determining how well he or she w i l l be able to participate in a conversation (Hine, et al., 2003). In all conversations, interlocutors are constrained by "the temporal  38  imperative" in which "the participants must provide a public account for the passage o f time in their individual parts o f that action" (Clark, 1996, as cited i n Higginbotham and Wilkins, 1999, p. 51). A public account includes actions such as talking, gesturing, or using an alphabet board. Higginbotham and Wilkins suggest that once there is a pause in the conversation, as occurs when an A A C user creates a message i n silence, there is no longer a public account o f actions. In English, pauses tend to be kept to a minimum and typically last no more than one second. However, A A C users are often required to pause for much longer than this when preparing a message, thus violating the constraint o f temporal imperative. Higginbotham and Wilkins (1999) state that these pauses can often lead to inattention from the conversation partner and may cause reluctance on the part o f the interlocutor to enter into conversations with someone who uses A A C . Thus, problems with timing can have very real social impacts on individuals who use A A C . It was previously discussed that language can be thought o f as serving two purposes or functions: transaction and interaction (Brown & Y u l e , 1983). B r o w n and Yule report that the majority o f everyday conversation consists o f mainly interactional rather than transactional language use. For example, interactional language is used to "negotiate role-relationships, peer solidarity, the exchange o f turns in a conversation, and the saving o f face o f both speaker and hearer" (Brown & Yule, 1983, p. 3). Hine, Arnott, and Smith (2003) report that a large part o f conversations involves sharing stories, expressing opinions, and gossiping, all o f which serve a social function. They suggest that as much as three-quarters o f conversation is devoted to these purposes. Clearly, the social function o f language plays an important role in conversation. Light (1988) proposes that communicative interactions accomplish four purposes, two o f these  39  "communication o f needs/wants, information transfer" are transactional and the other two "social closeness, and social etiquette" are interactional (p. 76). However, she notes that in the past the majority o f A A C technological developments have been concentrated in the areas o f communication o f needs and wants and information transfer. This suggests that there is a mismatch between what A A C systems have been designed to accomplish and the reality o f conversations in everyday life. While A A C systems may be effective in the transactional function o f language, the majority o f conversations serve an interactional function o f which social closeness and etiquette are important parts. It is evident that A A C users face a number o f challenges when conversing with others. N o t only must they contend with the different stages o f conversation and their requirements, they must also attempt to enter into those stages with appropriate timing. Furthermore, A A C users may be attempting to do all of this with a device that is not suited for conversations in everyday life.  1.5.2 Recent Advances in AAC  Technology  Designers in A A C technology have begun to recognize the need for devices which w i l l enable A A C users to participate in conversations at an interactional level. These recent advances in A A C technology provide users with a means o f communicating more than just their needs and wants. Currently there are several types o f computer-based A A C devices available for the comprehensive communicator, which have been designed . to support particular aspects o f speech that are difficult for these individuals. However, the A A C device should not be expected to replace speech entirely, as the comprehensive communicator has already developed strategies that aid in message transmission. One  40  example o f such a device is the Predictive Retrieval o f Story Extracts ( P R O S E ; Waller & Newell, 1997)). This program was designed to allow individuals with aphasia to tell narratives. With this program, users can retrieve different stories they have developed. The effect P R O S E had on conversations was investigated for one participant and it was found that with P R O S E , there was an increase in initiation and topic expansion, thus allowing the individual with aphasia to have more conversational control. Family members reported that the program made conversations easier as they knew the topic o f conversation. Furthermore, the reported frustration for both the individual with aphasia and the family members was decreased by use o f P R O S E . This illustrates that A A C devices have the ability to help people with aphasia become more active participants i n conversations. One limitation o f this device is the fact that the user must rely on others to input the narratives, and therefore, the user is restricted to telling narratives which are known by the people who are inputting them. Conversational control has also been examined with the use o f TalksBac, a predictive communication device designed for individuals with nonfluent aphasia (Waller, Dennis, Brodie, & Cairns, 1998). With TalksBac, a person close to the person with aphasia enters new utterances into the program and deletes old ones. The user is then able to select different conversational utterances by going through a series o f potential choices on the computer screen to come up with the appropriate utterance. Waller et al. examined whether TalksBac was able to augment communication by comparing unaided and aided conversations with both familiar and unfamiliar partners. Typically, conversations with an individual with aphasia are controlled by the nonimpaired partner. W i t h the use o f TalksBac, the individuals with nonfluent aphasia who  41  participated in the study demonstrated improvements in their conversational ability; there were increases in their ability to initiate new conversational topics and elaborate on topics, and fewer conversation breakdowns occurred compared to unaided interactions. Furthermore, there was less responding and less asking for confirmation from the partners without aphasia leading the researchers to conclude these partners were controlling the conversations less. The P R O S E and TalksBac studies illustrate that A A C devices have the ability to allow individuals with aphasia to become more active participants in conversations. Similar to P R O S E , users o f TalksBac are limited in that they are dependent on other people to input conversational material into the device.  1.5.3 Sentence Shaper: Further  Possibilities  Although not initially designed as an A A C program, there have been hints o f promise that Sentence Shaper could be used for more than just language remediation. Most o f the participants liked using Sentence Shaper, and one even used the program to give a speech to over a thousand people (Linebarger et al., 2000). Another person from the same study felt that Sentence Shaper would be helpful to use everyday. The participant used Sentence Shaper to produce this message: "I need verbalizing. I need it bad. The therapy's working, it's working. I enjoy therapy. I need it. I need the computer [to] record the message I want every day. I can't stand my stroke..." (p. 425). Furthermore, questionnaire results from Linebarger and Schwartz's study (2005) indicated that participants were willing to use Sentence Shaper in the long term and were interested in using it independently. Eighty-two percent o f the participants "expressed a lot o f interest in continuing to use the C S following completion o f the study" and 63%  42  "expressed the desire to work on the C S with little or no assistance" (p. 16). The question that remains is whether Sentence Shaper can be integrated into functional contexts as this w i l l be the main motivating factor for individuals with aphasia to continue using Sentence Shaper (Linebarger & Schwartz, 2005). Davies, Marcella, McGrenere, and Purves (2004) had a similar question regarding P D A (Personal Digital Assistant) technology. They explored this topic by using an ethnographic approach to learn how a man with nonfluent aphasia integrated a P D A with communication strategies already used. This type o f approach was important because, while previous studies such as those exploring P R O S E and TalksBac have investigated the effects each can have on conversations, they have not looked in detail at how individuals with aphasia are actually able to incorporate technology into their everyday lives and communication strategies. This time-intensive approach revealed important details regarding the strengths and limitations o f the P D A system and its usability in the real world. Conducting a similar ethnographic study with Sentence Shaper would be a beneficial addition to the current research on this processing prosthesis.  1.6 Research  Objectives  This study has two primary research objectives. The first objective is to partially replicate Linebarger, M c C a l l , and Berndt's (2004) language remediation study involving Sentence Shaper, extending their question o f the efficacy o f Sentence Shaper with respect to morphosyntax, to investigate the effects o f Sentence Shaper on narrative structure. The second objective is to explore in more depth the possible uses o f Sentence Shaper in everyday life, as this information is unknown at this time.  43  The rationale behind component one is that while previous research has demonstrated that Sentence Shaper can facilitate improvements in the morphosyntactic ability and informativeness and efficiency o f narratives, those studies did not explore the impact o f Sentence Shaper on story structure. Furthermore, as the previous studies were conducted with a relatively small number o f participants, documenting the language profile characteristics o f another user adds further to the understanding o f the impact o f this program on different types o f nonfluent aphasia. With respect to component two, the rationale is that previous studies o f Sentence Shaper have left unanswered and unexplored the question o f how this processing prosthesis can be integrated into the lives o f those living with aphasia. Because the I C F suggests that intervention should focus on the participation o f people i n contexts which are meaningful to them, investigating whether the Sentence Shaper can improve communicative effectiveness and diminish the negative impact o f a communication disorder on one's quality o f life w i l l be an important contribution. This is particularly relevant as technology is becoming increasingly available to those living with aphasia. This understanding w i l l be important for clinicians recommending this program to their clients in the future. Furthermore, while investigating the benefits Sentence Shaper may have for the communication o f an individual with aphasia, the study w i l l also explore the barriers faced when using the program. Knowledge o f these barriers w i l l be important for guiding further development o f the program and its use in different contexts. For example, currently Sentence Shaper is not fully portable as it needs to be run on a home computer or laptop. This study w i l l look at the implications this has for the user and the limitations  44  it may have on the contexts i n which he or she wishes to use the program. Researchers o f TalksBac found that even though the program was available on a laptop, this was not sufficiently portable for everyday situations away from home (Waller et al., 1998). Understanding the limitations w i l l help determine the usefulness o f putting the program on a handheld device. Moreover, Sentence Shaper is also an off-line program which means that users w i l l have to prepare messages in advance and then access them when they are having a conversation. The study w i l l look at what implications this may have on conversations and how a user can actually incorporate Sentence Shaper into his or her interactions with others. The following chapter w i l l describe the methodology as it was conceptualized at the outset o f the study. In a qualitative design, the unique characteristics o f the participants often shape the study, and therefore, these details are not known advance. Chapter Three w i l l provide a description of the study as it evolved following the identification o f the participants.  45  Chapter Two: Methodology  2.1 Introduction This chapter begins with a description o f case study methodology. From there the overview o f the study, the participant guidelines, ethical considerations, and communication profde assessments are explained. Following this, the data collection and analysis procedures are outlined separately for each component, namely, the language remediation component and the ethnographic component. In relation to the I C F framework discussed in the first chapter, the language remediation component explores changes at the impairment and activity levels o f functioning, whereas the ethnographic component seeks to investigate whether Sentence Shaper can facilitate change at the participation level o f functioning and to explore the contextual factors which impact the use o f Sentence Shaper.  2.2 The Case Study A case study approach was chosen for the study as this paradigm is particularly well suited to the exploratory nature o f the research objective. Stake (1995) defines the case study as "the study o f the particularity and complexity o f a single case, coming to understand its activity within important circumstances" (p. xi). A s the potential uses o f Sentence Shaper in everyday life have not been previously explored, relevant parameters contributing to its use were not known in advance, necessitating a descriptive approach. A single case study is beneficial in this circumstance as it allows for extremely broad and deep description o f the participant's use o f Sentence Shaper and the context in which it  46  took place. Stake distinguishes between two types o f case studies: intrinsic and instrumental. In the intrinsic case study, the case is already given and a choice is made to study that particular case. The instrumental case study, in contrast, begins with a research question and a case is found to investigate that question. This study was motivated by an initial research question; therefore it is considered an instrumental case study. Stake states that the purpose o f the case study is "particularization, not generalization" (p. 8). In this study, I examine the particular variables that influence the ways in which Sentence Shaper can be used in one person's unique life situation. This is not an attempt to explain how the program could be used for all people with aphasia, but is rather a description o f its use in a specific context. This information w i l l guide future research in this area and w i l l also provide insight for clinicians working with people who have nonfluent aphasia.  2.3 Overview This study was conducted over a four month period with an individual with aphasia and a communication partner. Each component i n the study addresses one o f the two objectives discussed at the end o f the first chapter. The first component was a partial replication o f a previous treatment study involving Sentence Shaper, extending their investigation o f the effects o f Sentence Shaper on morphosyntax to also examine the impact on story structure (Linebarger et al., 2004). In order to meet this first objective, a pre-post treatment comparison o f narratives with and without Sentence Shaper was undertaken. Quantitative and qualitative measures were used to investigate whether there were any improvements to language ability following use o f the Sentence Shaper, comparing them to those obtained by Linebarger et al. (2004) in their study o f Sentence  47  Shaper. The second component o f the study used a qualitative case study approach (Stake, 1995). I assumed the role o f a moderate participator in this study (Spradley, 1980), that is, I was an "insider" in the sense that I was interacting with the participants, but also remained an "outsider" in that I do not have aphasia, nor am I involved with a close family member with aphasia. The qualitative design o f component two fits well with a social model o f intervention as qualitative methods are particularly useful for obtaining the perspectives o f the participants. The purpose o f this component was to investigate how Sentence Shaper could be incorporated into the communication strategies of an individual with aphasia and used in his or her everyday life.  2.4 2.4.1 Initial Participant  Participants  Guidelines  Participants were recruited through local speech-language pathologists. The following guidelines were provided to assist the speech-language pathologists in identifying possible participants: 1. One participant should have non-fluent aphasia. This is defined as difficulty with grammatical language production, but with relatively spared comprehension. This was defined using criteria from Linebarger et al.'s (2000, 2001, 2004) research, a.  Participants had chronic nonfluent aphasia (i.e., agrammatic or nonagrammatic). Saffran, Berndt, and Swartz (1989) describe agrammatic speech as having omissions o f some bound and free-standing grammatical morphemes, while nonfluent, nonagrammatic speech does not have these same errors but is still structurally impoverished.  48  b. The participants in Linebarger et al. (2004) had a rate o f speech below 107 words/minute, which was considered to be characteristic o f nonfluent aphasia. 2. The participant with aphasia must have a close family member or friend with whom he or she communicates regularly (the second participant). 3. The primary language spoken in the home is English. 4. Both participants must live relatively close to the researcher to make repeated meetings possible. 5. The participant with aphasia must have a home computer on which Sentence Shaper could be installed. When the speech-language pathologist identified a potential participant, the individual with aphasia was given an information letter with an invitation to contact the researcher for a meeting with both the individual with aphasia and his or her proposed communication partner. This meeting served two purposes. First, on the basis o f listening to the speech o f the individual with aphasia, his or her language ability was screened to ensure appropriateness for use of Sentence Shaper. Second, the outline o f the study was explained and both participants were made aware o f the time commitment involved i f they decided to participate. If the language screening met the criteria, both the individual with aphasia and the communication partner were invited to participate in the study and were asked to give consent.  49  2.4.2. Ethical  Considerations  The ethical considerations involved in this study were two-fold. First, it was important to ensure that the participant with aphasia was giving informed consent. To facilitate comprehension o f the consent form, an aphasia-friendly consent form was prepared with larger font size, pictures, and simplified language. The participant with aphasia was also given help reading the consent form and asked for verification o f understanding throughout the process. Consent forms are found in Appendix A . The second ethical consideration was with regard to the confidentiality o f the participants. To ensure confidentiality, pseudonyms were used in all transcripts and fieldnotes presented in this thesis for the participants and any people mentioned in those documents. Furthermore, any mention o f specific identifying information was deleted or made more general. A third ethical consideration was with respect to introducing a system that the participant with aphasia might wish to incorporate into his or her communication strategies at the end o f the study. In order to prevent the removal o f a device which might prove useful for the participant, the participant with aphasia w i l l be offered the program upon completion o f the study. Approval for this research was obtained through the Behavioural Research Ethics Board at the University o f British Columbia. A copy o f the approval form is included in Appendix B .  50  2.4.3 Communication  Profile  After the individual with aphasia and family member agreed to participate in the study, further analysis o f the communication abilities o f the individual with aphasia and of the communication interaction between both participants was undertaken. The overall purpose o f the communication profde was to fully characterize the comprehension and production abilities o f the participant with aphasia, and was not intended as a measure to determine participant appropriateness for the study. Test selection was based on Linebarger et al.'s (2004) study to facilitate comparison o f the participants' language abilities. Comprehension was tested using subtests of the Psycholinguistic Assessments of Language Processing in Aphasia, in particular, the spoken word-picture matching, sentence-picture matching: written version, and written comprehension o f locative relations subtests ( P A L P A ; K a y , Coltheart, & Lesser, 1992) and o f the Boston Diagnostic Aphasia Examination, specifically, reversible possessives and embedded sentences subtests (Goodglass, Kaplan, & Barresi, 2001). Lexical retrieval was also tested using the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 2001) as well as the Noun/Verb Naming test (Zingeser & Berndt, 1990). Results o f these assessments helped to characterize the comprehension and lexical retrieval abilities o f the participant with aphasia. Participant observation facilitated a comprehensive description o f the communication strategies that the individual with aphasia and the communication partner were using before Sentence Shaper was introduced. The Quality o f Life Communication Scale ( Q C L ; Paul et al., 2004) and the Communicative Effectiveness Index ( C E T I ; Lomas, et al., 1989) questionnaires were also completed as part o f the communication profile assessment to gain some insight into  51  the perspectives o f both the participant with aphasia and the communication partner. However, unlike the other communication profile assessments, these questionnaires were also completed at the end o f the study to measure any changes that may have occurred. The Quality o f Communication Life Scale ( Q C L ) provides a measure o f "the impact o f a communication disorder on an adult's relationships; communication interactions; participation in social, leisure, work, and education activities; and overall quality o f life" (Paul, et al., 2004, p. 1). The Q C L is completed by the participant with aphasia and is designed and written in an aphasia-friendly manner. The Communicative Effectiveness Index consists o f a number o f questions detailing a person's communicative ability in day-to-day situations as perceived by the communication partner (Lomas, et al., 1989). The communication partner is asked to indicate the person's ability on a scale ranging from "not at all able" to "as able as before stroke".  2.5 Component One:  Procedures  2.5.1 Sentence Shaper Training Following baseline assessments, Sentence Shaper was installed on the aphasic participant's computer. Training o f Sentence Shaper was conducted only with the participant with aphasia and was based on the training procedure in Linebarger et al's (2001, 2004) studies. The participant was instructed on how to record sound files, replay them on Sentence Shaper, and save them. The participant was also shown how to use the side buttons which contain frequently used verbs and pronouns and the word-finding tool. The training took place over approximately five hours. In the language remediation component o f the study, the participant with aphasia was expected to practice recording  52  messages using Sentence Shaper over a course o f approximately ten weeks. During this time, the researcher met regularly with the participant and was available to assist with message production by giving feedback on how to use the program more efficiently.  2.5.2 Data Collection 2.5.2.1 Narrative samples. Language samples were elicited by having the participant narrate a familiar fairy tale and The Cowboy Story, an eight-picture story sequence (Joanette & Goulet, 1990). The Cowboy Story was chosen as an additional task as it provides a story structure analysis based on normative data. These narratives were completed at the beginning and end o f the study as a way to characterize the speech o f the participant with aphasia and to measure changes i n language ability following use o f Sentence Shaper. The participant was also asked to record the same narratives using Sentence Shaper at the end o f the study. The participant was free to use the word finder and side buttons when necessary. In order to reduce practice effects from repeated narrations o f the same story and to counterbalance the aided and unaided narratives, the participant with aphasia was asked to narrate The Cowboy Story unaided and the fairytale aided by Sentence Shaper on one day; the reverse was done three days later. These tasks allowed for a comparison between unaided and aided language characteristics  53  2.5.3 Data Analysis 2.5.3.1 Morphosyntactic  analysis.  A l l narrative samples were transcribed. Quantitative Production Analysis, a standardized analysis procedure, was utilized for morphosyntactic analysis (Saffran, Berndt, and Schwartz, 1989). Analysis began by determining the number o f words in the sample (at least 100 words) and from there determining the number o f narrative words or those words representing propositional speech. Following this, the narrative words were segmented into utterances and the utterances were coded into different syntactic types. Saffran et al. define an utterance as "words that appear to form a coherent unit" taking into consideration syntactic and prosodic markers, pauses, and semantics (p. 470). The utterances were then further analyzed by counting different types o f words found in each utterance (i.e. open class words, nouns, verbs pronouns, etc). After this step, a structural analysis o f each utterance examined aspects such as embeddings and constituents within sentences. Finally, the number o f words per minute was calculated. This analysis provided measures o f the morphological and structural characteristics o f the participant's speech production. Inter-rater reliability checks were conducted on all measures.  2.5.3.2 Informativeness  and efficiency  analysis.  To investigate the informativeness o f the narratives and efficiency with which informative content was communicated, a procedure developed by Nicholas and Brookshire (1993) was used. In this analysis, the words o f the narrative sample were counted excluding all nonword fillers and unintelligible words. Following this, correct information units (CIU) were identified. Nicholas and Brookshire define CIUs as "words  54  that are intelligible in context, accurate in relation to the picture(s) or topic, and relevant to and informative about the content o f the picture(s) or the topic" (p. 15). From these counts, the total percentage o f CIUs in the sample and the CITJs/minute were calculated. Inter-rater reliability was determined by having two raters score at least 15% o f the samples independently and then compare the results o f this scoring to ensure reliability. This procedure was completed for both the Q P A and C I U analysis.  2.5.3.3. Story structure  analysis.  A n additional analysis o f the Cowboy Story was completed to examine the story schema and informative content o f the participant's narrative (Joanette & Goulet, 1990). The Cowboy Story contains three components, that is, the setting, the complication, and the resolution, which must be included in the narrative in order to be complete. A description o f the setting o f the story should include mention o f a tired cowboy coming into town, getting off his horse, resting on a bench, and holding his horse's bridle. The complication component o f the story deals with a young boy exchanging a small wooden horse for the cowboy's horse and running away with the real horse. Finally, the resolution involves the cowboy waking up to find that he is holding the small toy horse and being surprised at the change in the horse. Joanette and Goulet's analysis o f the informative content contains a list o f 32 core propositions detailing the content expected to be in the narrative. The expected version o f The Cowboy Story is based on the authors' narratives and the narratives produced by groups o f individuals with right-hemisphere brain damaged and non-impaired individuals. This allowed for a comparison between the content included in the participant's narrative to the expected content.  55  2.6 Component Two:  Procedures  In the second component o f the study, participant observation was conducted with the individual with aphasia and the communication partner. Using a participatory design, I worked together with both participants to assess the strengths and weaknesses o f Sentence Shaper, as well as to investigate different ways the program could be used. A second purpose o f our meetings was to learn more from the participants about communicating with aphasia. I observed and described as comprehensively as possible the different communication strategies used by the participants in daily life and how Sentence Shaper could be incorporated into those strategies. The data emerging from the sessions allowed for a rich description o f verbal and nonverbal communication o f the participant and also detailed the environment in which Sentence Shaper could be used.  2.6.1 Data  Collection  2.6.1.1 Participant  observation.  Throughout the four months o f data collection, I participated in regularly scheduled meetings with the participants. Most sessions were recorded with a Sony M i n i D i s c player with an external bi-directional stereo microphone. O f the sessions that were recorded, only the first 80 minutes or the length o f one minidisk was recorded. The minidisk recorder was somewhat intrusive and changing the disks would have been disruptive; so to reduce the impact o f the recorder, only one disk per session was used. Very brief field notes were written during visits with participants. Detailed field notes were taken after every visit. Audio-recordings were also reviewed following each session to help with documenting the field notes.  56  2.6.1.2 Interviews. A t the beginning and end o f the data collection period, I conducted brief interviews with each participant. The purpose o f the interviews was to learn about the participants' perceptions o f the participant with aphasia as a communicator and to enrich understanding o f the communication interaction between the participants. The questions focused on communication challenges faced by the two individuals and communication strategies used. The final interviews included questions about Sentence Shaper, specifically, the benefits o f and barriers to its use. A t the beginning o f each interview, the participant with aphasia was encouraged to use whatever strategies were useful to facilitate communication (i.e., gesture and writing). The list o f interview questions is provided in Appendix C .  2.6.1.3 Sentence Shaper log. The participant with aphasia was asked to keep a log detailing the use o f Sentence Shaper. The participant was asked to record the date, length o f time Sentence Shaper was used, purpose o f use (practice, conversation, or e-mail), and presence o f others during use. It was important to know the total time the participant with aphasia spent using the program, first, in order to compare this to the time the participants in Linebarger et al.'s (2004) study used Sentence Shaper, and second, to facilitate description o f the ways in which Sentence Shaper was used.  57  2.6.1.4 Quality of Communication Life Scale and Communicative  Effectiveness  Index. These assessments were conducted at the beginning as part o f the communication profde and at the end o f the data collection period with the participant with aphasia and her communication partner. The final questionnaires were completed by the participants without having them first review their answers from the first questionnaire. This was done to help prevent a positive bias in the participants' answers. After the participants had finished the final questionnaires, their answers were compared to those from the first questionnaire, and they were asked i f they felt the changes were accurate. A t this point, they were able to alter any o f their responses. Comparisons o f initial and final questionnaires were carried out to explore any changes that may have occurred.  2.6.2 Data Analysis A n important aspect o f data analysis in this study was the integration o f different forms o f data as each played an important role in exploring the communication strategies used by the participant and the incorporation o f Sentence Shaper into these strategies. B y interpreting each type of data in the context o f all others, a more in-depth picture could be obtained. This analytic strategy o f exploring the data as a whole can uncover patterns and consistencies and inconsistencies within certain contexts, and this leads to correspondence,  which Stake (1995) defines as "the relationship o f one variable to  others" (p. 170). Furthermore, finding convergence among different sources o f data, or triangulating the data, contributes to the internal validity o f the study (Creswell, 1994).  58  The following sections detail how the different sources o f data for both components o f the study were analyzed.  2.6.2.1 Analysis of fieldnotes and interviews. The fieldnotes and interviews were analyzed for meaning using thematic analysis (Luborsky, 1994). A significant benefit o f this type o f approach is that it allows for a representation o f the participants' perspectives. It was important to record and analyze each participant's perspective as this interpretation was critical to the results o f this study. Luborsky makes a distinction between themes and patterns, suggesting that themes reflect the perspectives o f the participants, whereas patterns reflect the researcher's perspective. Luborsky proposes two approaches to theme identification. The first approach is to look for repeated statements made by the participants. The second is to identify statements which appear to be meaningful or significant to the participant. In contrast, patterns are repeated observations made by the researcher that may or may not be apparent to the participants. This differentiation helps identify the findings in the data that are from the perspective o f the participants versus those that reflect the researcher's perspective. Analysis o f the fieldnotes and interviews had several steps. Fieldnotes and interviews were initially bracketed into sections and given a code. Codes were not generated in advance, but rather emerged as the data was analyzed. Codes (e.g. "technology", "response to Sentence Shaper") were used to describe more generally what was happening in each section o f the fieldnotes or transcripts. Once the data was coded in its entirety, the codes were listed and then grouped into categories which allowed themes  59  and patterns to emerge from the data. For example, codes o f "customizing the word finder and side buttons", "creating Sentence Shaper messages", and "motivation to create messages" were grouped into the category o f "use o f Sentence Shaper". This grouping allowed for the pattern o f "the participant as an independent user o f technology" to emerge. Alternatively, larger codes were broken down into more specific codes. For instance, the code o f technology was coded more specifically as "independent use o f technology", "barriers to technology", and "current technology used by participant". In this way, coding o f the data was bidirectional, being both bottom-up and top-down. This process was repeated several times to ensure completeness o f coding. Finally, interaction of codes was examined to provide a clearer description o f how patterns and themes were connected and related. For example, a pattern which emerged from the data was the participant's dependence on communication partners. Codes relating to this were "difficulty generating ideas to talk about" and "the effort o f communication". In addition, the code, "helping behaviour demonstrated by the communication partner" was seen as interacting with these codes.  2.6.2.2 Transcription of conversations. A n additional source o f data for analysis was conversations recorded with the participants. A s the majority o f visits with the participants were recorded, many hours o f conversation were available for transcription. Because the primary purpose o f transcribing conversations was to convey the nature o f participants' communication and to aid in the description o f communication strategies used by the participants, it was not  60  necessary to transcribe the entire set o f recordings. So conversations were selected on the basis o f their contribution to the description o f the participants' communication. However, the process o f transcription presents several challenges for researchers. A s a rule, transcription is "a selective process reflecting theoretical goals and definitions" (Ochs, 1979, p. 44). It is selective in the sense that it is almost impossible to include every aspect o f a conversation. However, the details o f a transcript often provide important information and meaning. Clearly, a balance must be met so that an adequate representation is obtained without overwhelming the reader with unnecessary details. Transcription also becomes selective when an audio-recording is used. Obviously, this type o f recording is unable to capture a person's use o f gestures, which are often an important form o f communication for individuals with aphasia. In order to partially compensate for this, an effort was made to note any particularly important gestures as they occurred and these were later added to the transcripts. However, the limitations o f using fieldnotes for transcription purposes are acknowledged as some important details are inevitably lost in this process (Psathas & Anderson, 1990). Recorded conversations were transcribed using a subset o f codes from the transcription coding system developed by Gumperz and Berenz (1993). The codes chosen for this purpose were those which enhanced the meaning o f the conversations and contributed to analysis. For instance, audible exhalations were included in the analysis because the participant with aphasia often indicated frustration by an audible sigh. The list o f transcription codes used is given in Appendix D .  61  2.6.2.3 Analysis of QCL and CETI. The Q C L and C E T I were analyzed by comparing the scores on the pre and post measures. In order to interpret these findings, participants were asked to compare the pre and post questionnaires and to provide an explanation for items which had changed in score.  2.7 Relating Findings to the Research  Objectives  Data from fieldnotes, transcripts, and questionnaires was compiled during the analysis procedure. These findings were integrated or triangulated by reviewing each set of data separately for themes and patterns and then combining the evidence from each source of data to better support the themes and patterns which emerged. This analysis was then related back to the initial research objectives i n an attempt to explain the reasons behind the way i n which Sentence Shaper was used to augment communication. In addition, results from the qualitative investigation were used to support the findings in the treatment component o f the study.  62  Chapter Three: Emergent Findings  3.1 Introduction In a qualitative case study, many details o f the study begin to take shape only after the participants are identified. In this way, the study evolves from its initial conceptualization to include aspects o f each participant's unique characteristics and life situations. This chapter begins with a description o f the participants and the baseline assessment results o f the participant with aphasia. This is followed by an outline o f the study and a description o f the data set obtained from the four months o f data collection. The chapter ends with a discussion o f the challenges involved in interviewing an individual with expressive and receptive communication impairments.  3.2 Study Participants Nicole, a woman with aphasia, and Eileen, her mother, agreed to participate in this study. Nicole is a 31 year-old woman who had a stroke approximately four years prior to the study. Her stroke occurred during aneurysm surgery and resulted in right hemiparesis and non-fluent aphasia. Following her stroke, Nicole received speech therapy during her stay in acute care. She also had speech therapy for 3-4 months as an inpatient at a rehabilitation hospital and for three months as an outpatient. According to Eileen, Nicole's verbal output was limited to a single word ("bathroom") for an entire year following her stroke, and has since improved to her current state of one-to-two word utterances with some highly routine phrases.  63  Nicole was born in India and moved to Taiwan at age five. The family immigrated to Canada when Nicole was 12 years o f age. English and Mandarin were historically the languages spoken at home; however, Nicole has not spoken Mandarin since her stroke. Prior to her stroke, Nicole was employed at a bank and had a high school education with some post-secondary level courses. Following the stroke, Nicole's mother retired from her job in order to care for her daughter. Nicole lives with Eric, her partner of seven years, but spends most days at her mother's house. Nicole is active in a weekly stroke club in her community and spends several hours a week doing language related activities with volunteers who come to Eileen's home.  3.2.1 Communication Profile Results A t the beginning o f the study a detailed assessment of Nicole's language production and comprehension abilities was completed. Her rate o f speech ranged between 4-8 words/minute as measured from her narrative samples and is considered nonfluent when compared to the normal fluency range (107-232 words/minute). Using the Saffran et al. (1989) criteria, Nicole's speech is considered to be agrammatic in that there was an omission o f both bound and free-standing grammatical morphemes. In addition, her narratives had a complete absence o f story structure. Communication assessments included subtests from the P A L P A and B D A E . Nicole's scores are presented in Table 3.1.  64  Table 3.1 Nicole's assessment scores Assessment Tool P A L P A sentence-picture matching subtest (written version)  Score  •  Active sentences  0.67  •  Passive sentences  0.50  •  Reversible sentences  0.55  • Non-reversible sentences P A L P A spoken word-picture matching P A L P A locative relations subtest B D A E auditory comprehension o f reversible possessives B D A E auditory comprehension o f embedded sentences The Boston Naming Test Noun/Verb Naming Test • •  0.75 0.95 0.38 0.70 0.30 0.27 0.80 0.77  Nouns Verbs  In comparing Nicole's assessment scores with subjects from Linebarger et al.'s (2004) study, similarities and differences are noted. A s with Nicole, both subjects from Linebarger et al.'s study had relatively intact auditory comprehension o f single words. They were also similar in their ability to comprehend reversible sentences. Nicole demonstrated superior production o f verbs compared to the other participants and was also better in her production of nouns than the first participant from Linebarger et al.'s study. In contrast, Nicole's discourse ability demonstrated in the pre-treatment narratives was much more impoverished than the Linebarger et al.'s participants' narratives.  3.2.2 Nicole as a  Communicator  Results of Nicole's language assessments revealed impairments in both comprehension and expression. However, assessment scores cannot fully depict the communicative ability o f an individual with aphasia. In this section, a description o f Nicole's communication style is provided. This data was obtained through participant observation o f Nicole's interactions with Eileen and me prior to the introduction o f  65  Sentence Shaper, from the brief interviews conducted with both participants at the beginning o f the study, and through the Q C L and C E T I questionnaires. Observing the participants' communication interactions revealed important details about Nicole as a communicator. It was apparent that despite Nicole's relatively high score on the Noun/Verb Naming Test, she experienced difficulty with word retrieval in conversation. However, she compensated for this difficulty in various ways. Eileen mentioned that Nicole had several communication strategies, such as gesture, writing, and facial expression. Her use o f gestures as a strategy to augment her communication was illustrated when she used gestures and single words to tell me about the exercises she did at the pool. Nicole also used gestures to cue herself to say particular words. While doing the locative relation subtest o f the P A L P A , she often demonstrated the correct gesture for the preposition; however, she was unable to transfer this to help her comprehend the relations. A s the majority o f Nicole's utterances consisted o f one noun or a short phrase o f two-to-four words, she often relied on her communication partner to expand her utterances. A n example o f this occurred when Nicole said "plants" which prompted Eileen to offer me some o f their plants they were giving away (FN03/14: 3-4). A t the initial meeting, Eileen often spoke for her daughter when I asked a question; however, Nicole did not seem bothered by this behaviour. I also noticed that in conversation, Eileen often asked Nicole questions to which Eileen already knew the answers. For instance, when we were discussing Nicole's stroke, Eileen asked Nicole questions, such as "when did you have your stroke", which seemed to facilitate Nicole's word retrieval. If Nicole was unable to answer the questions Eileen had posed, Eileen answered for Nicole.  66  The Q C L and C E T I also provided useful information about Nicole's communication. There were several items on the Q C L which Nicole rated quite negatively (below 3 on a scale o f 1 to 5). For instance, "It's easy for me to communicate" and "I am confident that I can communicate" were both rated at 2. The Q C L also highlighted that Nicole's communication difficulties have affected her ability to understand television shows and movies and have made her feel less included in conversations with others, as indicated by the very l o w scores on those items. However, despite these difficulties, Nicole rated that she likes to talk with people quite positively. The statement "In general, my quality o f life is good" was scored as a 3 which suggests that she perceives her quality o f life as neither overly good nor overly bad, but rather average. Eileen's responses on the C E T I suggested that several aspects o f conversation were difficult for Nicole since her stroke. For instance, Eileen gave l o w ratings to Nicole's abilities to be involved in group or one-to-one conversations, to initiate conversations, and to discuss a topic in detail. These answers corresponded with the way Nicole had answered her questionnaire, suggesting that Eileen's perceptions o f Nicole's communicative ability are consistent with those o f her daughter's. Clearly, Nicole's aphasia has had an impact on her ability to communicate with others and her perception of her quality o f life.  3.3 Outline of Study The two components o f the study were carried out concurrently. Following the initial assessment and observation period, the training phase was conducted i n which Nicole was trained in the use o f Sentence Shaper and was given practice time to become  67  an efficient user o f the program. The training sessions were also used as an opportunity to interact with the participants and to explore what communication strategies they were currently using. The training period took approximately two weeks during which I spent about three hours per week with Nicole. Following the training period, I typically met with Nicole twice each week. It was Nicole and Eileen's preference that I meet with them on weekday afternoons. Meetings typically lasted for 1 A hours and we met for a total o f l  four months. Approximately 54 hours were spent meeting with Nicole. During the meetings, Eileen was usually present at the beginning and the end for discussion. There were two meetings in which Eileen spent the entire session with Nicole and me. Meetings most often occurred at Eileen's home with Nicole and I seated at a table with her laptop computer. I also had the opportunity to attend a local stroke club on two occasions and visit with Nicole at her apartment. We also went out for lunch with a mutual friend to a restaurant o f Nicole's choice. A schedule o f the meetings is given i n Appendix F . A t the end o f each meeting, Nicole was asked to make a message using Sentence Shaper before the next meeting. A t the beginning o f the study, I gave Nicole more specific topics for Sentence Shaper messages. A s the study progressed, I attempted to reduce these suggestions to more general ideas; however, I was not always able to do this. Nicole required more support in generating ideas for messages than originally anticipated at the onset o f the study. She also had a tendency to work with others, such as Eileen, when creating messages on Sentence Shaper. In the Sentence Shaper manual, the designers suggest that for maximum improvement, the program should be used independently (Linebarger, 2005), so I often suggested to Nicole and Eileen that Nicole should try to work independently when using Sentence Shaper. In this way, the role I had  68  originally envisioned changed with the characteristics o f the participants. I not only facilitated the use o f Sentence Shaper by suggesting ways to use the program more effectively, but I also helped generate ideas and encouraged independent use o f the program.  3.4 Data Set 3.4.1 Narrative Samples A s previously mentioned, unaided narrative samples were taken at the beginning and end o f the study. Nicole chose to tell The Three Little Pigs for the fairytale narrative, but had some difficulty beginning her narrative. Saffran et al. (1989) suggest that a picture book can be used when a story is not well known, so when Nicole spontaneously found the story in a workbook, I allowed her to look at the pictures, but asked her not to read any o f the words. For the final narrative, she did not use her workbook. Nicole was also asked to narrate The Cowboy Story picture sequence which seemed somewhat easier than the fairytale as the pictures could aid i n word retrieval. Following use o f Sentence Shaper, Eileen stated that during conversation, Nicole had been saying unaided some parts o f the messages she had recorded. In light o f this development, I asked Nicole to tell me about her trip to M e x i c o which had been the topic of a Sentence Shaper message recorded earlier. This added an additional narrative sample to the data set.  69  3.4.2 Interviews, Conversations,  and Sentence Shaper Messages  Both initial interviews with Nicole and Eileen and the final interview with Nicole were audio-taped and transcribed orthographically for further analysis. Eileen preferred not to have her final interview audio-recorded; therefore field notes were taken instead. In addition, a selection o f conversations from the audio-recordings o f each session was transcribed for further analysis. When possible, Eileen's comments were written down verbatim. A summary o f the transcribed conversations are provided in Table 3.3; interviews are listed in the summary o f data collection in Appendix E . From this point, quotations from interviews w i l l be coded as " I " and conversations as " C " with the appropriate line numbers. Direct quotes from participants and specific examples recorded in fieldnotes w i l l be identified by F N , the date they were written, and the line number. Sentence Shaper messages Nicole created are labeled with SS and the date they were recorded. Table 3.2 Summary, o f Transcribed Conversations Date Code Participants Apr. 11 CI Nicole, Eileen, and Erin Apr. 14 C2 Nicole, Eileen, and Erin Apr. 21 C3 Nicole and Erin Apr. 25 C5 Nicole and Erin Apr. 25 C6 ' Nicole and Erin Apr. 28 C7 Nicole, Eileen, and Erin Apr. 28 C8 Nicole and E r i n May 2 C9 Nicole and Shelley May 2 CIO Nicole, Eileen, Shelley and Erin M a y 18 Cll Nicole and Erin June 2 C12 Nicole and Erin June 6 C13 Nicole and Erin June 9 C14 Nicole, M i r a , and E r i n June 20 C15 Nicole and Erin June 22 C16 Nicole and Erin June 29 C17 Nicole making Sentence Shaper message  70  Setting Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's T V room Eileen's Eileen's Eileen's Eileen's Eileen's Eileen's Eileen's  T V room T V room T V room T V room kitchen kitchen T V room  Length 0:40 3:29 3:00 2:20 1:58 0:33 1:14 1:09 2:25 2:01 0:40 1:36 1:24 1:22 1:36  I had the opportunity to observe Nicole working with Shelley, one o f her volunteers. This created a group situation in conversation nine and provided an excellent example o f Nicole's interactional style in those contexts. The majority o f the conversations took place i n Eileen's T V room as this is where Nicole kept her laptop. In Appendix F , a description o f the Sentence Shaper messages Nicole chose to share with me is provided.  3.5 Challenges of Communicating with Nicole Although a key component o f this study was to understand the perspectives o f the participants, several challenges arose in my interactions with Nicole. Her limited verbal output made it difficult for her to express her opinions and perspectives as clearly as she probably would have liked. Furthermore, her comprehension deficit often made it difficult for her to understand the questions I was asking, even with simplification. Initially, I attempted to ask open-ended questions, but these were often too difficult for her to understand and answer. If I did not have success with an open-ended question, I often resorted to a yes/no question. However, I was aware that this created leading questions. In this way, I gained a somewhat limited picture o f Nicole's perspectives as her responses focused on issues I had deemed important, and not necessarily on issues that were important to Nicole. This situation is illustrated in the following exchange: Excerpt 3.1 11. 12. 13. 14. 15. 16.  E: N: N: E:  When when talking to people like what are some challenges or some hard things that you Hard things Hard things um um um um um arms arms arms and leg leg and leg um arms I mean um um moving moving moving Okay  71  17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.  N:  M o v i n g uh moving and brace no good I mean I mean um cha I mean um um E: So when when you're talking to people though N: Yeah E: When you're talking with like other people N: Yeah E: What are some hard things about that? N: That? um ..um E: About talking N: Talking? E: Yeah. What are some hard things about talking? N: Oh um ... talking um ... um ... mom and okay mom and um me uh um um pra practice no I mean I mean uh .. um okay I mean um um we we read it read it read and read it read and uh I'm I'm ( ) I mean um I mean uh I don't know ... I mean um .. um read and uh uh wor I don't know (looking through one o f her work books here) E: Okay Nicole do you find it hard to find the words? N: Words? um yeah yes yes yes E: D o you have trouble getting your message N: = = Message yes E: = = Across to people N: Yes yeah yes yes (11.11-39)  Furthermore, a strategy Nicole used in conversations was to have the conversation partner expand on her one or two word utterances. This strategy often requires the conversation partner to guess at the gist o f the message and then the individual with aphasia either confirms or disconfirms the interpretation. While this was a strategy I often used with Nicole, once again it often created a leading situation. M y elaborations were also problematic in that I was often unclear whether Nicole had understood my message. In getting to know Nicole over the course o f four months, I learned important information about her perspectives and life which helped me in analyzing the transcripts. In many cases, without this background information, I would have been at a loss in trying to interpret her meaning. For example, in the initial interview, I wanted to find out which situations were more difficult for Nicole's communication. This example demonstrates  72  how easy it was for communication to break down, and in this case a repair was not possible: Excerpt 3.2 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184.  E: N: E: N: E: N: E: N: E: N: E: N: E: N:  E: N: E: N: E: N: E: N: E: N: E: N: E: N: E: N:  Okay and .. you told me some the .. things you like to do like you like going to the stroke group .. Stro stroke group uh what Stroke group Yeah L i k e the Vancouver stroke group Vancouver stroke group yeah yeah So is that.. a a place that you is easy for you to talk Yes yes yes okay yes yes okay uh uh stroke group yeah yeah uh uh yes uh yes yeah So are there places that you go to or things that you do Yeah L i k e in the week In the week yeah That you find like more difficult Difficult yeah yeah I mean um diffic I mean um difficult uh uh I mean I mean uh uh okay I mean um um um um speak with uh uh uh speak with uh uh me and uh me and uh uh and uh uh Eric no no I mean I mean uh uh uh A l l i s o n A l l i s o n uh U m hmm I mean uh uh A l l i s o n yeah? She's one o f the volunteers? Volunteers yeah volunteer yeah and then uh uh bin bingo bingo names oh I mean um names yeah yeah U m hmm A n d then I mean um um um um um um um ho ho hockey hockey um um yeah hockey yeah no no What's with hockey? So is that hard Hard Something hard to talk about or U m yes yes yeah When do you like to talk about hockey like ...at at at bingo or at at the stroke group Stroke group I mean uh uh .. uh um .. like I mean um games games games U h mm Yeah Okay Games uh games uh um um um lunch lunch uh I don't know  73  185. 186.  E:  Okay so I think maybe we're talking about two different things right now (11.146-186)  Later I found out that Nicole plays hockey and bingo and has lunch at the stroke group. This knowledge gave me a better understanding o f what Nicole was trying to communicate to me in this passage. She was clearly not answering the question I had asked about difficult communication situations, but was rather telling me about the activities she does at the stroke group.  74  Chapter Four: Results  4.1 Use of Sentence Shaper: Overview Nicole was introduced to Sentence Shaper with a five-hour training phase. After this period, she demonstrated the ability to use the program independently to produce messages, although she often chose to work with others. These messages were made on topics o f Nicole's choice with some input from Eileen and me. Nicole was also able to independently use the side buttons and word finder to aid in message production. Throughout the study, Nicole and I worked together to customize the lexical support available in Sentence Shaper to better suit her needs. Over the course o f the four months, Nicole spent a total of 35.75 hours creating Sentence Shaper messages. O f this time, she worked independently for only 8.75 hours and the remainder was spent creating messages with Eileen or me. In comparison, the participants in Linebarger et al. (2004) were reported to have practiced using Sentence Shaper independently for approximately one-hour per day. Overall, Nicole practiced less than the other participants and typically used the program with help from other people.  4.2 Component One The first component o f the study investigated whether a change in the narrative abilities o f the participant with aphasia occurred following extended use o f Sentence Shaper. In relation to the I C F framework, the results o f component 1 reflect changes to the impairment and activity levels o f functioning and disability. Structural analysis o f the narrative samples was completed using the Quantitative Production Analysis (Saffran et  75  al., 1989). The narratives were further analyzed for content using the Correct Information Units (CIU) analysis (Nicholas & Brookshire, 1993). A n additional content and story structure analysis was completed on The Cowboy Story using the methodology described by Joanette & Goulet (1990). The results for these analyses for both the pre- and posttreatment unaided narratives and aided narratives w i l l be presented. A l l narratives are provided in Appendix G .  4.2.1 Quantitative Production  Analysis  4.2.1.1 Preparation of the data. Before an analysis o f the language structure within the narrative samples was carried out, a second rater and I independently identified the utterances to be analyzed in each sample. Results were compared and any discrepancies were reviewed and resolved by the two raters. In the Q P A procedures, Saffran et al. state that all repetitions should be eliminated from the sample prior to analysis. However, the nature o f the narrative task in this study made such a procedure problematic because each story contains repeated elements and because Nicole used repetition as a strategy for putting sentences together. It was often difficult to determine whether an utterance was a repetition o f old information or an attempt to describe the repeated parts o f the story. For example, in The Cowboy Story, it was possible that Nicole repeated "the dog is switching" because in multiple pictures, the dog (small horse) is still being switched by the boy for the cowboy's horse. In this case, multiple repetitions o f "the dog is switching" were included. If on the other hand, it appeared Nicole was still talking about the same picture, the repetitious utterances were not included. Therefore, i f an argument could be made  76  that a repeated utterance was actually referring to a later part o f the story, the utterance was included in the analysis. However, i f it was determined that the repeated utterance appeared to be a retelling o f what was previously said as a strategy for constructing sentences, it was eliminated from the analysis. Once utterances were agreed upon by both raters, an independent Q P A was done by the first rater o f the entire narrative sample and more than 15% o f the sample was analyzed by a second rater. Following a comparison o f the analyzed samples, inter-rater reliability o f the Q P A was demonstrated to be between 0.9 and 1.0.  4.2.1.2 QPA analysis of unaided  narratives.  Selections from the unaided narrative samples that characterize the productions are given in Table 4.1. F u l l narratives are given in Appendix G . Results o f the Q P A for The Three Little Pigs and The Cowboy Story narratives are presented i n Table 4.2, including the results from Linebarger et al.'s participants for comparison. Following use of Sentence Shaper, Nicole demonstrated improvement in her morphosyntactic ability in several areas. Increases in median utterance length and mean sentence length were found for both narratives. There was a striking increase in production o f sentences with Nicole demonstrating an increase greater than 80% in sentence production for both posttreatment narratives. This increase in the proportion o f sentences is partially related to the large improvement in verb usage. Furthermore, the proportion o f sentences which were well-formed was also seen to increase following use of Sentence Shaper. Morphological errors, such as improper verb inflection (i.e. blowed) and omission o f determiners, and other errors, such as verb-argument violations, were found to be the main causes o f  77  ungrammatical sentences. However, in spite o f these errors, verb inflection and use o f determiners also showed notable improvements in both narrative tasks. In comparison, the participants in Linebarger et al. also demonstrated morphosyntactic improvements in their narratives; however, SI demonstrated only minimal gains compared to S2. Furthermore, unlike Nicole, not all morphosyntactic measures showed improvement for SI and S2's narratives. For S2, significant gains in structural measures were reported; however, the majority o f morphological measures remained unchanged. In contrast, both the structural and morphological Q P A measures o f Nicole's narratives showed substantial increases.  Table 4.1 Excerpts from The Three Little Pigs pre and post narratives. Underlined portions are those words included in the Q P A . Pre-treatment Post-treatment Okay um .. um .. [hh] yeah yeah? yeah [hh] U m and the w o l f . . blew the okay the w o l f and then oh I mean um .. um um um um blewed uh uh blew the hay .. the w o l f blew um pig uh pigs pigs um um pigs um um th ( ) uh uh uh the okay the w o l f um um um blow uh uh blow blow them blow them the w o l f um is .. no okay the w o l f blew um yeah yeah and yeah and then and then I um hay .. stop mean uh two people two people and oh um A n d um um um uh the w o l f okay the pigs [hh] [heh-heh] um yeah I mean um um ... the pig the pig um um ran ran um ran away glass no I don't know I mean I mean um I ran away mean .. uh 11 don't know ( ) no yeah and Okay the um uh the uh okay okay the w o l f then I mean uh um ho [hh] and then um um um um blow .. uh the stick ... and and um .. um um uh ( ) um um uh three ( ) three and um um um three no one one two uh and and um um three um um um three and three no one one two uh three no one two um um three and um um brick brick = = two pigs uh uh away uh uh okay the okay wait the w o l f . . the uh the uh pigs pigs uh uh ran out ran out (Pigs2: 11-21) ( P i g s l : 43-49)  78  Table 4.2 Q P A analysis results for Nicole's unaided narrative samples pre/post use o f Sentence Shaper. The Three Little Pigs Pre  Post  The Cowboy Story Pre  Post  4  Linebarger et al. (2004) subjects SI S2 Post Pre Post Pre  Structural Median length of utterance  1  4  2  2  4  4  6  Mean sentence length Prop, o f utterances that are sentences  0  4.13  3.0  5.4  4.76  4.78  6.35  0  .80  .05  .89  .29  .41  .62  .80  Prop, o f words in sentences  0  .85  .08  .95  .50  .54  .73  .81  Number o f verbs  1  8  Prop, o f narrative words that are closed class  0.07  .38  .50  .39  .36  .50  .47  Prop, o f sentences that are well-formed  0  .75  .65  .36  .47  .39  .30  Prop, o f verbs that are inflected  0  .33  .50  .93  .32  .35  .41  .52  Prop, o f determiners used in obligatory contexts  0  .90  .19  1.0  9  4.35  19  Morphological  .28  0  4.2.1.3 QPA analysis of aided narratives. A t the end o f the study, Nicole was asked to create narratives about The Three Little Pigs and The Cowboy Story using Sentence Shaper. The Q P A results are presented in Table 4.3. The Sentence Shaper messages are presented in Table 4.4. Table 4.3 Q P A analysis results for aided narrative samples following use o f Sentence Shaper. The Three Little Pigs Unaided Aided Structural Median length o f utterance  The Cowboy Story Unaided Aided  4  6  4  4.13  6.25  4.35  5.00  Prop, o f utterances that are sentences  .80  1.00  .89  1.00  Prop, o f words in sentences  .85  1.00  .95  1.00  Morphological Prop, o f narrative words that are closed class  .38  .40  .50  .53  Prop, o f sentences that are well-formed  1.0  1.0  .65  .67  Prop, o f verbs that are inflected .33  1.0  .93  .90  Prop, of determiners used in obligatory contexts  1.0  1.0  1.0  Mean sentence length  .90  Table 4.4 Sentence Shaper Narratives The Three Little P i gs The w o l f tried to blow hay. The pig ran. The w o l f tried to blow sticks o f the pigs. The w o l f tried to blow the house.  (SS06/26)  5  The Cowboy Story The man is riding on the horse. The man is yawning. She is sleeping. The boy is switching around. The horse is moving around. The dog tried switching next by the dog. The boy is leaving. The cowboy switch the dog. The horse is leaving. (SS06/29)  80  A comparison o f The Three Little Pigs narratives in the aided and final unaided conditions reveals several interesting findings. The aided narrative had far fewer utterances (4), than the unaided (10); however, the proportion o f words in sentences was higher (1.0) in the aided versus the unaided condition (.85). Despite the fact that Nicole had access to the side buttons and word finder when producing the aided narratives, The Three Little Pigs aided narrative was more morphosyntactically complex on several but not all measures compared to the unaided narrative. The aided narrative was found to have a greater inflection index and elaboration o f verb phrases, a larger mean sentence length, and a greater proportion o f utterances that are sentences and words in sentences than the unaided narrative. The Cowboy Story final unaided and aided narratives showed a similar pattern to The Three Little Pigs narratives in that the unaided version was much longer (19 utterances), than the aided version (9 utterances). In contrast to The Three Little Pigs narratives, however, The Cowboy Story narratives were more comparable in terms o f morphosyntactic complexity.  4.2.2 Correct Information  Unit Analysis  4.2.2.1 CIU analysis of unaided  narratives.  The C I U analysis was completed to investigate any changes in the informativeness o f the narratives and the efficiency with which this information was communicated by the participant with aphasia (Nicholas & Brookshire, 1993). Again, inter-rater reliability measures were completed and agreement was found to be 100%.  81  Results o f the C I U analysis are presented in Table 4.5. Again, Linebarger et al.'s results are included for comparison. W i t h The Three Little Pigs narrative, Nicole demonstrated an improvement in the proportion o f CIUs to overall words and in efficiency o f production o f C I U s . For both, measures tripled from the pre- to post-treatment narratives. In contrast, similar improvement was not found in The Cowboy Story narratives, with pre and post measures remaining relatively unchanged. In comparison to the Linebarger et al. participants, both 51 and S2 demonstrated a higher proportion o f words that are CIUs than Nicole in both the pre- and post-treatment narratives. Again, S I showed minimal improvement, while 52 demonstrated a significant increase in proportion of words that are CIUs.  Table 4.5 C I U analysis results for Nicole's unaided narrative samples pre/post use o f Sentence Shaper. Results presented Linebarger et al. (2004) are provided for comparison. The Three Little Pigs Pre  Total time o f narrative Number o f words Number o f CIUs Prop, o f words that are CIUs ClUs/minute  3:10 186 13 .07 3.47  Pre Prop, o f words that are CIUs ClUs/minute  Post  3:57 172 39 .22 9.87  The Cowboy Story Pre  4:35 180 29 .16 6.33  Post  9:05 388 69 .15 7.04  Linebarger et al. (2004) participants SI S2 Post Pre Post  .32 19.0  .35 25.0  82  .46 41.0  .71 51.0  4.2.2.2 CIU analysis of aided narratives. The results o f the C I U analysis o f the aided narratives are provided in Table 4.6.  Table 4.6 C I U analysis results for Nicole's aided narrative samples. The time required to create the narratives is also included. The Three Little Pigs Total time to create narrative Total time o f narrative Number o f words Number o f CIUs Prop, o f words that are CIUs ClUs/minute  40 minutes 0:43 25 25 1.0 34.72  The Cowboy Story 30 minutes 1:47 45 33 .84 21.35  One o f the benefits o f Sentence Shaper is that it gives the user the opportunity to revise his or her utterances. In Nicole's case, it allowed her to eliminate the filler words and repetitions from her utterances which resulted in the higher C I U proportions in the aided narratives; however, as indicated by the time required to create the narratives, this was a lengthy process for Nicole.  4.2.3 Analysis of Story Schema and Informative Content in The Cowboy Story Unaided and Aided  Narratives  Using the methodology described by Joanette and Goulet (1990), the story schema and informative content o f Nicole's Cowboy Story narratives were analyzed. In the story schema analysis, the narrative was examined to determine whether each o f the main components, (i.e. the setting, the complication, and the resolution) were included. In all versions o f Nicole's Cowboy Story, she mentions a dog when referring to the toy horse. Clearly, she mistook the toy horse for a small dog when looking at the pictures. A s this error was the result o f a misperception and not a language error, mention o f the dog  83  in her narratives was not coded as incorrect. Table 4.7 presents the percentage o f core propositions mentioned for the setting, complication, and resolution in each unaided and aided narrative.  Table 4.7 The percentage o f core propositions found i n the setting, complication, and resolution o f The Cowboy Story unaided and aided narratives. Story Component  Unaided Pre  Setting ( n = 10)  2  Complication (n = 18)  Aided Post  ~~  3  3  5  4  4  Resolution (n = 4)  1  1  0  Total (n = 32)  8  8  7  ~  The utterances defined for the Q P A were used i n comparison to the core propositions developed by Joanette and Goulet. There was no change in the total percentage o f core propositions mentioned between the pre- and post-treatment unaided narratives. A slight difference occurred in the setting and complication in the posttreatment unaided narrative in which Nicole was able to add additional information to the setting, but failed to include one o f the complication concepts described i n the pretreatment narrative. In both unaided narratives a total o f only 25 percent o f the expected content was mentioned. Even though her utterances were much more grammatically complex, the amount o f informative content included in the narrative remained stable. W i t h regard to the post-treatment unaided and aided narrative there was no change in the amount o f informative content included. In some respects, the aided narrative appears more impoverished than the unaided ones. For instance, in her aided  84  narrative, Nicole failed to discuss the resolution o f the story even though she was able to describe parts o f the resolution in her unaided versions. Overall, Nicole's unaided and aided narratives include the three main components o f the story; however, her description of those components is only partially complete and much o f the information expected to be included in the narrative is missing.  4.2.4 Listeners'  Perceptions  The findings o f component one had contrasting results in that the Q P A measures indicated a substantial improvement, whereas for one narrative (The Cowboy Story), the C I U measures remained relatively unchanged. It was evident from the C I U analysis o f the unaided Cowboy Story narratives that although there was an increase i n the number o f CIUs, there was also a corresponding increase in filler words which resulted in a similar proportion o f CIUs in the pre- and post-treatment narratives. In addition, it was apparent that the narratives produced by Nicole with the aid o f Sentence Shaper did not have a more complex story structure than the post-treatment unaided narratives. In fact when The Three Little Pigs and The Cowboy Story final unaided and aided narratives are compared, more aspects o f the story are mentioned in the unaided compared to the aided versions. These findings prompted an additional investigation involving listeners' perceptions o f the pre- and post-treatment unaided narratives and the post-treatment unaided and aided narratives. These perceptions were obtained to help determine which differences, i f any, between the pre- and post-treatment unaided narratives are meaningful for listeners. In addition, the comparison o f the unaided and aided narrative  85  allowed for the effect o f Nicole's fdler words on the narratives to be examined. Four peer group listeners (i.e., female and similar in age to Nicole, but with no experience o f aphasia) and four speech-language pathologists were asked to make two judgments. First, they were asked to listen to The Three Little Pigs and The Cowboy Story pre- and postunaided narratives and give their perceptions as to which story they perceived as better. Listeners had the option of rating the narratives as the same i f neither narrative was deemed to be better. The presentation of narratives was counterbalanced to rule out order effects. In the second judgment task, listeners were asked to compare The Three Little Pigs and The Cowboy Story post-unaided narratives to the aided narratives. It was explained to listeners, prior to the second judgment task, that the delivery o f the aided narrative was clearly superior to the unaided, but that the question they were meant to answer was still which narrative was a better story. Perspectives from the two groups (i.e., the peer group and the speech-language pathologists) were obtained as it was possible that due to their training, speech-language pathologists may have listened for different aspects o f the narratives compared to the peer group.  4.2.4.1 Perceptions of pre- and post-treatment  narratives.  Table 4.8 presents the listeners' perceptions o f The Three Little Pigs and The Cowboy Story unaided narratives.  86  Table 4.8 Listeners' perceptions o f The Three Little Pigs and The Cowboy Story unaided narratives. Listeners  Pre-treatment is better.  Post-treatment is better.  Narratives are the same.  The Three Little Pigs Speech-language pathologists  0%  100%  0%  Peer group  0%  100%  0%  The Cowboy Story Speech-language pathologists  25%  50%  25%  Peer group  50%  50%  0%  It is clear that both the speech-language pathologists and peer group listeners perceived a difference between The Three Little Pigs pre-treatment and post-treatment unaided narratives with all listeners rating the post-treatment narrative as better. One peer group listener reported that the post-treatment narrative was "significantly better" and another listener stated that the ideas were clearer and that the story was narrated with more confidence. The peer group listeners also mentioned that the post-treatment narrative had more complete sentences. The speech-language pathologists made reference to the increase in content words, the decrease i n filler words, the increased grammatical complexity, and the more complete sentence structure in the post-treatment narrative. In contrast to the listeners' perceptions o f The Three Little Pigs, the same level o f agreement did not occur for their perceptions o f The Cowboy Story. In general, most listeners reported experiencing more difficulty rating The Cowboy Story narratives  87  compared to The Three Little Pigs narratives. One peer group listener noted that even though the sentence structure was more complex in the post-treatment narrative, the pretreatment narrative was easier to follow and better in terms o f conveying the ideas. Another peer group listener commented that the conclusion in the pre-treatment version was better stating that Nicole conveyed the cowboy's surprise i n the pre- but not the posttreatment narrative. In contrast, one peer group listener stated that even though more patience was required to listen to the post-treatment narrative, it was better in that Nicole had a larger vocabulary and used more complex language. One o f the speech-language pathologists also commented on difficulty listening to the post-treatment narrative because o f its length, but still found that version to be better as more content was communicated. She mentioned that the pre-treatment narrative seemed more efficient, but that without the pictures it would have been more difficult to understand the story in the pre-treatment versus the post-treatment version. The one speech-language pathologist who rated the narratives as the same reported that even though the post-treatment narrative had more complete sentence structure, it did not add more information than the pre-treatment narrative. Another speech-language pathologist commented on the speaker's effort involved in producing the more complex sentence structure found in the post-treatment narrative and suggested that this may not be the most efficient way o f communicating in conversation for Nicole.  4.2.4.1 Perceptions o f post-treatment unaided and aided narratives. Table 4.9 displays the results o f the listeners' perceptions o f The Three Little Pigs and The Cowboy Story post-treatment unaided versus aided narratives.  88  Table 4.9 Listeners' perceptions o f The Three Little Pigs and The Cowboy Story posttreatment unaided narrative versus the aided narrative. Listeners  Unaided is  A i d e d is  Narratives are  better.  better.  the same.  The Three Little Pigs Speech-language pathologists  75%  Peer group  75%  0% 25% The Cowboy Story  25% 0%  Speech-language pathologists  25%  50%  25%  Peer group  25%  75%  0%  In comparing The Three Little Pigs post-treatment unaided narrative to the aided narrative, the majority o f listeners rated the unaided version as better. In general, the peer group listeners reported that the unaided version had more information (i.e. discussed the three types o f houses), although one listener commented that the aided version was easier to understand. The speech-language pathologists also reported that the unaided narrative conveyed more o f the events o f the story and one remarked that the sequence o f events was clearer in the unaided narrative. The one peer group listener who rated the aided narrative as better reported that the aided narrative was clearer, whereas the unaided version was so long that she forgot parts o f the story. Similarly, the speech-language pathologist who considered the two narratives to be the same also commented on the length o f the unaided narrative and said that it was difficult to listen to the unaided version.  89  In The Cowboy Story comparison, the majority o f peer group listeners preferred the aided version stating that it was easier to follow; however, one listener found that even though it was harder to listen to the unaided version, it described the events more thoroughly and thus, was a better story. The one speech-language pathologist who rated the unaided version as better stated that although it was easier to listen to the aided version, the content was less clear and parts o f the narrative were incorrect (i.e. "the cowboy switched the dog"). One listener judged that the unaided version was better as the "punch-line" was conveyed, but she reported that this decision was difficult as there were many other aspects o f the unaided version which made it much worse than the aided narrative. The speech-language pathologists who rated the aided version as better reported that it was more coherent, sequentially easier to follow, and put much less stress on the listener. One speech-language pathologist rated the narratives as the same because she stated that the unaided version had more content, but the narrative was so long, that it was difficult to remember what had been said. She found the aided version to be easier to listen to, but stated that it contained less information. Overall, the many listeners commented on the effort involved in listening to Nicole's unaided narratives, and in some instances, this impacted listeners' perceptions of the narratives. However, the majority o f listeners were able to focus on the narratives and make their decisions based on the information conveyed in each narrative.  4.2 Component Two In the second component o f the study, the possible uses o f Sentence Shaper in everyday life were explored through four months o f participant observation with Nicole  90  and Eileen. In this section, the findings which emerged from the analysis o f fieldnotes, interviews, conversations, and questionnaires are presented. First, a description o f the ways in which Sentence Shaper was used is given, differentiating between uses I suggested versus those generated by Nicole and Eileen. Following this, an interpretation of the reasons for these differences are presented, including a description o f the communicative environment in which Sentence Shaper was used and the themes and patterns which emerged from the hours o f observation within this environment. Finally, the outcomes o f the study are described with regard to the Q C L and C E T I questionnaires and Nicole and Eileen's ideas for future use o f Sentence Shaper.  4.2.1 Participation Opportunities:  Uses of Sentence Shaper in Everyday Life  This study was motivated by the question o f how Sentence Shaper could be expanded beyond its initial conceptualization as a language therapy tool to a device able to augment everyday communication. A t the beginning o f the study, I envisioned two potential uses. First, the use o f Sentence Shaper messages for e-mail was a possibility as the messages can be sent as sound files attachments i n an e-mail. This is an attractive alternative as many individuals with aphasia experience difficulty writing. A second potential use for Sentence Shaper was in conversation, although it was difficult to predict how such a device, which is offline, could be used in an on-line conversation. However, I anticipated that while the program would not be appropriate for all forms o f conversation, it could be used to generate messages that a participant with aphasia might want to share with many people. Use o f Sentence Shaper in these ways moves beyond impairment and activity level intervention and has the potential to promote participation in social  91  contexts, thus addressing the level o f participation within the I C F framework. The following sections present an overview o f the training phase and Nicole's eventual uses of Sentence Shaper.  4.2.1.1 The training phase. The augmentative communication goals helped shape the structure o f the study and were explained clearly to the participants at the onset o f and throughout the study. However, before these ideas for Sentence Shaper could be explored, a training phase was needed to introduce the program to Nicole. Initially, Nicole made messages using Sentence Shaper about events that she had recorded in a scrapbook. Her scrapbook, developed as part o f an earlier speech therapy program, contained photos o f the event with a written description o f what had occurred. From the beginning, Nicole was encouraged to use the Sentence Shaper side buttons when creating her messages. When cued to use the side buttons, she was often successful in narrowing her search to one or two possibilities. A n example was noted in our first session with Sentence Shaper, when I cued her to listen to the side buttons; she was able to narrow her choice down to 'from' and ' w i t h ' and after saying the sentence for her using the two prepositions, she was able to choose the correct preposition. During the training period, I noticed that Nicole was tempted to read the entries she had written i n her scrapbook and record those sentences while using Sentence Shaper. I encouraged her to refrain from doing this as I explained that she would not always have written material with which to make messages. However, despite my requests, Nicole continued to use the written material to make messages as I noticed that many of the sentences in her messages were ones she had written in her  92  scrapbook. This prompted me to suggest that she use Sentence Shaper to tell me about things she had done on the weekend. This step was beneficial as it encouraged Nicole to generate messages without relying on written material. The word finder was also introduced to Nicole. She showed immediate interest in the word finder and spontaneously began listening to all the words and repeating them.  4.2.1.2 Nicole's strategies for creating Sentence Shaper messages. When making messages with Sentence Shaper, Nicole's strategy was to repeat and expand the utterance until it reached a level o f complexity that satisfied her. Instead of using the Sentence Shaper desktop as a work area, she tended to use that space as a prompt. For instance, she often recorded parts o f the sentence and listened to those recordings to help her say the whole sentence at once. This situation is illustrated in the following excerpt: Excerpt 4.1 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.  N:  The man the okay okay the horse ( R E C O R D O N ) the horse .. is uh .. uh a uh a go a go ( R E C O R D O F F ) no no okay okay uh the uh horse is um walking w u w u okay the horse is the horse is um the horse is .. the horse is um going ( R E C O R D O N ) the horse is going ( R E C O R D OFF) the horse is Sentence Shaper: T H E H O R S E IS G O I N G <7> Sentence Shaper: T H E H O R S E IS G O I N G ... Sentence Shaper: T H E H O R S E IS G O I N G N: The horse is the horse is leaving ( R E C O R D O N ) the horse is going ( R E C O R D OFF) no no okay the horse Sentence Shaper: T H E H O R S E IS G O I N G Sentence Shaper: T H E H O R S E IS G O I N G N: The horse is leaving ( R E C O R D O N ) the horse is leaving (RECORD OFF) Sentence Shaper: T H E H O R S E IS L E A V I N G (C17: 1-15)  93  This strategy was interesting as it differed from the way in which the designers o f Sentence Shaper had predicted individuals with aphasia would use the program. Instead of recording each word or few words o f the sentence and arranging those words i n the sentence shaping area, Nicole used an internal workspace to expand her utterances and with each attempt was often able to increase the length and complexity. She then recorded the entire sentence when she was satisfied with her production. When a sentence involved difficult words, I often suggested that she split the sentence i n two and record each part. However, when possible, Nicole tended to record one sentence per icon and arranged those icons in the sentence shaping area. Therefore, one narrative button could contain several sentences. A similar strategy o f repetition and expansion was observed for N i c o l e ' s final unaided narratives. Excerpt 4.2 demonstrates how she was able to achieve a target sentence incorporating the strategy she used when making messages with Sentence Shaper. Excerpt 4.2 Okay um pigs um okay um .. mm ... mm .. um ... pi uh there is uh no there is there is um ... uh uh there is .. there is um .. okay um there is .. there is uh .. .there is three pigs (Pigs2: 7-9) In this example, Nicole had several attempts before she was able to say "there is three pigs".  4.2.1.3 Use of Sentence Shaper with e-mail. Once Nicole had spent several hours (6.5 hours) practicing using Sentence Shaper, I decided to introduce the possibility o f using Sentence Shaper with e-mail. In the  94  first e-mail, Eileen and I supported Nicole by helping her with the parts o f an e-mail (i.e. introduction and closing) and with generating ideas for the e-mail. Following this, Nicole sent a number o f e-mails to family members, volunteers, and friends. Initial feedback for Nicole's e-mails was very positive. For instance, M i r a , a former volunteer who had worked with Nicole, wrote in an e-mail "Nicole it was nice to hear your voice" (FN04/14: 8-9), and one o f Nicole's family members wrote " W e just listened to your homework and loved it" (FN04/21: 19). It was apparent that Eileen was pleased with this feedback when she shared that the family was very impressed with the e-mails Nicole had been sending them. She reported that hearing Nicole's voice was quite emotional for one family member, saying that he had phoned Eric to say how much he liked the e-mail and had started crying. Nicole also responded positively to the e-mails she was sending. After we had sent one e-mail message, Nicole repeated some o f the message she had just sent and then said "nice...whole sentence". When asked i f she liked sending the e-mail, she said "yes" (FN04/18: 54-56).  4.2.1.4 Use of Sentence Shaper messages in conversation. The next step was to encourage Nicole and Eileen to use Sentence Shaper in conversation. There seemed to be some reluctance on their part to use the program in this way. After my encouragement, Eileen thought o f several people with whom Nicole could use Sentence Shaper, explaining to Nicole that she could use it with her volunteers, with people at the stroke group, and even with family members. However, Nicole rarely used the program spontaneously in this way. Part o f the problem may have been that she was not sure what to talk about when using Sentence Shaper. For instance, when I suggested  95  she use Sentence Shaper to share a message with the people at the stroke group, Nicole said "speak speak .. what happens", answering "yes" when I asked her i f knowing what to talk about was a problem (FN04/21: 62-63). However after making a message about her holiday in M e x i c o , Nicole was motivated to share her M e x i c o message with others, so we chose this as a good message to share at the stroke group and decided to play the message for M i r a as a trial run. From my perspective, this use o f Sentence Shaper was quite successful: Nicole played the M e x i c o message for everyone and as it was playing she spoke the words aloud as well. People in the group made comments as it was playing and laughed at the funny parts (i.e. the tequila). Nicole was smiling the whole time (FN06/15: 19-23).  4.2.1.5 Sentence Shaper as language therapy. In contrast to the uses I suggested o f Sentence Shaper as an augmentative communication device, many o f the ideas Nicole and Eileen generated were related to Nicole's language therapy. For instance, Nicole used a Sentence Shaper message to help her with a homework exercise she had from the speech-language pathologist at the stroke group. For homework, she was asked to write about her trip to M e x i c o ; she listened to the Sentence Shaper message about her trip and wrote down what she had said. After I had introduced the word finder to Nicole, we made a new category o f present-past tense verbs. She seemed very motivated to add to this list and often had new words to add. Nicole also came up with ideas for additions to other categories o f words in the word finder: She went to the word finder and went to the preposition section. She got out her workbook and showed me that the word finder did not have all the prepositions that she had in the workbook.... then Nicole went to the adjective section and  96  then the colour section and said "pink and then". She turned to another page in her workbook that had many colours listed (FN06/13: 60-68). The way that Eileen saw Sentence Shaper being used was as a device to improve Nicole's unaided speech. For example, not far into the study Eileen wondered i f Nicole could practice repeating the messages so that she could say them without Sentence Shaper. She said that she wanted to see how close Nicole could get to those messages. To her, the main goal o f working with Sentence Shaper was that she should be able to say the messages again and again without the help o f Sentence Shaper. A s described previously, Nicole spent a total o f 35.75 hours making messages using Sentence Shaper, o f which only 8.75 hours were independent use o f the program. Working with other people in my absence was something Nicole spontaneously chose to do from the start. I was cognizant o f the fact that the participants i n Linebarger et al.'s (2004) study had worked independently, and in partially replicating the design o f that study, I thought it was important to stay as close to their method as possible. For this reason, I encouraged Nicole to spend at least part o f her practice time recording messages by herself. Despite Eileen telling me that she thought it was better when Nicole worked by herself, she continued to help Nicole with the Sentence Shaper messages throughout the study.  4.2.2 Contextual  Factors  In the ICF, contextual factors, which contribute to a person's functioning and disability, are separated into personal and environmental factors. A description o f these factors is necessary as the context in which Sentence Shaper was placed strongly  97  influenced how it was used. M a n y o f these factors were not apparent at the beginning o f the study and only became clear through the course o f the study.  4.2.2.1 Personal factors: Nicole as a  communicator.  In describing the context into which Sentence Shaper was introduced, an in-depth description of Nicole as a communicator is required. This description expands on the initial communication profile which was based on observation before Sentence Shaper was introduced. In the initial interview with Eileen, she revealed that Nicole often required prompting to aid in word retrieval: "I found that there she can uh she can come up with lots o f words i f you ask ask the right question" (12: 48-49). Eileen mentioned that Nicole was more successful when talking about familiar activities. This was apparent in many o f the conversations Nicole initiated with me. A t most meetings, she repeated the same questions, such as "would you like tea or coffee" (FN03/14: 2) or "how was school" (CI 1: 1). This suggests that she has more success talking about regular activities with their own scripts (i.e, "a standard sequence o f events that describes a situation" as defined by Brown and Yule, 1983). In the final interview, Eileen was asked further about communicating with her daughter. It was apparent that Eileen felt that something was missing from the conversations she held with her daughter. She commented that she and Nicole do not really talk about anything other than Nicole's needs and wants. In Eileen's words, it is "not really conversations" (FN06/29: 93-95). Observing conversations between Nicole and Eileen was an important source o f information about Nicole as a communicator. From the beginning o f the study, it was clear that Eileen tended to cue her daughter to talk about certain things by asking her  98  questions and, when this strategy failed, she often spoke for her daughter. However, other strategies were observed as the study progressed. For example, when Nicole said a one word utterance, Eileen sometimes modeled the full sentence for Nicole to imitate. This strategy o f helping Nicole expand her utterances was a method that she seemed to like. A t one meeting, she asked me i f I would like a cookie by pointing to the cookies and saying "cookie". After this occurred, she asked me how she would say that in a sentence (FN04/14: 80-82). This strategy was also used by the organizers and volunteers at the stroke group. When Nicole was collecting everyone's nametags she said "name tag please". One o f the volunteers told Nicole to say "pass me your nametag please", and she was able to say this; however, when she needed help, she would look to me to repeat the phrase (FN06/15: 69-73). In addition to needing help with production, Nicole required assistance with comprehension. Eileen attempted to compensate for this difficulty by repeating what others had said. For example, when Nicole, Eileen, and I were talking, Eileen often repeated what I had said to make sure Nicole understood. Eileen also used gestures to help her daughter understand. One particular instance occurred when we were discussing which days I should meet with them; Eileen helped Nicole understand what days we were talking about by counting the days with Nicole on their fingers. While I was talking with Nicole, I attempted to help her understand by only saying part o f a sentence and waiting for her confirmation that she had understood before continuing with the rest o f my message, as illustrated in the following excerpt: Excerpt 4.3 15. 16.  E: N:  We did Father's Day on Saturday Yes yes  99  17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.  E: N: E: N: E: N: E: N: E: N: E: N: E: N:  Instead o f Sunday O h yeah yeah yeah M y parents were busy on Sunday Okay yeah yeah = = So we went out for breakfast O h break oh breakfast yeah yeah yeah U m .. to .. this restaurant called Havana Tovana? Havana O h yeah O n Commercial Drive ... on Commercial Commercial Commercial oh oh Yup O h oh oh (C15: 15-30)  In this example, Nicole confirmed her understanding by either saying "yeah" or by repeating part o f the utterance I said. Another strategy I used was writing down key words for Nicole when we were having a conversation. For example, when Nicole had a question about how Sentence Shaper could be used in conversations, I wrote down the steps she would need to take which helped her understand the process. M y perception of Nicole was o f a person with a strong desire for social connection. This was demonstrated in her very social nature, which was observed on a number o f occasions, but was particularly evident at the stroke group in the way in which she greeted the members as they arrived and in her involvement i n the group activities. Her social nature was also apparent i n the way she preferred to work with others, such as her mother and volunteers, on her language therapy activities. However, another important characteristic o f Nicole as a communicator was her dependence on her communication partner, i n these social situations. Eileen highlighted this issue i n the initial interview: " I ' m doing more talking than she is ... I'm asking questions and she just has to say yes or no" (12: 10-12). This comment conveys a sense o f the unequal balance  100  that occurrs in conversations with Nicole, in which the conversation partner often has to contribute much more to the interaction and also plays a greater role in controlling the direction o f the conversation. Throughout the study, Nicole was observed using strategies to augment her communication with others. She often used gestures to help when she was having difficulties with word retrieval. For example, she pointed to her computer desktop to indicate that she wanted Sentence Shaper to have a shortcut on the desktop. Another strategy Nicole incorporated into her communication was the use o f objects to start conversations. For example, she was able to initiate a conversation about her former job by showing me her old business card. She also used her daytimer to help with word retrieval, as on one occasion, when I asked her about what she had done on the weekend, she used her daytimer as a prompt. In addition, she had a notebook with questions and phrases written in it. Although I never saw her use this spontaneously in conversation, she did use it once when making a message about her partner, Eric, to help say the phrase "Eric and I have been together for a long time" (FN04/28: 56). Nicole also incorporated drawing as a communication strategy three times in my presence. For example, in telling me about a visit her family had made to the graveyard, she mentioned the food they left at someone's grave and then drew a picture o f an incense stick and started blowing. A t the stroke group, one o f Nicole's tasks was to call out the numbers for bingo. To facilitate her retrieval o f numbers, she had a page with all the numbers written down, and with this she was generally able to say the correct number aloud. She also facilitated her verbal expression o f numbers by first writing them down. This strategy was demonstrated several times; in one case, Nicole wrote down the numbers 1 A to help her say the l  101  number o f years she worked as a loans officer. Clearly, these strategies were an important form o f communication for Nicole, and they contributed greatly to my understanding o f what she was trying to communicate. Taken together, this description allowed for a deeper understanding o f the context in which Sentence Shaper was used. A number o f important issues are apparent, namely, Nicole and Eileen's style o f communication, Nicole's social nature, her dependence on conversational partners, and the strategies Nicole used to augment her communication. These issues had implications for the use o f Sentence Shaper and w i l l be discussed further with regard to the themes and patterns found in the data analysis.  4.2.2.2 Environmental factors: The social context. The I C F defines environmental factors as "the physical, social, and attitudinal environment in which people live and conduct their lives" ( W H O , 2001, p. 10). With respect to Nicole's use o f Sentence Shaper, the social environment was a particularly important factor. Nicole's weeks were structured with various activities. O n weekdays, she spent the day with her mother at her mother's house. In the evenings, she went home to her apartment which she shared with her partner, Eric. During the weekdays, Nicole did several activities with her mother: they worked out at a gym, went swimming, worked on scrapbooks together and, in Nicole's words, "study language" (FN03/10: 34). Once weekly, Nicole spent the morning at a stroke group. She had a number o f tasks she performed at this group. She handed out the nametags and collected them at the end, took the lunch orders and collected the money, and called out the numbers for bingo. Nicole also had the opportunity to work with a speech-language pathologist, A l l i s o n , at the  102  stroke group. In the two sessions I observed at the stroke group, A l l i s o n and Nicole reviewed written homework Nicole had completed at home and then A l l i s o n provided Nicole with new worksheets to do for next time. Eileen had also arranged for Nicole to practice her language skills with a number o f volunteers who came to the house. Nicole generally had one, or sometimes two, sessions with volunteers per day. The volunteers often helped Nicole complete her homework from the stroke group. It was evident that the majority o f this language practice focused on Nicole's reading and writing skills. The majority of her worksheets concentrated on choosing a correct preposition to describe a picture, answering yes/no questions, categorizing nouns, and putting a single word into a sentence. Nicole also had several programs on her laptop designed to provide her with either language therapy or support. For instance, she used K u r z w e i l 3000 to help her read the e-mails she received. She also had K P Typing Tutor for typing practice, Sights ' n Sounds, and Puzzle and Word Games. From Eileen, I learned that Nicole was very motivated to practice her language and that she spent a large portion o f her day working on her language skills. Clearly, "studying" language, as Nicole and Eileen referred to it, was very important to Nicole and her days were structured to allow her to do this. Another aspect of Nicole's social environment during this study was our relationship. I was acutely aware o f our respective roles. During our time together, I tried to maintain the role o f a researcher in a participatory design study. Initially, I had envisioned Nicole and me taking equal roles in guiding the course o f the study and in determining different ways in which Sentence Shaper could be used. However, this was not the way in which my role evolved. From Nicole and Eileen's perspective, I was a  103  clinician providing Nicole with language therapy. I distinctly felt the unequal power balance o f the clinician/patient roles in that I was often required to guide the structure o f the meetings in the same way as a clinician with a lesson plan. Nicole also connected me to the role o f a clinician by showing me the various language activities she did in her workbooks. A t one point in the process o f making an e-mail message, I asked Nicole what we did together. She responded by saying "homework and reading" (FN04/11: 93), which I considered a description o f the activities she does with her volunteers, clearly placing me in the role o f someone who helps her with her language. A tension existed between acting as a clinician and wanting to encourage Nicole's independent use o f Sentence Shaper. This tension was present throughout the study and is captured i n the following excerpt: Excerpt 4.4 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.  E: N: E: N: E: N: E:  That might be something you can = work on together = (referring to prepositions) = yeah yeah yeah yeah = yeah yeah Because I think um .. these are hard Yeah yeah A n d um .. .1 think that w i l l be something that would be helpful for you Nicole = =Yeah yeah = yeah = = to know = (C3: 1-10)  O n this occasion, I recognized that Nicole was having difficulty understanding prepositions, so I tried to minimize my role as a clinician, by suggesting that she work on prepositions with her volunteers. This description of Nicole's social environment illustrates the individuals with whom she has the opportunity to communicate on a weekly basis. Aside from her mother,  104  with whom she spends a great deal o f time, the majority o f people Nicole sees during the weekdays are there for the purpose o f language therapy. O n the weekends, she spends time with her partner, friends, and family; however, these interactions were never observed as Nicole and Eileen chose to schedule meetings on weekdays only.  4.2.3 Patterns and Themes: Interpretive  Support  This section w i l l describe the patterns and themes which emerged from the data. Throughout the study, the notion o f independence was an extremely influential factor i n determining how Sentence Shaper was used. The ways in which Nicole was independent versus dependent on others and the reasons for this can be seen to interact with all other identified patterns and themes. Thus, independence can be considered an overarching concept and the way it interacts with the other various themes and patterns w i l l be discussed.  4.2.3.1 Language remediation as meaningful activity. Within the I C F framework, the way someone is able to function or live with a disability is o f concern and warrants attention in assessment and intervention. In Nicole's case, she lives with her aphasia by focusing on language remediation, such that participating in language therapy for many hours every week can be considered a meaningful activity for her or, in other words, her occupation. This pattern resonated throughout the study in various ways. The extent to which language therapy occupied Nicole's life suggests that Nicole and her mother held a view o f Nicole recovering, and the activities Nicole engaged in during the weekdays were geared towards that goal.  105  Another important aspect o f Nicole's language remediation is that it allowed for Nicole to obtain a social connection with others. Throughout the study, Nicole expressed that she had difficulty talking to many people she knew before her stroke, stating that it was "so fast" and "frustrating" (FN06/02: 100), thus making it difficult to gain a social connection with them. Eileen elaborated on this by saying that others often dismiss her and do not help her understand what is being said in conversations. This sentiment was also apparent in Nicole's Q C L questionnaire in which she rated her answer to the question "people include me in conversations" very negatively. A s N i c o l e ' s occupation at this time was working on her language recovery, many people in her life were there for that particular reason (i.e., her volunteers). Eileen has helped Nicole to structure her life in ways that provide her with the opportunity to communicate with people who are willing to be patient, thus providing Nicole with an important social connection. A s an important part o f Nicole's life was language therapy, Sentence Shaper was often viewed purely as a language remediation tool. This was evident i n the fact that working with Sentence Shaper was often viewed as homework by Nicole and Eileen. Initially, I started by calling it homework, but attempted to move away from this label, as I hoped that Nicole would become motivated to use Sentence Shaper for communicating with others as well as for language therapy. Nicole's messages which she sent as e-mails could be seen to undergo a transformation from being purely done as homework to emails which actually served a communicative purpose. In the beginning Nicole tended to label her e-mails as 'homework': " H i Shelly. H o w are you? This is my homework. I have to e-mail somebody about Sentence Shaper. I am practicing Sentence Shaper. H o w do you like it? B y e . " (SS04/20: 1-2). Later her e-mails lacked this 'homework' label and  106  served a purpose o f asking a question or telling someone about an event that had happened. For example, Nicole came up with the idea for this e-mail and we worked on it together: " H i Michelle. H o w are you? H o w are Brian, Joanna, and Chris? Michelle how about Tuesday M a y 16 you and me go shopping together. D o you want to go to th  Metrotown? See you soon. B y e " (SS05/12: 1-3). However, even at the end o f the study, Nicole and Eileen still seemed to view Sentence Shaper as primarily a language remediation tool. For example, towards the end o f the study Eileen said that Nicole should do the M i r a message for "homework" (FN06/06: 77), clearly placing Sentence Shaper in a language therapy role. Nicole also sent me an e-mail in which she recited poems that she practiced for her language therapy: "Dear Erin. H o w are you? I am practicing my poem..." (SS06/15: 1). This e-mail did not seem to serve a communicative purpose other than language practice. However, in light o f Nicole's focus on and motivation for language therapy, viewing Sentence Shaper primarily as a language therapy tool makes sense in the context o f her life.  4.2.3.2 "Sentence Shaper is not real life ". A major focus o f the study was the exploration o f Sentence Shaper as a way o f augmenting communication. Perhaps because an important part o f Nicole's life was language therapy, utilizing a device to augment communication at this time did not correspond with her or her mother's goals and expectations o f recovery. They may have been willing to see use o f Sentence Shaper in everyday communication as a short-term goal, but Eileen was quite clear that the ultimate goal was for Nicole to not have to rely on it. This perspective remained throughout the study, and at the end o f the study, Eileen  107  clearly indicated that she did not think Sentence Shaper was useful for augmenting Nicole's communication. In her words Sentence Shaper is "not real life" and Nicole w i l l not always have her computer with her to play the messages" (FN06/26: 63-64). During the course o f the study, Nicole traveled to M e x i c o for one week, and when she returned, we made a message about her vacation using Sentence Shaper. Following this, Eileen noticed that her daughter was able to say parts o f the message unaided and was encouraged by this fact, illustrating her goal o f recovery for Nicole. To explore this ability, I asked Nicole to tell me about her trip. In this unaided narrative, Nicole was able to repeat certain parts of the narrative and an improvement was observed in her ability to speak in complete sentences occasionally. In addition, an increase was noted in the amount o f information she was able to communicate about her trip after first composing the message with Sentence Shaper, as compared to her description o f her trip before this practice, a description which was quite limited and conveyed in one or two word utterances. The Sentence Shaper message and corresponding narrative sample are presented in Table 4.10.  108  Table 4.10 Sample o f Sentence Shaper message and corresponding message unaided. Sentence Shaper message Sentence Shaper message unaided M a y 19, 2006. Whole family went to Las Oh oh um whole family um uh okay uh Cabo Mexico. We stayed at the house. The whole family um uh whole whole family house was beautiful and nice. We had uh um uh uh whole family um um whole swimming pool, backyard, and bbq. I went family uh uh uh big big the the whole shopping. I bought rosary, blanket, family um um big uh big uh rent a room souvenirs, jewelry, shirt, and glasses. Eric room okay rent a room rent a room = yeah bought six t-shirt. I also bought a beautiful = yeah mirror. We went to the beach. We went A n d um um uh uh Jacob uh Jacob and swimming and suntanning. We went on the Robert are swimming boat. We saw seals, dolphins, and fish. The A n d uh uh I mean uh uh uh uh Eric uh uh seals were lying down on the rocks. M a y uh Eric uh wearing wearing a hat 26 was Scott's birthday. Whole family Okay okay the hat the uh uh Eric uh Eric went out for dinner. Scott had lobster and um wearing a sombrero the hat ribs. He drank tequila. Scott wore a wig. It U h uh I bought blanket, uh a blanket, rose was funny. We had chocolate cake for his .. rosary, uh sovernir, uh uh a beautiful uh birthday. O n M a y 28 whole family left uh uh mirror and uh ... and uh uh t-shirt Mexico. Whole family was tired. I slept i n U h I uh uh ka uh uh uh uh uh a uh uh ai um plane. It was a beautiful ten days. Scott uh uh Scott uh Scott had a um um no (06/02: 7-16, 104-105) okay okay happy happy no kay no Scott uh uh birthday uh .. um uh okay okay my birthday no kay the Scott uh um .. some some cake no no .. [heh-heh] I don't know um um .. I don't know U m okay um .. oh um .. um um the house okay the the uh the boat is uh the boat is um the boat is um um um the boat is lying um uh lu lying the okay the the okay the boat uh uh the boat is fishing fishing .. yeah (Mexico: 1-33)  In general, Nicole and Eileen appeared to accept the use o f Sentence Shaper more for e-mail than for conversation. This was evident in the way they used Sentence Shaper to create and send e-mail messages on their own, in addition to the messages I had suggested. However, use o f the same messages in conversation was only done at my request. Even when opportunities to use Sentence Shaper in conversation presented themselves, Nicole only spontaneously used the program in this way once. This occurred after playing the M e x i c o message for M i r a , when Nicole chose to share with us a  109  message she had made about having lunch with M i r a the previous week. This Sentence Shaper message is presented in the following excerpt: Excerpt 4.5 June 2, 2006. M o m and I visited M i r a in North Van. It was a beautiful house. I gave her presents. M i r a thank you for the gift. M i r a boiled some tea. We watch pictures on the computer. M i r a looks beautiful. M i r a showed me her engagement ring. It is a big diamond ring. M i r a and Sam are engaged. We had lunch at M i r a ' s aunty and uncles's restaurant. I had eggplant, shrimp, and chicken. I had mango milkshake and tea. It's so good and delicious. (SS06/08: 1-6) When Nicole played Sentence Shaper messages in conversation she chose to play each narrative button one by one as opposed to playing the entire message all at once. This allowed for greater interaction between interlocutors as the pause in the message allowed for the listeners to comment on the story being told. Nicole also talked along with the recorded message as it was playing, effectively making the off-line message online. The conversation which occurred between M i r a , Nicole, and me while Nicole played the Sentence Shaper message is provided in the following excerpt: Excerpt 4.6 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.  SS: E: M: N: SS: N: SS: N: SS: N: SS: N: SS: N: SS:  (First narrative button) J U N E 2 = VISIT M I R A = O h this is the message  N D  2006 M O M A N D I  Oh =Yeah yeah = I N N O R T H V A N IT W A S = A B E A U T I F U L H O U S E = = Was a beautiful house = I GAVE HER = = her = = PRESENTS Presents MIRA = = Mira = THANK Y O U FOR = = Thank you for = the gift = T H E GIFT  110  17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.  SS: N: SS: N: SS: N: SS: N: M: SS: N: SS: M: SS: N: M: SS: E: M: SS: N: E: SS: M: SS: N: ' SS:  44. 45. 46.  SS: N:  (Second Narrative Button) M I R A Mira BOILED Oh SOME Some TEA Tea tea yeah yeah [heh-heh] W E = W A T C H = PICTURES = O N THE COMPUTER = = watch = on the computer = MIRA LOOKS BEAUTIFUL Oh thank you MIRA SHOWED M E HER ENGAGEMENT RING Ring yeah yeah yeah = I forgot it IT IS = A = BIG = DIAMOND RING = [heh-heh] = = I forgot my ring MIRA A M ) = = O h my G o d = = [heh-heh] = SAM ARE ENGAGED = = I forgot = WE HAD LUNCH A T = MIRA'S = = Mira's = AUNTY AND = UNCLE'S RESTAURANT = = Uncle's restaurant = I H A D E G G P L A N T SHRIMP A N D C H I C K E N I H A D M A N G O M I L K S H A K E A N D T E A I T ' S SO G O O D A N D D E L I C I O U S = = Delicious (CI 4: 1-46)  When this message was played for M i r a and me, it created a context for conversation to occur. In this way, the message initiated a social interaction that may not have occurred without it. Furthermore, while the message was playing, she used the recording as a cue to help her repeat the message. Sharing the Sentence Shaper messages in this way had the effect o f making it seem as though Nicole was sharing the message on-line and not from a recording.  Ill  4.2.3.3 Seeking perfection. Another pattern which emerged from the data was the tendency for both Nicole and Eileen to want the Sentence Shaper messages to be grammatically perfect. This desire for perfection was apparent with the messages Nicole sent as e-mails. Nicole and I had made an e-mail message together for her Aunt Diana. While making this message, I had attempted to keep my input to a minimum and encouraged Nicole to produce what she could on her own. The following excerpt is the e-mail message we created together: Excerpt 4.7 Dear Diana Aunty. H o w are you?. O n 2 3 Diana Aunty, and mom, and me let's go to lunch. Call us. Take care. rd  (SS04/18: 1-2) After finishing the e-mail and sending it, Nicole seemed pleased with what she had done saying "nice...whole sentence" (FN04/18: 54-55). However, in the following session with Nicole, I learned that she had redone the e-mail to Diana with Eileen. Her second attempt at the e-mail to Diana is presented in Excerpt 4.8. Excerpt 4.8 Dear Diana Aunty. H o w are you? A p r i l 2 3 . D o you want to go to lunch on Sunday? Call us. Take care. rd  (SS04/20: 1-2) Nicole's second e-mail to Diana is more grammatically complex with the use o f a "do" question. A s this example demonstrates, Nicole and Eileen focused on correct grammar; however, Nicole, on her own, was often unable to correct the ungrammaticality herself. Therefore, working with others on the Sentence Shaper messages allowed Nicole to receive the help she required to make the messages perfect. Nicole also shared this  112  preference with me after we had made a message together in which I had provided her with considerable support. Excerpt 4.9 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.  N: E: N: E: N:  E: N: E: N: E: N: E: N: E: N: E: N: E:  Yeah yeah I mean um j a um um Erin and me uh spend time spend time okay talk talk yeah yeah U m hm Yeah yeah alone alone one [fj oh my god .. I mean uh .. one .. I mean [hh] I mean uh When your working Working yeah yeah I mean yeah yes yeah I mean I mean I mean uh I mean okay uh uh uh I mean um uh uh co I mean uh uh I mean uh uh tal I mean um um computer yeah I mean um friends I mean uh uh Erin and me okay good good yeah yeah yes yeah yeah .. um you no I mean no I mean um I mean um talking talking talking [Nicole making disgusted sound] Hmm Yeah bad bad So when you're alone Alone yeah uh Erin and me yes yeah good one yeah yeah and then uh speak speak I mean uh uh s uh uh = = So you're pointing to this the recorder Yeah yes yeah yeah I mean uh one one only one .. only [Nicole making disgusted sound] no good no = no I mean = = Okay so = working on Sentence Shaper Yes B y yourself B y myself Y o u don't like that Yes yes and then uh talk I mean uh uh E r i n and you uh friends yeah I mean uh um yes yeah Okay (CI 6: 1-29)  In this transcript, Nicole indicates that she does not like working on her own and that she prefers to work with others. Even though Nicole became quite adept at using Sentence Shaper and was able to create messages with more grammatically correct and structured language than her unaided language, she clearly did not think this was good enough. In line 12, she made a "disgusted" sound demonstrating this dislike. A t one point, I asked  113  Nicole why she did not like working alone on the messages. She started mentioning the side buttons and the stroke group, and I realized that she was referring to when I had cued her to use the side buttons when making the message at the stroke group. Clearly, Nicole appreciated this type o f help and felt that it was beneficial to the messages she produced. Incidentally, Nicole's use o f "friend" in lines 9 and 27 reflects back to the pattern o f language therapy allowing for social contact. Eileen's preference for Nicole's Sentence Shaper messages to be grammatically correct was evident in the way she helped Nicole with the messages. Instead o f prompting her with questions or discussing ideas for the message, Eileen often provided Nicole with the words to use in the message. Later, Eileen stated that Nicole would often say one word and then not be able to expand on it. To help her daughter, Eileen often mouthed the words so that Nicole would be able to say the sentence. This helping behaviour often made it difficult to determine what parts o f the message Nicole had done independently.  4.2.3.4 Nicole as a dependent communicator. Nicole's dependent communication was another pattern which emerged from the data. Her dependence in communication emerged as a preference i n working with others when making messages with Sentence Shaper. Part o f this dependence most likely resulted from, in addition to her desire for grammatical perfection, a difficulty in generating ideas for topics o f conversation. This was often apparent when Nicole was making messages with Sentence Shaper to send as e-mails. She often began with her typical e-mail greeting, and then she did not seem to know what else to say. A t one point,  114  Eileen said that Nicole had required a lot o f help coming up with ideas about what to put in the e-mails. During our meetings, I was often responsible for coming up with topics to talk about. Once a topic was chosen, Nicole also demonstrated difficulties expanding on that idea. This was often apparent when she wanted to e-mail someone. She could decide on whom to e-mail, but had difficulty knowing what to say i n the e-mail. Once Nicole had an idea for a message, she often had difficulty expanding those ideas into sentences, as I noted in this fieldnote excerpt: I saw that she was jumping from one idea to the next and often came up with good sentences but never recorded them. She seemed to have a hard time expanding the one word ideas into sentences. (FN04/28: 70-75) Although, Nicole could occasionally generate ideas on her own, it was nonetheless a major source o f difficulty during the study. In order to support Nicole in message structure, I began asking her to tell me everything she could about a message she wanted to create. I then wrote down her ideas, which were typically represented as single words. She then worked from her list o f ideas putting them into sentences. I found this method particularly useful as it allowed Nicole to see the organization o f the story. The difficulties Nicole had when making a message with Sentence Shaper were both effortful and quite frustrating for her. Because this frustration was very evident, it is not surprising that Eileen wanted to help her daughter, even though she knew I preferred Nicole to work independently: Eileen said that she thought it was better when Nicole worked by'herself, but that it took her a very long time and it was frustrating for her. She said that it was much easier for Nicole when she works with her because she could cue her with gestures. (FN04/11: 37-44)  115  Because o f the effort and frustration involved in creating messages with Sentence Shaper, I began to notice Nicole becoming reluctant to make messages with me, which may have been due in part to the fact that I provided less support than her mother. Her expressions very clearly demonstrated this reluctance. O n one occasion, she made a face that showed me that she preferred not to work with Sentence Shaper, but in the end, agreed to make a message. In addition, Nicole began to voice her preference at working with Eileen on the messages. I suggested that she should try to work independently, but she repeated that she intended on practicing with her mother. I finally realized that as an equal partner in this project, Nicole had the right to make the Sentence Shaper messages as she saw best, and clearly, she felt that working with others to create the messages was the best situation for her. This realization allowed me to feel comfortable helping Nicole with her messages, even though this departed from the Linebarger et al. (2004) methodology. Being placed in a clinician's role was somewhat frustrating for me; however, once I accepted that this was how Nicole saw me and realized that it could not be changed, working with Nicole became much easier. B y providing Nicole more support with her messages, it created a better working relationship and decreased our frustrations. Towards the end o f the study, Nicole indicated that she wanted to make a Sentence Shaper message for her mother. Again, she seemed to have difficulty putting her ideas into words, so I wrote down her ideas for her. From her ideas, I realized that Nicole wanted the message to thank Eileen for everything that she does for her: I asked her i f she had difficulty saying what she really wanted to. She said yes. I could see that this was something that really affected her. She wanted to be able to express the extreme gratitude that she feels for her mom, but was unable to put it into words. (FN06/20: 155-160)  116  For this message, I increased the level o f support that I typically provided and helped Nicole come up with ideas and put them into words. When we were finished, Nicole seemed quite pleased with the message and said that she was going to play it for her mother. It was evident that Nicole had appreciated the help I had given her with the message, and as I was leaving, Nicole said "thank you everything". I realized that the message she had made for Eileen was meaningful to her and that she had appreciated the help I had given her (FN06/20: 174-177). Clearly, Nicole's frustration at not being able to put her ideas into words was a motivation for working with others on her messages.  4.2.3.5 Nicole as an independent user of technology. In contrast to her reluctance to make Sentence Shaper messages independently, Nicole was very motivated to customize the side buttons and word finder, which was something she often initiated herself. She seemed to enjoy adding more words to the different categories available in the word finder. When we were working on the word finder or side buttons, I felt that our roles underwent a change. I no longer felt the unequal power balance o f the clinician/patient roles that occurred when we were making messages with Sentence Shaper together. When customizing the word finder and side buttons, we became equal investigators o f Sentence Shaper, and this may have been a source o f motivation for Nicole. In this context, I was not seen as an expert i n Sentence Shaper but, rather, as someone learning the program along with Nicole. The fact that I was not an expert was very clear to Nicole in my struggles to figure out different aspects o f the program.  117  In this context, Nicole was observed taking steps to become an independent user o f the technology. She appeared to be very motivated to learn about the different technology available on her computer and, with everything she learned, she wanted to know how to do it by herself. For example, not long after learning how to use the program, Nicole was already independently renaming the messages on her own. When she was learning about her new C D burner from a volunteer, Nicole often repeated the steps that the volunteer had said and then did the actions herself. Another example occurred when Nicole and I were learning how to transfer the changes we had made in one section's word finder to all the other sections, as this was not done automatically. Nicole asked me to write down the steps so she would be able to do it herself. These examples illustrate Nicole's independence when using her computer and the steps she took to allow her to be independent. Her desire to be an independent user o f technology was in striking contrast to the dependence I observed when she was making messages with Sentence Shaper. Clearly, confidence with technology was not a factor which prevented Nicole from being able to make the Sentence Shaper messages independently.  4.2.4 Outcomes of Sentence Shaper This section w i l l present the various outcomes o f the ethnographic component o f the study, namely, the difference between Nicole's narrative and natural speech, the participants' final views o f Sentence Shaper, the Q C L and C E T I results, and the participants' ideas for future use o f Sentence Shaper.  118  4.2.4.1 Narrative versus conversational  speech.  A s discussed in the component one results, Nicole demonstrated structural and morphological improvements in her unaided narratives following use o f Sentence Shaper. This result was somewhat unexpected as similar morphosyntactic improvements were not readily apparent in her conversational speech. This trend was demonstrated well in a conversation with Nicole towards the end o f the study compared to her unaided narrative of the M e x i c o message. In excerpt 4.10, Nicole's natural conversational speech is presented. Excerpt 4.10 1. 2.  N:  Yeah yeah I mean um j a um um Erin and me uh spend time spend time okay talk talk yeah yeah (CI 6: 1-2)  In this sample, Nicole is responding to me thanking her for working so hard throughout the study. A s it illustrates, Nicole's conversational speech is much more morphosyntactically impoverished than her speech in the M e x i c o narrative. In this excerpt, which was quite typical o f her conversational speech, there is limited use o f closed class words and her utterance does not constitute a full sentence. In contrast, her narrative speech contained sentences with inflected verbs and closed class words. However, it is evident in this excerpt that Nicole was generating her own thoughts and attempting to express them, which is noteworthy given her previously described difficulty in generating ideas.  119  4.2.4.2 Final participant responses to Sentence Shaper. The final interview with each participant focused on her views o f Sentence Shaper with regard to the benefits o f and barriers to its use. To avoid leading questions as much as possible, in interviewing Nicole, I began with a very open-ended question: " T e l l me what it's like to use Sentence Shaper" (13: 3), encouraging Nicole to tell me anything she could. One response which she mentioned several times throughout the interview was "talk more" (13: 13). I interpreted Nicole's utterance as meaning that she spends more time talking since she began using Sentence Shaper. I asked Nicole i f she liked the word finder and side buttons. She said that she did and also mentioned "other words and and like this ... write wrote and um go and went no yeah yeah yeah" (13: 82-83). B y saying "write wrote", Nicole made reference to the present and past tense verbs category which we had added to the word finder. Nicole and I also discussed the use o f Sentence Shaper with e-mail, and this was also something Nicole had liked about the program: "E-mail yeah I mean uh uh uh Michelle uh yeah yeah me uh me and um Michelle uh uh me mom uh uh Michelle like like yeah" (13: 123-124). In this context, Nicole's mention o f Michelle, to whom she had sent several Sentence Shaper e-mails, suggests that being able to e-mail her friend was meaningful for Nicole. In the interview with Eileen, she revealed that "Sentence Shaper is helping N i c o l e " in that she "is able to repeat her messages unaided" (FN06/29: 111-114). Again, this demonstrates Eileen's view that Sentence Shaper should be used for the purpose o f improving Nicole's unaided speech. When asked what she thought the benefits o f Sentence Shaper were, Eileen responded by saying that she likes that Nicole is not reading or writing, but thinking about what is in her mind and saying it, suggesting the  120  benefits o f the program as a tool to aid in message generation. Eileen was also asked i f there was anything that she did not like about the program. She stated that Nicole became frustrated by the word finder and did not know what words are found i n the different categories, which further illustrates the effort and frustration Nicole experienced while making messages with Sentence Shaper. Eileen also noticed that Nicole was missing verbs and prepositions in her recordings and that she was "unable to connect the sentence" (FN06/29: 128-130). When I asked Eileen i f she thought there had been a noticeable change in Nicole's ability to communicate since using Sentence Shaper, she responded by saying that Nicole could communicate better about conversational topics used in Sentence Shaper messages, but that her communication was unchanged for topics which had not been practiced with Sentence Shaper.  4.2.4.3 The participants'  ideas for future use of Sentence Shaper.  In the final interview, each participant was asked how she saw Sentence Shaper being used in the future. In the interview with Eileen, she responded by saying that she would like Nicole to continue using Sentence Shaper so that she could make messages which she would later be able to repeat without Sentence Shaper. This idea relates to Eileen's perspective throughout the study that Nicole's goal should be communicating without the aid o f an A A C device. Eileen also highlighted the use o f Sentence Shaper as a language remediation tool by suggesting different ideas for messages that Nicole could make (e.g., about her volunteers and daily activities). It was clear that the purpose o f these messages was not for use in conversation, but rather as practice for Nicole so that she would be able to say them unaided. I also asked Eileen about Nicole's use o f  121  Sentence Shaper with e-mail in the future. She responded by saying that there were not many people whom Nicole could e-mail, but began to think o f possibilities. Interestingly, the people she mentioned were all volunteers o f Nicole's. Her emphasis on volunteers as the only possible people whom Nicole could e-mail suggests that Eileen continued to regard the function o f Sentence Shaper within the context o f language therapy. However, it is also possible that she recognized the social contact that the volunteers provided for Nicole and considered the Sentence Shaper messages as a way for Nicole to remain in contact with the volunteers. Nicole responded quite differently to the question about how she would use Sentence Shaper in the future. Her answer to the question is presented i n the following excerpt: Excerpt 4.11 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143.  E:  N: E: N: E: N: E: N: E: N: E: N:  So Nicole do you want to use continue using Sentence Shaper .. like in July and August September. D o you wanna continue using it? U h yes = yes yes yes = = Keep using it = Yes A n d what do you wanna use it for? L i k e what are you gonna do? U m um um I mean um um uh e-mail E-mail? Yeah e-mail um and then uh uh um um um .. e-mail yeah Y o u wanna use it for e-mail? Yeah = yeah = = yeah = Yeah (13: 130-143)  The answer o f e-mail was somewhat unexpected from Nicole. Although use o f Sentence Shaper messages with e-mail had been quite successful from my perspective, Nicole often had difficulty thinking o f people to e-mail. From this excerpt, it appears as  122  though Nicole could envision herself using Sentence Shaper in the future for e-mail despite this difficulty. Nicole also stated that she planned on working with her mother in the future when making messages with Sentence Shaper: Excerpt 4.12 149. 150. 151. 152. 153. 154. 155. 156.  N:  E: E: N: E: N:  U m okay mom and me uh uh uh mom and me uh uh uh I mean um mom and me uh uh shape okay uh Sentence Shaper um talk about talk = talk talk = yeah yeah yeah =um hmm = U m hmm Yeah Okay .. so you like to use it with your mom Yes (13: 149-156)  It is not surprising that Nicole suggested continuing to work with her mother or volunteers on Sentence Shaper messages, as this was her preference throughout the study.  4.2.4.4 QCL and CETI outcomes. The Q C L was completed by Nicole and the C E T I by Eileen both at the beginning and at the end o f the study. Scores on the questionnaires are presented in Appendix H . The participants were not given their original questionnaires to view before completing the second questionnaire. This allowed for an absolute rating on the questionnaires, as it is unlikely that the participants remembered their original scores. These results were then compared to those obtained as part o f the initial communication profile. Some interesting changes were noted on both participants' questionnaire answers. In order to explore these changes further, each participant was shown her pre- and post-study questionnaires and asked i f they felt the changes were accurate. If they did not consider the answers to be an accurate reflection o f change, the participants were allowed to alter their answers. For the  123  Q C L , Nicole was only asked about items which differed by two or more points from the first to the final questionnaire on the 16 items and also the final question about overall quality o f life. O n the final Q C L , Nicole had several items which she rated much higher than she had previously. N o items decreased i n score. It was interesting to note that on the first Q C L , no items had received a score o f five, the highest rating possible, but on the final Q C L , Nicole had given 7 out o f 17 items a score o f five. Prior to completing the questionnaire, respondents are asked i f it is an especially good, average, or especially bad day, and on both questionnaires, Nicole had indicated that she was having an average day. Therefore, it is unlikely that the increase in high scores on the final Q C L was a reflection o f a more positive mood. The first item which received a two point increase was "It's easy for me to communicate". Initially, Nicole had rated this statement as a two, and on the final Q C L , it was given a four. When I asked her i f she felt that it was easier now, she said "yes" (FN06/29: 44-44). "People include me in conversations" also showed a two point increase from one to three. When asked about the accuracy o f this change, Nicole said "conversations conversations friends friends" and then said "Michelle talk more" (FN06/29: 48-49). She also mentioned the lunch that M i r a and I had with her. From her answers it seemed as though there were more social opportunities for her with people who knew how to include her in conversations. The largest increase occurred for the item "I use the telephone" which had a three point increase from two to five. Nicole reported that this change was accurate stating that she calls Eric and her mother; however, she was not able to describe how this had changed in the past four months. The item "I get out o f the house and do things" changed from a score o f three to five. This  124  improvement was understandable as she had recently returned from her vacation in Mexico. For the item "I have household responsibilities", Nicole had initially rated the item as a one and in the final questionnaire scored it as a three. I asked Nicole i f she thought she had more responsibilities now, but she was unable to provide a clear description o f the differences and decided to lower the final score by one. Finally, I asked Nicole about the increase from three to four in her response to the item "In general, my quality o f life is good". When I asked her i f she was feeling better about life, she said "yes.. .speak more" (FN06/29: 64-65). Overall, Nicole reported that the changes she had indicated were accurate aside from the one question regarding household responsibilities. Clearly, Nicole's answers on the final questionnaire demonstrate an increase i n her perceived quality o f life and communication ability. Eileen's answers on the final C E T I also indicated several areas o f perceived improvement. A s with Nicole, Eileen was asked to verify the accuracy o f her responses and explain i f possible what changes had occurred in Nicole's communicative ability to warrant increases in score. Again, she was allowed to change the score i f she thought it was inaccurate. One o f the largest increases in score occurred for the item "Having a oneto-one conversation with you". Eileen had originally scored this item as 4.5/10 (measured on a visual analog scale o f 10cm) and on the final questionnaire scored it at 8.3. When questioned about this change, Eileen said that Nicole had improved in that she was able to say more things when having a conversation; however, initially she was not able to be any more specific about what these things were. A s we talked further, she said that Nicole and she do not really talk about anything and stated that their conversations focus more on Nicole's needs and wants, adding that it is "not really conversations" (FN06/29:  125  95). A t this point, she decided to change her answer and lowered the score on that particular item to 6.7 which, though somewhat decreased, still showed an improvement from the initial score. When asked about the item "Having a spontaneous conversation", Eileen actually changed her final score to indicate greater improvement increasing her rating by 1.3 points, commenting that Nicole is "more spontaneous now" (FN06/29: 100101). Another item which showed a large improvement was "Understanding writing". Initially, Eileen had scored this at 4.4 and in the final questionnaire she gave it a score o f 7.0; she decided to change the final score to 8.6 to indicate even greater improvement. O n the final item which asks about the "overall language and communication skill during the previous week", Eileen had indicated an improvement from the initial to final questionnaires o f 2.5 points. When questioned i f this difference was accurate, she did not change her answer but was unable to describe what changes had occurred to warrant such an improvement in score. Overall, the scores on both questionnaires indicate some positive changes at the end o f the four month period. These findings w i l l be interpreted i n the context o f overall results in the discussion.  126  Chapter Five: Discussion  5.1 Introduction This study involved two components. The first component sought to examine the impact o f an extended use o f Sentence Shaper on the participant's morphosyntax, informativeness, efficiency, and story structure. With regard to the I C F which has been used as a framework for this study, a difference in the structural and morphological characteristics o f the participant's language reflects a change at the impairment and activity levels o f functioning. In component two, a broader view o f Sentence Shaper and the program's potential uses was taken compared to previous studies, allowing for the participation opportunities created from use o f Sentence Shaper to be explored as well as the contextual factors relating to the participant's use o f Sentence Shaper. The case study of Nicole and Eileen provides important findings for the use o f Sentence Shaper both as a language therapy tool and as an A A C device. Furthermore, findings from this study contribute important insights on treatment informed by a social model o f care within the framework o f the ICF. A s described in Chapter One, input from both the clinician and client is regarded as equally important in a social model o f healthcare. The qualitative methodology used in this study was particularly useful in gaining the participants' input and understanding their perspectives. The results presented i n the previous chapter w i l l now be discussed in terms o f the findings gained from the perspective o f the researcher (i.e., an outsider's perspective) versus those learned from the perspectives o f the participants (i.e., an insider's perspective) and w i l l be related to literature in this area.  127  Following this a discussion o f the limitations o f the study, implications for clinical practice, and directions for future research w i l l be provided.  5.2 The Researcher's  Perspective  5.2.1 Comparison of Nicole's Performance to Linebarger et al. 's (2004) Participants In replicating Linebarger et al.'s (2004) treatment study, it is important to discuss similarities and differences between the participants at the onset o f the study as well as following use o f Sentence Shaper. One important caveat in comparing the participants from the two studies is that different elicitation techniques were used i n obtaining the narrative samples. Linebarger et al. presented participants with a short silent video which the participants were then asked to retell. In contrast, Nicole was asked to tell a fairytale and narrate a short picture sequence story. It is interesting to note that Nicole's narrative ability in The Cowboy Story versus The Three Little Pigs was not equal. The majority o f scores in both the pre and post Cowboy Story narratives were superior to those o f The Three Little Pigs narratives. This trend can be explained by examining the task demands each narrative task presents. The picture sequence presented for The Cowboy Story created a situation which had fewer task demands than the telling o f The Three Little Pigs, which was not supported by pictures. A s Nicole had difficulty with story structure and generating ideas, it is probable that support offered by pictures allowed her to allocate more resources to sentence production, resulting in the production o f more complex language. Given the apparent effect o f stimulus on production, it is possible that she may have performed differently with the task used i n Linebarger et al. (2004). Therefore, it is with caution that the participants' narratives are compared.  128  A s previously described, a comparison o f Nicole's performance on the initial assessments to those completed by the participants i n Linebarger et al.'s study indicates that Nicole's scores were similar and in some cases better than the other participants. However, it is interesting to note that Nicole's initial narrative samples were much more impoverished than those o f the other participants. Except for the proportion o f inflected verbs in Nicole's initial Cowboy Story narrative, all other Q P A measures and the C I U results o f the initial narratives are well below those reported for the other participants. A s her other assessment scores were similar to Linebarger et al.'s participants, one might expect her narrative samples to follow this trend. Clearly, something other than her ability to retrieve and understand single words was impairing her ability to produce narratives. Before looking at what this might be, it is important to look further at the differences between Nicole and Linebarger et al.'s participants. Following use o f Sentence Shaper, the results o f the Q P A indicated that Nicole's narratives had impressive gains in morphosyntax. In comparison, the participants in Linebarger et al.'s study also demonstrated improvements in their narratives following use o f Sentence Shaper; however, unlike Nicole, not all morphosyntactic measures increased. S i ' s narratives were reported to show only minimal improvement compared to S2's. Linebarger et al. explained that S i ' s lexical impairment limited his ability to produce narratives unaided, but when given the lexical support available on Sentence Shaper (i.e., the side buttons and word finder) SI was found to be quite successful in producing narratives. S2 had noticeable gains in the structural measures o f the Q P A , but the majority o f morphological Q P A measures remained stable. In contrast, both the structural and morphological Q P A measures o f Nicole's narratives showed large  129  increases. A t this point, it is important to identify differences between the participants which could provide an explanation for their different morphosyntactic outcomes following use o f Sentence Shaper. It is possible that the difference in response could be based on whether the participants were agrammatic or nonfluent, nonagrammatic. S i ' s speech was reported to have been agrammatic, whereas S2's speech was determined to be nonfluent, nonagrammatic. Following the initial assessments with Nicole, her speech was found to be agrammatic. Because both Nicole (agrammatic) and S2 (nonfluent, nonagrammatic) demonstrated improvements, it appears as though the difference between agrammatic and nonfluent, nonagrammatic speech w i l l not explain the varying responses to Sentence Shaper. It is interesting to note that Nicole's post-treatment unaided narratives and aided narratives were characteristic o f a nonfluent, nonagrammatic speaker in that she was able to include both bound and free-standing morphemes in her speech. This change demonstrates a continuity between the categories of agrammatism and nonfluent, nonagrammatism and provides evidence that nonfluent, nonagrammatism may in fact be a milder form o f agrammatism, as put forth by Saffran et al. (1989). Furthermore, i f there is continuity between these two apparent groups, it suggests that distinct categories are not warranted. Although the structure and morphology o f Nicole's speech improved in her narratives, an additional factor was apparently preventing a similar improvement in the informativeness and efficiency o f her narratives. Again, comparison with the participants in Linebarger et al.'s study can offer some insight into this problem. In their study, S2 demonstrated impressive increases in informativeness and efficiency, whereas S I , like Nicole, did not. Linebarger and her colleagues posited that this lack o f improvement was  130  due to S i ' s significant lexical impairment. However, the initial lexical assessments indicated that Nicole had superior scores to the other participants i n lexical retrieval, which suggests that this was not the source o f her impairment. Clearly, other components of the study must be examined to shed light on these findings.  5.2.2 Possible Explanations for these Differences Data collected i n the ethnographic component o f the study provided insight into the differences noted between Nicole's and Linebarger et al.'s (2004) subjects' narrative performances. A s discussed in the results chapter, Nicole often had difficulty generating ideas o f what to talk about in her Sentence Shaper messages and expanding her ideas beyond a one-word utterance. These observations suggest that in addition to impairments at the functional and positional levels in Garrett's model o f speech production, Nicole had a deficit in the conceptual level defined by Fromkin & Bernstein (1998) as the level responsible for message generation. Impairment at this level could explain the difficulty Nicole had both in generating novel sentences and in producing narratives, which would not be captured in the initial assessment scores. Marshall and Cairns (2005) discuss the possibility that the conceptual level, which they describe as "thinking for speaking" (p. 1009) may be impaired i n aphasia. Drawing on Levelt's theory o f speech production, Marshall and Cairns suggest that messages produced at the conceptual level have both propositional structure and perspective before being mapped onto language. Therefore, i f this level o f speech production can be impaired in aphasia, some individuals with aphasia "may be unable to build grammatically principled, constrained schematizations o f events that can be  131  mapped onto their language" (p. 1011). Marshall and Cairns suggest that evidence for this locus o f impairment can be found from nonverbal event judgment tasks and from the way in which some individuals with aphasia verbally describe events. For instance, when asked to name the entities taking part in an event, some individuals with aphasia w i l l name both important and minor entities. In contrast, non-impaired individuals typically name only key elements (i.e., the agent, theme, and instrument). Furthermore, the order in which entities are named can be different. Non-impaired individuals generally name the entities as they would occur in a sentence, whereas, some individuals with aphasia demonstrate a random order to their naming. Marshall and Cairns explored this theory with a treatment experiment aimed at targeting thinking for speaking. After therapy which focused on highlighting the roles o f people and objects i n events, their participant was reported to have improvements i n production o f verbs and word order i n picture descriptions, although similar improvements did not occur in narrative production. The pattern o f impairment that they describe parallels Nicole's initial assessment scores and narrative performance. While she was able to access verbs when presented with a picture of the action, her ability to produce verbs and sentences in her initial narratives was grossly impaired. Marshall and Cairns propose that i f thinking for speaking is impaired, expressing multiple events or perspectives, as is required in a narrative, w i l l be much more difficult for the individual with aphasia than expressing single events (i.e., items found on the Noun/Verb naming test). Further evidence for a deficit at the conceptual level is Nicole's performance on her final unaided narratives versus her narrative performance aided by Sentence Shaper. Although the C I U score on the aided narratives was higher than that for the unaided  132  narratives, the aided narratives did not actually contain more elements o f the story. In fact, the unaided version o f The Three Little Pigs contained more aspects o f the story than the aided version. This comparison illustrates that while Sentence Shaper may have supported more complex language production, it did not facilitate improved story structure for Nicole. It is possible that when using Sentence Shaper, Nicole increased her focus on morphological structure which perhaps decreased the processing resources available to the production of story structure. However, there is no evidence to support this claim as there were no instances in which Nicole demonstrated good story structure. Alternatively, it is possible that Nicole's "thinking for speaking" was itself impaired and the processing support provided by Sentence Shaper was not designed to facilitate this, therefore, organizing events and expressing them still posed a difficulty for Nicole. O f course, personal and environmental factors could also be a source o f Nicole's difficulty with narrative production. Nicole was a dependent communicator who spent a great deal o f time on language therapy focused on written rather than spoken language. Furthermore, her social environment was one i n which she was often not expected to generate narratives and also involved individuals who were willing to speak for her when she was unable to do so herself. It is possible that after four years o f dependent communication and a focus on reading and writing she had lost some o f her ability to produce narratives. Croteau and Le Dorze (2006) discuss the idea o f overprotection and "speaking for" behaviour and the effects this behaviour can have on conversations with people with aphasia and their spouses. In the relationships o f people with aphasia and their spouses, some spouses have been found to often be overprotective. Croteau and L e Dorze found  133  that the level o f the spouses' self-rating o f overprotection was related to their tendency to speak for their spouse with aphasia. Furthermore, the presence o f "speaking for" behaviour was found to limit the participation o f the people with aphasia i n conversation. Croteau and Le Dorze caution that i f overprotection and "speaking for" behaviours by the spouse continue, language rehabilitation could be negatively affected for the individual with aphasia. While this study involved individuals with aphasia and their spouses, it is possible that these results could also apply to conversations between Nicole and her mother. Eileen was observed often to speak for her daughter in conversations. This behaviour may have limited Nicole's participation in conversation and decreased the number o f opportunities she had in developing the skills she practiced in speech and language therapy and applying them in real conversation.  5.2.3 Implications for Processing versus Linguistic Approaches to Intervention of Nonfluent  Aphasia  Linebarger and her colleagues (2004) propose that Sentence Shaper's processing support helps to alleviate the impact of "slowed retrieval and/or rapid decay o f linguistic information" (p. 269). In Linebarger et al.'s (2000) study, they demonstrated that the processing support provided by Sentence Shaper without the lexical support o f the word finder and side buttons allowed for greater structural and morphological complexity in the participants' aided narratives. Although Linebarger et al.'s (2004) study included the use o f lexical support in the form o f the word finder and side buttons, they claim that the significant structural improvements in one participant's aided productions is also evidence that processing support can improve the morphosyntactic ability o f individuals  134  with nonfluent aphasia. This current study was based on the methodology used in Linebarger et al. (2004). The structural and morphological improvements Nicole demonstrated in her aided narratives add further support to the benefits o f processing approaches o f intervention. Nicole's performance also provides an interesting example o f K o l k ' s (1998) activation theory. A s previously described, Nicole often repeated and expanded her utterances until she was satisfied with the sentence, which she then recorded using Sentence Shaper. In this way, Nicole was seen to use Sentence Shaper more as a prompt than an actual workspace for creating sentences. This strategy o f restarts was also evident in Nicole's final narratives. K o l k suggests that these restarts allow the speaker to benefit from activation remaining from previous attempts at the utterance, thereby reducing the time needed to reach the activation threshold. In reflecting on the way i n which Nicole used Sentence Shaper, the use o f fillers and false starts in her speech may have been reinforced by her strategy o f repeating and expanding her utterances, as this often included the repetition o f many filler words and false starts. O n the other hand, Nicole's practice o f saying full sentences as opposed to chunking the utterance into separate parts may have contributed to the production o f morphosyntax in her unaided narratives; however, this was at the cost o f increased production o f filler words and false starts which detracted from the overall informativeness o f the narrative. The listener perception task indicated that although listeners were sensitive to changes in morphosyntax, improvement in morphosyntax did not necessarily contribute to their perceptions o f a better story. For example, Nicole's Cowboy Story post-treatment narrative contained more complex morphosyntax which  135  most listeners commented on; however, many listeners did not use this factor i n their judgment about which story was better. For listeners, it appeared to be the content o f the stories which impacted their ratings. However, it was evident that despite the processing support provided by Sentence Shaper, Nicole continued to have difficulties with certain aspects o f language. Nicole's initial assessment o f locative relations demonstrated that this was an area o f weakness for her. Many o f her answers had the arguments in reverse order. For example, when presented with a picture o f a chicken in an egg, Nicole chose the sentence "the egg is in a chicken". This ability did not appear to improve over the course o f the study with the use of Sentence Shaper. Nicole often seemed to realize that her sentence was missing a preposition, but had difficulty choosing the correct one from the side buttons. Furthermore, Nicole demonstrated a deficit with verb-argument structure. This was particularly evident in her Cowboy Story narratives. The verb switch was often used incorrectly by Nicole in both the unaided and aided narratives. It was apparent that she had difficulty with the thematic role mapping for this rather complex verb. Interestingly, these deficits have a commonality in that they all involve establishing relations. Whether these deficits are best explained by reduced processing resources or specific linguistic impairments remains unclear; however, it is evident that the processing support provided by Sentence Shaper did not alleviate these difficulties. Even though Sentence Shaper provides processing support, the program does not entirely eliminate the demands o f putting a sentence together. Memory demands placed on the user when organizing the icons and the user must still establish the relations in the sentence, whether it is verbargument or prepositional relations. The question remains then whether direct  136  intervention targeting these areas o f deficit would have beneficial effects on Nicole's language ability.  5.2.4 Generalization to Conversational Speech Past research has shown that the speech characteristics o f individuals with nonfluent aphasia can vary depending on the task demands (Heeschen & Schegloff, 2003). A s discussed in Chapter One, the speech samples obtained from picture description or fairytale retelling are often quite different from those obtained from conversations. Therefore, the finding that the improvements Nicole demonstrated in her final unaided narratives did not appear to generalize to her natural, conversational speech is not surprising. After a series o f experiments, Heeschen and Schegloff reported that the use o f telegraphic speech by people with agrammatism is "not an all-or-none phenomenon" (p. 242). They found that use o f telegraphic speech, or non-finite utterances, occurred in normal interactions, whereas in non-interactional storytelling contexts, there was an increase in the use o f finite verbs (For further discussion, see K o l k , 1995, 2001). Even when task demands were increased in the storytelling condition, telegraphic speech was not used. From this, Heeschen and Schegloff concluded that the use o f telegraphic speech is an adaptive strategy which is "interactionally motivated" (p. 249). Nicole's superior morphosyntactic performance in the narratives versus that in conversation provides support for this finding. Perkins (2003) drawing on Clark and Schaefer's model o f conversations, provides another explanation that is relevant to the findings in this study. This model o f conversation suggests that there are two components within a conversation: the  137  presentation stage and the acceptance stage. The presentation phase occurs when an utterance is presented by one interlocutor. The acceptance phase is then initiated by the second interlocutor and establishes that the utterance has been sufficiently understood. In this way, conversations are collaborative, and both interlocutors are responsible for the message transmission. Another important principle o f this model is that interlocutors work to minimize the effort spent i n either phase o f the conversation. Thus, in conversations with individuals with aphasia, the dyad's effort is minimized i f the unimpaired partner takes a greater responsibility for message generation. This pattern was evident in the conversations with Nicole. Less effort was expended i f Nicole's utterances were expanded by the conversation partner than i f she was required to produce the entire message herself. In contrast, the narrative condition was non-collaborative, and the responsibility for the content o f the message rested solely with Nicole. In this context, her language was more structured and grammatically complex, although, this situation was much more effortful for her. Clearly, the context was an important determiner o f Nicole's language characteristics.  5.2.5 The Integration of Sentence Shaper into Conversation: Lessons from Nicole It is difficult to comment on precisely how Sentence Shaper was used in conversation as there was only one instance o f spontaneous use o f the program i n this way. The conversation in which Sentence Shaper was used spontaneously occurred during a situation which I had created to explore use o f the program in conversation. In this situation, Nicole had already played one message for M i r a and following this spontaneously chose to play another message. In this way, Nicole was seen to initiate a  138  new topic o f conversation by playing the Sentence Shaper message. However, it is not clear how Sentence Shaper could be successfully integrated into ongoing conversation. Higginbotham & Wilkins (1999) state that use o f A A C in conversation is at risk o f not following the "temporal imperative" (p. 51) or, in other words, causing a disruption in the temporal flow o f conversation. In the case o f Sentence Shaper, the time it takes to turn on the computer and choose a message would violate the temporal imperative. Nicole's use o f Sentence Shaper messages to initiate conversation was quite successful as evidenced by the conversation with M i r a and the sharing o f the M e x i c o message at the stroke group. A s previously described, while a message was playing, Nicole talked along with the recording. This strategy had the effect o f making the off-line Sentence Shaper message on-line. The participant in Waller and N e w e l l ' s (1997) study on P R O S E was also found to use this strategy. " H M reads through the text on the screen while pointing at the words as i f she is 'telling' the story to her communication partner" (p. 295). Clearly, there is motivation for users o f off-line devices to attempt to make the messages on-line in conversation.  5.3 The Participants' 5.3.1 Changes in Communication  Perspectives  Ability  Although the structural and morphological changes o f Nicole's conversational speech did not appear to improve following use o f Sentence Shaper, the questionnaire data and feedback from Nicole and Eileen indicate that some positive change i n Nicole's ability to communicate was noticed by both participants. T w o factors made it difficult to describe more precisely what these changes were. I was unable to provide an objective  139  measure o f change in conversational abilities as my relationship with Nicole changed throughout the study. Therefore, changes I noticed in conversation could not confidently be attributed to improvement in language skill as they may have reflected a change in our relationship. For instance, Nicole's ability to initiate conversation was seen as improving throughout the study; however, this may have be due to the fact that we knew more about each other and she felt more comfortable conversing with me. A second factor which made it difficult to assess change is the fact that conversations are context dependent. There are many factors which contribute to one's performance in a conversation; the communication partner, topic o f conversation, and fatigue are a few examples. Nicole's conversational skills varied depending on these different factors. From Eileen's perspective, her interactions with Nicole were not considered to be real conversations. She stated that the majority o f their interactions focused on Nicole's needs and wants. These types o f interactions are considered as serving a transactional function which is defined as the exchange o f information and content (Brown & Yule, 1983). Apparently, the interactional function o f language was missing from Eileen's conversations with Nicole. This is the language function responsible for establishing and maintaining relationships. B r o w n and Yule report that it is interactional language, rather than transactional language, that forms the majority o f our everyday interactions. It is understandable then, that Eileen should feel that something was different about the conversations she had with her daughter. The important social connection we gain through interactional language was often absent from their conversations. Another interesting aspect o f Nicole and Eileen's conversations was the way in which Eileen often spoke for Nicole. A s described previously, Croteau and L e Dorze  140  (2006) found that this behaviour can decrease the participation i n conversation o f the person with aphasia, and that "speaking for" behaviours had the tendency to occur in relationships where the non-impaired partner perceived him or herself as overprotecting the person with aphasia. In Nicole and Eileen's situation, it was apparent that neither o f them considered this behaviour to be overprotective. Nicole often asked her mother to speak for her and Eileen readily provided this help when requested. Whether or not this adversely affected the generalization o f structural and morphological gains seen in Nicole's narratives to her conversational speech is unknown; however, it is clear that having Eileen speak for her was something that both Nicole and Eileen perceived as valuable.  5.3.2 Acceptance of Sentence Shaper as an AAC Device Obviously, acceptance o f A A C had a considerable impact on the ways in which Sentence Shaper was used in everyday communication by Nicole. Clearly, Nicole and Eileen were reluctant to use Sentence Shaper as an augmentative device. This lack o f acceptance w i l l now be discussed in respect to Lasker and Bedrosian's (2001) A A C Acceptance M o d e l , which outlines three contributing components: milieu, person, and technology. The area o f milieu, which includes the communication partner and environment, was a major factor in the way in which Sentence Shaper was used. Use o f the program as an A A C device clearly did not fit with Eileen's expectations o f Nicole's language recovery, and she did not encourage Nicole to use Sentence Shaper in this way. Furthermore, Nicole's dependence in communicating created an environment which did not facilitate the use o f Sentence Shaper for communication purposes. Other personal  141  characteristics also impacted the acceptance o f Sentence Shaper as A A C . Nicole was very much still focused on language recovery and was not motivated to accept an A A C device at this time. Despite the success Nicole experienced with e-mail and the pride she clearly had in her narratives, the effort and frustration she experienced while making messages with Sentence Shaper further decreased her motivation to use the program. The technology component was a positive factor in this equation. Nicole was a very successful user o f technology and seemed to enjoy this aspect o f Sentence Shaper. Overall, however, the costs o f Sentence Shaper outweighed the benefits, and this in turn led to the lack o f acceptance o f Sentence Shaper as an A A C device. Nonetheless, both Nicole and Eileen requested a copy o f Sentence Shaper at the end o f the study, suggesting that they are "partial/reluctant" users o f A A C , according to Lasker and Bedrosian's model. Past research has shown that the attitudes families have towards A A C can hinder its use. A study done by Lasker (1997, as cited in H u x et al., 2001) demonstrated that despite the fact that families rated the high-tech A A C option as the easiest to understand, family members preferred listening to natural speech. The belief that language recovery would happen with further therapy and effort was seen as the cause o f this preference. This sentiment, although not explicitly stated by Eileen, resounded in many o f my interactions with her. She very much wanted to see her daughter communicating without the aid o f Sentence Shaper. In Waller et al.'s (1998) investigation o f the TalksBac system, one participant was found to be reluctant to use the program and refused to use it with his wife even though it was successful i n augmenting his communication. This situation is comparable to  142  Nicole's use o f Sentence Shaper. Even though she was capable o f using the program to create more informative and complex narratives than she was able to with her unaided speech, she was reluctant to use the program in conversation. It is evident from these examples, that despite an individual with aphasia being capable o f using an A A C device it may not necessarily be accepted by that person.  5.4 An Integration of Perspectives Throughout the study a tension existed between wanting to promote N i c o l e ' s independence and her dependence on others. It is paradoxical that with Nicole and Eileen's focus on Nicole's language therapy and goal o f language recovery that many o f their behaviours seemed to promote Nicole's dependence on her communication partners. This tension is exemplified by Croteau and L e Dorze's (2006) idea o f overprotection and "speaking for" behaviours versus Perkins's (2003) principle o f economy o f effort. Croteau and L e Dorze suggest that overprotection o f individuals with aphasia and speaking for them w i l l lead to reduced participation in conversation and may in fact limit the effects o f therapy. O n the other hand, according to Perkins's principle, effort may be reduced in conversations with people with aphasia i f the conversational partner "speaks for" the individual with aphasia. Clearly there is a tradeoff between minimizing the effort expended in conversation and promoting the independence o f the individual with aphasia in conversation. In the case o f Nicole and Eileen, it appears as though they have chosen to minimize their effort by having Eileen "speak for" Nicole. The question that remains is what effect this behaviour has on Nicole's language recovery.  143  5.5 How Do the Principles of a Social Model Apply to this Study? Results o f the study provide insights into the use o f a social model in the treatment o f aphasia. There are several areas o f the study met the principles o f the social model philosophy outlined by B y n g and Duchan (2005), and there were also many instances in which these principles were not met. The first principle o f the social model philosophy described by B y n g and Duchan (2005) states that the healthcare service users and providers should work together collaboratively rather than having a relationship in which the person with the disability is dependent on the healthcare provider. The importance o f this principle was clearly illustrated throughout the study in the power dynamics o f the relationship between Nicole and me. Initially, I resisted Nicole's preference to work with others on her Sentence Shaper messages. However, I finally realized that as an equal contributor to this study, Nicole had the right to decide how she was going to make her messages. Eventually, I came to respect Nicole's decision to work with others as this clearly made sense in the context o f her life. The second principle proposes that people with disabilities should be involved authentically in all aspects o f the decision making process in regard to their own health and healthcare in general. It is interesting to note that while Nicole agreed to participate in this study, she was never asked whether working on narrative production with a software program was a goal she was interested in pursuing. Furthermore, she was never consulted in regard to whether she was interested in using a high-tech A A C device. If given the choice, Nicole may have never chosen to work on her narrative ability, and this could possibly explain her reluctance to make messages with Sentence Shaper. In  144  addition, it is unlikely that Nicole would have chosen an augmentative communication goal for therapy. It was clear throughout the study that Nicole enjoyed customizing the side buttons and word finder. This situation was one in which Nicole could become authentically involved in the decision making process. A s previously described, the power dynamics shifted i n this situation to an equal relationship in which Nicole and I were both able to suggest ideas o f words to be added or changed. B y n g and Duchan describe the third principle o f the social model philosophy as the need to create therapies which are engaging experiences for people with disabilities. It was evident from Nicole's reluctance to use Sentence Shaper to make messages, that this was not an engaging experience for her. However, sharing the messages as e-mails and in conversation appeared to be something that Nicole valued. Unfortunately, the processing costs o f making the messages outweighed the benefit she received from sharing the messages. This point illustrates that real-life activities, such as sharing stories with others, can be very engaging, and that therapy which is able to achieve this can be very motivating for people with aphasia. In contrast, experiences which do not relate to real-life (i.e., making narratives alone on a computer) may not be engaging or motivating. The final two principles are related to how people with disabilities should be involved in the decision making process in healthcare organizations and that these organizations should be accountable to the people they serve. These principles address issues o f working with people with disabilities at a higher level than the individual client and clinician relationship explored in this study. However, a social model is certainly relevant to researchers and designers o f A A C technology. Clearly, it is useful and necessary to work with people with disabilities when designing devices which w i l l  145  ultimately be used by them. In this way, consumers o f A A C technology have a voice in the products which are developed and the issues they feel still need to be addressed by technology.  5.6 Implications for Clinical  Practice  The results o f this study suggest that practice in narrative production with the processing support provided by Sentence Shaper can lead to increases in the morphosyntactic complexity o f unaided narratives. In another study investigating the effects o f a treatment program focused on story retelling, similar results were found (Peach & Wong, 2004). Following a 10-week program involving narrative production, the participant demonstrated improved expressive syntax. However, it was apparent in Nicole's case that the improved morphosyntactic skills she demonstrated in her narratives did not generalize to her conversational speech. Clearly, clients w i l l need to be supported in developing these new skills in a conversational context. In respect to the use o f Sentence Shaper as an A A C device, this study highlights the importance o f considering contextual factors when recommending an A A C device to a client. Personal factors such as attitudes, expectations, and communication style, and the environment in which the device w i l l be used w i l l contribute greatly to the acceptance of an A A C device. Attention should also be given to the communication partners, as they can also influence the ways in which an A A C device w i l l be used. Furthermore, the study illustrates important aspects o f the social model and how this model can be applied to treatment for people with aphasia. A s previously described, Nicole was never asked whether she was interested in working on narratives, and this  146  may have explained her lack o f motivation to work with Sentence Shaper. Therapy is costly and often requires many hours of work for both the speech-language pathologist and person with aphasia. If therapy goals are not viewed as meaningful and valuable to the client, it is likely the client w i l l not be a motivated participant in therapy.  5.7 Limitations of the Study A substantial limitation to this study was that a stable baseline was not established prior to the beginning o f the study. Therefore, changes in language measures and questionnaire scores can not with confidence be attributed solely to use o f Sentence Shaper. The fact that Nicole was working on language therapy worksheets in addition to the use o f Sentence Shaper was also a limiting factor in this study. It is unclear what impact these worksheets had on Nicole's language ability and whether this type o f language therapy had any effect on Nicole's narrative ability. In regard to the replication o f Linebarger et al.'s (2004) study, an additional limitation is the fact that the narrative elicitation task differed from the tasks used in this study, limiting the comparisons which can be made across the participant' narratives.  5.8 Directions for Future Research Results from this study suggest that Sentence Shaper has the potential to augment communication in social contexts; however, the participants' personal and environmental factors limited the use o f the program in this way. Further studies with a greater number of participants with aphasia would be beneficial to explore how Sentence Shaper could be used as an A A C device in different contexts. This would also allow for further  147  investigation o f the impact o f Sentence Shaper on conversational speech. Structural and morphological characteristics of Nicole's language in conversation did not appear to improve with use o f Sentence Shaper. A s this was not measured in Linebarger et al.'s (2000, 2001, 2004) studies, it is not known whether Sentence Shaper had a significant impact on those participants' conversational speech. Clearly confirming whether use o f Sentence Shaper can improve conversational speech is a necessary step in determining its potential as a language therapy tool. Finally, as suggested by Linebarger and colleagues (2001), Sentence Shaper could be used in various ways depending on the participant's language impairment. The way in which users o f Sentence Shaper are able to construct utterances may shed light onto different underlying impairments and could add further to the literature on sentence processing.  5.9 Conclusion This study set out to investigate the effects o f extended use o f Sentence Shaper on language structure and explore the ways in which the program could be used in everyday life to augment communication. The four-month case study with Nicole and Eileen revealed several interesting findings. Following use o f Sentence Shaper, Nicole's unaided narratives were found to be more structurally and morphologically complex, however, a similar improvement was not demonstrated in the informativeness o f these narratives. While use o f Sentence Shaper increased the C I U measures due to the lack o f filler words, it did not facilitate the production o f a more complex story structure i n the final aided narratives.  148  This study highlights that benefits can occur from working on narratives in a therapy setting; however, in Nicole's case, the language improvements seen i n her narratives did not appear to generalize to her conversational speech. In the second component, Sentence Shaper messages were used by Nicole with e-mail and in conversation. From my perspective, these uses were quite successful; however, Nicole and Eileen were reluctant to accept Sentence Shaper as a device to augment communication in conversation. The I C F provided a useful framework for analyzing the impact o f Sentence Shaper on Nicole's communication. The qualitative measures conducted in this study facilitated the use o f a social model o f intervention. Within this model, people with disabilities are seen as experts in their own conditions and their input into healthcare services is viewed as a crucial component. Gaining the perspectives o f the participants in this study provided an important source o f data. Use o f the I C F framework and social model principles allowed for a deeper investigation o f the factors affecting the treatment impact o f Sentence Shaper and its use in everyday communication. This type o f investigation was useful in providing answers, but also in generating additional questions which have the potential to be explored further in research and in clinical settings.  149  References Beukelman, D . R . , & Mirenda, P. (1998). Augmentative and alternative Management of severe communication  communication:  disorders in children and adults (2  nd  Ed.).  Baltimore: Paul H . Brookes Publishing C o . Bird, H . , & Franklin, S. (1996). Cinderella revisited: A comparison o f fluent and nonfluent aphasic speech. Journal ofNeurolinguistics,  9, 187-206.  Brown, G . , & Y u l e , G . (1983). Discourse Analysis, Eds. Comrie, B . , Fillmore, J., Lass, R., Lightfood, D . , Lyons, J., Matthews, P . H . , Posner, R., Romaine, S., Smith, N . V . , & Vincent, N . Cambridge: Cambridge University Press. Byng, S., Nickels, L . , & Black, M . (1994). Replicating therapy for mapping deficits in agrammatism: Remapping the deficit? Aphasiology, 8 (4), 315-341. Byng, S., & Duchan, J.F. (2005). Social model philosophies and principles: Their applications to therapies for aphasia. Aphasiology,  19, 906-922.  Chapey, R., & Hallowed, B . (2001). Introduction to language intervention strategies in adult aphasia. In R. Chapey (Ed.), Language Intervention Strategies in Aphasia (4 Ed.) (pp. 3-17). Baltimore: Lippincott, Williams, & Wilkins. th  Code, C . (2003). The quantity o f life for people with chronic aphasia. Neuropsychological  Rehabilitation,  13, 379-390.  Creswell, J.W. (1994). Research design: Qualitative & Quantitative  Approaches.  Thousand Oaks, C A : S A G E Publications Inc. Croteau, C , & L e Dorze, G . (2005). Overprotection, "speaking for", and conversational participation: A study o f couples with aphasia. Aphasiology,  150  20, 327-336.  Davies, R., Marcella, S., McGrenere, J., & Purves, B . (2004). The ethnographically informed participatory design o f a P D A application to support communication, Proceedings  of ACM ASSETS 2004, 153-160.  Doyle, P.J., Goldstein, H . , & Bourgeois, M . S . (1987). Experimental analysis o f syntax training in Broca's aphasia: A generalization and social validation study. Journal of Speech and Hearing Disorders, 52, 143-155. Friedmann, N . , & Grodzinsky, Y . (1997). Tense and agreement in agrammatic production: Pruning the syntactic tree. Brain and Language,  56, 397-425.  Fromkin, V . A . , & Bernstein Ratner, N . (1998). Speech production. In J. Berko Gleason, & N . Bernstein Ratner (Eds.), Psycholinguistics  (2  nd  Ed.) (pp. 310-346). Belmont,  C A : Wads worth/Thomson Learning. Garrett, K . , & Beukelman, D . (1998). Adults with severe aphasia. In D . Beukelman, & P. Mirenda (Eds). Augmentative and alternative communication: severe communication  disorders in children and adults (2  nd  management of  Ed.) (pp. 465-479).  Baltimore: Paul H . Brookes Publishing C o . Gleason, J.B., Goodglass, H . , Green, E . , Ackerman, N . , & Hyde, M . (1975). The retrieval of syntax in Broca's Aphasia. Brain and Language, 2, 451-471. Goodglass, H . , Kaplan, E . , & Barresi, B . (2001). The nature o f the deficits. In: The assessment of aphasia and related disorders (3  rd  edition) (pp. 5-13). Philadelphia:  Lippincott Williams & Wilkins. Gumprez, J.J., & Berenz, N . (1993). Transcribing conversational exchanges. In J . A . Edwards & M . D . Lampert (Eds.), Talking data: Transcription  and coding in  discourse research. Hillsdale, N J : Lawrence Erlbaum Associates.  151  Hagiwara, H . (1995). The breakdown o f functional categories and the economy o f deprivation. Brain and Language, 50, 92-116. Heeschen, C , & Schegloff, E . A . (2003). Aphasic agrammatism as interactional artifact and achievement. In C . Goodwin (Ed.), Conversation and brain damage (pp. 231-282). Oxford: Oxford University Press. Helm-Estabrooks, N . , Fitzpatrick, P . M . , & Barresi, B . (1981). Response o f an agrammatic patient to a syntax stimulation program for aphasia. Journal of Speech and Hearing Disorders,, 46, 422-427. Helm-Estabrooks, N . , & Ramsberger, G . (1986). Treatment o f agrammatism in long-term Broca's aphasia. British Journal of Disorders of Communication,  21, 39-45.  Hewitt, A . , & B y n g , S. (2003). From doing to being: from participation to engagement. In S. Parr, J. Duchan, & C . Pound (Eds.), Aphasia inside out (pp. 51-63). Open University Press. Higginbotham, J.D., & Wilkins, D . P . (1999). Slipping through the timestream: Social issues o f time and timing in augmentative interactions. In: D . Kovarsky, J. Dunchan, & M . M a x w e l l , Constructing (in)competence: Disabling evaluations in clinical intervention and society (pp. 49-81). Mahwah, N J : Lawrence Erlbaum Associates. Hinckley, J.J. (2002). Vocational and social outcomes o f adults with chronic aphasia. Journal of Communication  Disorders, 35, 543-560.  Hine, N . , Arnott, J.L., & Smith, D . (2003) Design issues encountered i n the development of a mobile multimedia augmentative communication service, Universal Access Information Society, 2, 255-264.  152  Hux, K . , Manasse, N . , Weiss, A . , & Beukelman, D . R . (2001). Augmentative and alternative communication for persons with aphasia. In R . Chapey (Ed.), Language intervention strategies in aphasia and related communication  neurogenic  disorders, (4 Ed.) (pp. 675-687). Baltimore: Lippincott Williams th  & Wilkins. Jacobs, B . , Drew, R., Ogletree, B . T . , & Pierce, K . (2004). Augmentative and alternative communication ( A A C ) for adults with severe aphasia: Where we stand and how we can go further. Disability and Rehabilitation,  26, 1231-1240.  Joanette, Y . , & Goulet, P. (1990). Narrative discourse in right-brain-damaged righthanders. In Y . Joanette & H . H . Brownell (Eds.), Discourse ability and brain damage: Theoretical and Empirical Perspectives (pp. 131- 153). N e w York, N Y : Springer- Verlag. Jones, E . (1986). Building the foundations for sentence production i n a nonfluent aphasic. British Journal of Disorders of Communication,  21, 63-82.  Kagan, A . (1995). Revealing the competence o f aphasic adults through conversation, a challenge to health professionals. Top Stroke Rehabilitation,  2, 15-28.  Kaplan, E . , Goodglass, H . , & Weintraub, S. (2001). Boston Naming Test (2  nd  Ed.),  Baltimore: Lippincott Williams & Wilkins. K a y , J., Coltheart, M . , & Lesser, R. (1992). PALPA  - Psycholinguistic  Assessment of  Language Processing in Aphasia, Hove, U K : Psychology Press. K o l k , H . (2001). Does agrammatic speech constitute a regression to child language? A three-way comparison between agrammatic, child, and normal ellipsis. Brain and Language, 77, 340-350.  153  K o l k , H . (1998). The malleability o f agrammatic symptoms and its implications for therapy. In E . G . Visch-Brink & R. Bastiaanse (Eds), Linguistic Levels in Aphasiology  (pp. 193-209). San Diego: Singular.  K o l k , H . (1995). A time-based approach to agrammatic production. Brain and  Language,  50, 282-303. Lasker, J.P., & Bedrosian, J.L. (2001). Promoting acceptance o f augmentative and alternative communication by adults with acquired communication disorders. Augmentative and Alternative  Communication,  17, 141-153.  Le Dorze, G . , & Brassard, C . (1995). A description o f the consequences o f aphasia on aphasic persons and their relatives and friends based on the W H O model o f chronic diseases. Aphasiology,  9, 239-255.  Lesser, R. (1990). Aphasia: Theory-based intervention. In M . Leahy (Ed.), Disorders of Communication:  The Science of Intervention.  London, N J : Whurr.  Light, J. (1988). Interaction involving individuals using augmentative and alternative communication systems: State o f the art and future directions. Augmentative and Alternative Communication,  4, 66-82.  Linebarger, M . C . (2005). A guide to installation and use o f the SentenceShaper communication system. Retrieved M a r c h 3, 2006, from www.SentenceShaper.com. Linebarger, M . C , M c C a l l , D . , & Berndt, R . S . (2004). The role o f processing support in the remediation o f aphasic language production disorders, Cognitive Neuropsychology,  21, 267-282.  154  Linebarger, M . C , & Schwartz, M . F. (2005). AAC for hypothesis-testing and treatment of aphasic language production: Lessons from a 'processing prosthesis'. Aphasiology, 19, 930-942.  Linebarger, M . C , Schwartz, M.F., & Kohn, S.E. (2001) Computer-based training of language production: An exploratory study, Neuropsychological  Rehabilitation,  77,57-96. Linebarger, M . C , Schwartz, M.F., Romania, J.R., Kohn, S.E., & Stephens, D.L. (2000). Grammatical encoding in aphasia: Evidence from a "processing prosthesis". Brain and Language,  77,  416-427.  Lomas, J., Pickard, L., Bester, S., Elbard, H., Finlayson, A., & Zohhaib, C. (1989). The Communicative Effectiveness Index: Development and psychometric evaluation of a functional communication measure for adult aphasia. Journal of Speech and Hearing Disorders, 534, 113-124.  Luborsky, M.R. (1994). The identification and analysis of themes and patterns. In O.F. Gubrium & A. Sunkar (Eds.), Qualitative methods in aging research (pp. 189210), Thousand Oaks, CA: Sage. Marshall, J., & Cairns, D. (2005). Therapy for sentence processing problems in aphasia: Working on thinking for speaking. Aphasiology, 19, 1009-1020. Marshall, J., Pring, T., & Chiat, S. (1998). Verb retrieval and sentence production in aphasia. Brain and Language, 63, 159-183. Mollica, B.M. (1999). Emerging technologies in augmentative and alternative communication: Restorative and compensatory approaches to acquired disorders of communication. NeuroRehabilitation,  155  12, 21-11.  Nicholas, L . E . , & Brookshire, R . H . (1993). A system for quantifying the informativeness and efficiency o f the connected speech o f adults with aphasia. Journal of Speech & Hearing Research, 36, 338-351. Ochs, E . (1979). Transcription as theory. In E . Ochs & B . Schieffelin (Eds.), Developmental pragmatics (pp. 185-207), N e w Y o r k , N Y : Academic Press. Paul, D.R., Frattali, C M . , Holland, A . L , Thompson, C . K . , Caperton, C.J., & Slater, S . C (2004). Quality of Communication  Life Scale, Rockville, M D : American Speech-  Language-Hearing Association. Peach, R . K . , & Wong, C M . (2004). Integrating the message level into treatment for agrammatism using story retelling. Aphasiology,  18, 429-441.  Penn, C . (2005). Forum: W h o ' s tired o f the W H O ? A commentary on Ross and Wertz, "Advancing appraisal: Aphasia and the W H O " . Aphasiology,  19, 875-879.  Penn, C , & Clearly, J. (1988). Compensatory strategies in the language o f closed head injury patients. Brain Injury, 2, 3-17. Perkins, L . (2003). Negotiating repair in aphasic conversation: Interactional issues. In C . Goodwin (Ed.), Conversation and brain damage (pp. 147- 162), Oxford: Oxford University Press. Psathas, G . , & Anderson, T. (1990). The "practices" o f transcription in conversation analysis. Semiotica, 78, 75-99. Ross, K . B . , & Wertz, R . T . (2002). Relationships between language-based disability and quality o f life i n chronically aphasic adults. Aphasiology,  16, 791-800.  Ross, K . B . , & Wertz, R.T. (2003). Quality o f life with and without aphasia. 77,355-364.  156  Aphasiology,  Ross, K . B . , & Wertz, R . T . (2005). Forum: Advancing appraisal: Aphasia and the W H O . Aphasiology,  19, 860-870.  Saffran, E . M . , Berndt, R.S., & Schwartz, M . F . (1989). The quantitative analysis o f agrammatic production: Procedure and data. Brain and Language, 37, 440-479.  Schwartz, M . F . , Fink, R . B . , & Saffran, E . M . (1995). The modular treatment o f agrammatism. Neuropsychological  Rehabilitation,  5, 93-127.  Schwartz, M . F . , Linebarger, M . C . , Saffran, E . M . , & Pate, D . S . (1987). Syntactic transparency and sentence interpretation in aphasia. Language and Cognitive Processes, 2, 85-113. Shadden, B . B . (2005). Aphasia as identity theft: Theory and practice. Aphasiology,  19,  211-223. Simmons-Mackie, N . (2001). Social approaches to aphasia intervention. In R . Chapey (Ed.), Language Intervention Strategies in Aphasia (4 Ed.) (pp. 246-266). th  Baltimore: Lippincott, Williams, & Wilkins. Simmons-Mackie, N . N . , & Damico, J.S. (1997). Reformulating the definition o f compensatory strategies i n aphasia. Aphasiology,  11, 761-781.  Spradley, J.P. (1980). Participant Observation. Orlando, F L : Harcourt Brace Jovanovich Inc. Stake, R . E . (1995). The art of case study research. Thousand Oaks, C A : S A G E Publications Inc.  157  Stuart, S., Lasker, J.P., & Beukelman, D . R . (2000). A A C message management. In D . R . Beukelman, K . M . Yorkston & J. Reichle (Eds.) Augmentative and  alternative  Communication for adults with acquired neurologic disorders. Baltimore: Paul H . Brookes Publishing C o . Thompson, C . K . (2001). Treatment o f underlying forms: A linguistic specific approach for sentence production deficits in agrammatic aphasia. In R. Chapey (Ed.), Language Intervention Strategies in Aphasia (4 Ed.) (pp. 605-628). Baltimore: th  Lippincott, Williams, & Wilkins. V a n de Sandt-Koenderman, M . W . M . E . (2004). High-tech A A C and aphasia: Widening horizons? Aphasiology,  18, 245-263.  V a n de Sandt-Koenderman, W . M . E . & Bonta, E . (1998). Agrammatism: Theory and therapy. In E . Visch-Brink & R. Bastiaanse (Eds.) Linguistic Levels in Aphasia (pp. 211-227). San Diego, C A : Singular. Waller, A . , Dennis, F., Brodie, J., & Cairns, A . Y . (1998). Evaluating the use o f TalksBac, a predictive communication device for nonfluent adults with aphasia, International Journal of Language & Communication  Disorders, 33, 45-70.  Waller, A . , & Newell, A . F . (1997). Towards a narrative-based augmentative communication system, European Journal of Disorders of Communication,  32,  289-306. World Health Organization (2001), International  Classification  of  Functioning  Disability and Health, World Health Organization. Zingeser, L . , & Berndt, R . S . (1990). Retrieval o f nouns and verbs in agrammatism. Brain and Language, 39, 13-32.  158  APPENDICES  159  Appendix A : Consent Forms Consent Form for Person with Aphasia Consent Form Investigating Sentence Shaper: Can a processing prosthesis augment communication? Principle Investigator: Erin Albright, U B C Masters Student Faculty Advisor: Barbara Purves, M . S c , S L P ( C ) , Clinical Professor Purpose: This research project is part o f M s . Albright's Masters' thesis. The purpose o f this study is to learn about how the Sentence Shaper, a story-telling tool, can be used i n everyday life by individuals with aphasia, and to find out whether improvements i n language ability can be seen following extensive use o f the Sentence Shaper. Y o u are being asked to take part in this study because you have aphasia. Study Procedures: If you agree to take part i n this study, M s . Albright w i l l be assessing your language ability. After that, she w i l l visit you one to two times per week over approximately three to four months. These visits w i l l take place i n your home or in another place o f your choosing. Each visit w i l l last for one to two hours at a time that is best for you. A t the beginning o f each visit, M s . Albright w i l l make sure that you are still willing to take part in the research. In the first few weeks o f meetings, M s . Albright w i l l be asking questions to you and your communication partner to learn more about communicating with aphasia. During this time, you and your communication partner w i l l work out with M s . Albright when she can visit over the next three to four months. During the fourth and fifth weeks, training o f the Sentence Shaper w i l l commence. Y o u w i l l be instructed on how to use the Sentence Shaper on your home computer. Y o u w i l l be given a log book to record how often you are using the Sentence Shaper, with whom you are using the sound files, and what the purpose is. Following the training period, you w i l l be asked to practice using the Sentence Shaper for approximately eight weeks. M s . Albright w i l l continue to visit you during this time to discuss how the program is working for you and to explore different ways the Sentence Shaper could be used. M s . Albright w i l l also be talking to you and your communication partner to find about what you feel the strengths and weaknesses o f the Sentence Shaper are. For some o f the visits, M s . Albright w i l l ask for permission to record the conversation with either a video camera or a tape-recorder. Everyone who takes part in the conversation w i l l be asked for their permission to be recorded.  160  Consent: I understand that my consent in this study is entirely voluntary and that I may withdraw from the study or refuse to participate at any time. I have received a copy o f this consent for my own records. I consent to participate in this study.  Subject Signature  Date  Consent to Show Videotapes and Sound Files for Scientific and Educational Purposes (to be completed in the course of study) I have had the opportunity to look at videotapes made o f me in the course o f the study and to listen to sound fdes created by me on the Sentence Shaper. I consent to the showing o f these videotapes and sound fdes for scientific and educational purposes. I understand that no additional identifying information w i l l be given with these tapes during such exhibition.  Subject Signature  Date  162  Consent Form for the Communication Partner Consent Form Investigating Sentence Shaper: Can a processing prosthesis augment communication? Principle Investigator: Erin Albright, U B C Masters Student Faculty Advisor: Barbara Purves, M . S c , S L P ( C ) , Clinical Professor Purpose: This research project is part o f M s . Albright's Masters' thesis. The purpose o f this study is to learn about how the Sentence Shaper, a story-telling tool, can be used in everyday life by individuals with aphasia, and to find out whether improvements in language ability can be seen following extensive use o f the Sentence Shaper. Y o u are being asked to take part in this study because you communicate regularly with someone who has aphasia. Study Procedures: If you agree to take part in this study, M s . Albright w i l l visit you multiple times over approximately three to four months. These visits w i l l take place in your home or in another place o f your choosing. Each visit w i l l last for one to two hours at a time that is best for you. During the first visit, M s . Albright w i l l ask you to complete a questionnaire that looks at how effectively you feel you are able to communicate with someone who has aphasia. She w i l l also ask you a few questions about what it is like to converse with the person with aphasia. A t the beginning o f each visit, M s . Albright w i l l make sure that you are still willing to take part in the research. In the first few weeks o f meetings, M s . Albright w i l l be asking questions to you and the person with aphasia to learn more about communicating with aphasia. During this time, you and the person with aphasia w i l l work out with M s . Albright when she can visit over the next three to four months. During the fourth and fifth weeks, training o f the Sentence Shaper with the person with aphasia w i l l commence. Y o u w i l l not be expected to take part in these sessions. Following the training period, the person with aphasia w i l l be practicing using the Sentence Shaper. M s . Albright w i l l continue to visit you during this time to discuss with you and the person with aphasia how you feel the program is working and to explore different ways the Sentence Shaper could be used. M s . Albright w i l l also be talking to you to find about what you feel the strengths and weaknesses o f the Sentence Shaper are. For some o f the visits, M s . Albright w i l l ask for permission to record the conversation with either a video camera or a tape-recorder. Everyone who takes part i n the conversation w i l l be asked for their permission to be recorded.  163  Computer Technology  V  computers  Investigating how the Sentence Shaper can be used in everyday life. We want to know if the Sentence Shaper can help you.  09  Does it help you?  166  How often? 1-2 Sessions per week. A minimum of 2 hours per week with the researcher For 3-4 months.  Where?  When? to be arranged  167  What  can you expect?  Weeks 1-3: A researcher will assess your language skills and talk with you to find out more about communicating with aphasia.  168  Weeks 4-5: Training with Sentence Shaper and helping the researcher learn more about communicating with aphasia.  The researcher will show you how to use the Sentence Shaper on your computer.  Talking with the researcher about how you communicate.  169  Weeks 6-14: Practice with the Sentence Shaper. The researcher, you, and a close family member or friend will work together to find out different ways to use Sentence Shaper and talk about what you like and don't like about the program.  Sentence Shaper  you  researcher  170  Weeks 15-16: Final assessment of languag skills to find out if there is any change.  At this point you can keep Sentence Shaper if you want it.  171  The researcher may audiotape or videotape some sessions.  172  Right to Withdraw  You can stop at any It is your  173  Potential Risks  3C  There is NO danger in participating in this study  Everything is  i/  confidential  Will this harm you?  174  |\JO  Concerns  not satisfied If you have concerns about your treatment or rights, you can call the U B C Office of Research Services.  Research Subject Information Line: 604-822-8598  176  Project Title: Investigating Sentence Shaper  Project Consent: The information on the previous 12 pages has been e x p l a i n e d to me. I a g r e e to participate in this research project.  I have been given a c o p y of this form.  Signature of Participant  Date  177  Appendix C : Interview Guide These questions were asked at the beginning o f the study to the individual with aphasia: 1. What are some challenges you have when talking? 2. D o you use any strategies to help get your message across? What ones work best? - i f this is hard I can ask more yes/no questions such as do you ever write things down, do you use gestures, etc. 3. What situations are most easy for you to communicate in? 4. What situations are hardest for you to communicate in? 5. What are some activities that you like doing? 6. H o w often do you use your computer? 7. What do you use your computer for? 8. D i d you use a computer before your stroke? These questions were asked at the end o f the study to the individual with aphasia: 9. What do you think the benefits o f the Sentence Shaper are? 10. What things about the Sentence Shaper make it hard to use? 11. D o you think the Sentence Shaper has helped you talk better? 12. D o you like using the Sentence Shaper? 13. Would you like to continue using the Sentence Shaper? 14. D o you think the Sentence Shaper has helped in your communication with (communication partner's name)? 15. Would you change anything about the Sentence Shaper?  These questions were asked at the beginning o f the study to the communication partner: 1. What challenges do you face when trying to talk to name)?  179  (individual with aphasia's  1. What situations do you think are easiest for  to communicate i n and which  situations are hardest? 2. What do you think  finds most difficult about having aphasia?  3. What things do you and  do that make it easier to understand each other? D o  you have any specific strategies that make things easier? These questions were asked at the end o f the study to the communication partner: 1. What do you think the benefits o f the Sentence Shaper are? 2. Is there anything about the Sentence Shaper that you don't like or would change? 3. D o you think the Sentence Shaper has made a noticeable change to to communicate?  ability  4. D o you think there are any barriers to using the Sentence Shaper? 5. D o you like it when uses the Sentence Shaper i n conversation? (this question w i l l only be asked i f the individual with aphasia actually uses the program i n conversation)  180  Appendix D : K e y to Transcription Timing features  <n> = ==  Pauses o f <.5 second Pauses o f < 2 seconds Duration o f pauses > 2 seconds Indicates overlap, placed at beginning and end o f overlapped segments Latched utterances (i.e., immediately following but not quite overlapping preceding utterance)  Prosodic features o f talk: Speech rate and stress : :  Lengthened utterance  Strategies for marking how particular features are related to talk {[ ]} [ ]  Nonlexical phenomena (vocal and nonvocal) that overlay talk Nonlexical phenomena (vocal and nonvocal) that interrupt talk  Non-talk vocal features that accompany talk [hh] [.hh]  Audible exhalation Audible inhalation  [heh-heh]  Indicates phenomenon, not the quality, o f laughter  Analyst's strategies for denoting problematic stretches o f talk ( )  Unintelligible speech  181  Appendix E : Schedule o f Data Collection DATE/TIME  PEOPLE  NOTES  PLACE  PRESENT  Erin, Dr. Purves, Nicole and Eileen Nicole, Eileen, and Erin Nicole, Eileen, and Erin  Eileen's house  Initial meeting  Eileen's house  Assessment  Eileen's house  Mar. 17, 10:30-12:30 N o recording Mar. 22,4:00-5:20 N o recording Mar. 24, 2:00-3:30 N o recording Mar. 29,3:45-5:15  Nicole, Erin Nicole, Erin Nicole, Erin Nicole, Erin  Eileen, and  Eileen's house  Assessment and interview with Nicole and Eileen Assessment  Eileen, and  Eileen's house  Eileen, and  Eileen's house  Eileen, and  Eileen's house  Mar. 31,2:00-3:45 N o recording  Nicole, Eileen, and Erin  Eileen's house  Apr. 4, 3:45-5:00 N o recording  Nicole, Eileen, and Erin  Eileen's house  Apr. 7, 2:00-4:00 N o recording  Nicole, Eileen, and Erin  Eileen's house  Apr. 11,2:00-3:30  Nicole, Eileen, and Erin  Eileen's house  Apr. 14, 11:00-12:30  Nicole, Eileen, and Erin  Eileen's house  A p r i l 15  Email from Nicole to Erin Nicole, Eileen, and Erin  Eileen's house  Nicole, Eileen, and Erin  Eileen's house  Mar. 3, 1:00-2:30 Mar. 10, 10:30-12:30 Mar. 14, 1:00-3:15  Apr. 18,2:00-3:30  Apr. 21,2:00-4:00  182  Eileen's house  Tried to install SS but had old C D ; assessment 2 failed attempt at installing SS Installed SS; messages about scrapbook (Nicole's computer lessons) Scrapbook messages (Nicole's birthday party) Message about visiting grandmother on weekend Introduced word finder; practiced making sentences with side buttons Conversation with Nicole and Eileen; emailed M i r a Received e-mail back from M i r a ; e-mail to Dr. Purves n d  Resent e-mail to Dr. Purves; email to Aunt Diana Nicole showed me emails she had done to Shelley, Ella, Diana, and Scott;  Apr. 25, 3:00-4:30 A p r . 28, 11:00-12:00 M a y 2, 1:30-3:00 M a y 4, 9:15-1:00 N o recording  Nicole, Eileen, and Erin Nicole, Eileen, and Erin Nicole, Eileen, Shelley, and E r i n Nicole, Eileen, Erin, and stroke group members  M a y 9, 1:30-3:15  Nicole, Eileen, and Erin  M a y 12, 2:00-3:30 N o recording M a y 18, 2:30-4:00  Nicole and E r i n  M a y 30, 3:15-4:30 Tape did not record June 2, 3:00-4:45  Nicole, Eileen, and Erin Nicole, Eileen, and Erin Nicole, Eileen, and Erin  June 6, 3:00-4:30  June 8 June 9, 1:00-3:00  Nicole, Eileen, and Erin  E-mail to E r i n from Nicole Nicole, Eileen, M i r a , and E r i n  Eileen's house  message about M i r a ' s engagement and Eric Message about stroke  Eileen's house  Message about Eric  Eileen's house  Discussion about C D burner  Nicole's apartment, Erin's car, and Stroke group Eileen's house  Nicole's apartment Eileen's house  Eileen's house Eileen's house Eileen's house  Eileen's house Restaurant  June 13, 3:00-4:30  Nicole, Eileen, and Erin  Eileen's house  June 15,9:15-1:00  Nicole, Eileen, Erin, and stroke group members  Nicole's apartment; Erin's car; stroke group  183  Nicole played e-mails she made to Uncle W i l s o n and Michelle and messages she made about Eric and Michelle; customized side buttons E-mail to Michelle Nicole played message she had made about Mother's Day; customized word finder Made message about Mexico Finished M e x i c o message Video recorded conversation with Nicole; wrote down ideas for M i r a message Joke e-mail Nicole played M e x i c o and M i r a messages for M i r a ; Nicole, M i r a , and E r i n went to lunch Nicole played message about Maria's birthday; customized word finder Nicole played M e x i c o message for people at stroke group  June 15 June 19 June 20, 2:00-4:00  June 21 June 22, 3:30-5:00 June 26, 3:30-5:00 June 29,2:00-5:00  E-mail to E r i n from Nicole E-mail to E r i n from Nicole Nicole, Eileen, and Erin E-mail to E r i n Nicole Nicole, Eileen, Erin Nicole, Eileen, Erin Nicole, Eileen, Erin  Nicole reciting poems  Eileen's house  from and  Eileen's house  and  Eileen's house  and  Eileen's house  184  E-mail about Father's day Customized word finder; thank you message to Eileen E-mail about Nicole's haircut Eric's birthday message Final assessments and interview with Eileen Final assessments and interviews with Nicole and Eileen  Appendix F : List o f Sentence Shaper Messages Date M a r c h 29 A p r i l 3-4  A p r i l 5-6 April 7 April 8 April 9 A p r i l 11 A p r i l 13 A p r i l 14  A p r i l 14 A p r i l 18 A p r i l 20  A p r i l 21 A p r i l 23 A p r i l 25-26 A p r i l 28 May 3 May 7 May May May June June June  12 16 302 8 12  Subject Message about computer lessons from scrapbook Messages about events from scrapbook Message about Eileen and Eric Sentences using side buttons Visit with family Visit with friends; Jessica's birthday E-mail to M i r a (past volunteer) Stroke group activities Eric's birthday from scrapbook; e-mail to D r . Purves E-mail to E r i n E-mail to Aunt Diana; sentences using side buttons E-mail to Aunt Diana, E l l a (volunteer), Shelley, and M a r k (Eric's brother); M i r a ' s engagement; Valentine's D a y gift from Eric; G y m exercises Sentences using side buttons E-mail to E r i n Stroke message  Length o f Time to M a k e Message 30 minutes  People Involved i n M a k i n g Message Nicole and E r i n  2 hours  1.5 hours  Nicole and Eileen; Nicole and volunteer Nicole and Eileen  2 hours 45 minutes 45 minutes  Nicole and E r i n Nicole Nicole  45 minutes 45 minutes 1 hour  Nicole, Eileen, and Erin Nicole Nicole and E r i n  30 minutes 1.5 hours  Nicole Nicole and E r i n  1.5 hours  Nicole and Eileen  1 hours 30 minutes 2.5 hours  Start o f Eric message E-mail to Uncle W i l s o n and Ella Eric message; Michelle (friend) message E-mail to Michelle Mother's D a y M e x i c o Message  1 hour 45 minutes  Nicole Nicole Nicole Nicole Nicole Nicole  1 hour  Nicole  1 hour 45 minutes 2.75 hours  Visit with M i r a Maria's birthday  1 hour 1 hour  Nicole and E r i n Nicole Nicole and Erin; Nicole Nicole Nicole  185  and and and and and and  Erin Eileen Erin; Eileen Erin Eileen  June 15 June 18 June 20 June 21 June 22 June 26 June 29  E-mail to E r i n E-mail to E r i n (Father's Day) Thank you message to Eileen E-mail to E (Nicole's haircut) Birthday message for Eric Three Little Pigs Cowboy Story  45 minutes 45 minutes  Nicole and Eileen Nicole and Eileen  1.5 hours  Nicole and E r i n  45 minutes  Nicole and Eileen  1 hour 40 minutes 30 minutes  Nicole and E r i n Nicole Nicole  186  Appendix G : Transcribed Narratives Pre-Treatment Cowboy Story Unaided 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.  N: E: N: E: N: E: N:  E: N:  E: N: E:  U m talk talk I mean um um I don't know okay u m <4> uh <6> u m horses .. h horse ... um uh <2> I don't know. Just do your best to explain what's going on in the pictures U m ... um <3> a man and um a man <2> and um . . . a man horse okay Perfect Yeah ... um and then um um .. hor horse and um um um and a man um .. um .. um.. um this one [heh-heh] I don't know Uhhuh U m and then I mean um uh sleeping sleeping .. .sleeping ..and um um .. um b boy and um . . . uh um gir and um man uh no no um boy boy and um . . . boy and .. um <2> u m oh a boy and um no a boy um and um um dog yeah and then [heh-heh] um sleeping sleeping sleeping sleeping and then yeah and then and then u m u m oh . . . um um um um uh horse horse I mean um sleep swish swip switch switch this I mean switch this I mean swiptch U m h m um hrnm A n d then and then I mean um um um girl uh no uh dog and dog and sleeping and um um um this one and um . . . um swish switch and I mean um I mean um um .. um um ... um um hor horse leave leaving leave them and then and then um um boy u m um um yeah and then um okay uhhuhh I mean um sleeping and then um um um dog yeah yeah this one and then um um um .. boy and and boy and uh um horse oh oh oh gone gone and then and then [h] ohhh [hehheh] um uh and then switch switch I mean I mean um um um um wo wo woke up woke up and then um .. um .. um um d dog and m man uh um .. don't know I don't know don't know yeah yeah Perfect. D o you want to say anything else? A r e you done? N o no no Okay perfect that was great Nicole. Thank you.  Pre-Treatment Three Little Piss Unaided  1. 2. 3. 4. 5. 6. 7. 8.  N: E: N: E: N:  W o l f w o l f no um um hmm Wolf wolf Yup W o l f okay okay okay um um hay hay okay I mean um no no no oh my G o d uh uh um uh um um uh pig pig um um u m new ho new house ne ne new house yeah yeah and then .. and then and then um:: oh um um um um ha I mean um um w o l f and uh no no no I  187  9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.  E: N:  E: N: E: N:  E: N:  E: N: E:  mean um I mean [hh] uh uh ( ) I mean no no um I mean um hay hay yeah hay um um um oh m y G o d [hh] um Keep going Nicole you're doing really well ( ) Okay um .. um .. [hh] yeah yeah? yeah [hh] and then oh I mean um .. um um um um um pig uh pigs pigs um um pigs um um blow uh uh blow blow them blow them yeah yeah and yeah and then and then I mean uh two people two people and oh um [hh] [heh-heh] um yeah I mean um um ... glass no I don't know I mean I mean um I mean .. uh 11 don't know ( ) no yeah and then I mean uh um ho [hh] and then um um .. um um uh ( ) um um uh three ( ) three and and um um three um um um three and um um three and um um brick brick = = = = um hm = = brick = = = = um hm = = A n d then and then um oh um um um um um um um um brick okay and um ... um .. .and then oh uh safe safe safe ... okay sorry [hehheh] safe safe [heh-heh] Okay Safe [heh-heh] um safe yeah and then uh uh [hh] [hh] and then uh uh [hh] [hh] and then [hh] [hh] [hh] and then and then uh brick brick yeah uh yes yeah I don't know ... [hh] U m hm are you finished? Okay yeah no no no no uh finish finish I don't know Okay perfect thank you so much  Post-Treatment Cowboy Story Unaided 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.  N:  E: N:  N: E: N: E: N:  Okay wait okay uh <4> um okay <2> the .. um the um .. um the um the [throat clear] the um um uh um man is um um uh wearing .. no oh okay okay the .. the okay .. the um um the um um la okay um the lady no the man .. is riding .. riding ... uh the horse Okay keep going good one Okay .. um okay .. okay um <2> okay um okay .. the .. um k h the ... um man is yawning .. uh tir uh okay the the .. um man is yawning .. yawning .. and yawning yawning and and sleeping sleeping okay and and okay okay .. okay uh oh .. she she sssshe no uh .. he he .. is .. lying .. down no okay she is tired and oh um .. uh and um and um and uh sleeping okay okay Andum:: D i d you look at this one? O h m y god [heh-heh] Okay okay okay okay okay okay okay the boy .. and the okay okay oo okay okay sh sh okay okay um the uh um <2> the uh .. um uh .. girl the um girl the um boy okay he he is .. switching seats no no  188  19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.  E: N:  E: N:  E: N: E: N:  E: N:  E: N:  E: N: E: N:  no she oh okay .. I mean ..okay she okay oh ... okay okay okay the .. oh um um the um um the um um .. the horse is .. um .. the horse is um um tr the horse is um .. um ...switching switching yeah the horse is switching over there no okay okay the uh um the horse is um switch ..switch switch switch Okay Okay okay okay um ... and and uh um and um and and um and um .. uh and um and um and um and .. um um um hor and um .. and um um horse is Um ... um switching um and um and um ... swi okay um the horse is um um .. um leaving and um .. the dog is um .. the dog is um switching Okay um hum [hh] okay the man okay .. okay the the um dog is um the is um the dog is [throat clear] the dog is ... um um um um ... the dog is switching .. uh o over o no no no no okay the okay okay the dog is oh okay yeah the oh okay oh okay the um um the horse is uh uh leaving leaving yeah U m hum U m yeah and and um the dog is uh swit switching um switching um .. um .. the uh uh um um man U m hum A n d <2> and okay um okay and um the girl the the uh the ma the uh uh .. boy is um leaving leaving the girl is the the uh the uh uh ma the uh uh gi um the the um boy the boy um is um uh uh si um um quiet quiet U m hmm Oay and um [throat clear] and um okay the oh oh yeah ... leaving no kay kay leaving the the uh the girl the boy is uh quiet the boy the boy is .. on the okay the boy is .. um .. uh um qu .. um on the um .. on the um .. um .. no no okay the man the gir the boy is qu uh quiet U m hm um hm A n d and oh um the uh the um .. the um uh um um the .. um the um man is ... the man is um .. um where do you where where .. um .. where is the um .. where is the um dog? O h no no okay the okay w a i t . . . the the man is um okay wait okay the dog no no the uh uh okay switch uh no no no no .. the um ... um the um .. the um boy .. the um uh dog is .. um ... okay okay the horse is gone and um and .. and the uh dog is .. um [throat clear] um um switching ... switching .. um .. switching U m hm Okay Finished Yeah  189  Post-Treatment Three Little Piss Unaided 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.  N:  E: N: E: N:  E: N:  E: N: E: N:  E: N:  E: N:  E:  Okay [yawn] oh G o d [hh] um ... um ... pigs and w o l f okay um the uh um ... um .. um I don't know okay wait I mean .. go back yeah yeah yeah (Nicole points to Sentence Shaper) Nope No No [heh-heh] okay um pigs um okay um .. m m ... m m .. um ... p i uh there is uh no there is there is um ... uh uh there is .. there is um .. okay um there is .. there is uh .. .there is three pigs Umhmm U m and the w o l f . . blew the okay the w o l f blewed uh uh blew the hay .. the w o l f blew th ( ) uh uh uh the okay the w o l f um um um the w o l f um is .. no okay the w o l f blew um um hay .. stop Okay A n d um um um uh the w o l f okay the pigs the pig the pig um um ran ran um ran away ran away Umhmm Okay the um uh the uh okay okay the w o l f um um blow .. uh the stick ... and and um and um um um three no one one two uh three no one one two uh three no one two two pigs uh uh away uh uh okay the okay wait the w o l f . . the uh the uh pigs pigs uh uh ran out ran out U m hmm Ran out and and uh and uh and uh um w o l f um w o l f um um um um um sh um oh okay the w o l f um uh uh the w o l f um .. um the w o l f um .. the w o l f uh .. blowed the um brick U m hmm Brick and .. and okay uh uh three one two three three um um pigs um um um uh run no uh uh ran no uh okay okay the okay the pigs stay stay no more Okay  M e x i c o Message Unaided 1. 2. 3. 4. 5. 6. 7. 8. 9.  E: N:  E: N: E: N:  Can you tell me what you did in Mexico? O h oh um whole family um uh okay uh whole family um uh whole whole family uh um uh uh whole family lira um whole family uh uh uh big big the the whole family um um big uh big uh rent a room room okay rent a room rent a room = yeah = yeah - um hmm = A n d um um uh uh Jacob uh Jacob and Robert are swimming U m hmm A n d uh uh I mean uh uh uh uh Eric uh uh uh Eric uh wearing  190  10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.  E: N: E: N: E: N:  E: N:  E: N: E: N: E:  wearing a hat U m hmm Okay okay the hat the uh uh Eric uh Eric um wearing a sombrero the hat U m hmm U h uh I bought blanket, uh a blanket, rose .. rosary, uh sovernir, uh uh a beautiful uh uh uh mirror and uh ... and uh uh t-shirt U h hmm U h I uh uh k a uh uh uh uh uh a uh uh ai um Scott uh uh Scott uh Scott had a um um no okay okay happy happy no kay no Scott uh uh birthday uh .. um uh okay okay m y birthday no kay the Scott uh um .. some some cake no no .. [heh-heh] I don't know um um .. I don't know Keep going U m okay um .. oh um .. um um the house okay the the uh the boat is uh the boat is um the boat is um um um the boat is lying um uh lu lying the okay the the okay the boat uh uh the boat is fishing fishing .. yeah U m hmm The mou the the um tunnel through the um um um .. water U m hmm The water I don't know I don't know no no .. the um the okay uh uh stop Okay  191  Appendix H : Q C L and C E T I Scores The Quality o f Communication Life Scale (Paul, Frattali, Holland, Thompson, Caperton, & Slater, 2004) Post Post Item Pre Comparison 4 5 5 1.1 like to talk with people. 4 4 2. It's easy for me to communicate. 2 5 4 5 3. M y role in the family is the same. 3 4.1 like myself. 3 3 Not Not 5.1 meet the communication needs o f m y Not applicable applicable applicable job or school. 5 6.1 stay i n touch with family and friends. 4 5 3 1 3 7. People include me in conversations. 3 3 8.1 follow news, sports, and stories on 1.5 TV/movies. 5 2 5 9.1 use the telephone. 3 3. 10.1 see the funny things i n life. 3 4 4 4 11. People understand me when I talk. 4 5 5 12.1 keep trying when people don't understand me. 4 5 5 13.1 make m y own decisions. 2 3 3 14.1 am confident that I can communicate. 5 5 15.1 get out o f the house and do things. 3 2 1 3 16.1 have household responsibilities. 4 4 17.1 speak for myself. 3 4 4.1 Mean score overall for items 1-17 2.8 4 4 18. In general, my quality o f life is good. 3  192  The Communicative Effectiveness Index (Lomas, Pickard, Bester, Elbard, Finlayson, & Zohhaib, 1989). Post Post Pre Item Comparison (scale rangingfrom1-10; not at all able to as able as before stroke) 8.4 8.4 8.4 1. Getting someone's attention 2.8 2.8 0.9 2. Getting involved in group conversation about him/her 8.6 8.6 6.2 3. Giving yes and no answers appropriately 7.3 7.1 4. Communication his/her emotions 7-1 7.2 7.2 6.2 5. Indicating that he/she understands what is being said to him/her 4.3 4.5 4.3 6. Having coffee-time visits and conversation withfriendsand neighbours (around, beside, or at home) 6.7 8.3 7. Having a one-to-one conversation with .4.5 you 9.1 6.0 9.1 8. Saying the name of someone whose face is in front of him/her 9.2 9.2 9. Communicating physical problems 7,4 such as aches or pains 5.8 4.5 3.5 10. Having a spontaneous conversation (i.e. starting the conversation and/or changing the subject) 7.3 7.3 7.7 11. Responding to or communicating anything (including yes or no) without words 2.7 2.7 12. Starting a conversation with people 1.3 who are not close family 8.6 4.4 7.0 13. Understanding writing 1.5 14. Being part of a conversation when it 1.0 1.5 is fast and there are a number of people involved 1.3 1.3 15. Participating in a conversation with 1.0 strangers 1.0 1.0 16. Describing or discussing something 0.9 in depth 5.5 5.5 Overall language and communication 3 skill during the previous week (scale rangingfrom1-7; extremely poor to excellent)  193  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092659/manifest

Comment

Related Items