“Lg-:16 so z". i 3. W (9 m «ml/09” This is to certify that the dissertation entitled AN EXPLORATORY STUDY OF THE ASSOCIATIVE RELATIONSHIPS BETWEEN FAMILY PARADIGMS ANDAUGMENTATIVE AND ALTERNATIVE COMMUNICATION SATISFACTION IN FAMILIES WITH YOUNG CHILDREN presented by Mary Josephine Cooley Hidecker has been accepted towards fulfillment of the requirements for the PhD. degree in Audiology and Speech Sciences (“fl/aka; >1§am¢a Major Professor's Sfipéture ' Q4437 Usl/t /é~; 076704 Date MSU is an Affirmative Action/Equal Opportunity Institution 0 C ---C C o O -. f -. w A- -4 - LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJClRC/DateDuepss-pt 5 ———-—.————.. AN EXPLORATORY STUDY OF THE ASSOCIATIVE RELATIONSHIPS BETWEEN FAMILY PARADIGMS AND AUGMENTATIVE AND ALTERNATIVE COMMUNICATION SATISFACTION IN FAMILIES WITH YOUNG CHILDREN By Mary Josephine Cooley Hidecker A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Audiology and Speech Sciences 2004 A5 ALI imponm satisiac Cm'irof mndon and m; IEID fc innox Para: need Strat Mil (/l (/r Te ir ABSTRACT AN EXPLORATORY STUDY OF THE ASSOCIATIVE RELATIONSHIPS BETWEEN FAMILY PARADIGMS AND AUGMENTATIVE AND ALTERNATIVE COMMUNICATION SATISFACTION IN FAMILIES WITH YOUNG CHILDREN By Mary Josephine Cooley Hidecker Family involvement in augmentative and alternative communication (AAC) is important, especially with a toddler or preschooler. This study examined AAC family satisfaction within the framework of family paradigms, which states that families view their environments in paradigmatically different ways. The view from a paradigm, (i.e., closed, random, open, and synchronous), guides families in using resources of time, space, energy, and material in the pursuit of their goals of control, affect, meaning, and content. Closed paradigms involve strategies that follow traditions, conserve resources, and reinforce loyalty to one's family. Random paradigms involve strategies that seek innovation, expend resources with zest, and encourage individual freedom. Open paradigms involve strategies that create collaborative action, balance group and individual needs for resources, and communicate practicality. Synchronous paradigms involve strategies that ground family constancy, maintain resources, and connect with the timeless universals of life. More than one paradigm can be used by a family. In this study, participants were 54 primary caregivers from 27 different states. The 55 children, aged 15 to 75 months, had not yet started kindergarten, had used the then current MC system for six to 24 months, and had a variety of developmental disorders resulting in a severe communication disorder. Most had multimodal AAC systems, which included both aided and unaided components although 11 had only the unaided AAC system compo lO—que was ad;- Ihe .U paradig: and 10°.- combinir I1 Paradigm jUngDent However~ families w increasgd I’ Continued 1 The Families “'1 Ideally (I65 1' I” cC’mrast system of sigi language. More than half of the AAC systems included a voice output component. Sign language was the most frequently used AAC within the child's home. The primary caregivers, all mothers, completed an 81-question case history and a IO-question instrument, the AAC F amily-Paradigm Assessment Scale (AA CF —PAS), which was adapted from earlier scales that identified family paradigms that guide family behavior. The AACF-PAS results from 42 families revealed that 33% frequently used a closed paradign, 48% frequently used a random paradign, 76% frequently used a open paradign, and 10% frequently used a synchronous paradign, with one half of the families frequently combining family paradigns. In contrast to those results, closed (36%), random (57%), and open (55%) paradigns were most frequently remembered when caregivers provided a retrospective judgnent of their family functioning fi'om the period of time before the child's diagiosis. However, after the child's diagiosis, closed and open paradigns were used by more families while random paradign decreased. Afier AAC implementation, families reported increased reliance on random and open while decreasing closed strategies. This trend continued when families identified a hypothetical, ideal method of family functioning. The primary caregivers were generally satisfied with the child's AAC system. Families who started with random paradigns before the child was diagnosed and/or who ideally desired random paradigns were more likely to be satisfied with their AAC system. In contrast, families who more frequently used a closed paradign at the time of the study were more likely to be dissatisfied with their AAC system. Additional demographics, future research directions, and clinical implications are also discussed. Cepyfight by MARY JOSEPHINE COOLEY l-HDECKER 2004 DEDICATION T o my husband Jefl who always believed. member outstanc especial apprecia gracious paradign| endeavor I I AAC and ACKNOWLEDGNIENTS I want to acknowledge the guidance and assistance provided by the faculty members on my dissertation committee. Rebecca Jones, my chair and advisor, was an outstanding mentor throughout my doctoral program. Jill Elfenbein supported my growth especially in teaching and in research dissemination. John Eulenberg instilled in me an appreciation for AAC history as well as for innovative technological advances. David Imi g graciously shared his knowledge and experience, especially in the area of family paradigns. Ida Stockman kept me focused that language and communication underlie AAC endeavors. Francisco Villarruel encouraged me to think about the connections between AAC and family paradigns. Gratitude is extended to the generous family members and to the professionals and organizations who participated or otherwise assisted in this research, to Dianne Angelo who shared her questionnaires from her AAC family research, and to Betsy Bizot, Pam Harris, and Gail O'Leslcy-Villarruel for their insights. In addition, Rana Al Khamra, Amyn Amlani, Johanna Boult, David Medei, Diane Ogiela, Robin Pulford, Greg Robinson, and Deborah Stuart provided critical comments in their role as members of a graduate student writing group. Finally, a heartfelt thank you to all the faculty, staff, fellow students, family, and friends who supported me throughout my doctoral education. This project was supported in part by a predoctoral fellowship from the National Institutes of Health-National Institute on Deafness and other Communication Disorders (NIH-NIDCD 5-F31-DC05443-02) as well as an Incentive Fellowship and a Dissertation Completion Fellowship by Michigan State University. vi TABLE OF CONTENTS LIST OF FIGURES ................................................................................................................. ix LIST OF TABLES .................................................................................................................... x CHAPTER 1 ............................................................................................................................. 1 Introduction and Review of the Literature ............................................................................... 1 The Importance of Families ......................................................................................... l The Theory of Family Paradigns ................................................................................ 3 AAC and Family Paradigns ...................................................................................... 13 Research Questions .................................................................................................... 17 CHAPTER 2 ........................................................................................................................... 19 Research Design and Methods ............................................................................................... 19 Participants .................................................................................................................. 19 Instruments ................................................................................................................. 20 Procedures ................................................................................................................... 26 Analysis ...................................................................................................................... 28 CHAPTER 3 ........................................................................................................................... 32 Results ..................................................................................................................................... 32 Demographics ............................................................................................................. 32 Post-AAC Paradigns ................................................................................................. 39 Changes in Paradigns After-Diagtosis and Post-AAC4O Paradign and AAC satisfaction ................................................................................. 46 Paradigm Resources and Goals .................................................................................. 55 CHAPTER 4 ........................................................................................................................... 57 Discussion ............................................................................................................................... 57 Discussion of Results ................................................................................................. 57 Clinical Implications .................................................................................................. 66 Limitations of This Study ........................................................................................... 73 Future Research Needs ............................................................................................... 75 Conclusions ................................................................................................................. 78 APPENDIX A ......................................................................................................................... 81 Tables and Figures .................................................................................................................. 81 APPENDIX B ....................................................................................................................... 110 Case History .......................................................................................................................... 110 APPENDIX C ....................................................................................................................... 121 AAC Family - Paradign Assessment Scale (AACF-PAS) ................................................. 121 vii APPE.\ .uCF—l APPEX' Rules It" REFER APPENDIX D ....................................................................................................................... 128 AACF-PAS Calculations ....................................................................................................... 128 APPENDIX E ....................................................................................................................... 131 Rules to Correct Raw AACF-PAS Scores ............................................................................ 131 REFERENCES ..................................................................................................................... 133 viii 2. Ch; LIST OF FIGURES 1. Basic processing of inputs and creation of outputs by a family system ........................ 108 2. Changes in family paradigns across the 4 judgnents. .................................................. 109 "J - 0rd; . Reas ' Alder - FIEQL Chili Q‘VPP’NE" N 10. 11. 12. I3. 14. 15. 16. 17. 18. 19. 20. 21. LIST OF TABLES Comparison of resources by paradign. .......................................................................... 82 Comparison of goals by paradign .................................................................................. 83 Age, gender, race, and ethnicity of AAC users and caregivers ..................................... 84 Caregivers' education level and employment ................................................................. 85 Family size, language(s) spoken, income, and geographical location ........................... 86 Child’s diagnosis, the severity of disability on family's daily activities, and child's educational setting ........................................................................................................... 87 Frequency of child's age when problem suspected, diagnosed, and AAC received ..... 88 Length of time fi'om when problem suspected, diagnosed, and AAC received ............ 89 Type of AAC system currently available ....................................................................... 90 Usage of communication modes at home ....................................................................... 95 Ordinal AAC-use and AAC satisfaction ........................................................................ 97 Child's language skills with or without AAC ................................................................. 98 Reasons why the child does not use aided AAC ............................................................ 99 Aided AAC components and funding sources ............................................................. 100 Frequency and percentage of paradigns by judgnent .................... A ............................ 101 Summary of multiple regression analysis to predict changes in family paradign as a function of the Before-Diagiosis (BD) paradign ........................................................ 102 Summary of multiple regression analysis to predict changes in family paradign as a function of Afier-Diagtosis (AD) paradign ................................................................ 103 Summary of multiple regression analysis to predict changes in family paradign as a fimction of Post-AAC (PA) paradign .......................................................................... 104 Summary of multiple reg'ession analysis to predict AAC satisfaction as a function of family paradign: A priori and main effects models .................................................... 105 Summary of multiple regression analysis to predict AAC satisfaction as a function of family paradign, Combining significant main effects in final model ......................... 106 Average paradign cluster score by element across judgnents ..... . .............................. 107 CHAPTER 1 Introduction and Review of the Literature Many have recognized the importance of family involvement in augnentative and alternative communication (AAC) intervention (Angelo, 2000; Angelo, Jones, & Kokoska, 1995; Cress, 2004; Parette, VanBiervliet, & Hourcade, 2000; Sweeney, 1999). However, Parette and Angelo (1996) suggested that, historically, AAC practices have focused more on the role of service providers and less on the role of families in assisting AAC users to become competent communicators. In order to change this mindset, professionals need to understand AAC users within the context of their families' beliefs and preferences, but little is known about which family features contribute to successful and satisfied AAC users and families. Therefore, examining theories of family functioning may provide organizing frameworks to explore this area of research. Hence, the field of AAC may benefit by using family theories to identify and test potentially-important family variables (Sweeney, 1999). The Importance of Families Families play critical roles in children's early communication development (Andrews & Andrews, 2000; Chapman, 2000; Donahue-Kilburg, 1992; Gillette, 2000). Early social interactions often require parents and other communication partners to ascribe meaning to the child's behavior. However, a child's developmental delays, due to cogiitive and/or physical limitations, can interfere with the communication partner's responsivity (Harwood, Warren, & Yoder, 2002; Siegel & Cress, 2002). The partner may be unsuccessful in recogrizing and/or interpreting the child's behavior and vocalizations. DC‘CrE'u prevet‘. come the chill tools en Speech- and meet used to c into the .- and recor (Judge 8,; Ofincreas dlSCOnIInL 50mg F686;: [initiated F A (COHSIami’r I a 31,, 2000 the family ' centered‘ fa; Decreasing these communication breakdowns between the child and his or her parent may prevent or lessen the severity of the communication disorder. Some professionals have advocated using AAC to clarify and expand the child's communicative repertoire in order to increase the communication partner's responsivity and the child's communication successes (Cress, 2002; Light, Parsons, & Drager, 2002). AAC tools encompass a wide assortment of symbols, aids, strategies, and techniques (American Speech—Language—Hearing Association, 1991) that are matched to the AAC user's strengths and needs. This process of feature-matching AAC components to an AAC user is often used to develop a multimodal AAC system (Glennen & DeCoste, 1997). Inserting a step into the AAC evaluation, in which family features are also matched to AAC components and recommendations, might enhance successful AAC implementation within the family (Judge & Parette, 1998). F eature-matching to the family may also assist in the prevention of increased family stress (Judge & Parette, 1998) and subsequent technology discontinuance by the AAC user and family (Phillips & Zhao, 1993). Believing that there is only one ideal version of family functioning is a mistake that some researchers and clinicians make (Kantor & Lehr, 1975). This belief can lead to frustrated professionals and disheartened families. Alternatively, some professionals (Constantine, 1986; Crais, 1991; Imig, 2000a; McWilliam, Tocci, & Harbin, 1998; Parette et al., 2000) believe that they are obligated to assist a family in ways that are important to the family. This fundamental principle underlies numerous terms, including family- centered, family-focused, family-directed, family-enabled, and family-empowered (Crais, 1991), that have been used in the literature of early intervention and of communication disorders. (Note: F amily-centered will be used in this paper to represent the principle.) 2 Given 111C305 _ CEDICIC The 77; 1975) h functior this chat. shows h (Constar reSource: Importan may then rsquire r} Kamor & H and SJTIch Given the complexity of family dynamics, professionals must find effective and efficient means to determine what a particular family views as important if the principle of family- centered practices is to become widespread. The T heory of Family Paradigms The theory of family paradigns (Constantine, 1986; Irnig, 2000a; Kantor & Lehr, 1975) has been used by family science professionals to determine individual family functioning. Its variables may be relevant to AAC intervention as will be discussed later in this chapter. Family paradigns are based on qualitative and quantitative research, which shows how people view the world in very different ways, leading to different behaviors (Constantine, 1986; Irnig, 2000a; Kantor & Lehr, 1975). While all families use their resources to obtain their desired outcomes, each family can differ in its view of what is important and how to accomplish its goals. Using this theoretical perspective, professionals may then help a family integrate AAC recommendations into its current paradign and not require the family to change its view of what is important (Constantine, I986; Irnig, 2000a; Kantor & Lehr, 1975) for the sake of the AAC recommendations. Historically, four family paradigns have been described as closed, random, open, and synchronous (Constantine, 1986, 1993a; Imig, 2000a; Kantor & Lehr, 1975). In order to introduce this theory, each of the four paradigms will be described as a distinct category, although, in many cases, families may live as a combination of paradigns. Additional details about the specific types of resources and goals that families use will be presented in later sections of this chapter. (See Figure l for a diagram of family functioning. Note that all tables and figures are in Appendix A.) Families that prefer stable, predictable use of their resources to obtain traditional, time-honored goals are employing a structured or closed family paradign. These families follow the directions of the head of the family. They prefer routines and predictable, steady, efficient use of their resources. Families that prefer spontaneous, individual use of their resources in order to obtain individual, innovative goals are applying a spontaneous or random paradign. These families value individuals pursuing interesting, inventive goals. Families that prefer practical use of their resources to achieve consensus goals are implementing a negotiated or open paradign. These families will decide on goals by discussion leading to consensus. Families that understand each other so well that words are not needed are said to be unconsciously connected or synchronous. These families appear to have little need for overt communication as each one "just knows" what to do in order to achieve successful family functioning. In all families, the resources of time, space, energy, and material are constantly being expended by family members to accomplish their goals of control, affect, meaning, and content (see Figure 1). The results of these expenditures are evaluated by the family member(s) through feedback mechanisms. Gathering and evaluating feedback about their resource use and goal attainment will assist families in rebalancing those elements. For example, one family may believe that volunteering to help at a homeless shelter is a meaningful goal. The family members will monitor and decide on how much time, energy, and material donations should be expended. If the meaning of this activity is very important and does not interfere with other family goals, the family is likely to expend more 4 resour accom might I family creatin' followe Kantort actions ( prefer to later in tl I in which {Bubolz a that are In consul“:IE natural ent cOnsmlcte and techno Cmoms, 1a resources. However, if the family feels that other important goals are not being accomplished or that it does not have enough resources for all its goals, then one family might devote less time, energy, and/or material goods to the homeless shelter. Another family might recruit more families and friends to use their resources for the shelter, thus creating additional resources for the homeless shelter. Family members will use one or more of the four player parts, mover, challenger, follower, and bystander, within their personal interactions (Constantine, 1986; Irnig, 2000a; Kantor & Lehr, 1975). Movers initiate actions. In contrast, challengers tend to resist the actions of the movers. Followers tend to follow the actions of the movers. Bystanders prefer to reflect on the various actions. More details about player parts will be presented later in this chapter. The family's behavior is also dependent on the larger contexts of the environments in which they live, including natural, human-constructed, and social-cultural venues (Bubolz & Sontag, 1993). The natural environment includes physical or biological forms that are unaltered by humans such as air, climate, plants, animals, and water. The human- constructed environment consists of all things humans have built or transformed hour the natural environment in order to achieve human goals. Some elements of the human- constructed environment include buildings, farmland, cities, roads, pollution, electricity, and technology. The social-cultural environment includes how people live in community together; what types of economies are created; and what types of culture, such as laws, customs, language, and social norms, are developed. Thus, within these multiple environments, the dynamic family system will by to adjust its use of goals and resources in order to better match its preferred methods of functioning. 5 the g0 an ela from ti amoun feeding love in less timr microwa IIICXpenS Pizza del resCurses be an app the Physic In Hoyt-even less approg achievjn g t Fa, colleq and Consider the decision-making process that a family may implement to accomplish the goal of feeding the children. In one case, a family member could spend all day cooking an elaborate meal. Depending on the natural location, supper might be animals and plants from the family's natural environment. The farmly member may decide that expending that amount of time and energy is an appropriate use of family resources since the outcome of feeding children homegrown and home-cooked meals is viewed as an important labor of love in the family. In another family, however, a can of soup might be heated up for supper, requiring less time as long as the family has the human-constructed appliances of a can-opener and a microwave. This meal of soup would be viewed as appropriate if the family valued inexpensive, quick feeding of its members. In a third case, a family member could have a pizza delivered. The cost of the pizza would have to come from the family's material resources but would require less immediate family energy expenditure. Pizza would have to be an appropriate food in the family's culture, and pizza delivery would need to exist within the physical proximity of the family's house. In each of these cases, the family goal of feeding the children was accomplished. However, families with different paradigns will evaluate the same sh‘ategy as a more or less appropriate use of their resources that will bring them closer to or further from achieving their desired goals (Imig, 2000a; Villarruel, Imig, & Kostelnik, 1995) Family Paradigm Resources. The theory of family paradigms proposes that people collect and expend their resources of time, space, energy, and materials (Constantine, 1986; Imig, 2000a). See Table 1 for a summary of resources by paradign. conce’,‘ found family ' story, a omflh. agendzc apprOpr mi ght st bed at 8 and alter: meeting I SPeCial te day Pressi knowledg P. 5)_ For Q Changing l bed at Suns The (Consumer, distance am OUI {Or the e The resource of time refers to how the family schedules its actual time. Using the concept of the closed paradign, a family looks to the past for the stability and guidance found in following the family's traditional routines or methods. For example, the closed family might use the same routine with the children (e.g., 7 pm. bath, 7:30 pm. bedtime story, and 7:45 pm. lights off) that the mother followed when she was growing up. On the other hand, the random paradign is "organized as an aggegate of individual agendas". . .carried out "according to whose individual priorities. . .[are] most relevant or appropriate at that particular momen " (Imig, 2000a, chap. 5, p. 3). For example, a child might stay up until midnight one night to work on an important school project, yet go to bed at 8 pm. on another night because she is tired. The open paradign concept adapts to the consensual decision of what is practical and attempts to balance individual and family issues. For example, the family may hold a meeting to decide if a child should be allowed to stay up past her bedtime to watch a special television show. The concept of the synchronous paradign "transcends the day-to- day pressures of life" by following "a plan characterized by simplicity and implicitly shared knowledge among family members" and flows with nature’s rhythm (Imig, 2000a, chap. 5, p. 5). For example, the synchronous farme might subconsciously adjust bedtimes with the changing length of daylight. Without overt communication, the child would know to go to bed at sunset. The resource of space includes the use of both physical and interpersonal space (Constantine, 1986; Imig, 2000a). Physical space can be used to increase or decrease the distance among people. For example, being in the same room or leaving the family to go out for the evening is a physical space issue. Interpersonal space includes decisions on what 7 are acceptable ideas. The closed paradign may draw strong physical boundaries between those in the family and those outside the family, and thus, certain rooms in the house are not to be shown to "outsiders." Closed families may have certain topics that are considered unacceptable to bring up in the family. The random paradign may have fluid boundaries where anyone in the physical proximity is included in family activities. The random family would likely be encouraging of any topic. The open family can decide to change the use of physical space but only after discussion. Most topics are appropriate as long as the family can reach a consensus opinion. The synchronous family has implicit boundaries and acceptable topics. The family knows its membership by those who share its common understandings. The energy resource, which is the intensity used for one's actions, must be obtained and dispersed to accomplish daily activities (Constantine, 1986; Imig, 2000a). The closed paradign looks for steady, efficient use of available predictable energy. The random paradign allows for fluctuating energy. Thus, sometimes energy will be abundant and, at other times, energy will be low because the family is overextended. The open paradign seeks to create flexible energy. If the family needs to expend more energy, they will need to find more energy supplies. The synchronous paradign's energy flows from its connection with nature. A synchronous energy example is the Chinese concept of chi, in which a person's energy is believed to be connected with the universe (Imig, 2000a, Ch. 5, p. 11) Material goods raise questions about the roles and consumption of possessions and belongings of the family (Constantine, 1986; Imig, 2000a). The closed paradign views material goods as rewards for effective effort. The random paradign prefers to avoid material goods because of their potential to complicate relationships. The open paradign 8 views material goods as resources to use in its pursuit of individual and familial goals. The synchronous paradigm values material goods that are perfect and timeless, which unconsciously reminds each family member of his or her connection with each other and with nature. For example, a synchronous family could display a pristine conch shell in a place of honor in its home, symbolizing the family's shared ocean connections. Family Paradigm Goals. The theory of family paradigns proposes that people try to achieve the goals of control, aflect, meaning, and content (Constantine, 1986; Imig, 2000a). See Table 2 for a summary of goals by paradign. The goal of control is to achieve the family's desired outcomes. The closed family develops plans that can be counted on in all situations. The random family allows individuals to determine what, when, and how to complete necessary tasks. The open family will discuss what, when, and how to best work together to accomplish the necessary actions. The synchronous family implicitly carries out the needed activities. The goal of affect is for the families to express feelings including affection, caring, and support (Constantine, 1986; Imig, 2000a). The closed family prefers traditional, private expressions of caring. The random family seeks spontaneous and passionate expressions of caring. The open family desires authentic and responsive expressions of caring. The synchronous family considers explicit expressions of caring as unnecessary displays of what its members already unconsciously know. The goal of meaning is to determine the value placed on different aspects of life (Constantine, 1986; Imig, 2000a). The closed family follows the "traditional" values of its culture. The random family values each individual member finding and following each one's vision. The open farmly constructs a shared meaning through an open discussion and 9 consensus of its values. The synchronous family shares a common vision that arises from the members' similar views of their common experiences. The goal of content is to make sense of the family's reality (Constantine, 1986; Imig, 2000a). The closed family believes reality is defined, objective, and absolute. The random family considers reality as individually-constructed and subjective. The open family views reality as combining the objective and subjective views in order to create a practical or useful truth. The synchronous family believes that individuals will each find the path to the sMe inevitable conclusion. Compromise and compound family paradigms. Although family paradigns have thus far been presented as separate categories, family style is generally more complex than a single paradign (Constantine, 1986; Imig, 2000a). F amilies may blend two or more paradigns, creating a compromise system, or each family member may rely on a different paradign, yielding a compound system (Constantine, 1986, pp. 352-366; Imig, 2000a, Ch 9). For instance, a family may employ an open paradign for time and energy issues with a random paradign for space and material resources. If members agree with their use of specific paradigns for particular tasks, they are using a random-open compromise paradign. On the other hand, if family members disagree on whether to use an open or a random paradign to utilize resources, they would find their open and random paradigns in competition for each task, resulting in a random-open compound paradign. Enabled and disabled family paradigms. Within their configuration of paradigns, families may have differing levels of success in meeting their individual and goup goals with their available resources. If most of the family needs and goals are being met, the family operates with an enabled paradign (Constantine, 1986; Imig, 2000a). If the family 10 is not successful in its resource allocation and goal attainment, the family uses a disabled paradign. Families under stress may rely increasingly on their paradigns, repeating the same behaviors which have already caused negative functioning (Constantine, 1986, pp. 202- 205; Imig, 2000a, Ch. 10, pp. 1-10). For example, closed families might fiirther tighten their boundaries against outside influences and impose dictatorial rule. Random families could be hampered by each individual competing for all of the family's resources and goals. Open families might be paralyzed by unending negotiations and schisms. Synchronous families may expel members who do not understand and follow the family's unstated way of functioning. If this negative functioning persists, the family's functioning is disabled. A disabled family may benefit from counseling to improve its functioning. The . counseling would be undertaken, not to change the family’s worldview, but to increase the effectiveness of its current paradign (Imig, 2000a). In other words, the purpose of counseling is to change a disabled paradign into an enabled paradign. For example, I counseling a family with a random, disabled paradign would not require the family to change its paradign. Instead, counseling would focus on ways to enhance its strategies to accomplish what the family deems important within its random worldview. See Constantine (1986) for family therapy approaches to use with different paradigns. Four Player Parts within the Family Paradigm. Kantor and Lehr (1975) outlined four possible behaviors within the family system: initiating, opposing, supporting, and observing. These behaviors are represented by the player parts of mover, challenger, follower, and bystander, respectively, which are incorporated within every relationship. In addition, the player part pattems vary by paradign. ll In a closed family paradign, the mover initiates a time-honored strategy that the follower will support (Constantine, 1986; Imig, 2000a; Kantor & Lehr, 1975). The challenger offers constructive criticism to improve the strategy. The bystander acts as the conscience for the mover and requires all family members to conform to the time-honored strategy. In a random family paradign, the mover initiates a strategy but does not expect anyone else to follow it (Constantine, 1986; Imig, 2000a; Kantor & Lehr, 1975). The challenger will support the mover's right to initiate a strategy as long as the strategy does not impinge upon other family members' resources. The follower will choose to support the mover or the challenger and does not automatically follow the mover’s strategy. The bystander ensures healthy competition for the resources and goal priorities among the mover, the challenger, and the follower. In an open family paradign, the mover proposes a possible action while the challenger offers suggestions to improve the proposed action (Constantine, 1986; Imig, 2000a; Kantor & Lehr, 1975). The follower supports this dialogue. The bystander is important to encourage consensus-building to create the actual strategy that will be used by the entire group. In a synchronous family paradign, the mover begins an activity that naturally flows from the environment and context (Constantine, 1986; Imig, 2000a; Kantor & Lehr, 1975). No follower or challenger is needed because everyone already knows what to do. Everyone is a bystander because each knows that this activity is required by the context of daily life. 12 AAC and Family Paradigms AAC researchers and clinicians increasingly have focused on the concerns and views of the AAC user and his or her family (Angelo, 2000; Angelo et al., 1995; Angelo, Kokoska, & Jones, 1996; Blackstone, Williams, & Joyce, 2002; Cress, 2004; Jinks & Sinteff, 1994; Jones, Angelo, & Kokoska, 1998; Schlosser, 1999; Sweeney, 1999; VanBiervliet & Parette, 1999). Parette and Angelo (1998) summarized some of the research on how assistive technology; including AAC systems of sigi language, communication pictures, communication books, and voice output communication aids (VOCAs); may affect the family of young children with disabilities. For example, at least one caregiver will likely need to use time and energy to implement, learn, and/or troubleshoot the AAC system. The researchers suggested that professionals should ask the family how physical and interpersonal space within the home as well as material resources are likely to be affected by the addition of different AAC components. The environments of home, school, and other community locations must accommodate the differing physical space needs (e.g., weight, size, and shape) of the selected AAC system. In addition, many families want additional information about different types of AAC and how they work in other farnilies' homes. The researchers also noted that families may not have realistic expectations of how AAC will affect their daily lives. Professionals must also try to understand how the family's values affect its perception of different AAC components. While this research (Parette & Angelo, 1998) cites family systems theory, it does not use family paradigns as its theoretical framework. Through the organizing framework of family paradigns, these AAC issues could be viewed in the context of how the family 13 uses the resources of time, space, energy, and materials to achieve its goals of control, affect, meaning, and content. Families will have to decide how much time and energy they should devote to AAC issues. AAC use may have a negative effect on the regulation of interpersonal space (e. g., family cohesion) if it results in fewer resources being available for other family members' needs. Material issues related to the AAC system might include the role of technology within the family structure and whether possessions are individually or communally used. This may lead to decisions on where and how AAC may be used in the family. For example, some families will post communication pictures throughout the house, while other families will reject this as inappropriate home decorating. Likewise, in some families, the AAC system will be viewed as something that only the AAC user should touch while, in other families, the AAC system will be used by multiple family members. In addition, AAC may alter families' views of accomplishment or control, of affect or love, of value or meaning, and of reality or content by changing the roles and the expectations of family members. Complicated AAC systems may challenge the family's feeling of control. Affect for each other may be strained by changing demands on the family's resources. Professionals may not understand the family's dreams or values. The person without AAC may have been viewed as a non-communicator, which was part of the family's reality or content. After successful AAC implementation, the family's content goal must change, as the AAC user becomes a better communicator of his or her thoughts. Any changes in the family’s use of resources and goals will likely lead to increased stress in the family system. Professionals need to recogiize the increased family stress and I4 help families find the least stressful ways to integrate AAC into their lives (Parette & Angelo, 1998). Current AAC practices often rely on integated and predictable communication opportunities (Beukelman & Mirenda, 1998; Calculator & Jorgensen, 1991; Goosens', Crain, & Elder, 1992; Zangari & Kangas, 1997). As a result, these practices appear to favor open and closed family paradigns. Since open families value communication fi‘om all members, they may be eager to expend family resources to ensure a more elaborate voice from the AAC user. Closed families rely on routines that may allow for predictable communication opporttmities for beginning communicators, although closed families may not be comfortable with the nontraditional mode of communication. They may prefer to wait longer before implementing AAC to see if their child will develop speech, which many would consider a more traditional mode of communication. Matching AAC systems to random and synchronous family paradigns appears more difficult than open and closed views when considering current AAC service delivery models. Random families value spontaneous communication, which is difficult to accomplish with AAC systems that need to be created or programmed before communication opportunities arise. Different members of random families may value different communication methods, thus creating a multimodal AAC system. Synchronous family members do not need overt communication to expend its resources or to accomplish its goals. Therefore, they may deem AAC unnecessary to successful synchronous fiinctioning. AAC recommendations may be most successful in a closed family paradign when AAC systems require limited change in the daily routines and recognize the hierarchical 15 communication pattern. For example, if the person currently communicates by eye gazing at desired objects, perhaps the initial AAC system would include eye gazing at picture symbols in order to choose desired activities and objects. With the closed family's use of routines, predictable choices are known and could be placed on communication boards. However, closed families may not be comfortable with nontraditional modes of communication and wait longer before implementing AAC in the hope that their child will develop speech. In contrast, the random family may fluctuate in the amount of resources expended for AAC integration. This may be negatively perceived by professionals as inconsistency and a lack of follow-through by the family. Each family member may have developed a different way of communicating with the AAC user. Thus, the AAC system may be truly multimodal in this family paradign. For example, the AAC user may vocalize with the mother, use a communication board with the father, and use a voice output system in the community. One difficulty with AAC systems operating within a random paradign is ensuring spontaneous communication messages. Simple digitized speech devices that allow for quick recording of messages may be useful in such circumstances. On a positive note, random families may be more interested in trying innovative AAC systems. An open family paradign requires all family members to achieve consensus on a practical, flexible plan. AAC that provides the geatest communication opportunities may be most valued. Family members will work together to implement and integrate AAC because of their desire for the AAC user's full participation in the family. This family type may be more successful than the other paradigns in integating AAC recommendations within its daily life. 1 6 In contrast, a synchronous family shares an implicit, unstated plan. Explicit communication is not needed for family firnctioning because everyone "just knows" what to do. The professionals may feel that the family is uncooperative and non-communicative. However, to the family, collaboration with the professional team may not be seen as necessary to the successful function of the family. Synchronous communication may already be reported between the major caretaker and the potential AAC user so the intervention goals might be better focused on communication outside of the synchronous family system or subsystem. The family members may be willing to facilitate the AAC user's communication with peers, teachers, and other community members rather than within the family. Family paradign theory provides a theoretical fiamework including potential variables from which to research AAC intervention. While professionals recogrize the family’s critical role in an AAC user’s communicative success (Huer & Lloyd, 1990), professionals have difficulty predicting the family’s responses to AAC. According to family paradign theory, each family decides, based on its paradign, which resources and goals to use towards changing the person’s lack of intelligible communication. AAC professionals may not have questioned their own assumptions about "appropriate" paradigns and strategies. AAC professionals and families have not considered explicitly that families' different ways of using their time, energy, space, and materials to achieve control, affect, meaning, and content may influence their decisions regarding AAC. Research Questions This research explored the applicability of family paradign variables to early AAC intervention by asking parents of young AAC users to complete three instruments: a case 17 history, an AAC satisfaction rating scale, and a family paradign scale. The objectives of this study were to 1) identify the then current paradigns of the family, 2) determine if the family believed that its use of the paradigns had changed from a. the time before the child was diagiosed as having a severe communication problem and b. after the diagrosis but before AAC was obtained, and 3) explore satisfaction with the family's paradign and the AAC system. 18 CHAPTER 2 Research Design and Methods Participants Fifty-four primary caregivers of young AAC users participated in this study. Primary caregivers self-identified themselves as the adults in the families who spent the most time communicating with the AAC user and/or who had more child care and home responsibilities. The young AAC users were toddlers and preschoolers (i.e., had not begun kindergarten) with severe commrmication disorders. Severe communication disorders occur when “gestural, speech, and/or written communication is temporarily or permanently inadequate to meet...communication needs” (American Speech-Language-Hearing Association, 1991, p. 10). Some of the AAC users in this study used vocalizations, gestures, speech, and/or speech approximations at least some of the time. None of the AAC users in this study had hearing loss as the primary cause of their communication disorder. The AAC children in this study, ages one to six, had used their current AAC system between 6 months and 2 years. This time period was selected for two reasons: 1) to allow time for families to adjust to the AAC system and 2) because assistive technology abandonment is believed to be greatest within the first three months of obtaining new technology (Scherer, 1996, p.115). However, no level of family adjustment was presumed, and caregivers were asked in the case history about use frequency of different AAC system components at home. See Appendix B for questions 20 to 24 in the case history. 19 Instruments Case History Questionnaire. The case history contained 81 questions, which are listed in Appendix B. The AAC user section contained 45 questions about the child’s birth date, gender, race, ethnicity, disability, acquisition and use of multimodal AAC components, communication skills, and educational setting. The family section contained eight questions regarding the language(s) spoken at home, number of siblings, geographical location, and income. The primary caregiver and secondary caregiver sections contained 28 questions that asked for caregivers' ages, gender, relationship to the AAC user, marital status, education, employment status, occupation, race, and ethnicity. The wording of the questions was refined with input from two mothers of AAC users and ten professionals in education or commtmication disorders. AAC F amily-Paradigm Assessment Scale. The AAC F amily-Paradign Assessment Scale (AA CF -PAS) was specifically developed for this study and can be found in Appendix C. The AACF-PAS was adapted from the R-PAS, the Relational-PAS, (Imig, 2000b), which determines the paradign orientation and player part for each of the resource and goal elements. (Note: The player part data was not analyzed for this research.) The AACF-PAS retained the order of the 20 questions from the R-PAS, and the wording was only minimally changed (i.e., "relationship" was replaced by "family"). The statement "Please check that each column has one and only one 10 in it" was added to the last page of the paper version after several participants had difficulty completing the AACF—PAS. The R-PAS required judgnents of current and ideal functioning for each item. For the AACF-PAS, participants also provided two retrospective judgnents of family functioning: before the child was diagrosed with a severe communication problem and 20 after the diagiosis but before AAC was implemented. They also indicated how they would prefer their families to function if they lived in an ideal world. Thus, four judgnents of family paradigns were computed fiom the AACF-PAS: 1) The paradign before the child was diagrosed with a severe communication problem, as perceived retrospectively (Before- Diagiosis paradign); 2) The paradign after the diagrosis but before AAC was implemented, as perceived retrospectively (After-Diagnosis paradign); 3) The paradign existing with the then current AAC system (Post-AAC paradigm); and 4) The paradign desired, hypothetically, in an ideal world (Ideal paradigm)- While the retrospective judgnents were a new application of the PAS, they offered the potential for comparing the current and desired paradignatic orientation to the person’s perceived past orientation. This comparison could provide useful insights into how the farme viewed its behavioral changes over time. In the AACF-PAS, caregivers self- identified these two retrospective points when they were asked to think back to those times. Some early intervention research suggests that parents are able to identify stressfiil events (Bailey, 1988). These events likely include the diagiosis of a disorder and transitions of services. Consequently, parents are likely to view the diagiosis of a severe communication disorder as a stressful event and the AAC recommendations as both a change in the child’s communication mode and as a loss of typically-developing speech and language. In 21 addition, families may rely more on their paradigns during periods of stress (Imig, 2000a) so that family functioning may be clearly remembered from those times. The original works of Kantor and Lehr (1975) and Constantine (1986) were based on a qualitative methodology of extensive field observations for each family. This extensive observation of families within their multiple environments does not translate easily to clinical use. To rectify this difficulty, the PAS (Imig, 1993; Imig & Phillips, 1992) was created as a paper-and-pencil instrument to use in intervention counseling. The item construction for the original PAS was based on Kantor's, Lehr's, and Constantine's qualitative research. During research to develop the Family Regime Assessment Scale (FRAS), an earlier version of the PAS, participants who identified with different paradigns commented on both the appropriateness and the acceptability of the wording of the PAS (Imig & Phillips, 1992). The scoring of the PAS required a scaling procedure, based on MultiAttribute Utility Technology (MAUT), where respondents assigred numerals 0 to 10 to each statement (Edwards & Newman, 1982). A 0 represented the perceptual judgnent that this attribute was not characteristic of how the family functioned, while a 10 indicated that this attribute was exactly how the family firnctioned. The MAUT procedure transformed these raw scores into complex coefficients, which were list-ordered fi'om the family's highest to the lowest attributes. These results were then compiled into quartile and cluster scores (Imig, 2000a). Details about these calculations can be found in Appendix D. All versions of the PAS appear to have construct validity, since they were specifically created to measure the structures and player parts described in family paradign theory (Constantine, 1993a; Imig, 1993, 2000a; Imig & Phillips, 1992). Family experts and participants have felt that the FRAS and the R-PAS are representative of family functioning, 22 which is a component of content validity (Imig, 1993, 2000a; Imig et al., 1996; Imig & Phillips, 1992; Pate, 1994; Pegorraro, 1999). The FRAS is recogtized as a measure within the discipline of farme science (Touliatos, Perlmutter, & Straus, 1999, pp. 50-51 ). Criterion-related validity has not been established due to the fact that the PAS is the only instrument currently available for measuring family paradigns. However, professionals and family members have found the results useful in research and clinical activities (Imig, 1993, 2000a; Imig et al., 1996; Imig & Phillips, 1992; Pate, 1994; Pegorraro, 1999; Villarruel et al., 1995; Ward, 1997). Different versions of the PAS have been used to assess paradigm orientation in several research studies, such as post-divorce adjustment of divorced single-mother families (Pate, 1994); individuals', couples', and families’ behavior under stress (Imig, 1993; 2000); families’ decisions to home-school their children (Pegorraro, 1999); factors affecting participation in men’s groups (Imig et al., 1996); and factors affecting transfer of family businesses within the family (Imig et al., 1996). Test- retest reliability studies are still needed for the PAS. Variables. The independent variables were based on the child's age, the severity of the child's disability, the amount of AAC use at home, and the overall paradign cluster score for each paradign. The dependent variables included the overall paradign difference score and the primary caregiver's satisfaction with the child's AAC system. The child's age in months was calculated as the difference between the birth date and the case history date completion for the Internet version or between the birth date and the date the paper version was returned. The severity of the disability variable was ordinally—based on the parent’s severity rating of the effect of the child’s physical and/or cogritive disability on the family's daily activities: 23 1 = mild, 2 = moderate, and 3 = severe. (See Appendix B for questions 7 in the case history.) The ordinal AAC-use variable was based on the AAC components that were used “often” or “always” at home: 0 = no AAC components used, 1 = unaided AAC components, 2 = no technology AAC, 3 = low technology AAC, 4 = direct select computer-based, 5 = non-direct select computer-based, 6 = multimodal combinations. This new variable was created hour the answers of case history questions 15 through 25. (See Appendix B for question wording.) The family ordinal variable represented the family’s resource of income by income category. (See Appendix B for Question 53 in the case history.) Originally, the number of adults in the household was also going to be used in the creation of the family ordinal variable. (See Appendix B for Question 50in the case history.) However, 50 of the 54 families had only two adults living in the house so that variable was not included in the family ordinal variable due to the data set's limited variability. Family paradigm was operationalized as a rating of 5 on the AACF-PAS overall cluster score for categorical descriptives. Within the multiple regression analysis, each 24 family paradign was captured as an ordinal variable from a 0, meaning the family never used those paradign strategies, to a 5, meaning that the farmly always or almost always used those paradign strategies. The dependent variable of paradign change was calculated as the structural difference score for the overall cluster score between two judgnents. First, the absolute value of each paradign difference score was calculated. For example, the formula to calculate the closed paradign difference score from the Before-Diagiosis judgment to the Afier-Diagrosis judgnent was: Closed paradigm difl'erence score = | Closed paradigm cluster score fi'om Before-Diagnosis - Closed paradigm cluster score from After-Diagnosis I This calculation was repeated three times, replacing closed with random, then with open, and finally with synchronous paradign cluster scores. In order to calculate the overall paradign difference score, the closed, random, open, and synchronous paradign difference scores were totalled. The dependent variable of AAC satisfaction was created from the primary caregiver's rating from question 12 or 14 in the case history: 1 = strongly dissatisfied, 2 = dissatisfied, 3 = neutral, 4 = satisfied, 5 = strongly dissatisfied. (See Appendix B for question wording.) Satisfaction ratings from question 14 were used except for those caregivers who responded that the child was still using the first AAC 25 Ii] system. In those cases, the satisfaction rating from question 12 regarding the first AAC system was used. Procedures Few demographics have been collected on young AAC users. In the United States, children with severe disabilities have been estimated as comprising 0.4% of the under 3- year age group, with males (0.5%) numbering more than females (0.1%), and 0.7% of the 3- to 5-year age group, with males (0.9%) numbering more than females (0.4%) (Aron, Loprest, & Steuerle, 1996, p. 19). Children with severe communication disorders are likely some part of this category; however, the exact number of AAC users is not known for these age groups. No large pool of young AAC users and their families was found during this study. This made random sampling from the population of families with young AAC users problematic. Therefore, a convenience sample was used for this initial study. The families were recruited nationally through networks of AAC professionals, parent support groups, disability rights organizations, and AAC consumer g'oups. More than 500 advertising contacts were made by email, phone calls, mail, and face-to-face appeals over a 14-month interval. Where appropriate, advertising contacts were repeated every three to four months. Announcements about the study were posted on websites and e- mailed on listservs; some readers forwarded the postings to other professionals and family members. Advertisements reached people throughout the United States. Interested people e-mailed or phoned the researcher who determined project eligibility, explained the participant tasks, and answered any questions via phone or e-mail. Caregivers were advised of their rights through a written consent form approved by the Michigan State University Institutional Review Board and were paid $15 for 26 participation. The primary caregiver completed the case history questionnaire, the AACF- PAS, and the AAC in Families survey. (Note: The results from the AAC in Families survey will be analyzed at a later date.) Estimated completion time for all three questionnaires was 60 to 90 minutes. The caregiver had the choice of either completing the paper version and sending it through the mail or completing the electronic version through a secure Internet site. If the caregiver completed the paper version, the researcher copied the caregiver's answers onto the electronic version. Caregivers were asked to complete the questionnaires within two weeks. If questionnaires were not completed and returned after three weeks, the researcher used e-mails and/or phone calls to determine any problems and to encourage the participant to complete the questionnaires. Fifty-four primary caregivers participated in this project. Nineteen (35%) respondents completed the paper version and 35 (65%) of the respondents completed the electronic version. (Note: One of the respondents switched to paper after having difficulty accessing the electronic version.) On average, the families using the electronic version had higher incomes and lived in more densely populated areas than the families using the paper version. Fourteen additional primary caregivers enrolled in the study but did not complete any of the instruments: nine gave no reason, three reported time constraints, and two were caring for new babies. Forty-two primary caregivers completed all three instruments. In addition, 12 caregivers completed the case history but did not complete the AACF-PAS. Reasons given for the incomplete AACF—PAS included difficulty understanding the instructions, intricacy of the rating task, and encountering time constraints. In addition, eight primary caregivers did not correctly complete the paper version of the AACF—PAS. Seven of the eight AACF- 27 PAS were used after rules for incomplete or incorrect data were applied. (See Appendix E for the rules used.) Results were imported into a Microsoft Access database. Multiple response questions were recoded into binary categories. Data entry was checked for accuracy. In addition, the caregivers' responses were checked for errors and inconsistencies, which were corrected when possible. For example, Question 49 of the case history asked for the number of children in the family, including the AAC user, who had disabilities. Responses of 0 were changed to 1 as all of the families had at least one child using AAC. Analysis Data analysis was completed using the statistical software package of SPSS 1 2. 0 for Window. An a priori decision of a = .05, with two-tailed tests as applicable, was used for all statistical tests. The main purpose of this initial study was to describe possible relationships among family, child, and AAC variables. These relationships will need to be replicated in future research. Therefore, an experimental alpha was not adjusted for the multiple statistical tests in this exploratory study. Research Question 1. Identifir the paradigmatic orientations of AA C families after AAC was obtained for their young children. A fiequency distribution of paradigns for the Post-AAC judgnent was compiled and compared to the AAC user, family, and caregiver demographics using Speannan rank correlational analysis. Research Question 2. Determine if changes in the family's paradigm occurred when (a) parents became aware of the child is severe communication disorder and (b) AAC was obtained. For part a, the Before-Diagrosis judgnent was compared to the Afier-Diagrosis judgnent, and, for part b, the After-Diagnosis judgnent was compared to the Post-AAC 28 judgnent. According to Imig (2000a), a cluster structure that showed a difference value of 2 or larger for a paradign is noteworthy. For example, a cluster score of 5 results from a caregiver's judgnent that a particular strategy is used always or almost always, which is indicated by ratings of 9 or 10 on the AACF-PAS. A cluster score of 3 results fi'om a caregiver’s judgnent that a particular strategy is used sometimes which is indicated by ratings of 6 or 5 on the AACF-PAS. Thus, a cluster score change fi'om 5 to 3 indicates that the caregiver had noted that the paradigm strategy decreased from being used always or almost always to being used sometimes. Multiple regression was used to compare the paradign difference score with the original family paradign while statistically controlling for children, the AAC system, and family variables. The additive multiple regression model was: change in family paradigm (total cluster score difi'erence) = intercept +fl1closed cluster + ,Bzrandom cluster + ,B3open cluster + Asynchronous cluster + ,B5severity of child ’s disability + flgAAC ordinal variable + ,67family resources. Research Question 3. Determine satisfaction with family paradigm and the current AAC system used at home. Comparing the Post-AAC judgnents and the Ideal paradign judgnents from the AACF-PAS provided data regarding whether the primary caregiver was happy with the current paradign orientation. If the difference score between the paradigns showed a difference value of 2 or greater (Imig, 2000a, Ch. 11), then the caregiver desired some changes in his or her current paradign use. 29 Multiple regression was used to compare the difference scores and the primary caregivers' satisfaction with the farmly paradigns while statistically controlling for some of the child, the AAC system, and family variables. The first additive multiple regession model, using the Post-AAC paradign cluster scores to predict the difference between the j udgnents and Ideal paradign judgnents, was: change in family paradigm (total cluster score difi’erence) = intercept +fl1closed cluster + flzrandom cluster + ,63open cluster + Asynchronous cluster + ,85severity of child 's disability + ,B6AAC ordinal variable + ,67family resources. In addition, caregivers' satisfaction with the AAC system being used can be compared to the paradign judgnents. Multiple regression was used to compare the caregivers' satisfaction with the family paradign while statistically controlling for some of the child, the AAC system, and family variables. The second additive multiple regression model was: primary caregiver satisfaction with AAC system = intercept + ,6, closed cluster + ,Bzrandom cluster + ,B3open cluster + ,64synchronous cluster + ,65severity of child ’s disability + flaAAC ordinal variable + ,Byfamily resources. A priori Power Analysis for Research Questions 2 and 3. The exploratory nature of this project required adopting a medium effect size of r = .36 and R2 = .13 (Cohen, 1988) as is a typical convention of social sciences research. A sample size of 37 (Hintz, 2001) achieved 70% power (Stevens, 1999) to detect a medium effect of family paradigm 30 variables using an F-test (ct = .05). Tested variables were adjusted for an additional three independent variables with medium effects (R2 = .40). 31 CHAPTER 3 Results The results are presented in four sections: 1) demographics, 2) Post-AAC paradigns, 3) changes in paradigns After-Diagrosis and Post-AAC, and 4) paradign and AAC satisfaction. Demographics Fifty-four U.S. families with 55 AAC youngsters participated. One family had twin AAC users. Families lived in 27 different states with only Arizona (n = 7) and Minnesota (11 = 6) contributing more than 10% of the sample. The demographics were also computed for the subset of 42 families with 43 AAC users who completed the AACF-PAS. Tables 3 through 14 contain demogaphics for the full sample of 54 families and for the subset of 42 families. Independent t tests were run between the demogaphics for the 42 families with AACF-PAS data and the 12 families without AACF-PAS data Only the variable of "satisfaction with the first AAC system" was statistically different, t(52) = 5.0, p < .0005. Those who did not complete the AACF-PAS were more satisfied with their first AAC system (M = 4.5) than those who did complete the AACF-PAS (M = 3.4). (See Table 11 for details on the satisfaction ratings.) As seen in Table 3, the age of the child ranged from 15 months to 75 months with an average age of 47 months. Sixty-five percent of the children were boys and 35% were girls. All of the children were identified as Caucasian. A few children also were identified as American Indian or Alaskan native; Asian, Native Hawaiian or other Pacific Islander; or 32 yr {'7}. Hispanic or Latino. The primary caregivers were all female, with an average age of 36 years and ages ranging fiom 23 to 53 years. The secondary caregivers were all male, with an average age of 38 years and ages ranging from 26 to 58 years. F ifty-three of the primary caregivers were Caucasian, one was American Indian or Alaskan native, and one Asian, Native Hawaiian or other Pacific Islander. Fifty-one of the secondary caregivers were reported to be Caucasian; one was reported to be American Indian or Alaskan native; two were was reported to be Asian, Native Hawaiian or other Pacific Islander; and one was reported to be Hispanic. None of the children or caregivers were Afiican American. Non- Hispanic and Caucasian people were over-represented in this sample when compared to the 2003 general population of the US: 14% Hispanic; and 82% Caucasian, 13% Afiican American, 5% Asian, 2% American Indian or Alaskan native, and less than 1% Native Hawaiian or other Pacific Islander (U .8. Census Bureau, 2004) Almost all of the families were two-parent intact families. F ifty-three primary caregivers were married to the child's other parent while only one family reported the caregivers being separated or divorced. Fifty-two of the primary and the secondary caregivers were the biological parents of the AAC user while two sets of the primary and the secondary caregivers were adoptive parents. In contrast, 69% of US. children under the age of 18 lived with two parents; 88% of which were two biological parents; 9% were one biological and one stepparent; 1% was adoptive parents; and 1% was other combinations (US. Census Bureau, 2001, p. 6-2) As seen in Table 4, the caregivers were generally well-educated with only 9% of the primary caregivers and 13% of the secondary caregivers having a 12th grade education or lower. Indeed, 43% of the primary caregivers and 36% of the secondary caregivers had 33 \V Lo) ho; ant. taken or completed graduate level studies. In contrast, females in the US, aged 25 to 49 years, listed their educational attainment as 4% junior high or less, 37% high school, 29% some college, 21% college degree, and 8% advanced degree. Similarly, males in the US, aged 25 to 49 years, listed their educational attainment as 4% junior high or less, 40% high school, 26% some college, 20% college degree, and 9% advanced degree. [Notez Educational attainment levels for females and males were calculated from a US. Census data set (U .8. Census Bureau, 2002).] Sixty-one percent of the primary caregivers were stay-at-home caregivers, 20% were part-time employees, 19% were full-time employees, 9% were business owners or self-employed, 6% were students, and 2 % were unemployed as listed in Table 4. In contrast, 4% of the secondary caregivers were stay-at-home caregivers, 4% were part-time employees, 76% were full-time employees, 13% were business owners or self-employed, 2% were students, and 0% were unemployed. (Note: Respondents could indicate multiple responses for employment level so employment percentages could exceed 100%.) Family characteristics are listed in Table 5. The AAC user was the only child in 30% of the families as compared to 21% of US. families having only one child (U .8. Census Bureau, 2001 , p. 6-2). Thirty-five percent of the AAC families had two children, and 35% had three to five children including the AAC user. Eighty-seven percent of the families had only one child with a disability. English was spoken in all of the homes; in one home Spanish was also spoken. Sixteen percent of families reporting made under $40,000 annual gross income; 57% made $40,000 to $99,999; and 27% made $100,000 or more per year. Six percent of the 54 families did not report their income. The families in this sample had generally higher income when compared to married-couple families in the general US. 34 population of whom 30% made under $40,000 annual gross income; 51% made $40,000 to $99,999; and 19% made $100,000 or more per year (U .S. Census Bureau, 2000). Fifteen percent of families lived in rural areas, 22% lived in towns smaller than 50,000 people, and 63% lived in larger metro areas. Table 6 shows the frequency of different diagroses, the effect of the child's disability on the family's daily activities, and the child's educational settings. Most of the children had multiple diagroses. Of the 55 AAC users, 60% had a developmentally- delayed label, 33% had apraxia, 33% had mental retardation, and 29% had autism or pervasive developmental disorder. When the primary caregiver indicated the impact of the child's disability on the family's daily activities, 15% chose mild, 56% chose moderate, and 29% chose severe. The children often had multiple educational settings with 44% receiving home visits, 49% attending self-contained preschools (where all the children had disabilities), and 42% attending inclusive preschools (where at least 50% of the children had no disabilities). Table 7 lists the ages when the child's problems were suspected and then diagrosed. It also lists the ages when the first and the latest AAC systems were acquired. Nearly half of the children were less than one year old and all of the children were under three when a problem was suspected. Thirty-one percent of the 55 children were diagiosed before their first birthday, 58% before their second birthday, and 89% before their third birthday. Two of the children did not yet have a firm diagrosis. The first AAC system was acquired for 16% of the children before their first birthday, 29% between their first and second birthday, 33% between their second and third birthday, 18% between their third and fourth birthday, and 4% between their fifth and sixth birthday. Twenty-seven percent of the 55 children are 35 still using their first AAC system. In contrast, 7% received their latest AAC system between their first and second birthday, 26% between their second and third birthday, 18% between their third and fourth birthday, 20% between their fourth and fifth birthday, and 2% between their fifth and sixth birthday. Table 8 lists the amount of time between when the problem was first suspected or diagrosed to when AAC was acquired. On average, one year passed between the first suspicion of the problem to actually receiving the first AAC system, although some families waited as long as 5 years. In a few cases, baby sigr, which has been used with typically-developing infants, was implemented before the parents suspected that the child had a problem. That is why the minimum length of time between suspecting a problem and implementing the first AAC system was negative. (See Acredolo, Goodwyn, & Abrams, 2002, for a description of baby sigr.) On average, a half year passed between the child's diagrosis and the acquisition of the first AAC system. For those 40 AAC users who were not using their first AAC system, an average of 1.2 years passed with a range from less than a year to 4 years between the first and the latest AAC system. Sigr language, communication pictures or boards, communication books, simple VOCAs, and complex VOCAs were most often available in various combinations. Table 9 lists the frequencies of these multimodal combinations. Eighty-two percent of all 55 children had sign language available with 20% of the children having sigr language as their only mode of AAC. Sixty-seven percent of all the children had communication pictures and/or boards available with only one child having only communication pictures and/or boards. Twenty-four percent of the children had communication books available with none of the children having only communication books. Thirty-three percent of the children had 36 simple VOCAs available with none of the children having only simple VOCAs. Thirty-five percent of the children had complex VOCAs available with two of the children having only complex VOCAs. Some of the communication pictures were described as PECS, an acronym for the Picture Exchange Communication System (Frost & Bondy, 2002). Different models and manufacturers were listed for the simple and complex VOCAs. Table 10 lists the fi'equency of use of different communication modes in the home environment. Vocalizations were used often or always by 58% of the 55 children; 4% never vocalized at home. Speech or speech approximations were used often or always by 31% of the 55; 22% did not use any speech at home. Pointing or gesturing was used often or always by 56% of the 55 children; 13% did not use pointing or gesturing at home. All of the children who had sigr available used it at least part of the time. Sigr language was used often or always by 67% of the 45 children reported to have sigr language available. Communication pictures or boards were used often or always by 24% of the 37 children reported to have communication pictures or boards which is in contrast to 11% who did not use their communication pictures or boards at home. Communication books were not used often or always by any of the 13 children reported to have communication books. Instead, about half used their books rarely and half used their books sometimes at home. Simple VOCAs were used always by 24% of the 18 children reported to have simple VOCAs available which is in contrast to 29% who did not use their simple VOCAs at home. Complex VOCAs were used always by 42% of the 19 children reported to have complex VOCAs available, which is in contrast to 11% who did not use their complex VOCAs at home. Thus, unaided AAC (i.e., sigt language) was used by a higher percentage of children 37 than aided AAC (i.e., complex VOCAs, communication pictures or boards, simple VOCAs, and communication books). Table 11 lists the frequency of the different categories of the ordinal AAC-use variable, which was based on which AAC components were used often or always at home. Twenty-six percent of the 55 children did not use any of the AAC components often or always. F orty—seven percent used unaided AAC components (i.e., sigi language) often or always, and 9% used non-voice output aided AAC components (i.e., communication pictures, boards, and/or books) often or always. Four percent used direct select complex VOCAs often or always, and 15% used more than one AAC component (i.e., multimodal) often or always. Satisfaction data with the AAC system are also contained in Table l 1. Sixty percent were satisfied with the first AAC system, and 75% were satisfied with their then current AAC system, which included the 15 first AAC systems that had not been changed. Table 12 outlines the children's language skills with or without using the AAC system. Twenty-two percent of the children responded but did not initiate communication, and 26% did not spontaneously communicate. Thirty-six percent of the children produced more than 50 words independently with or without AAC. Thirty-three percent produced 10 to 50 words, 27% produced 1 to 9 words, and only 4% produced no words independently with or without AAC. (Note: Children could have skills in several categories so the following percentages do not add to 100%.) F orty-six percent of the children followed requests and directions without assistance, and 66% made spontaneous requests. F orty- four percent of the children responded to comments appropriately and directions without assistance, but only 26% made spontaneous comments. Thirty-five percent of the children 38 answered questions appropriately, but only 15% asked spontaneous questions. In terms of utterance length, only 18% of the children used three-word or longer utterances, and only 5% of the children constructed grammatically-correct sentences. Table 13 lists reasons why the 11 unaided children did not have any aided AAC components available at home. The most frequently-selected reason was that the child talked, sigred, or gestured (64% of the 11 children). This was followed by the explanation that the family was not given any reason for not using aided AAC (46% of the 11 children). In Table 14, the subset of aided AAC components is described. Color line drawings, color photographs, printed words, and black and white line drawings were used by more than half of the aided AAC users (11 = 44). F orty-one of the 44 aided AAC users were direct selectors; that is they used a finger or another body part to point to or touch the AAC symbol. (See the table for additional selection methods.) More than half of the AAC users had voice-output types of AAC available. Education agencies were the highest sources of AAC funding, followed by the family. Post-AAC Paradigms Forty-two primary caregivers completed the AACF-PAS. The criterion for assigrment to paradign categories was an AACF-PAS cluster score of 5, which indicated that the family used that particular paradign often or always. A given family could have a 5 in more than one paradign, indicating that the family frequently uses multiple paradigns. Fourteen families (33%) used a closed paradign, 20 families (48%) used a random paradign, 32 families (76%) used an open paradign, and 4 families (10%) used a synchronous paradign. 39 As shown in the Post-AAC column of Table 15, 22 families (53%) indicated using primarily one paradign: 5 closed, 5 random, and 12 open. Fourteen families (33%) indicated using a combination of 2 paradigns: 3 closed-open, 10 random-open, and 1 open- synchronous. Six families (14%) indicated using a combination of 3 or 4 paradigns. Thus, an open paradign was the most frequently-used paradign in the Post-AAC judgnent: Twenty-nine percent of the families reported primarily using an open paradign, 33% of the families reported the combination of open with another paradign, and 14% of the families reported open within a combination of 3 or 4 paradigns. The most frequent combination, used by 24% of the families, was a random paradign in conjunction with an open paradign. The Post-AAC paradign was not correlated with the child's age, the child's disability level, the ordinal AAC-use variable, family income, or caregiver education level. Primary and secondary caregivers' ages were sigrificantly correlated (p = .36, p = .028, two-tailed, and p = .31, p = .045, two-tailed, respectively) with an increased reliance on a closed paradign. In other words, families with older caregivers were more likely to identify a closed paradign. Changes in Paradigms After-Diagnosis and Post-AAC Forty-two primary caregivers also provided two retrospective judgnents of family functioning: before the child was diagrosed with a severe communication problem (Before-Diagrosis) and after the diagrosis but before AAC was implemented (After- Diagrosis). Figure 2 shows how each of the paradigns varied across the 3 judgnents of Before-Diagrosis, After-Diagnosis, and Post-AAC. (Note: The Ideal judgnent will be discussed in the next section.) Reports of closed paradign slightly increased from 15 40 families Before—Diagrosis to 18 families After-Diagnosis but then declined to 14 families Post-AAC. Over half the families (11 = 24) perceived their starting paradign to be random, but only one third of the families (n = 14) used random after the child's diagrosis. However, this decrease was temporary as nearly half the families (n = 20) reported random strategies in their Post-AAC paradigns. The open paradign (n = 23) was as frequently used as the random paradign during the time before the child's diagrosis. However, the number of families reporting an open paradign increased at each successive time point (n = 25 After-Diagrosis, n = 32 Post-AAC). The synchronous paradign was least used in each of the judgnents (n = 3 After-Diagrosis, n = 4 Post-AAC). Table 15 shows the frequencies of families using different combinations of paradigns across the four judgnents. Families showed the most diversity in the Before- Diagrosis judgnent, reporting 11 different paradign configurations (closed, random, open, synchronous, closed-random, closed-open, closed-synchronous, random-open, open- synchronous, closed-random-open, closed-random-synchronous). Random, open, and random-open paradigns were reported by over half the families (n = 24). In the After- Diagiosis judgnent, families reported only eight paradign configurations with a decrease in random paradigns and an increase in closed paradigns (closed, random, open, synchronous, closed-open, random-open, closed-random—open, closed-open-synchronous). The increase in closed paradigns consisted of the following: The purely closed paradign was reported by three families in the Before-Diagrosis judgnent and increased to nine families in the After-Diagrosis judgnent; the number of families reporting the closed-open paradign combination rose from two families in the Before-Diagnosis judgnent to seven families in the After-Diagrosis judgnent. In the Post—AAC judgnents, nine paradign 41 C01 su syr mo Jud. Dia oit Be configurations were reported (closed, random, open, closed-open, random-open, cpen- synchronous, closed-random-open, closed-open-synchronous, closed-random- open- synchronous). From those nine configurations, cpen and random—open paradigms were most used (11 = 22). (The Ideal column will be discussed in the next section on paradign satisfaction.) Even though the data above describe overall group changes, changes for individual families can be tracked as a paradign difference score between cluster scores of two judgnents. For example, the closed paradign difference score between the Before- Diagrosis judgnent and the After-Diagrosis judgnent was computed as the absolute value of the difference of the After-Diagrosis closed paradign cluster score subtracted from the Before-Diagnosis closed paradign cluster score: closed paradigm difference score = I closed paradigm cluster score from Before-Diagnosis - closed paradigm cluster score from After-Diagnosis | As discussed in the previous chapter, difference scores of 2 or more are considered noteworthy (Imig, 2000a). Each family had four paradign difference scores (i.e., one for each paradign of closed, random, open, and synchronous) per judgnent category (i.e., Before-Diagrosis, After-Diagrosis, Post-AAC, and Ideal). Thus, for this set of 42 families, a total of 168 difference scores (i.e., 4 difference scores multiplied by 42 families) existed per judgnent category. Most of the cluster score differences were 0, representing no change; or 1, representing a slight change. Comparing the Before- and the After-Diagnosis judgnents, 17 families reported only 26 difference scores of 2 or geater. Five were increased closed strategies, six were decreased random strategies, five were decreased synchronous 42 Sll Sig strategies, and the remaining difference scores did not group into a pattern. Comparing the After-Diagnosis judgnents and the Post-AAC judgnents, 15 families showed only 18 sigrificant difference scores with no clear paradign patterns among the families. In order to determine if certain paradigns were more likely to change, simultaneous multiple regression (Licht, 1995) was used to compare the paradign difference score with the original family paradign while statistically controlling for child, AAC system, and family income. The total cluster score difference was the sum of the closed, random, open, and synchronous paradign difference scores. The a priori hypothetical model was: change in family paradigm (total cluster score difl'erence) = intercept +Aclosed cluster + Arandom cluster + Aopen cluster + Asynchronous cluster + Aseverity of child ’s disability + AAAC ordinal variable + Afamily resources. This model was the starting point for the following two analyses: 1) changes in paradigns from Before- to After-Diagrosis and 2) changes in paradigns from After-Diagrosis to Post-AAC. Changes in paradigms from Before- to After-Diagnosis. First, the a priori model was evaluated. Seven predictors were used: one for each paradigm and one for each of three possibly-confounding variables: child's disability, AAC use, and family income. The Before-Diagrosis paradign cluster scores explained nearly one-third of the variance in the difference scores, R2 = .34. The overall model was statistically significant, F (7, 31) = 2.30, p = .05. (See the a priori model column of Table 16 for the reg'ession estimates of this model.) As indicated by their nonsigrificant regression coefficients, the potentially- 43 di 01'- col 116 If: (1’ confounding child, AAC, and family income variables did not contribute to explaining the differences in cluster scores. Therefore, these variables were not used in the subsequent multiple regression modeling for this analysis. Next, a main effects model was evaluated with the predictors of Before-Diagiosis closed, random, open, and synchronous paradigms. The Before-Diagnosis paradign cluster scores explained nearly one-third of the variance in the difference scores, R2 = .28, and this overall model was statistically sigrificant, F (4, 37) = 3.57, p = .015. (See the main effects column of Table 16 for the regression estimates of this model.) As indicated by their negative sigrificant regression coefficients, the Before-Diagnosis cluster score of random and open paradigns independently decreased .76 and .93, respectively, for every 1 point increase in the difference score, after partialling out the effect of the other variables. Since the random and open variables were sigrificant in the main effects model, an interaction term of random by open was added to the third model to determine if the interaction was needed to explain the changes in difference scores. The interaction variable did not improve the model as indicated by no change in the proportion of variance explained, R2 = .28. This overall model was also statistically sigrificant, F (3, 38) = 4.96, p = .005. (See the interaction column of Table 16 for the regression estimates, none of which were significant predictors of the cluster score change.) Multicollinearity (i.e., highly-correlated predictors) was indicated by the combination of a sigrificant overall model paired with non-sigrificant regression coefficients that had large standard errors. Therefore, this interaction model was rejected due to multicollinearity. The final and most parsimonious model included only the Before-Diagnosis cluster scores for random and open paradigns and explained 26% of the variance in the difference 44 me Th: mt :- tot score fi'om before to after the diagnosis, R2 = .26. The overall model was statistically significant, F (2, 39) = 6.67, p = .003. (See the final column of Table 16 for the regression estimates.) Although the residuals were somewhat skewed, the model generally appeared to meet the required error assumptions. The fitted model was: change in family paradigm (total cluster score drflerencefiom Before- to After-Diagnosis) = 10.21 - .73random cluster fiom the Before-Diagnosis judgment - .93open cluster from the Before-Diagnosis judgment. Thus, this final model shows that the Before-Diagrosis cluster score for random and open paradigns independently decreased .73 and .93, respectively, for every 1 point increase in the difference score while controlling the other variables' effect. Changes in paradigms fiom After-Diagnosis to Post-AA C. First, the a priori model mentioned at the beginning of this section was evaluated. Seven predictors were used: one for each paradign and one for each of three possibly-confounding variables, which were severity of child's disability, AAC use, and family income. After-diagnosis paradign cluster scores explained 31% of the variance in the difference scores. However, the overall model was not statistically sigrificant, F(7, 31) = 2.02, p = .08. (See the a priori model column of Table 17 for the regression estimates of this model.) As indicated by their nonsigrificant reg'ession coefficients, the potentially-confounding child, AAC, and income variables did not contribute to the difference in cluster scores. Therefore, these variables were not used in the subsequent multiple regression modeling for this analysis. Next, a main effects model was evaluated with the predictors of After-Diagnosis closed, random, open, and synchronous paradigns. The After-Diagrosis paradign cluster scores explained one-fourth of the variance in the difference scores, R2 = .24, and this 45 513 1'3 overall model was statistically sigrificant, F (4, 37) = 2.89, p = .04. (See the main effects column of Table 17 for the regression estimates of this model.) As indicated by its significant negative regression coefficients, the After-Diagnosis cluster scores for the open paradign decreased .73 points for every 1 point increase in the difference scored from the After-Diagrosis judgnent to the Post-AAC judgnent after partialling out the other variables. The final and most parsimonious model included only the After-Diagrosis cluster scores for the open paradign and explained only 14% of the variance in the difference score fiom after the diagrosis to Post-AAC, R2 = .14. The overall model was statistically sigiificant, F(1, 40) = 6.41, p = .015. (See the final column of T able 17 for the regression estimates.) Although the residuals were somewhat skewed, the model generally appeared to meet the required error assmnptions. The fitted model was: change in family paradigm (total cluster score diflerence from After- Diagnosis to Post-AA C) = 5. 75 - .79open cluster fiom After-Diagnosis Therefore, this final model showed that the After-Diagnosis cluster score of open paradigns independently decreased .79 for every 1 point increase in the difference score. Paradigm and AAC satisfaction Ideal paradigms. Families were asked to rate the hypothetical paradign strategies that they would prefer to use in an ideal world. Based on their responses, families could indicate 1, 2, 3, or all 4 paradigns as their ideal. Identification of an ideal paradign was determined by a given family having a cluster score of 5 for a particular paradign. The results were that five families (12%) chose a closed paradign, 25 families (60%) chose a 46 5n clu l9" :1 1.1 Of.) or.) A‘s 80 C01 bi) fal random paradign, 37 families (88%) chose an open paradign, and 2 families (5%) chose a synchronous paradign. As shown in the Ideal column of Table 15, twenty of the 42 families (48%) had a cluster score of 5 for one paradign: five random and 15 open. Eighteen families (43%) had cluster scores of 5 for two paradigns: two closed-open and 16 random-open. Four families (9%) had cluster scores of 5 for three or four paradigns. An cpen paradign was most frequent with 88% of the 42 families having a cluster score of 5 for that paradign. This 88% subdivided into the following categories: 36% were only an open paradign, 43% were a combination of open with another paradign, and 9% were an open paradign within a combination of three or four paradigns. The next most frequent paradign was random with 60% of the families: 12% selected a random paradign, 38% selected a combination of random with another paradign, and 9% selected a random paradign within a combination of three or four paradigns. The most frequent combination, selected by 16 of the 42 families, was a random paradign in conjunction with an open paradign. Figure 2 shows how each of the paradigns varied across the four judgments of Before-Diagnosis, After-Diagrosis, Post-AAC, and Ideal. The Ideal data is indicated by the hatched bars in Figure 2. (Note: The Before-Diagrosis, After-Diagrosis, and Post-AAC judgnents were presented in the previous section.) Reports of closed paradign decreased to only five families in the Ideal judgnent. Over half of the families (n = 25) preferred a random paradign, rebounding to the pre-diagnosis level. The open paradign continued to increased at each successive time judgnent to a high of 88% (n = 37). In contrast, the synchronous paradigm decreased to 5% (n = 2). 47 putt its open. open- Open. Table 15 shows the frequencies of families reporting different combinations of paradigns across the four judgnents. Families showed the least diversity in the Ideal judgnent, reporting only seven different paradign configurations (random, open, closed- open, random-open, closed-random-open, random-open—synchronous, and closed-random- open-synchronous). Nearly all of the 42 families (n = 36) cited a preference for random, open, or random-open paradigns in the Ideal judgnent, as indicated by a cluster score of 5. As discussed in the previous section, changes for individual families can be tracked as a paradign difference score between cluster scores of two judgnents. Each of the 42 families had four paradign difference scores per j udgnent category, resulting in a total of 168 difference scores per category comparison for this set of 42 families. Difference scores between the Post-AAC and Ideal judgnents should be zero, if families are satisfied with, their family fimctioning; difference scores of 2 or more suggest that these families want a major change in their current paradign(s) (Imig, 2000a). Families showed few difference scores of 2 or more between the Post-AAC and Ideal judgnents. Most of the cluster score differences were 0, representing no desired change, or 1, representing a slight desired adjustment (Imig, 20003). Only 9 of the 42 families reported difference scores of 2 or greater between the Post-AAC and Ideal judgnents. Those nine families showed 15 changes involving difference scores of 2 or more: 4 decreased closed strategies, 2 increased closed strategies, 3 increased random strategies, 2 increased closed strategies, and 4 decreased synchronous strategies. Changes in paradigms fiom Post-AAC to Ideal. In order to determine if certain paradigns were more likely to predict changes in difference scores, simultaneous multiple regression (Licht, 1995) was used to compare the paradign difference score with the Post- 48 AAC family paradign cluster score while statistically controlling for child, AAC system, and income variables. The total cluster score difference is the sum of the closed, random, open, and synchronous paradign difference scores from the Post-AAC judgnents to the Ideal judgnents The following a priori hypothetical model was the starting point for this analysis: change in family paradigm (total cluster score difi’erence) = intercept +Aclosed cluster + Arandom cluster + Aopen cluster + Asynchronous cluster + Aseverity of child 's disability + AAAC ordinal variable + Afamily resources. Seven predictors were used: one for each paradign and one for each of three possibly- confounding variables: child's disability, AAC use, and family income. Post-AAC paradign cluster scores explained more than one-half of the variance in the difference scores, R2 = .57. The overall model was statistically sigrificant F(7, 31) = 5.82, p < .0005. (See the a priori model column of Table 18 for the regression estimates of this model.) As indicated by their nonsigrificant regression coefficients, the potentially-confounding child, AAC, and income variables did not contribute to explaining the difference in cluster scores. Therefore, these variables were not used in the subsequent multiple regression modeling for this analysis. Next, a main effects model was evaluated with the predictors of Post-AAC closed, random, open, and synchronous paradigns. Post-AAC cluster scores explained one-half of the variance in the difference scores, R2 = .50, and this overall model was statistically significant, F(4, 37) = 9.37, p < .0005. (See the main effects column of Table 18 for the 49 regression estimates of this model.) As indicated by their sigrificant negative regression coefficients, the Post-AAC cluster scores for the open paradign independently decreased 1.81 and the Post-AAC cluster scores of synchronous paradigns independently increased .52 for every 1 point increase in the difference scores item the Post-AAC judgnent to the Ideal judgnent after partialling out the other variables' effects. Since the open and synchronous variables were sigrificant in the main effects model, an interaction term of open by synchronous was added to the third model to determine if the interaction was needed to explain the difference score changes. The interaction variable did not improve the model as it indicated little change in the proportion of variance explained, R2 = .47. This overall model was also statistically sigiificant, F (3, 38) = 11.34, p < .0005. (See the interaction column of Table 18 for the reg'ession estimates.) The interaction term of open by synchronous was not statistically sigrificant and was not used in subsequent modeling for this analysis. The Post-AAC synchronous variable also lost statistical sigrificance in this model, probably due to multicollinearity with the interaction term as indicated by their sigrificant bivariate correlation of .96. A simple regression model with Post-AAC open paradign predicting the change fiom the Post-AAC judgnents and the Ideal judgnents was also calculated, although the results from this interim model were not listed in Table 18. This simple model captured 37% of the variance in the outcome variable of cluster difference score. However, adding the Post-AAC synchronous paradign captured an additional 9% of the variance. Thus, Post-AAC cluster scores for open and synchronous paradign explains 46% of the variance in the difference score fiom the Post-AAC judgnent to the Ideal judgnent, R2 = .46. The overall model obtained statistical sigrificance, F (2, 39) =16.34, p < .0005. (See the final 50 column of Table 18 for the regression estimates.) The model generally appeared to meet the required error assumptions although the residuals are somewhat skewed and a few of the residuals violate heteroscedasticy. The fitted model was: change in family paradigm (total cluster score difi'erencefrom Post-AAC to Ideal) = 8.85 - 1 . 71 open cluster from the Post-AAC judgment + .4 7synchronous cluster from the Post-AA C judgment. Thus, this final model shows that the Post-AAC cluster score of open paradigns independently decreased 1.71 and the Post-AAC cluster score of synchronous paradigns independently increased .47 for every 1 point increase in the difference score while controlling the other variables' effects. Satisfaction as a function of paradigm. In this satisfaction analysis, the primary caregiver's satisfaction rating (i.e., strongly dissastisfied, dissatisfied, neutral, satisfied, strongly satisfied) was used to create a 5-point ordinal outcome variable. The following a priori hypothetical model was the starting point for this analysis: primary caregiver's satisfaction with child 's AAC = intercept +Aclosed cluster + Arandom cluster + Aopen cluster + Asynchronous cluster + Aseverity of child 's disability + AAAC ordinal variable + Afamily resources. Seven predictors were used in this model: one for each Post-AAC paradign and one for each of three possibly-confounding variables: child's disability, AAC use, and farme income. This model explained more than one fourth of the variance in the satisfaction scores, R2 = .28. However, the overall model was not statistically sigrificant, 51 F (7, 31) = 1.74, p = .14. (See the a priori model column of Table 19 for the regression estimates of this model.) As indicated by their nonsignificant regression coefficients, the potentially-confounding child, AAC, and family variables did not contribute to explaining the difference in AAC satisfaction. Therefore, these variables were not used in the subsequent multiple regession modeling for this analysis. Next, a main effects model was evaluated with the predictors of Post-AAC closed, random, open, and synchronous paradigns. The Post-AAC cluster scores explained approximately one fourth of the variance in the satisfaction ratings, R2 = .26, and this overall model was statistically significant, F(4, 37) = 3.19, p = .024. (See the main effects using the PA column of Table 19 for the regression estimates of this model.) As indicated by its significant negative regression coefficients, the Post-AAC cluster scores for the closed paradign decreased .63 for every 1 point increase in the primary caregiver's satisfaction with the child's AAC system after partialling out the other variables' effects. Satisfaction with the then current AAC system might also be related to the family's original, retrospective, Before-Diagnosis paradign. Accordingly, a main effects model was evaluated with the predictors of Before-Diagnosis closed, random, open, and synchronous paradigns. Before-Diagrosis cluster scores explained more than one fourth of the variance in the satisfaction ratings, R2 = .28, and this overall model was statistically sigrificant, F(4, 37) = 3.58,p = .015. (See the main effects using the BD column ofTable 19 for the regression estimates of this model.) As indicated by their significant negative regession coefficients, the Before-Diagnosis cluster scores for the closed paradign independently decreased .38 and the Before-Diagnosis cluster scores of the random paradign 52 independently increased .60 for every 1 point increase in the primary caregiver’s satisfaction with the child's AAC system after partialling out the other variables' effects. Satisfaction with the current AAC system could be related to the family's Ideal paradign in the same way. Accordingly, a main effects model was evaluated with the predictors of Ideal closed, random, open, and synchronous paradigns. The ideal cluster scores explained nearly one third of the variance in the satisfaction ratings, R2 = .30, and this overall model was statistically significant, F (4, 37) = 3.90, p = .01. (See the main effects using the I column of Table 19 for the regession estimates of this model.) As indicated by its significant regession coefficients, the Ideal cluster scores for the random paradign increased .62 for every 1 point increase in the primary caregiver’s satisfaction with the child's AAC system after partialling out the other variables' effects. The next model combined the statistically significant predictors fi'om the previous main effects models. These predictors included the cluster scores from Before-Diagnosis closed and random, from the Post-AAC closed, and from the Ideal random paradigns. These cluster scores explained over one half of the variance in the satisfaction ratings, R2 = .52, and this overall model was statistically significant, F(4, 37) = 9.83, p < .0005 . (See the combination column of Table 20 for the regession estimates of this model.) The Before-Diagnosis closed variable also lost statistical significance in this model, probably due to multicollinearity with the Post-AAC closed variable as suggested by their significant bivariate correlation. As indicated by their significant regession coefficients, the Post- AAC closed scores independently decreased .37, the Before-Diagnosis and the Ideal cluster scores of the random paradign independently irncreased .42 and .62, respectively, for every 53 1 point increase in the primary caregiver's satisfaction with the child's AAC system after partialling out the other variables' effects. Since the Post-AAC closed, the Before-Diagnosis random, and the Ideal random variables were significant in the combination model, four interaction terms (i.e., Post-AAC closed by Before-Diagnosis random, Post-AAC closed by Ideal random, Before-Diagnosis random by Ideal random, and Post-AAC closed by Before-Diagnosis random by Ideal random) were added to the next model to determine if the interaction terms were needed to explain the difference score changes. The interaction variables did not improve the model as indicated by the small change in the proportion of variance explained, R2 = .53. This overall model was statistically significant, F (7, 34) = 5.56, p < .0005. (See the interaction column of Table 20 for the regession estimates.) None of the interaction terms were statistically significant, and thus, were not used in subsequent modeling for this analysis. The final and most parsimonious model included only the significant predictors from the combination model described above. Consequently, the predictors included the cluster scores fi'orn the Before-Diagnosis Random, fi'om the Post-AAC Closed, and horn the Ideal Random paradigns. This final model of 3 predictors explained one half of the variance in the AAC satisfaction column, R2 = .50. The overall model was statistically significant, F(3, 38) = 12.53, p < .0005. (See the final column of Table 20 for the regession estimates.) Although the residuals were somewhat skewed, the model generally appeared to meet the required error assumptions. The fitted model was: primary caregiver's satisfaction with child '3 AAC = 1.21 - .4 7cIosed cluster from the Post-AA C judgment + .38random cluster fiom the Before- Diagnosis judgment + .65random cluster fiom the Ideal judgment. 54 Thus, this final model showed that the Post-AAC cluster scores for the closed paradign independently decreased .47 and the Before—Diagnosis and Ideal cluster scores for the random paradign independently increased .38 and .65, respectively, for every 1 point increase in the primary caregiver's satisfaction with the child's AAC system while controlling for the other variables' effects. Paradigm Resources and Goals The overall paradign cluster scores can also be broken into cluster scores for the resources, for the goals, and for the individual elements of time, space, energy, material control, affect, meaning, and content. Each individual element has its own cluster scores ranging from 0 to 5 for closed, random, open, and synchronous. (Table 21 lists the average paradign cluster scores across the four judgnents.) For the Before-Diagnosis judgnent, the average cluster scores for random and open paradigns were generally higher than closed and synchronous paradigns for the resource elements of time, space, energy, and material. The paradigns for the goal elements showed more variability: synchronous was highest for control, random was highest for affect, closed was highest for meaning, and open was highest for content. When comparing the Before-Diagnosis judgnent to the After-Diagnosis judgnent, the highest average cluster score for time shifted from open to closed. In addition, the cluster score for random decreased between the two judgnents. The highest average cluster score for space shifted slightly from random to open. While the highest cluster score for energy remained in the open paradign, random and synchronous paradigns showed decreasing average cluster scores. The goal elements showed less change between the two judgnents. While the highest average cluster score shifted fi‘om synchronous to random for 55 control and fi'om random to closed for affect, the highest average cluster scores for closed meaning and open content simply increased their size. Random meaning also saw a sizeable decrease from the Before-Diagnosis to the After-Diagnosis judgnent. The Post-AAC judgnents of the individual elements showed a trend towards values more similar to the Before-Diagnosis judgnents. Random and cpen were generally the highest cluster scores for the resource while the goal elements showed synchronous control; closed, random, and open affect; closed meaning, and open content. The Ideal judgnents included an average preference for open time, random space, open energy, and random material. For the goal elements, closed and open control; random and open affect; closed, random, open, and synchronous meaning; and open content were the highest average cluster scores. 56 CHAPTER 4 Discussion The chapter is divided into a discussion of the results, clirnical implications, study limitations, firrther research needs, and conclusions. The research involved an initial investigation of family paradign variables (Imig, 2000a) with a set of families of young AAC users. Caregivers were primarily white, non-Hisparnic, and married to the child's other biological parent. The families showed some diversity in caregiver education level, occupation, geogaphical location, and income but were more highly educated than the general population and, thus, had higher incomes. Children had a variety of diagnoses but the largest proportion of diagnoses (e. g., developmentally-delayed, mental retardation, and autism or pervasive developmental disorder) involved cognitive challenges for the AAC user. The heterogeneity of the children's language skills is shown in Table 12. Discussion of Results Post-AA C paradigms. The first research objective focused on the identification of family paradign(s) that were being used by the families at the time of the study. The range of AACF -PAS cluster scores from 0 to 5 represents the family's increasing use of that particular paradign from never to always. Most families used each of the four paradigns at least part of the time. In other words, very few families had a paradign cluster score of zero. The results in Table 15 and Figure 2 are limited only to those paradigns used almost always or always by the family (i.e., paradigns with cluster scores equal to 5). The Post-AAC results in Table 15 and Figure 2 revealed that the geatest percentage of families relied on open strategies, followed by random, then closed, and 57 finally synchronous strategies. Given the expectation that the open paradign might be most consistent with AAC recommendations, the high use of an open paradign within these AAC families was anticipated. It was originally thought that the closed paradign would be more frequent than the random paradign due to the compatibility of predictable routines in AAC intervention with closed strategies. However, that prediction was not supported by these results. Instead, the random paradign, which was employed by almost one half of the families, was the second most frequently reported paradign. In this sample, random was found most often in combination with other paradigns, the most common combination being a random-open paradign. Explanations for this might be that, regardless of which types of communication modes were used, the random-open family would be able to implement the necessary changes for AAC while continuing to value the unique voice of the child. Another explanation nnight be that the AAC families simply reflected the general population in the United States, for which a random-open paradign is the most likely paradign combination (Imig, 2000a, Ch. 9, p. 3) The closed paradign, which was less frequently used by families in this study, may be linked to unease with the use of nontraditional communication modes, such as sign, pictures, and VOCAs, as the primary communication vehicles. In addition, these closed families might have solved a misalignment between their closed paradign strategies and the AAC recommendations by decreasing their reliance on closed and increasing their use of other paradign strategies that they felt were a better fit. Few families indicated the use of a synchronous paradign, which was always in combination with at least one other paradign. The Iirrnited used of a synchronous paradign is consistent with Innig's report 58 (2000a, Ch. 9, p. 4) that synchronous elements are employed by a small portion of the general population in the United States. Another possibility is that synchronous families, who tend not to rely upon spoken communication for successful family functioning, would be less likely to seek out speech-language pathology services. It is unclear whether the paradign distribution found in this research is related to AAC factors, is simply a reflection of the paradign distribution for the general US. population, or exists for other reasons. While the theory of family paradigns is gounded in the assumption that the family's paradigns will affect all aspects of family's life including AAC, research is needed to confirm and expand this line of study. An example of this type of research might be a qualitative study where family members from different paradigns describe their reactions to different AAC recommendations and their subsequent decision- making processes about whetlner and how to employ AAC witlnin the child's life. Nearly half of the families in this study were combirning two or more paradigns used always or almost always by them. The four paradigns can be viewed as varying along two scales: 1) cohesiveness, running from connected to separate, and 2) adaptability, running from structured to flexible (Imig, 2000a, Ch. 2, pp. 1-3, Ch. 9). Closed paradigns are connected and structured while random paradigns are separate and flexible, making these direct opposites. Likewise, open paradigns are connected and flexible while synchronous paradigns are separate and structured, making these direct opposites of each other. The lack of commonality in the two scales gives the combinations of closed- random paradigns and open-synchronous paradigns the geatest tension in merging paradignatic views when compared to other paradign combinations (lrnig, 2000a, Ch. 9, 59 pp. 2, 4). Contrast that to paradigns that share a similar location on one of the scales. For example, combining closed (i.e., connected and structured) and cpen (i.e., connected and flexible) paradigns results in the farmly being compatible in its pursuit for feeling connected but having tension in finding its appropriate level of adaptability, from structured to flexible. Similarly, combirning random (i.e., separate and flexible) and synchronous (i.e., separate and structured) paradigns provides the family compatibility along the cohesiveness scale but creates tension along the scale of adaptability. F arnilies will typically need to expend more resources to accomplish their goals when their combinations of paradigns are further apart from each other (Nugent & Constantine, 1988). See Imig (2000a, Ch. 9, pp. 1-12) for his discussion regarding each possible paradign combination in compromise and compound family systems. Changes in paradigms After-Diagnosis and Post-AA C. The second research objective was to explore whether families perceived any changes in their paradigns over time. Most paradign difference scores were less than 2, and many of the paradign changes appeared to be an adjustment in how the families balanced their paradigns rather than an outright rejection of their initial paradign(s). For example, a family may have a difference score of l for random, representing a change from a cluster score of 5 to a 4. This would have indicated that the family switched fi'orn "always or almost always" using those strategies to "often" using those strategies. In order to increase our understanding of these numerical shifts in cluster scores, future research could ask families with children of different ages for examples of when and why they are making these changes. Some possible patterns could include 1) that families with young children often go through this paradign adjustment, 2) that this adjustment is 60 unique to families with children who have special needs or to families in crisis, or 3) that these changes are not clirnically significant. If families with children often make this adjustment, it could represent typical alterations that arise in the course of child rearing. If only families in crisis showed this change, it may represent changes imposed by the additional stressors of having a child witln severe disabilities. If families are unable to identify examples of changes, then the changes in paradign scores may not be related to clinically significant events in the family's life. In Table 15 and in Figure 2, the pattern of paradigns with a cluster score of 5 shows that reliance on random views decreased while reliance on closed views increased as the family moved from not having a diagnosis to having a diagnosis for their child. Then, as the family moved to having their most recent AAC system, the frequency of closed paradigns decreased as random and open paradigns increased. This shift is consistent with Imig's hypothesis (2000a, Ch. 3, pp. 12-14) that closed strategies, which are efficient and based on past successes, might be initially employed when confronting a problem. However, if the problem becomes long-term and the established solutions did not achieve the desired outcomes, then the adaptive strengths from random and/or open strategies might be more successful. Random strategies may spontaneously create new and innovative ways from which the individual family members can explore the problem. Open strategies would include seeking out information from all possible avenues and requiring communication and consensus from all family members on the next plan of action. These changes in family paradigns are consistent with a roller-coaster model of family crisis (Boss, 1987; Hill, 1949; Koos, 1946), which suggests that families become "disorgarnized" after a crisis, hit bottom, and then recover family functioning. The degee of 61 recovery or reorganization may be due to external factors, such as the type and severity of the crisis, and to internal factors, such as the family's paradignatic worldview, cohesiveness, and/or available resources (Constantine, 1986, pp. 181-205; Duis, Summers, & Summers, 1997; Haveman, van Berkum, Reijnders, & Heller, 1997; Imig, 2000a, Ch. 9, pp.1 1-12, Ch.10, pp. 1-26; McCubbin, McCubbin, & Thompson, 1993; McCubbin & Patterson, 1983). The family's reorganization may be reflected in a change of cluster scores, representing a rebalancing of the family's use of different paradigns. The pattern of retrospective judgnents from before to after the child's diagnosis (see Table 16) shows that families who started with higher random and open paradign scores had less overall change in their use of the four paradigns than those who innitially relied less on random and open views of the world. In addition, results from Table 17 suggest that families using open strategies after the child's diagnosis reported smaller difference scores at the point of receiving their most-recent AAC system. These results may mean that families find open and random strategies compatible with the transition into special education. Families with open paradigns would likely be comfortable with the consultative, collaborative teaming process that underlies much of special education and AAC intervention (American Speech-Language-Hearing Association, 2004). Those with random paradigns would likely be accepting of the child, regardless of the child's strengths and weakrnesses. In contrast, families with closed paradigns, looking for efficient and effective restoration to "normal," may be more reticent to accept the child's diagnosis and the subsequent changes in having "outsiders" participate in family decisions. 111 a similar way, those using synchronous paradigns may not see any need to change based on judgnents from those outside the family's view. Consequently, families with closed or 62 synchronous views may feel misalignment between their paradigns and the practices of educational systems. Further research is needed to explore these possibilities. Paradigm and AAC satisfaction. The third research objective focused on measuring family paradign satisfaction and AAC satisfaction as a function of family paradign. As shown in Table 18, families who used an open paradign after receiving their then current AAC system were less likely to want to change their paradign, as indicated by lower difference scores. In contrast, families who used a synchronous paradign were more likely to have higher difference scores, suggesting that they would like to change their Post-AAC paradign. This is consistent with the earlier suggestion that AAC recommendations may fit open family furnctioning better than synchronous family functioning. Predicting AAC satisfaction was the focus of the results in Tables 19 and 20. Families, who relied on random strategies before the child was diagnosed and/or who would prefer to use random strategies in an ideal world, reported more satisfaction with the then current AAC system. In contrast, families who used closed strategies at the time of the then current AAC system reported less satisfaction with that system. Although further research is needed to replicate and explore these differences, the random and closed satisfaction differences may reflect how the family perceives the acceptability of alternative communication modes. As seen in Table 11, three fourths of the primary caregivers indicated satisfaction with the child's AAC system. In transcripts from interviews and focus goups (V anBiervliet & Parette, 1999), family members have expressed the hope that the AAC system will allow their children to communicate more with the important people in their lives. However, it is difficult to knnow how quickly family members expect the AAC user's communication to 63 improve. Comparing family members' expectations and possible benchmarks with their family paradign may firrther clarify this issue. For example, closed families might expect faster changes and be more quickly disappointed if the changes did not occur in a timely manner. On the other hand, open families might find incremental improvement encouraging and talk about practical ways to adjust the plan. AAC early intervention practices. Although not a primary focus of this study, these results provide some interesting observations about AAC practices in early childhood. The parents in this study generally suspected problems early in their children's lives. In fact, nearly half of the children were under 1 year of age when symptoms of a disability arose. The subsequent diagnosis generally occurred during the first, second, or third year of the child's life. This raises the question of whether early AAC intervention was primarily implemented for children with obviously severe disorders or if this was simply an artifact of the convenience sample of this study. The length of time for the acquisition of the first AAC systems varied. The average of one half year from diagnosis to first AAC acquisition was negatively skewed by families who used baby sign with their child before any problems were diagnosed. Other families saw five years pass between the diagnosis and the first AAC system. However, the median was still less than one year, indicating that some type of AAC system was often available within a relatively short period of time. However, because speech and language undergo a period of rapid gowth during early childhood, the effect of any delay in implementing various AAC systems could be considerable. This issue requires further research. In this study, sign language was the AAC mode used most often or always at home, although most of the AAC users had multimodal systems, including vocalization, gestures, 64 pictures, and simple and complex VOCAs. Even though more than half of the AAC users had voice output AAC systems, less than half of those with voice output devices reported fiequent use at home. It is unclear from this study if the families were satisfied with that level of use or if they were expecting increased use as the child gew more proficient with their language and AAC competencies. The lack of consistency of AAC components and modes recommended for young AAC users is not surprising, since little evidence-based information currently exists to guide clirnicians in choosing various AAC components. Furthermore, there is also limited evidence on various AAC components' resulting effects on speech and language development (Schlosser, 2003). Recommendations for using unaided or aided AAC as well as the role of voice output continue to be debated and researched (Bedrosian, 1997; Bondy & Frost, 2002; Kumin, 2003; Locke & Levin, 1999; Mirenda, 2003; Romski & Sevcik, 2003; Schlosser & Blischak, 2001; Sigafoos & Drasgow, 2001; Smith & Grove, 2003). Eleven of the families used only unaided AAC in the form of sign language. When asked to give reasons why they were not using aided AAC, four of the answers were troubling: 1) child not ready and/or working on skills, 2) can't afford AAC or AAC not available, 3) not told, and 4) don't krnow why. The idea of a lack of aided AAC readiness implies that some prerequisite skills for AAC are necessary. This idea has generally been rejected in the AAC field (American Speech-Language-Hearing Association, 2004; National Joint Committee for the Communication Needs of Persons With Severe Disabilities, 2002). In addition, the lack of funding or adequately-trained personnel cannot be used as justification by educational systems for not providing appropriate AAC services (National Joint Committee for the Communication Needs of Persons With Severe 65 Disabilities, 2002) although families and professionals may disagee on what constitutes appropriate AAC services. As a standard of practice, family members should receive information about the various aided and'unaided AAC options. This information may be especially valued by open families. In conversations with the author, several primary caregivers expressed dismay at the length of time that passed before they becanne aware of the variety of aided AAC options that exist. Clinical Implications Family paradigns theory has been found useful in couples and family therapy (Constantine, 1986, 1993a; Constantine & Israel, 1985; Irrnig, 2000a). After the PAS is completed by family members, the results are used as a starting point from which the professional will ask for farmly verification and examples of the different paradign strategies (Imig, 2000a, Ch. 9, p. 6, Ch. 11). In a similar way, a speech-language pathologist could ask a couple or a family to each complete the AACF-PAS. By looking for patterns in the resulting paradign cluster scores, the speech-language pathologist has a focus for the initial discussions on how different AAC systems and strategies may complement or be in conflict with the family's paradign(s). This process will not supplant the need to talk with family members. Although the family paradign scores provide the professional with insight on possible family issues to address, the farme members will always be asked to validate or correct those initial insights with their own interpretation and examples. Consider this illustration: When completing the AACF—PAS, a mother indicated two cluster score changes from her current to ideal functiong: Random decreased from 5 to 4 while open irncreased ham 4 to 5. The speech-language pathologist, recognizing the pattern 66 of a possible decreased emphasis in individual change to an increased emphasis in goup change, would explain this result and ask the mother if that was how she felt. The motlner might explain that she has been searching the Internet and talking to many other mothers who are in a similar situation (Note: use of a random strategy), but that she feels her husband does not want to discuss what she has learned (Note: rejection of an open strategy). Thus, the speech—language pathologist would ask if the mother feels that it is time to bring people together to make a decision (Note: use of an open strategy). If the mother agees, the mother may be signaling a need to move fiom random paradign information gathering to open paradign consensus-building. At that point, the mother probably does not desire additional information from the professional but may want the professional to facilitate a decision-making process. Of course, before proceeding, the speech-language pathologist would need to know if the husband as well as other important team members are willing to adopt this open paradign strategy of goup consensus. In this example, the speech-language pathologist continued to perform typical professional duties. However, due to the patterns raised by the AACF-PAS, the speech- language patlnologist was able to explore areas suggested by family paradigns that were likely to be important to this specific family member. Consequently, with family paradigns, the professional is able to provide family-centered intervention targeted to each family's individual pattern of resources and goals. By considering the goodness-of-fit for the AAC system within a particular family's manner of functioning, families and professionals may be able to improve their informed decision-making and obtain more compatible AAC systems for children with severe communication disorders. 67 Paradigm resources and goals. Although not the target of a specific research objective for this study, the paradign changes for individual elements of time, space, energy, material, control, affect, meaning, and content may be useful in further specifying the effects of particular AAC recommendations on individual families. In this study, families' overall paradign scores did not predict the importance of the individual elements for each family. Consequently, families and professionals will need to discuss each family member's scores from the elements to determine their possible effect on AAC recommendations. Imig (2000a, Ch. 11-12) has provided detailed instructions and examples for using the individual element scores fi'om the R-PAS to help couples better understand their relationships. Using Irnig's methodology, AAC case studies could be employed to document and evaluate similar procedures during AAC evaluations and implementations. For example, imagine that a speech-language pathologist is working with a family whose 3-year-old has autism, resulting in a severe communication disorder. If the father indicated a high cluster score on the elements of synchronous meaning and on closed time (see Tables 1 and 2 for a summary of paradignatic elements), the speech-language pathologist would want to check with him and other family members on how those elements are combirned within the family. The speech-language pathologist would recognize that the pattern of synchronous (i.e., individual-continuity) and closed (i.e., goup-continuity) suggests that this father has more movement fi'om separate to connected on the cohesiveness scale but is likely on the structured side of the adaptability scale. Therefore, the speech-language pathologist might speculate that the synchronous meaning score suggests that this family implicitly knnows what it values, while closed time suggests 68 that this family member wants time to be used in an efficient manner as determined by the head of the family. If these speculations are confirmed by the family, several possible AAC implementation topics would need exploration in light of this family's paradignatic view of the world. First, questions about the meaning of the child's severe communication disorder to the family would be important to discuss. The difficulty with this task is that the family members may see little reason to discuss their synchronous meaning with each other (because they already implicitly in ageement) and with the speech-language pathologist (who likely will be considered outside the family's boundaries). However, the speech- language pathologist realizes that the family is less likely to use their resources and goals for AAC implementation if the fannily does not have concerns about the child's communication skills. Thus, the speech-language pathologist must decide with the family if it is willing to look at its view of meaning. If not, they would then need to discuss the family's willingness to implement AAC at home and/or its preferences for AAC implementation within contexts outside of the family such as in the child's preschool. See Constantine and Israel (1985) for a case study that describes some of the challenges, opportunities, and actions they encountered with a synchronous family in family therapy. Second, if the family is willing to implement AAC at home, the closed time element will need to be explored, especially with whomever is the head of the household (i.e., the mover). Using family paradign theory, the speech-language pathologist will suspect that this mover is the gatekeeper for decisions about time use within the family. The mover will need to be asked about whether and how AAC recommendations could fit within existing routines without requiring an excessive amount of the family's time 69 resources. This process of identifying specific elements and discussing how they may influence AAC decision-making would continue especially for those elements that are higher ranked by the family. [See Innig (2000a, Ch. 12) for details on using ranked paradign data within vector charts] The speech-language pathologist would never suggest that a synchronous (or a closed, random, or open) family will be unable to implement AAC. Instead, the professionals and the family members need to investigate where tensions could exist between the AAC recommendations and the fannily's preferred functioning. Being able to recognize any misalignment between the AAC implementation plans and the family paradigns should lead to trying to minimize the misalignment. If that is not possible, then, at least, everyone should be aware that the tensions from the misalignment will likely feel difficult to the farme and will require more of the family's goals and resources to manage. If the original AAC implementation is not successful, the AAC team would ask the family for examples of the problems in order to further understand and minimize misalignments between the family's preferred paradign(s) and the AAC system. Note also that the speech-language pathologist's role in these examples is not to try to change the family's way of firnctioning, which is a difficult to impossible task. Instead, the role is to use the krnowledge of communication disorders intervention and of family paradigns to work with a particular farmly and its child to uncover how AAC and communication irntervention fit within what is already important to them, the very definition of being a farnily-centered professional. If this process is successful, any resulting AAC recommendations for the child should then be more naturally tailored to the 70 family's real world and, thus, be more likely to be implemented. Clinical case studies would provide important insight into the outcomes of this type of approach. Family paradigms' effects on professional decision-making. Constantine (1986, pp. 390-392) proposed that professionals should become aware of their own paradign preferences in order to guard against negatively judging other people's paradign strategies. Some people may mistakenly believe that their own family furnctioning or paradign should be adopted or avoided by their clients as the correct or incorrect model of family behavior. Professionals should also consider whether they use a particular paradign as an example of "appropriate" professional behavior. For example, some professionals may prefer a more expert or medical model where they diagnose a problem and prescribe a treatment. This may be consistent with closed paradign behaviors. Otlner professionals may prefer an individualistic approach where the evaluation and treatment emerges as a creative product of this particular session. Such an approach appears to be consistent with a random paradign. Still other professionals may prefer a more collaborative model where they assist communication and consensus among team members. This would be consistent with an open paradign. Synchronous professionals would already be in tune with their clients without the need for overt communication, creating a harmonious and peaceful connection. The most effective professional is likely to be someone who, while recognizing his or her own paradignatic preferences, is able to shift his or her strategies to validate and align with the clients' preferred paradigns. Constantine (1991; 1993b) has also written about paradignatic orientations within orgarnizations and teams. This area of research could also be applied to the different environments and orgarnizations with which the AAC user and the family interacts. 71 Similarly, Villarruel et a1. (1995) proposed that early childhood educators should be educated about and subsequently validate the different paradignatic views that are held by fannilies. In addition, educators are then faced with finding ways to embrace and encourage all paradignatic viewpoints that their students bring into the classroom. For example, one preschool classroom could rely on more closed paradign strategies such as explicit rules and structured routines. A different preschool classroom might employ random paradign strategies such as encouraging each child to follow his or her interests. Another preschool classroom might use open paradign strategies such as collaborative planning of the day's activities and the use of negotiation to resolve disageements. In contrast, a preschool classroom employing synchronous strategies might have few written policies but would expect the children to fit into the natural rhythm of the place. If the parents' and child's paradign is different (i.e., misaligned) from the teacher’s paradign, more stresses are likely to occur. Without understanding that the stresses are coming from holding different views of the world, the family and teachers may find the tensions difficult to resolve. However, with an understanding of family paradigns, the teacher could better understand his or her own teaching style as emanating fi'om a particular paradignatic view and then be able to consider and implement practices that support and value all of her students' paradigns (V illarruel et al., 1995). More research is needed on the interaction effects of child, family, professional, and organizational paradigns within clinical intervention and special education practices. 72 Limitations of This Study Although the present research demonstrated that family paradigns accounted for a significant amount of the variances in paradign difference scores and AAC satisfaction ratings, several linnitations need to be considered. Caution in interpreting multiple regession models is prudent as multiple regession will indicate only associative, not causal, relationships. Each of the final models described in Chapter 3 is underspecified. Therefore, additional variables need to be evaluated in order to improve the models to better predict changes in paradign preferences and AAC satisfaction. Looking at different variables and using additional analysis tools such as structural equation modeling, hierarchical linear modeling, and logistic regession may improve the statistical models and increase our understanding of the relationships between family variables and AAC intervention. However, adding variables generally will require larger sample sizes, which are difficult and time-consuming to obtain on low-incidence populations such as AAC users. The use of three long questionnaires, requiring 60 to 90 minutes of time, likely limited those who decided to participate in this study. For example, stay-at-home caregivers may feel they have more time to participate in research and may be over- represented in this sample. Conversely, some researchers have proposed that, if possible, a parent, usually the mother, will try to stay home when the child(ren) are young (Heller, Hsieh, & Rowitz, 1997, p. 407; Huston & Vangelisti, 1995, p. 158). If that is true, then stay-at-home caregivers may be the majority, especially in two parent families of young children. Additional demogaphic research of AAC families could clarify this issue. 73 Some caregivers did not complete the AACF-PAS, which limited the paradign results. Reasons included finding the directions confusing and the measure too tinne- consuming. These points should be considered in developing future versions of the AACF- PAS. Although the results from this study suggest that many families perceive paradign changes fi'om prior to and after the child's diagnosis, the AACF-PASs methodology is limited by its retrospective nature for the Before- and After-Diagnosis judgnents. Those results could be verified by finture longitudinal studies. For example, a longitudinal study could follow at-nisk children (e.g., neonatal intensive care survivors) from birth through the preschool years to determine any changes in family paradigns if communication disorders and/or AAC intervention occur. A longitudinal study would also be useful for detailing changes in the type and the use fiequency of different communication modes during the birth to kindergarten time. The lack of demogaphics regarding young AAC users, their families, and their environments limits knnowing how representative this convenience sample is of the general population of young AAC users. It also limits generalization. Demogaphic studies that would increase our knowledge about young AAC users and their natural, human- constructed, and social environments continue to be needed. The resulting demogaphics would, then, improve sampling strategies for additional studies about these young children and their families. Little is known about paradign changes in families with typically-developing young children. Perhaps families with any young children demonstrate similar perceptions of paradign change that were seen in this study. Designs that include and compare families 74 who have typically-developing young children and those who have young children with AAC needs would firrther clarify this issue. Future Research Needs Much work is needed in extending and evaluating this initial investigation. Some of the needs are specific to testing interactions between family paradigns and AAC practices, some suggestions are focused on improving our tools and krnowledge of family paradigns, and some proposals are specific to improving our understanding of the demogaphics of families with young AAC users. In addition to the qualitative and quantitative research ideas discussed earlier, the influence of the player parts of mover, challenger, follower and bystander (which was introduced briefly in Chapter 1) needs to be considered and researched in AAC decision- making. Interactions between paradign type and player part within the family is one likely area of study. Another potential line of research is whetlner an individual maintains his or her paradign and/or player part witlnin different contexts such as the home, in medical or educational settings, and in other community environments. Cultural and social class influences on family paradigns are another area that needs further study. The samples gathered in this study were generally fi'om white, non-Hispanic females who had more college education and higher incomes than the general US. population. Although this sampling bias was not intended, it reflected the type of persons who responded to the survey. Therefore, future research can be driven by the question of whether the family paradigns used in this study are equally applicable to or generalizable to families in other cultural goups and social strata. 75 Villarruel et a1. (1995) discussed some of the contextual effects of blending culture and fannily paradigns. One cannot predict a family's paradign by simply observing the family's behavior because the same behavior could be motivated by different underlying reasons or different behaviors could result from the same underlying reason. They proposed that although family paradigns can exist within every culture, the family's choice of behaviors resulting from a given paradign may differ due to cultural influences. One of their examples contrasts two families, both using a closed paradign to gieve the deatln of a family member. The European American family members were struggling not to cry while the Lebanese family members were wailing loudly. While both families employed traditional gieving behaviors emanating fi'om their closed paradign, the resulting behaviors appeared different due to their different cultural backgounds. Although culture and family paradigns are integally connected, Stocknnan, Boult, and Robinson (2004) cautioned that cultural influences are seldom monolitlnic. Instead, people from different cultural goups may share some experiences and values but not others. Similarly, people from different social classes may share some features but not others. A common example is that two people may share a particular ethnic backgound but have differing educational experiences. Another example is that two people can have a similar social class but different values about the role of money in their lives. In light of the complexity of defining culture and social class, researchers need to recognize the multiple bases on which people form shared connections when exploring whether family paradign theory as applied in the present study has cross-cultural and social relevance. In addition, professionals must be sensitive to multiple contextual influences, including the culture in which family paradigns exist, that affect a child's development. This contextual approach 76 to culture and family paradigns needs to be integated with the gowing body of literature on culture and AAC. (See the concept paper contained in VanBiervliet & Parette, 1999, for a review of some of the cultural issues in AAC.) In order to increase the firnctionality of the AACF-PAS, several areas should be addressed, including improving its ease of use, measuring the test-retest reliability, simplifying instructions and possible scale items, and creating easily-understood results summaries for family members. Having multiple caregivers within the same family complete the AACF-PAS should improve the understanding of that particular family's functioning. Getting all the caregivers to complete the surveys can be difficult logistically. More research is needed on how to quantify if the family system is compromise or compound. Conducting demogaphic research that describes the characteristics of young AAC users, their families, and their environments could provide important insights into the use of AAC within early intervention as well as improve sampling strategies for future research projects. For example, determining the percentage of different communication modes recommended and implemented for young children with different diagnoses as well as the underlying rationales would further expand our understanding of current practices. Although using multimodal systems is a widely accepted AAC practice (American Speech-Language-Hearing Association, 2004; Sigafoos & Drasgow, 2001), limited research exists on which AAC components to combine. Irn addition, little is knnown about the effect of order and timing when introducing new communication modes. As seen in Table 8, three fourths of the AAC users in this study had acquired a new AAC system with an average of a year between the first and the most recent systems. However, the time 77 varied fiom less than a year to four years between the two events. Unfortunately, details about any AAC introduced but subsequently discontinued was not collected in this study. In addition, this research did not capture the AAC acquisition order and underlying rationale for recommending specific AAC systems, which would be useful to consider in future research. Possible options include a) starting with one mode and adding to it as the child shows communicative competence, b) starting with one mode and adding to it if the commnunication environment warrants a different approach, c) moving from unaided to aided modes, (1) moving from aided to unaided modes, e) discontinuing one mode and starting another mode, and 0 starting several modes at the same time. In addition, families with young AAC users could be included in studies that also collect data from farrnilies of typically-developing young children. Wlnile increasing our knowledge about the range and change of family functioning in general, the data could also be compared and contrasted to families whose young children exhibit severe commrunication disorder. Conclusions As seen in this initial research, family paradign theory may provide a useful framework witlnin which to research and clinically intervene with families of young AAC users. The different worldviews of closed, random, open, and synchronous may lead to differing familial strengths, barriers, needs, and supports. For example, finding ways to increase AAC satisfaction, especially with closed and synchronous paradigns, may be necessary if the models from this initial study are replicated. The influence of different family elements of time, space, energy, material, control, affect, meaning, and content need to be further explored within AAC families arid their daily lives. 78 AAC research and clinical intervention require the considerations of multiple factors including the characteristics of the AAC user, the AAC system, the family, other commtunication partners, and communication environments. Family paradigns theory extends the family-centered literature by providing an efficient framework for discussing family structure and for improving family-centered practice. AAC clirnical interventions that employ the theory of family paradigns also need to be explored. In conclusion, the recognition that families make decisions based on different, but equally valid views of the world allows the multi-pardignnatic professional to tailor recommendations to a specific family. Ultimately, identifying the paradigns of professionals, clients, their families and significant others, and their environments has the potential to improve clinnical services for all people with communication disorders, including AAC users. 79 APPENDICES 80 APPENDIX A Tables and Figures 81 Table 1. Comparison of resources by paradign. Paradign Closed Random Open Synchronous elements Resources Time Plan and follow Spontaneous and Balance Subconsciously schedule individually individual and understood changing goup needs Space Structured with Flexible with Practical; Integated and only traditional any and all ideas exploring ideas seamless ideas accepted acceptable as long as any conflict is resolved Energy Constant, Fluctuating; Changing and Peaceful and predictable flow enthusiastic flow adapting to harmonious individual and flow goup needs Material Important and Avoided when Practical and Preserved for valued, possible because useful their inherent symbolizes they can value achievement and complicate status relationships Note: Descriptions based on Constantine (1986) and Imig (2000a). .82 Table 2. Comparison of goals by paradign Paradign Closed Random Open Synchronous elements Goals Control By organization Individually- Discussion Innplicitly know and structure, determined leading to how and when discipline consensus action to complete Affect Private, Spontaneous, Sensitive, Understood conventional demonstrative responsive without words and actions Meaning Found within Found within Found by doing Found by “traditional” following one's what is effective following values own instincts and pragnatic timeless and path universals Content Time-tested rules Individually- Ask, share, and Absolutes, created, relative agee, "what is, is" constructionist Note: Descriptions based on Constantine (1986) and Imig (2000a). 83 SEQ: 35 =: 80.8 8 3x8 225 3:88:81 .0820 .3”: 5 women. was :va flonBeU 5:83 =<= 5:38 05 .5.“ 338.. can 95 .0m 05 Co 8: we own 3.. 8:38: o: a 2 0.5 .0829 .38 588 £5 3238 5s, 8:25 a case as 8.865 widows. daze o _ c c o o 3:88. oz o a o o c _ osaflafiaaz .v _ m 3. mm 9 m m 323E§8o=ao 58:25. o o o o c o 5o£<§u£m cogs—m— 2203 550 :o 5:93: m N _ a m m 2:233 ”seesaw 03:: :33? 8 SEE: _ _ _ _ N N SEE— :aotog need me 3 o c nm cm 222 o c we em 2 e. BEE...— 5:50 who» av Emu» am can» 3 23% mm £69: :. 332: we 83::er memo» cm E3» om Eco» mm Eco» mm 332: 2 «5:2: 2 8:832 mas» o Eco» N. E8» m So» e «5:2: 2 280:. 2 5:35: 33.5% was» on fine» on fine» mm E8» mm 332: 3. «6:2: 3 5:52 ow< Eco» hm was» am who» mm Emu» em «5:2: me £89: 3 :82 a Nvuz mvnz mvnz Om kuz on. 3.1.2 «:8: U<< nmnz a 69 Boxwoao e an: Eozwoeeo Bum: U<< aid .65. Secosm =< aid 2% SE: .2 widest .2 horse ozea> £03380 :5 £8: U<< mo DEE? e5. doe: aegow .ow< .m Bank 84 Table 4. Caregivers' education level and employment Variable Category All Primary AA C F -PAS All Secondary AA C F -PAS Caregivers (PC) 3 Caregivers (SC) 3 N=54 PC N=54 SC N=42 N=42 Highest Grade school only 0 0 O 0 educational Some high school 0 0 2 1 level attained High school graduate 5 4 5 5 Vocational school 1 l 0 0 graduate Some college 12 9 ll 7 College graduate 13 11 l6 14 Some gaduate 9 7 3 3 school Advanced degee 14 10 16 12 No response 0 0 l 0 b Business owner/ 5 3 7 5 Employment Self-employed Full-time employee 10 7 _ 41 33 Part-time employee 11 7 2 1 Stay—at-home 33 27 2 l caregiver Student ~ 3 3 l l Unemployed l l 0 0 No response 0 0 l 0 a Note. AACF-PAS indicates the subset of fannilies with completed family paradigm data. b Note. Respondents were asked to select all that apply. 85 Table 5. Family size, language(s) spoken, income, and geogaphical location Vanable Category All Fanulnes AA C F- P A S Farrnilies a N=54 N=42 Number of siblings Zero l6 13 One - 19 13 Two 13 l 1 Three 5 Four 1 1 Number of children with One 47 37 a disability Two 5 3 (including AAC user) Three 2 2 Number of adults living One 1 l in the house Two 50 39 Three 2 2 Four 1 0 Languagf spoken English 54 42 Spanish 1 0 at home Other 0 0 Family income Under $ 10,000 0 0 $10,000 to 19,999 3 3 $20,000 to 39,999 5 5 $40,000 to 59,999 14 10 $60,000 to 99,999 15 9 $100,000 to 149,999 10 8 Over $150,000 4 4 Prefer not to answer 3 3 Geogaphical location Farm 1 0 (population density) Rural, but non-farm 7 7 Town, not part of a 12 12 metro area (under 50,000) Small city 10 6 (50,000 to 99,999) Medium city 10 6 (100,000 to 299,999) Large city 14 l 1 (over 300,000) a Note. AACF-PAS indicates the subset of families with completed family paradign data. b Note. Respondents were asked to select all that apply. 86 Table 6. Child's diagnosis, the severity of disability on family's daily activities, and child's educational setting Variable Category All AA CF -PAS AAIE=Islgers AAC Users a - N=43 Diagnosisb Angelrnan syndrome 2 2 Apraxm 18 14 Autism or Pervasive developmental disorder l6 l4 Cerebral palsy 9 7 Developmental delay 33 25 Down syndrome 9 5 Dual sensory irrnpairment (deaf-blind) 0 0 Dysarthria l 1 Hearing impairment-temporary 7 4 Hearing impairment-permanent l 0 Medically fragile 6 2 Mental retardation 18 13 Rett syndrome 2 2 Rubenstein Taybi syndrome 3 3 Seizure disorder 7 6 Sensory integration disorder 2 2 Traumafic brain injury 0 0 Visual impairment not corrected by glasses 2 1 Cause unknown 3 3 Other neurological disorder 4 4 Severity of the Mild 8 6 child's disability as it Moderate 31 25 affects the family's Severe 16 12 daily activities Educational settingb Home visits by early interventionists 24 20 Self-contained preschool 27 23 Integated preschool 23 18 None 1 1 Other 8 7 a Note. AACF-PAS indicates the subset of families with completed family paradign data. b Note. Respondents were asked to select all that apply. 87 Table 7. Frequency of child's age when problem suspected, diagnosed, and AAC received Variable Category All AA CF -PA S AAC AAC Users Users ' N=55 N=43 Child's age Less than 1 year old 27 21 when problem 1 year through 1 year, 11 montlns 22 20 first suspected 2 years through 2 years, 11 months 6 2 3 years through 3 years, 11 months 0 O 4 years through 4 years, 11 months 0 0 5 years through 5 years, 11 months 0 0 6 years through 6 years, 11 months 0 0 Child's age Less than 1 year old 17 11 when diagnosed 1 year through 1 year, 11 months 15 14 2 years through 2 years, 11 montlns 17 14 3 years through 3 years, 11 months 3 2 4 years through 4 years, 11 months 1 0 5 years through 5 years, 11 months 0 0 6 years through 6 years, 11 months 0 O No response 2 2 Child's age Less than 1 year old 9 4 when AAC 1 year through 1 year, 11 montlns l6 14 film acquired 2 years through 2 years, 11 months 18 16 3 years through 3 years, 11 months 10 8 4 years through 4 years, 11 months 0 0 5 years through 5 years, 11 months 2 l 6 years through 6 years, 11 months 0 0 Child's age Still using first AAC system 15 10 when current Less than 1 year old 0 0 AAC acquired 1 year through 1 year, 11 months 4 3 2 years through 2 years, 11 months 14 ll 3 years through 3 years, 11 months 10 8 4 years through 4 years, 11 months 11 10 5 years through 5 years, 11 months 1 1 6 years through 6 years, 11 months 0 0 a Note. AACF-PAS indicates the subset of fannilies with completed family paradigm data. 88 Table 8. Length of time from when problem suspected, diagnosed, and AAC received Variable All AA CF -PAS AAC AAC Users Users Length of time from when N 55 43 problem first suspected to first Mean 1.1 years 1.2 years AAC system . Median 1.0 year 1.0 year Standard deviation 1.1 years 1.1 years Minimum -1 year -1 year Maximum 5 years 5 years Length of time from when N 53 41 diagnosed to first AAC system Mean 0. 5 year 0.6 year Median < 1 year < 1 year Standard deviation 1.1 years 1.1 years Minimum -2 years -1 year Maximum 5 years 5 years Length of time fiom when AAC N 40 33 first acquired to current AAC Mean 12 years 1.2 years ”Stem Median 1.0 year 1.0 year Standard deviation 0.9 years 1.0 years ' Minimum < 1 year < 1 year Maximum 4 years 4 years Length of time from when N 40 33 problem first suspected to current Mean 22 years 2.2 years AAC system Median 2.0 years 2.0 years Standard deviation 1.1 years 1.1 years Minimum 1 year 1 year Maximum 5 years 5 years Length of time fiom when N 38 31 diagnosed to current AAC system Mean 15 years 1, 5 years Median 2.0 years 1.0 years Standard deviation 1.1 years 1.2 years Minimum < 1 year < 1 year Maximum 4 years 4 years a Note. AACF-PAS indicates the subset of farrnilies with completed family paradigm data. 89 Table 9. Type of AAC system currently available Variable Category All AA CF -PAS AAC AAC Users U a sers N=55 N=43 b Sign Language Alone 1 I 8 In combination with communication 7 4 pictures and/or boards In combination with communication books 0 0 In combination with simple VOCA l 1 In combination with complex VOCA 3 3 In combination with communication 4 3 pictures, boards and/or books In combination with communication 9 7 pictures, boards, and/or books AND with simple VOCAs In combination with communication 6 6 pictures, boards, and/or books AND with complex VOCAs In combination with simple VOCAs AND 1 1 with complex VOCAs In combination with communication 3 3 pictures, boards, and/or books AND with simple VOCAs AND with complex VOCAs Total sign language users 45 36 a Note. AACF-PAS indicates the subset of families with completed family paradigm data. b Note. Respondents were asked to select all that apply. 90 Table 9. (continued) Variable Communication picturels; and/or boards Category Alone In combination with sign In combination with communication books In combination with simple VOCA In combination with complex VOCA In combination with sign AND communication books In combination with sign AND with simple VOCAs In combination with sign AND with complex VOCAs In combination with communication books AND with simple VOCAs In combination with simple VOCAs AND with complex VOCA In combination with sign AND with communication books AND with sirrnple VOCAs In combination with sign AND with communication books, AND with complex VOCAs In combination with sign AND with simple VOCAs AND with complex VOCAs In combination with communication books AND with simple VOCAs AND with connplex VOCAs In combination with sign AND with communication books AND with simple VOCAs AND with complex VOCAs All AAC Users N=55 l 7 1 AA CF -PAS AAC 3 Users N=43 l 4 l Total communication pictures and/or boards users 37 29 aNote. AACF-PAS indicates the subset of fannilies with completed family paradigm data. b Note. Respondents were asked to select all that apply. 91 Table 9. (continued) Variable Category All AA CF -PAS AAC AAC Users Users a N=55 N=43 Communication Alone 0 O b°°k5 In combination with sign 0 o In combination with communication 1 1 pictures and/or boards In combination with sirnnple VOCA 0 0 In combination with complex VOCA 0 0 In combination with sign AND with 4 3 communication pictures and/or boards In combination with conununication 1 1 pictures and/or boards AND with simple VOCAs In combination with sign, communication 4 4 pictures and/or boards, AND with simple VOCAs In combination with communication 1 1 pictures and/or boards AND with simple VOCAs AND with complex VOCAs In combination with sign, communication 2 2 pictures and/or boards AND with simple VOCAs AND with complex VOCAs Total communication book users I 3 12 a Note. AACF-PAS indicates the subset of families with completed family paradign data. b Note. Respondents were asked to select all that apply. 92 Table 9. (continued) Variable Simple VOCAs Category Alone In combination with sign In combination with communication pictures and/or boards In combination with communication boards In combination with complex VOCA In combination with sign AND with communication pictures and/or boards In combination with sign AND with complex VOCAs In combination with communication pictures and/or boards AND with books In combination with communication pictures and/or boards AND with complex VOCAs In combination with sign, communication pictures and/or boards, AND with books In combination with sign, communication pictures and/or boards AND with complex VOCAs In combination with communication pictures, boards, and/or books AND with books AND with connplex VOCAs In combination with sign AND with communication pictures, boards, and/or books AND with books AND with complex VOCA All AAC Users N=55 0 l AACF—PAS AAC a Users N=43 0 l l Total simple VOCA users [8 l6 a Note. AACF-PAS indicates the subset of families with completed family paradigm data. b Note. Respondents were asked to select all that apply. 93 Table 9. (continued) Variable Complex VOCAs Category Alone In combination with sign In combination with communication pictures and/or boards In combination with communication boards In combination with simple VOCA In combination with sign AND with communication pictures and/or boards In combination with sign AND with simple VOCAs In combination with communication pictures and/or boards, AND with simple VOCAs In combination with sign, communication pictures and/or boards, AND with books In combination with sign, AND with communication pictures and/or boards AND with simple VOCAs In combination with communication pictures and/or boards AND with books AND with simple VOCAs In combination with sign AND with communication pictures and/or boards AND with books AND with simple VOCA All AAC Users N=55 2 AA CF -PAS AAC a Users N=43 l Total complex VOCA users l9 l6 a Note. AACF-PAS indicates the subset of families with completed family paradigm data. b Note. Respondents were asked to select all that apply. 94 Table 10. Usage of communication modes at home Communication Used at home? All AAC Users AA CF -PAS AAC Users a Mode N=55 N=43 Vocalizations Never 2 2 Rarely 5 4 Sometimes l6 l3 Often 24 18 Always 8 6 Not available 0 0 Speech or speech Never 12 10 approximations Rarely 10 7 Sometimes 16 13 Often 16 13 Always l 0 Not available 0 0 Pointing or Never 7 7 gesturing Rarely 7 5 Sometimes 10 9 Often 24 17 Always 7 5 Not available 0 0 Sign Language Never 0 0 Rarely 5 4 Sometimes 10 8 Often 26 21 Always 4 3 Not available 10 7 Communication Never 4 4 pictures and/or Rarely 9 6 boards Sometimes 15 ll Often 8 7 Always 1 1 Not available 18 14 a Note. AACF-PAS indicates the subset of families with connpleted family paradign data. 95 Table 10. (continued) Communication Used at home? All AAC Users AA CF -PAS AAC Users a Mode N=55 N=43 Communication Never 0 0 books Rarely 6 6 Sometimes 7 6 Often 0 0 Always 0 0 Not available 42 31 Simple VOCAs Never 5 5 Rarely 5 5 Sometimes 3 2 Often 4 3 Always 0 0 No Response 1 1 Not available 37 27 Complex VOCAs Never 2 2 Rarely 2 1 Sometimes 7 6 Often 6 6 Always 2 1 Not available 36 27 a Note. AACF-PAS indicates the subset of families with completed family paradign data. 96 Table 11. Ordinal AAC-use and AAC satisfaction Variable Category All AA CF -PAS AAC Users AAC Users a N=55 N=43 Ordinal AAC-use No AAC components used 14 10 variable Only unaided AAC components (sign language) 26 20 (based on AAC Only no technology AAC 5 5 components “53d (communication pictures, boards, books) 33:; r always at Only low technology AAC (simple VOCAs) 0 0 Only direct select connputer-based 1 (complex VOCAs) Only non-direct select computer-based O 0 (connplex VOCAs) Multimodal combinations 8 7 Satisfaction with Strongly dissatisfied 4 4 first AAC system Dissatisfied 3 3 Neutral 15 15 Satisfied 22 16 Strongly satisfied 1 1 5 Satisfaction with Still using only first AAC system 15 10 current AAC system Strongly dissatisfied 1 l Dissatisfied 5 4 Neutral 6 5 Satisfied l7 l3 Strongly satisfied 11 10 Satisfaction with Strongly dissatisfied 1 first and/or current Dissatisfied 5 4 AAC system Neutral 8 7 Satisfied 22 17 Strongly satisfied 19 14 a Note. AACF-PAS indicates the subset of fannilies with connpleted family paradign data. 97 Table 12. Child's language skills with or without AAC Variable Category All AA CF -PAS a MSJSJEHS AAC Users N=43 Overall Follows requests and directions without 25 20 language skills assnstance Makes requests spontaneously 36 29 Responds to comments appropriately 24 18 Makes comments spontaneously 14 12 Answers questions appropriately 19 17 Asks questions spontaneously 8 6 Responds to communication by others but 12 7 does not initiate communication Does not spontaneously communicate 14 10 Number of None 2 t 2 Words produced Few (approximately 1 to 9) 15 12 independently Several (approximately 10 to 50) 18 13 Many (more than 50) 20 16 Utterance Uses single words 36 28 length Uses two-word utterances 22 16 Uses three-word or longer utterances 10 7 Constructs gammatically correct sentences 3 3 Does not communicate in any of these ways 4 4 a Note. AACF-PAS indicates the subset of families with completed family paradign data. b Note. Respondents were asked to select all that apply. 98 B l Table 13. Reasons why the child does not use aided AAC Variable Category All AA CF -PAS AAC Users AAC Users a n=ll n=8 Re as o nsb Child talks, signs, or gestures 7 6 Child not ready and/or working on skills 2 2 Can't afford AAC or AAC not available 1 1 Child quit using 0 0 Waiting for AAC assessment or in AAC 1 1 trial period AAC ordered/Waiting for delivery 0 0 AAC is broken/Waiting for repairs 0 0 Not told 5 3 Don't know why 2 l a Note. AA CF -PAS indicates the subset of families with completed family paradign data. b Note. Respondents were asked to select all that apply. 99 Table 14. Aided AAC components and funding sources Variable Category All AA CF -PAS AAC “15:21:63 Users a N=43 Type of Printed words. . . 27 19 picture(s) Black and white line drawnngs 25 20 Color line drawings 32 27 or symbol(s) Color photographs 31 25 Black and white photographs 5 3 Raised or textured symbols 3 3 Parts of actual objects 5 2 Full-sized actual objects 3 2 Miniature objects 7 4 Selection Points or touches pictures with finger or 41 32 metlnod(s) other body part Looks at the picture (eye gaze) 14 11 Uses a joystick 0 0 Uses 1 switch 2 2 Uses 2 or more switches 3 2 Uses optical pointer, light beann, 0 0 or head mouse Uses head stick or chin stick 1 0 Uses standard keyboard 2 2 Uses alternate keyboard 1 1 Uses touch screen 10 9 Uses encoding 0 0 Otlner 0 0 Voice output Yes 31 26 No 24 17 Funding By the school or other educational agency 25 19 source(s) By a non-profit organization 3 3 By health insurance 4 4 By Medicaid or Medicare 9 7 By our family 15 15 By community fundraisers 0 0 Other 3 2 a Note. AACF-PAS indicates the subset of families with completed family paradigm data. b Note. Respondents were asked to select all that apply. 100 Table 15. Frequency and percentage of paradigms by judgnent #Paradigms Paradigma Before- After- Post- Ideal per Family Diagnosis Diagnosis AAC n % n % n % n % One Closed (C) 3 7% 9 21% 5 12% 0 0% Random (R) 9 21% 6 14% 5 12% 5 12% Open (0) 8 19% 9 21% 12 29% 15 36% Synchronous (S) l 2% 2 5% 0 0% 0 0% Two CR 2 5% 0 0% 0 0% O 0% CO 2 5% 7 1 7% 3 7% 2 5% CS 2 5% 0 0% 0 0% 0 0% R0 7 17% 7 17% 10 24% 16 38% RS 0 0% 0 0% 0 0% 0 0% OS 2 5% 0 0% l 2% 0 0% Three CRO 4 10% 1 2% 3 7% 2 5% or four CRS 2 5% 0 0% 0 0% 0 0% R08 0 0% 0 0% 0 0% 1 2% COS 0 0% 1 2% l 2% 0 0% CROS O 0% 0 0% 2 5% l 2% Total 42 100% 42 100% 42 100% 42 100% Note. Based on the subset of 42 families with completed famijly paradign data. a Note. Paradigm category is defined as a cluster score of 5. 101 Table 16. Summary of multiple regession analysis to predict changes in family paradign as a function of the Before-Diagnosis (BD) paradign Regession models to predict difference scores from Before- to After-Diagnosis , A priori Main Interaction" Final" P‘ “11°F“ Modelt Effects' BD paradigns: Closed BD -.40 -.23 m --- (-22) (-.12) Random 80 -.98* -.76' -135 -.73' (-.36) (-.30) (-130) (-29) Open ED -105” -93" -3.61 -93" (-51) (-.46) (-l.78) (-.46) Synchronous BD .28 .19 m «- (.19) (.14) Random BD" --- --- .59 --- Open BD (1.56) Statistically-controlling for: Severity of -.40 --- --- --- disability (-. 14) Ordinal AAC use -.05 m --- --- ' (-05) Income -. l6 --- --- «- (-.12) Intercept 14.21 10.63 22.12 10.21 R2 .34 .28 .28 .26 Adjusted in2 .19 .20 .22 .22 Note. Unstandardized regession coefficients are reported for predictor variables with standardized coefficients shown in parentheses underneath the unstandardized regession coefficient *p s .05. "p s .01. "*p s .001. 102 Table 17. Summary of multiple regession analysis to predict changes in family paradign as a finnction of After-Diagnosis (AD) paradign Regession models to predict difference scores from After-Diagnosis to Post-AAC . . . . 1 1...... its “a. “‘3‘ AD paradigms: Closed AD -.71’ -.61 --- (-.36) (-.3 1) Random AD -.52 -.56 (-.26) (-.28) Open AD -81' 473' -.79' (-.39) (-.34) (-.37) Synchronous AD .06 .09 m (.03) (.06) Statistically-controlling for: Severity of .61 m --- disability (.20) Ordinal AAC use .02 m --- (.02) Income -. 19 m --- (-.15) Intercept 10.27 9.99 5.75 R2 .31 .24 .14 Adjusted R2 .16 .16 .12 Note. Unstandardized regession coefficients are reported for predictor variables with standardized coefficients shown in parentheses underneath the urnstandardized regression coefficient. 'p s .05. "p 5.01.1"p s .001. 103 Table 18. Summary of multiple regession analysis to predict changes in family paradign as a function of Post-AAC (PA) paradign Regession models to predict difference scores from Post-AAC to Ideal . A riori Main Interactionm Finer" Predictors M £61m Effects'" PA paradigns: Closed PA -. 10 -.01 --- --- (-.05) (-.004) Random PA -.38 -.51 m (-.16) (-.22) Open PA -1721" -1 .811" -1291 -171” (-.70) (-.74) (-.53) (-.70) Synchronous PA .29 .52' 1.47 .47' (.17) (.35) (.98) (.31) Open PA" m -, 22 Synchronous PA (-.75) Statistically-controlling for: Severity of .55 --- --- --- disability (.18) Ordinal AAC use -.23 --- --- --- (-.23) Income .04 ..- -.. -.. (.03) Intercept 10.46 1 1.33 7.01 8.85 R2 .57 .50 .47 .46 Adjusted R2 .47 .45 .43 .43 Note. Unstandardized regession coefficients are reported for predictor variables with standardized coefficients shown in parentheses underneath the unstandardized regession coefficient *p g .05. "p s .01. "*p s .001. 104 Table 19. Summary of multiple regession analysis to predict AAC satisfaction as a function of family paradign: A priori and main effects models Regession models to predict AAC satisfaction Pr di t A priori Model ‘ Main Effects Main Effects Main Effects e c ors using PA using PAf using BDt using I" Paradigms: Closed -.561 -.63" -.38* -.43 (-.46) (-.50) (-.35) (-.32) Random .24 .10 .60" .62' (.19) (.08) (.39) (.44) Open -.04 -.05 -.04 .20 (-.03) (-.04) (-.03) (-.06) Synchronous .10 .13 -.004 . 10 (.11) (.15) (-.004) (.11) Statistically- controlling for: Severity of .12 --- --- «- disability (.07) Ordinal AAC -.06 --- --- --- use (-.10) Income -.05 --- --- «- (-.07) Intercept 5.17 5.86 2.87 1.47 R2 .28 .26 .28 .30 Adjusted R2 .12 .18 .20 .22 Note. Unstandardized regession coefficients are reported for predictor variables with standardized coefficients shown in parentheses underneath the unstandardized regession coefficient. 'p s .05. "p s .01. *"p s .001. 105 Table 20. Summary of multiple regression analysis to predict AAC satisfaction as a function of family paradigm, Combining significant main effects in final model Regression models to predict AAC satisfaction Predictors Combination" Interactionm Final’" Paradigms: Closed PA -.37* -6.93 -.47* (-.29) (-S.50) (-.37) Closed BD -. l7 --- --- (-.16) Random BD .42' -5.47 .38' (.27) (-3.53) (.25) Random I .62'" -335 .65’" ( .44) (-2.38) (.46) Closed PA"l 1.58 -- Random BD (5.95) Closed PA“ 1,15 --- Random I (5.21) Random BD" 1.05 --- Random I (4.55) Closed PA‘ -,29 --- Random BD" (-5.99) Random I Intercept 1.40 24.95 1.21 R2 .52 .53 .50 Adjusted R2 .46 .44 .46 Note. Unstandardized regression coefficients are reported for predictor variables with standardized coefficients shown in parentheses underneath the unstandardized regression coefficient. *p s .05. "p s .01. *"p s .001. 106 Table 21. Average paradigm cluster score by element across judgments Elements Paradigm Before- Diagnosis After- Diagnosis Post- AAC Ideal M c1al M c133l M CIa M c1a Resources Closed 3.5 $.32 3.9 $.34 3.6 $.29 3.0 $.34 Random 4.2 $.29 3.5 $.40 4.1 $.26 4.8 $.13 Open 4.5 $.23 4.4 $.35 4.6 $.28 4.6 $.22 Synchronous 2.8 $.43 2.0 $.37 2.2 $.40 2.5 $.34 Time Closed 3.3 $.48 4.4 $.33 4.2 $.32 4.1 $.35 Random 3.8 $.48 2.6 $.47 3.2 $.40 4.3 $.31 Open 4.2 $.37 3.9 $.46 4.4 $.31 4.8 $. 13 Synchronous 2.6 $.53 2.1 $.47 2.1 $.46 2.1 $.50 Space Closed 2.5 $.52 2.7 $.50 2.4 $.50 1.9 $.50 Random 4.1 $.36 3.9 $.37 4.1 $.38 4.8 $.20 Open 4.0 $.44 4.3 $.39 4.0 $.44 4.1 $.36 Synchronous 2.2 $.48 2.0 $.52 1.8 $.52 2.0 $.52 Energy Closed 3.6 $.42 3.6 $.50 3.8 $.39 3.7 $.40 Random 4.1 $.35 3.6 $.52 3.9 $.38 4.4 $.30 Open 4.3 $.30 4.2 $.37 4.5 $.31 4.8 $.13 _ Synchronous 2.7 $.53 2.0 $.47 2.6 $.44 3.7 $.40 Material Closed 3.5 $.51 2.6 $.55 2.9 $.52 2.5 $.48 Random 4.0 $.42 4.4 $.38 4.5 $.32 4.9 $.08 Open 4.1 $.42 4.0 $.44 4.2 $.36 3.7 $.40 Synchronous 2.7 $.48 2.1 $.46 2.2 $.49 2.1 $.51 Goals Closed 3.9 $.32 3.9 $.40 3.9 $.32 4.1 $.27 Random 4.3 $.25 3.9 $.34 3.8 $.33 3.6 $.33 Open 3.9 $.37 3.9 $.43 4.1 $.39 4.8 $.15 Synchronous 3.2 $.37 3.3 $.42 3.5 $.41 2.8 $.42 Control Closed 3.1 $.50 2.7 $.53 3.2 $.47 4.5 $.30 Random 3.6 $.44 3.8 $.44 3.2 $.47 2.3 $.51 Open 3.1 $.48 2.7 $.56 3.5 $.50 4.7 $.16 Synchronous 3.8 $.44 3.3 $.51 3.8 $.40 3 .5 $.47 Affect Closed 3.4 $.47 3.6 $.50 3.6 $.47 3.4 $.44 Random 3.9 $.40 3.1 $.52 3.6 $.47 4.6 $.26 Open 3.7 $.39 3.5 $.45 3.6 $.45 4.2 $.34 Synchronous 2.2 $.52 2.9 $.57 2.7 $.52 2.3 $.54 Meaning Closed 4.2 $.34 4.6 $.29 4.3 $.26 4.4 $.32 Random 3.8 $.44 2.9 $.50 3.3 $.38 3.9 $.38 Open 2.9 $.46 3.3 $.54 3.7 $.42 4.2 $.27 Synchronous 3.5 $.40 3.0 $.49 3.5 $.43 3.9 $.42 Content Closed 3.7 $.42 3.2 $.49 3.6 $.41 3.7 $.36 Random 3.6 $.51 3.7 $.39 3.4 $.46 2.7 $.48 Open 4.0 $.39 4.2 $.42 4.3 $.33 4.9 $.16 Synchronous 2.5 $.48 2.4 $.50 2.5 $.48 2.2 $.46 a . . Note. Confidence interval (CI) is based on a 95% two-sided level. Bolding within each element indicates the highest average cluster score for that judgment (based on output data with 2 decimal points). 107 Family Ecosystem Concepts Resources are transformed to achieve goals. Control How are important things accomplished? =-;.__ Time What rhythm is used? Family The family Space members Affect How are physical and may act How are affection, caring, interpersonal space used? individually, and support expressed? in small groups, or as . one may, I Energy according ‘0 Meaning How much effort is used to their family What do you value? get things done? paradigm Material How are material possessions and belongings viewed? Content How do you determine what is real? Feedback Mechanism | These processes occur within the natural, human-constructed, and social-cultural environments. Figure 1. Basic processing of inputs and creation of outputs by a family system (Based on Bubolz & Sontag, 1993; Imig, 2000a; Villarruel et al., 1995) 108 100.0% 1 M DBefore Diagnosis 800% ‘3 DAfter Diagnosis A IPost AAC 3: Ildeal g 60 0°/ ' a . o '1 o l h o , e r 8 = 40.0% 1 E 1 ~— «I it. 20.0% +1 1 s 0.0% “_—_ I , Closed Random Open Synchronous Paradigms Figure 2. Changes in family paradigms across the 4 judgments (N=42 families).[MJCHll 109 APPENDIX B Case History 110 Date ID Case History Tlrank you for your participation. Please answer the questions below. "Child" refers to your preschool child who is using augmentative and alternative communication (AAC). AAC is the way that your child communicates (other than with natural speech or gestures). AAC includes sign language, communication pictures, communication boards, communication books, and voice output devices. AAC user 1) When is your child's birthday? Please write as mm/dd/yyyy. For example: 04/12/1999 2) What is your child's gender? _Female _Male 3) What is your child's racial/ethnic background? Please select all that apply. _American Indian/Alaskan Native _Asian _Native Hawaiian or Other Pacific Islander _Black/African American _Caucasian/White _Hispanic/Latino _Prefer not to answer _Other (specijy in Question 4) 4) If you checked Other in Question 3, please specify. 5) Please indicate all of your child's diagnoses. Please select all that apply. _Apraxia _Autism/Pervasive developmental disorder (PDD) _Cerebral palsy _Developmental delay _Dual sensory impairment (Deaf-blind) _Dysarthria _Hearing impairment-temporary such as with ear infections _Hearing impairment-permanent _Medically fragile _Mental retardation _Seizure disorder _Traumatic brain injury _Visual impairment not corrected by glasses _Cause unknown _Other (please specifi» in Question 6) 6) If you checked Other in Question 5, please specify. 111 7) How would you rate the severity of your child's disabilities as they affect your family's daily activities? Mild Moderate Severe 8) How old was your child when you first suspected that he or she had a communication disorder? Less than 1 year old 1 year through 1 year, 11 months 2 years through 2 years, 11 months _3 years through 3 years, 11 months _4 years through 4 years, 11 months _5 years through 5 years, 11 months 9) 6 years through 6 years, 11 months How old was your child when he or she was diagnosed with a communication disorder? Less than 1 year old :1 year through 1 year, 11 months _2 years through 2 years, 11 months 3 years through 3 years, 11 months _4 years through 4 years, 11 months _5 years through 5 years, 11 months _6 years through 6 years, 11 months 10) Who made this diagnosis (e.g., speech-language pathologist, family physician, teacher, neurologist)? 11) How old was your child when he or she first received an AAC system (i.e., some way for your child to communicate other than with your child's natural speech or gestures)? _Less than 1 year old _1 year through 1 year, 11 months _2 years through 2 years, 11 months _3 years through 3 years, 11 months 4 years through 4 years, 11 months _5 years through 5 years, 11 months __6 years through 6 years, 11 months 12) How satisfied were you with this first AAC system? _Strongly satisfied _Satisfied Neutral ___Dissatisfred _Strongly dissatisfied _Did not have access to it at home , 13) How old was your child when he or she received the most recent AAC system (i.e., some way for — ~— — _— your child to communicate other than with your child's natural speech or gestures)? Still using only the first AAC system (Skip to Question 15) Less than 1 year old 1 year through 1 year, 11 months 2 years through 2 years, 11 months 3 years through 3 years, 11 months 4 years through 4 years, 11 months 5 years through 5 years, 11 months 6 years through 6 years, 11 months 112 ll 14) How satisfied are you with the current AAC system? _Strongly satisfied _Satisfied _Neutral _Dissatisfied _Strongly dissatisfied _Do not have access to it at home 15) What types of AAC components are available for your child to use at home, at school, and/or in the community? Please select all that apply. _Sign language _Communication pictures/boards _Communication books _Sirnple voice output device (limited to one page of messages and one button equals one word, phrase, or sentence.) _Complex voice output device (permits multiple pages or screens and/or the same buttons can be combined to create different words-semantic compaction.) _Other (please describe in Question I 6.) i 16) Please describe other type(s) of AAC components if you checked Other in Question 15. 17) How often does your child VOCALIZE (uses voice that doesn't sound like words) to communicate at home? _Never _Rarely _Sometimes _Often _Always 18) How often does your child use SPEECH or SPEECH APPROXIMATIONS (uses voice that does sound like words) to communicate at home? _Never _Rarely _Sometimes _Ofi‘en _Always 19) How often does your child communicate by POINTING or GESTURING at home? _Never _Rarely _Sometimes _Ofien _Always 20) How often does your child use SIGN LANGUAGE to communicate at home? _Never _Rarely _Sometimes _Often _Always 113 21) How often does your child use COMMUNICATION PICTURES (single pictures) or BOARDS (single page of pictures) to communicate at home? _Never _Rarely _Sometimes _Ofien _Always 22) How often does your child use COMMUNICATION BOOKS (multiple pages) to communicate at home? _Never _Rarely _Sometimes _Ofien _Always 23) How often does your child use a SIMPLE VOICE OUTPUT DEVICE (limited to one page of messages and one button equals one word, phrase, or sentence) to communicate at home? _Never _Rarely _Sometimes _Often _Always 24) How often does your child use a COMPLEX VOICE OUTPUT DEVICE (permits more than one page or screen of messages and/or the same buttons can be combined to create different words - semantic compaction) to communicate at home? _Never _Rarely _Sometimes _Ofien _Always 25) Please list any other ways your child communicates at home and the frequency with which your child uses them (rarely, sometimes, ofien, always). 26) If you know the brand names of your child's AAC components, please list them. If your child has communication pictures, communication board, communication book, and/or voice output device, please continue with Question 29. If your child does NOT have communication pictures, communication board, communication book, and/or voice output device, please answer Question 27 and then skip to Question 35. 27) Why doesn't your child use communication pictures, communication boards, communication books, and/or voice output devices? Please select all that apply. _Child talks/signs/gestures _Child not ready and/or working on skills _Can't afford AAC or AAC not available _Child quit using _AAC ordered/Waiting for delivery _AAC is broken/Waiting for repairs _Not told _Don't know why _Other (please describe in Question 28) 114 28) If you checked Other in Question 27, please describe. 29) If your child uses communication pictures, communication boards, communication books, and/or voice output devices, please describe the pictures/symbols. Please select all that apply. _Printed words _Black and white line drawings _Color line drawings _Color photographs _Black and white photographs _Raised or textured symbols (e. g., sandpaper, cloth, glue-gun outlined) _Parts of actual objects _Full-sized actual objects _Miniature objects 30) How does your child select messages from his or her AAC system? Please select all that apply. _Points or touches pictures with finger or other body part _Looks at the picture (eye gaze) _Uses a joystick _Uses 1 switch _Uses 2 or more switches _Uses optical pointer, light beam, or head mouse _Uses head stick or chin stick _Uses standard keyboard _Uses alternate keyboard (addition to or replacement for a standard keyboard) _Uses touch screen _Uses encoding _Other (please describe in Question 31) 31) Please describe any other ways that your child selects messages. 32) What body part or parts does your child use to point to or activate the AAC system (e.g., finger, hand, toe, eye blink, eye gaze, head turn)? 33) How was the AAC system funded? Please select all that apply. _By the school or other educational agency _By a non-profit organization _By health insurance _By Medicaid or Medicare _By our family _By community fundraisers _Other (Please describe in Question 34) 34) If you checked Other in Question 33, please describe. 35) What is your child's educational setting? Please select all that apply. _Home visits by early intervention professionals _Self-contained preschool (all kids have disabilities) _Integrated preschool (at least 50% of kids have no disabilities) _None _Unsure _Other (please specijy in Question 36) 115 36) If you checked Other in Question 35, please describe. 37) Has your child and/or any other family member received assistance in learning to use the AAC system? Please select all that apply. _No (Please continue with Question 41) _Yes, at home _Yes, at preschool _Yes, other (please describe in Question 38.) 38) If you checked Other in Question 37, please describe. 39) What type of AAC assistance has been received? Please select all that apply. Learning how to: _Maintain and/or troubleshoot the AAC system _Program or create new communicative messages _Use AAC to communicate in daily activities _Access messages (learning to use switches, scanning, joystick, picture pointing, etc.) _Other (please describe in Question 40) 40) If you checked Other in Question 39, please describe. 41) Please describe your child's overall language skills with or without AAC. Please select all that apply. _Follows requests and directions without assistance _Makes requests spontaneously _Responds to comments appropriately _Makes comments spontaneously _Answers questions appropriately _Asks questions spontaneously _Responds to communication by others but does not initiate communication _Does not spontaneously communicate 42) How many words or messages does your child produce independently through speech, sign, and/or AAC? _None _Few (approximately 1-9) _Several (approximately 10-50) _Many (more than 50) 43) How does your child communicate using AAC, his or her own speech, and gestures? Please select all that apply. _Uses single words _Uses two-word utterances (e.g., more milk, no go, Mommy car) _Uses three-word or longer utterances (e.g., Dog ran fast, Not on box) _Constructs grammatically correct sentences (e.g., I am playing with the ball.) _Does not communicate in any of these ways 44) In an average week, approximately how many hours does someone read to your child at home? 45) In an average week, approximately how many hours does your child draw, scribble, color, or write at home? (This could be done with pencils, crayons, computer programs, or other adaptations if needed.) 116 Family 46) What is the primary language spoken at home? Please select all that apply. _English _Spanish _Other (Please list in Question 47) 47) If you checked Other in Question 46, please describe. 48) Please list the gender and age of any other children in your family (even if they live with you only part of the time): Please put each child on a new line. (For example, "male, 2" would indicate a 2 year-old brother. "female, 12" "male, 8" "female, 1" would indicate that your child has a 12 yearoold and 1 year-old sister as well as an 8 year-old brother.) 49) How many children with disabilities are in your family (including the AAC user)? 50) How many adults live in your house full-time (including yourself)? 51) Where do you live? _F arm _Rural but not a farm _Town (population up to 49,999) which is not part of a metro area _Small city (population of 50,000 to 99,999) _Medium city (population of 100,000 to 299,999) ____Large city (population over 300,000) 52) What is your zip code? 53) What is your family's combined total gross (before taxes) yearly household income? _Under $10,000 _$10,000 - $19,999 ___$20,000 - $39,999 __$40,000 - $59,999 ____$60,000 - $99,999 _$100,000 - $149,999 _Over $150,000 _Prefer not to answer Primary Caregiver NOTE: The term "primary caregiver" refers to the adult who spends the most time communicating with the child and/or who has more child care and home responsibilities. 54) Are you the primary caregiver? _Yes _No 55) What is the primary caregiver's age? 56) What is the primary caregiver's gender? _Female _Male 117 57) What is the primary caregiver's relationship to the child? _Biological parent _Foster parent _Step-parent _Legal guardian _Adoptive parent _Other (please specijy in Question 58) 5 8) If you checked Other, please specify the relationship to the child. 59) What is the educational level of the primary caregiver? _Grade school only _Some high school _High school graduate _Some college _College graduate ____Some graduate school _Advanced degree 60) What is the primary caregiver's marital status? _Single _Married (to child's other parent) _Married (but not to child's other parent) _Separated or divorced _Widowed _Other (please explain in Question 61) 61) If you checked Other in Question 60, please describe. 62) Please describe the occupational status of the primary caregiver. Please select all that apply. _Business owner _Full-time employee _Part-time employee _Stay-at-home caregiver _Student _Retiree _Other (please specify in Question 63) 63) If you checked Other, please describe the occupational status. 64) What is the occupation of the primary caregiver (e. g., clerk, teacher, salesperson, stay-at-home caregiver)? 65) Where does the primary caregiver work (e.g., discount store, high school, home-based business)? 66) What is the primary caregiver's racial/ethnic background? Please select all that apply. _American Indian/Alaskan Native _Asian _Native Hawaiian or Other Pacific Islander _Black/African American _Caucasian/White _Hispanic/Latino _Prefer not to answer _Other (specify in Question 67) 118 67) If you checked Other in Question 66, please specify. Secondary Caregiver NOTE: The term "secondary caregiver" refers to another adult who shares in child rearing responsibilities. 68) Is there a secondary caregiver? _Yes (If yes, please continue with Question 69.) _No (If no, you areflnished with this case history questionnaire.) 69) What is the secondary caregiver's age? 70) What is the secondary caregiver's gender? _Female _Male 71) What is the secondary caregiver's relationship to the child? _Biological parent _Foster parent _Step-parent _Legal guardian _Adoptive parent _Other (please specify in Question 72) 72) If you checked Other, please specify the relationship to the child. 73) What is the educational level of the secondary caregiver? _Grade school only _Some high school _High school graduate _Some college _College graduate _Some graduate school _Advanced degree 74) What is the secondary caregiver's marital status? _Single _Married (to child's other parent) _Married (but not to child's other parent) _Separated or divorced _Widowed ___Other (please explain in Question 75) 75) If you checked Other in Question 74, please describe. 76) Please describe the occupational status of the secondary caregiver. Please select all that apply. _Business owner _Full-time employee _Part-time employee _Stay-at-home caregiver _Student _Retiree _Other (please specifil in Question 77) 119 77) If you checked Other, please describe the occupational status. 78) What is the occupation of the secondary caregiver (e.g., clerk, teacher, salesperson, stay-at-home caregiver)? 79) Where does the secondary caregiver work (e. g., discount store, high school, home-based business)? 80) What is the secondary caregiver's racial/ethnic background? Please select all that apply. _American Indian/Alaskan Native _Asian _Native Hawaiian or Other Pacific Islander _Black/African American _Caucasian/White _Hispanic/Latino _Prefer not to answer _Other (specify in Question 8]) 81) If you checked Other in Question 80, please specify. 120 APPENDIX C AAC Family - Paradigm Assessment Scale (AA CF -PAS) 121 AAC Family - Paradigm Assessment Scale (AACF-PAS) The purpose of this questionnaire is to assess your perception of how your family functions. Instructions: Column 1 labeled C: Complete the questions on the following screens by assigning a value of 10 to the ONE choice (A, B, C, D) in the C column which most accurately describes your CURRENT (C) understanding of your family. From the three remaining choices in the C column assign a value ranging from 0-9 to the 2nd most descriptive choice. 0 indicates that this is NOT how your family functions. As the number increases, the characteristic comes closer to how the family operates. Repeat for the 3rd and 4th choices. All values (0-9), except for the number 10 may be repeated any number of times. Remember, that there must be and should be only one 10 in the C column (see Sample Question below). Column 2 labeled I: Family life is not always what we would like for it to be. Please repeat the process as described above for the column marked I . Assign a value of 10 to the ONE choice (A, B, C, D) that most IDEALLY (1) represents how you would like for this aspect of your family to be. As before, assign values to the three remaining choices. All values except for 10 may be repeated any number of times. Column 3 labeled B: Now, think back to the time BEFORE (B) you were aware that your child had a communication disorder. Please rate how your family functioned. Again, assign a 10 to the choice that most closely represented your family. Then, assign values from 0-9 to the other 3 choices. Column 4 labeled A: Now, think back to the time IMMEDIATELY AFTER (A) your child was diagnosed with a communication disorder. Please rate how your family functioned. Below is an example question with the numbers filled in. Again, note that there is one 10 per column, but not more than one 10 per column. EXAMPLE our we to each other in the . - In a manner - a direct manner - a manner - a and manner 4 Relationships also involve behavior. The four types of behaviors are described below. Initiating-Moving: In any relationship someone or something starts, initiates, causes or determines what will take place and when. Questioning-Challenging: In your relationship you or your partner may comment on the action taking place and have a range of suggestions for how things might be changed for a variety of reasons. Reflecting-Commenting: Sometimes someone doesn't initiate, challenge or support any actions taken, but insteads acts as a kind of guide and conscience by providing a balanced, accurate and nonblaming sense of reality, insight and wisdom about what they have observed. Agreeing-Supporting: Finally, someone may agree with or confirm the behaviors of one, any or all of the other behaviors (initiating - moving, questionning - challenging, or reflecting - commenting). Following the same instructions provided above, please assign CURRENT (C), IDEAL (1), BEFORE (B), and IMMEDIATELY AFTER (A) your child's communication disorder diagnosed values. Please see the below. EXAMPLE you to communication in Think back 122 1. In your family, how do the important firings that Think back must geitdone, get done? C l B A A - We just know what needs to get done & how to do it my being well organized, using successful & structured routines, and perhaps most importantly having a plan we can count on C - Each person does what they think needs to get done and how to do it D - By regularly discussing and agreeing with each other what needs to get done and how “best” to work together get things done you to what has to done? Think back 2. How do you show your affection, care 8. support Think back for one another? C I B A A - We share our love 8affection for each other in an intimate, expressive, emotionally shared and somewhat private manner B - We demonstrate our love &fiection for one another in a somewhat conventional, regulated, modest and always private manner C - We show our love & affection for each other in a playful, spontaneous, uninhibited and sometimes public _Lnanner g D - We share our affection in an unspoken manner - because we just know without saying it that we deeply love 8. care for each other What behaviors do you use when showing affection & carinig in your family? lhlnk back C I B A Initiating-Moving Questioning-fiChaIIenging Agfeeing-Supporting Reflecting-Commenting 123 3. Within your family, what contributes to providing Think back you with a sense of purpose and meaning? C l B A A - By each of us having the freedom 8 afihomy to engage in a personal journey of growth. exploration & self-discovery B - By valuing the family more than ourselves as individuals, making decisions that benefit our common good, and valuing the virtues of organization, discipline 8- responsibility C - When our personally unique experiences 8. insights result in a shared, implicit 8. unspoken sense of unity, harmony & way of knowing _ D - By working together in our family to ”go beyond" what has always been to create new and different ways of living life you and has Think back 4. How in your family do you go about “making Think back sense" out of what you experience in life ? C I B A A - By sharing our ideas with each other, by asking each other questions, and listening to the opinions 8. thoughts of others B - Each of us subjectively relies on ourselves and our own ideas to personally make sense out of what it is we experience in life C - By using “the” time-tested & established rules & truths of life, and by having learned how to look at any _situation in an objective and factual manner D - We just seem to know without much discussion how to understand and make sense out of what we experience in life What behaviors do you play when trying to make sense out of life experience? _Think back C l B A Initiating-Moving Questioning-Challenging Agreeing-Supporting Tieflecting-Commenting 124 5. From a relational polnt-of-vlew, what emphasis do you feel is being placed on the following areas? Think back A A - The importanEe of our being able to undo—rstand & make sense out of our life experiences in an accurate 8. realistic way ' B - That our family is guided by a greater sense of purpose and meaning in life C - That in our family we provide each other with the amount & kind of affection, caring, love & support wanted & needed D - That the important & necessary things that need to get done in order to have a quality family get done What behaviors do you play in determining the importance of these areas? _ Think back C A Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting 6. In your family, how is time generally used? Think back A A - In a flexible & adaptive manner - it can be changed as needed B - In a planned, scheduled & organized manner C - In a spontaneous manner so that opportunities for unplanned, interesting 8. creative experiences can happen D - Without hardly any discussion, in our family we just seem to know how time is to be used What behaviors do you play in determining how time W be used in your family? _ Think back C A Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting 125 Tin your family, how are questions and ideas handled? l ink back In; A - In our family we can ask any questions. We can say anything to each oflrer, no matter how personally intimate, confronting orjust plain silly. Its OK to ask any questions - no matter what! B - Certain issues and topics are rarely discussed in our family because they are inappropriate. Discussions are usually constructive 8. conducted with mutual respect C - Within reason, most questions can be asked and ideas can be discussed - but differences causing conflict are to be resolved D - There doesn't seem to be any real need for us to ask questions of each other, we just seem to understand most things in the same way What behaviors do you play In deterrnlning how Ideas 8: questions will be handled? Think back C I A Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting 8. How Is energy and effort used In your family? Think back a A A - In a steady, consistent, regulated and controlled manner B - In a dynamic, enthusiastic, spirited and vigorous manner C - In an peaceful, calm, serene and tranquil manner D - In a flexible, adaptive, changeable 8. accommodating manner What behaviors do you play in determining how effort 8- energy will be used? Jhink back C l A Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting 126 9. In your family, how do you relate to your possessions and belongings - the “things" of life? Wk back A A - “Things” are valued bacause we worked hard to get them, and for us they represent the “just” & deserving rewards of life B - “Things” aren’t what’s really important in life - it’s experiencing & living life that’s important - things often jiflget in our way C - “Things” are useful in life because we can use them to get other more important things done & to make life more convenient D - “Things" are to be valued and respected because of the personal meaning that they represent. Because of their importance they should be protected & kept as perfect as possible What behaviors do you play In determining how to relate to possessions and belongings? Think back C I Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting 10. What emphasis is placed in your family on the following areas? Think back A A - The importance of possessions and belongings B The importance of effort & energy in our family C - The importance of time 8. how it will be used D - The importance of ideas, questions & information What behaviors do you play in determining the comparative Importance of these areas? LThInk back C A Initiating-Moving Questioning-Challenging Agreeing-Supporting Reflecting-Commenting Please check that each column has one and only one 10 in it. Thank you for participating! 127 APPENDIX D AA CF -PAS Calculations 128 AACF-PAS Calculations Goals Resources Question Paradigm Question Paradigm 1 Control 6 Time 1 A Synchronous 6 A Open 1 B Closed 6 B Closed 1 C Random 6 C Random 1 D Open 6 D Synchronous 2 Affect 7 Space 2 A Open 7 A Random 2 B Closed 7 B Closed 2 C Random 7 C Open 2 D Synchronous 7 D Synchronous 3 Meaning 8 Energy 3 A Random 8 A Closed 3 B Closed 8 B Random 3 C Synchronous 8 C Synchronous 3 D Open 8 D Open 4 Content 9 Material 4 A Open 9 A Closed 4 B Random 9 B Random 4 C Closed 9 C Open 4 D Synchronous 9 D Synchronous 5 Goals 10 Resources 5 A Content 10 A Material 5 B Meaning 10 B Energy 5 C Affect 10 C Time 5 D Control 10 D Space The table above shows which AACF-PAS questions capture different elements and paradigms. Each question in each section (e. g., 1A, 1B, 1C, 1D in the control section) is given a rating ranging fi'om 0 to 10. These four raw scores are totalled. A simple coefficient for an element's paradigm is calculated by dividing the raw score for that element by the total of the four raw scores for that element. (For example, to calculate the simple coefficient for synchronous control, the raw score for 1A is divided by the sum of 1A, 1B, 1C, and 1D. To calculate the simple coefficient for closed control, the raw score for 1B is divided by the sum of 1A, 1B, 1C, and 1D.) This procedure is repeated to calculate a simple coefficient for each paradigm of the eight elements as well as for the goals (question 5) and the resources (question 10). 129 A complex coeflicient for an element's paradigm is calculated by multiplying its simple coefficient by the appropriate goal or resource simple coefficient. [For example, to calculate the complex coefficient for synchronous control, the simple coefficient for synchronous control is multiplied by its appropriate simple goal coefficient for control (5D/(sum(5A, 5B, 5C, 5D))). To calculate the simple coefficient for closed material, the simple coefficient for closed material is multiplied by its appropriate simple resource coefficient for material (10A/(sum(10A, IOB, 10C, 10D)))]. This is completed for all 32 of the paradigm elements (i.e., 4 paradigms by 8 elements.) The resulting 32 complex coefficients for the individual elements are then ranked from highest to lowest in order to create quartile scores. - To determine cluster scores for individual elements, a complex cluster coefiicient is calculated by dividing the complex coefficient for a paradigm element by the maximum complex coefficient for that element. Then a cluster score is assigned based on the cluster complex coefficient: 1 to .9 = 5; .89 to .7 = 4; .69 to .4 = 3; .39 to .2 = 2; .19 to .10 = l; and < .1 = O. [For example, assmne the complex coefficients for closed, random, open, and synchronous material are .035, .042, .057, and .071. In this example, the maxirntun complex coefficient for that element is .071 synchronous material. Thus, to calculate the complex cluster coefficient for closed material, its complex coefficient (.035) is divided by the maximum complex coefficient (.071), resulting in a cluster complex coefficient (.49). The cluster score of 2 is then assigned for closed material] To determine cluster scores for goals, each of the complex cluster coefficients for the 4 paradigms for the goal elements are summed and then divided by the paradigm with the highest sum. Then cluster scores are assigned based on the cluster complex coefficients: 1 to .9=5; .89to .7=4; .69to .4=3; .39to .2=2; .19to .10= 1;and<.l =0. [For example, each paradigrn's four goal complex coefficients will be summed (first, the closed goal sum of closed control, closed affect, closed meaning, and closed content; then the random goal sum of random control, random affect, random meaning, and random content; then the open goal sum of open control, open affect, open meaning, open content; and finally the synchronous goal sum of synchronous control, synchronous affect, synchronous meaning, synchronous content). To calculate a cluster score for open goals, the sum (.256) of open control, open affect, open meaning, and open content is divided by whichever sum of goal complex coefficients is highest (in this case, .298 synchronous goal) resulting in a cluster complex coefficient (.86). The cluster score of 4 is then assigned for open goal]. Resources cluster scores are determined in a similar way by replacing the goal elements with the resource elements. To determine overall cluster scores, an overall complex cluster coefficient must be calculated. All 8 of the complex coefficients for a paradigm are summed (e.g., closed control, closed affect, closed meaning, closed content, closed time, closed space, closed energy, and closed material). Each sum is divided by whichever sum (i.e., closed, open, random, or synchronous) of the 8 complex coefficients is highest. Then the overall cluster score is assigned based on the cluster complex coefficient: 1 to .9 = 5; .89 to .7 = 4; .69 to .4=3;.39to.2=2;.19to.10=1;and<.1=0. 130 APPENDIX E Rules to Correct Raw AACF—PAS Scores 131 -4‘- 1, i Rules to Correct Raw AACF-PAS Scores Note: These were used only for the paper version since the computer version required that numbers be completed correctly: 1) If a column does not contain a 10, take the highest number and make it a 10. 2) If there is a tie of the highest number, randomly break the tie and make that number a 10. 3) If a number is missing, look for a logical pattern and then obtain inter examiner agreement. If no logical pattern, then put in 55 to go for no strong preference. 4) If missing a coltunn of data, assume the null of no change. 5) If missing element data, look at Question 5 or Question 10, whichever is appropriate, to see if element is rated high. If the element is rated high, then the person's paradigm data cannot be used. If element is rated low, put in null data or logical data since the element will have limited impact on the rankings. 132 :11 1 ~! REFERENCES Acredolo, L. P., Goodwyn, S., & Abrams, D. (2002). Baby signs: How to talk with your baby before your baby can talk. Chicago: Contemporary Books. American Speech-Language-Hearing Association. (1991). Report: Augmentative and alternative commrmication. Asha, 33(Suppl. 5), 9-12. American Speech-Language-Hearing Association. (2004). Roles and responsibilities of speech language pathologists with respect to augrnentative and alternative communication: Technical report [Electronic version]. ASHA Supplement 24, 1-18. Andrews, J. R., & Andrews, M. A. (2000). Family based treatment in communication disorders: A systemic approach (2nd ed.). DeKalb, IL: Janelle Publications. Angelo, D. H. (2000). Impact of augrnentative and alternative communication on families. Augmentative and Alternative Communication, 16, 37-47. Angelo, D. H., Jones, S. D., & Kokoska, S. M. (1995). Family perspective on augrnentative and alternative communication: Families of young children. Augmentative and Alternative Communication, 1 I, 193-201. Angelo, D. H., Kokoska, S. M., & Jones, S. D. (1996). Family perspective on augrnentative and alternative communication: Families of adolescents and young adults. Augmentative and Alternative Communication, 12, 13-20. Aron, L. Y., Loprest, P., & Steuerle, C. E. (1996). Serving children with disabilities: A systematic look at the programs. Washington: Urban Institute Press. Bailey, D. B. (1988). Assessing critical events. In D. B. Bailey & R. J. Sirneonsson (Eds), Family assessment in early intervention (pp. 119-138). Columbus, OH: Merrill. Bedrosian, J. L. (1997). Language acquisition in young AAC system users: Issues and directions for firture research. Augmentative and Alternative Communication, 1 3, 1 79-1 85. 133 Beukelman, D. R., & Mirenda, P. (1998). Augmentative and alternative communication: Management of severe communication disorders in children and adults (2nd ed.). Baltimore: Paul H. Brookes. Blackstone, S. W., Williams, M. B., & Joyce, M. (2002). Future AAC technology needs: Consumer perspectives. Assistive Technology, 14, 3-16. Bondy, A., & Frost, L. (2002). A picture's worth: PECS and other visual communication strategies in autism. Bethesda, MD: Woodbine House. Boss, P. (1987). Family stress. In M. B. Sussman & S. K. Steinmetz (Eds), Handbook of marriage and the family (pp. 695-723). New York: Plenum Press. Bubolz, M. M., & Sontag, M. S. (1993). Human ecology theory. In P. G. Boss & W. J. Doherty & R. LaRossa & W. R. Schumm & S. K. Steinmetz (Eds), Sourcebook of Family Theories and Methods: A Contextual Approach (pp. 419-448). New York: Plenum Press. Calculator, S. N., & Jorgensen, C. M. (1991). Integrating AAC instruction into regular education settings: Expounding on best practices. Augmentative and Alternative Communication, 7, 204-214. Chapman, R. S. (2000). Children's language learning: An interactionist perspective. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41(1), 33-54. Cohen, J. (1988). Statistical power analyszs for the behavioral sciences (2nd ed) Hillsdale, NJ: L. Erlbaum Associates. Constantine, L. L. (1986). Family paradigms: The practice of theory in family therapy. New York: The Guilford Press. Constantine, L. L. (1991). Fitting intervention to organizational paradigm. Organization Development Journal, 9(2), 41-50. Constantine, L. L. (1993a). The structure of family paradigms: An analytical model of family variation. Journal of Marital and Family Therapy, 19, 39-70. 134 Constantine, L. L. (1993b). Work organization: Paradigms for project management and organization. Communications of the Association for Computing Machinery (ACM), 36(10), 34-43. Constantine, L. L., & Israel, J. T. (1985). The family void: Treatment and theoretical aspects of the synchronous family paradigm. Family Process, 24, 525-547. Crais, E. R. (1991). Moving from "parent involvement" to family-centered services. American Journal of Speech-Language Pathology, 1, 5-8. Cress, C. J. (2002). Expanding children's early augmented behaviors to support symbolic development. In J. Reichle & D. R. Beukelrnan & J. C. Light (Eds), Exemplary practices for beginning communicators: Implications for AAC (pp. 219-272). Baltimore: Paul H. Brookes. Cress, C. J. (2004). Augmentative and alternative communication and language: Understanding and responding to parents' perspectives. Topics in Language Disorders, 24, 51. Donahue-Kilburg, G. (1992). F amily-centered early intervention for communication disorders: Prevention and treatment. Gaithersburg, MD: Aspen Publishers, Inc. Duis, S. S., Summers, M., & Summers, C. R. (1997). Parent versus child stress in diverse family types: An ecological approach. Topics in Early Childhood Special Education, 1 7(1), 53-73. Edwards, W., & Newman, J. R. (1982). Multiattribute evaluation. Beverly Hills: Sage Publications. Frost, L., & Bondy, A. (2002). Picture exchange communication system training manual (2nd ed.). Cherry Hill, NJ: PECs. Gillette, Y. (2000). Collaboration with families in communication intervention. In L. R. Watson & E. R. Crais & T. L. Layton (Eds), Handbook of early language impairment in children: Assessment and intervention (pp. 137-171). Albany, NY: Delmar Thomson Learning. 135 Glennen, S., & DeCoste, D. C. (1997). The handbook of augrnentative and alternative communication. San Diego: Singular. Goosens', C., Crain, S., & Elder, P. (1992). Engineering the preschool environment for interactive, symbolic communication. Birmingham, AL: Southeast Augmentative Communication Conference Publications. Harwood, K., Warren, S. F., & Yoder, P. (2002). The importance of responsivity in developing contingent exchanges with begimiing communicators. In J. Reichle & D. R. Beukehnan & J. C. Light (Eds.), Exemplary practices for beginning communicators: Implications for AAC (pp. 59-95). Baltimore: Paul H. Brookes. Haveman, M., van Berkum, G., Reijnders, R., & Heller, T. (1997). Differences in service needs, time demands, and caregiving burden among parents of persons with mental retardation across the life cycle. Family Relations, 46, 417-425. Heller, T., Hsieh, K., & Rowitz, L. (1997). Maternal and paternal caregiving of persons with mental retardation acorss the lifespan. Family Relations, 46, 407-415. Hill, R. (1949). Families under stress: Adjustment to the crises of war separation and reunion. New York,: Harper. Hintz, J. (2001). NCSS and PASS number cruncher statistical systems. Kaysville, UT: www.NCSS.com. Huer, M. B., & Lloyd, L. L. (1990). AAC users' perspectives on augrnentative and alternative communication. Augmentative and Alternative Communication, 6, 242- 249. Huston, T. L., & Vangelisti, A. L. (1995). How parenthood affects marriage. In M. A. Fitzpatrick & A. L. Vangelisti (Eds), Explaining family interactions (pp. 147-176). Thousand Oaks: Sage. Imig, D. R. (1993). Family stress: Paradigms and perceptions. Family Science Review, 6, 125-136. 136 Imig, D. R. (2000a). A conversation about interpersonal relationships, family systems and paradigms. Venice, CA: ETEXT.net Electronic Textbook Publishing. Imig, D. R. (2000b). The relational paradigm assessment scale (R-PAS). Michigan State University: Author. Imig, D. R., Pate, S. M., Mitchell, M. M., Davis, D. A., Pegorraro, C., Barton, E. R., Meloy, S., Villarruel, F. A., Hoedel, J ., Phillips, R. G., Buthelezi, S., & Bayes, T. G. (1996). Paradigmatic family systems theory: Applications and praxis. Paper presented at the 58th Annual Conference Education & Enrichment Section, Kansas City, MO. Imig, D. R., & Phillips, R. G. (1992). Family theory: The family regime assessment scale (F RAS). Family Science Review, 5, 217-234. Jinks, A., & Sinteff, B. (1994). Consumer response to AAC devices: Acquisition, training, use, and satisfaction. Augmentative and Alternative Communication, 10, 184-190. Jones, S. D., Angelo, D. H., & Kokoska, S. M. (1998). Stressors and family supports: Families with children using augmentative & alternative communication technology. Journal of Children's Communication Development, 20, 3744. Judge, S. L., & Parette, H. P. (1998). Assistive technology decision-making strategies. In S. L. Judge & H. P. Parette (Eds), Assistive technology for young children with disabilities: A guide to family-centered services (pp. 127-147). Cambridge, MA: Brookline Books. Kantor, D., & Lehr, W. (1975). Inside the family: Toward a theory of family process. San Francisco, CA: Jossey-Bass. [(005, E. L. (1946). Families in trouble. New York: King's Crown Press. Kumin, L. (2003). Early communication skills for children with down syndrome: A guide for parents and professionals (2nd ed.). Bethesda, MD: Woodbine House. 137 Licht, M. H. (1995). Multiple regression and correlation. In L. G. Grimm & P. R. Yarnold (Eds), Reading and understanding multivariate statistics (lst ed., pp. 19-64). Washington, DC: American Psychological Association. Light, J. C., Parsons, A. R., & Drager, K. (2002). Developing interactions for social closeness with beginning communicators who use AAC. In J. Reichle & D. R. Beukehnan & J. C. Light (Eds), Exemplary practices for beginning communicators: Implications for AAC (pp. 187-218). Baltimore: Paul H. Brookes. Locke, P., & Levin, J. (1999). Making connections: A practical guide for bringing the world of voice output communication to students with severe disabilities. Minneapolis, MN: AbleNet, Inc. McCubbin, H., McCubbin, M., & Thompson, A. I. (1993). Resiliency in families: The role of family schema and appraisal in family adaptation to crises. In T. H. Brubaker (Ed.), Family Relations: Challenges for the Future (Vol. 1, pp. 153-177). Thousand Oaks, CA: SAGE Publications, Inc. McCubbin, H. I., & Patterson, J. M. (1983). The family stress process: The double ABCX model of adjustment and adaptation. In H. I. McCubbin (Ed.), Social Stress and the Family (pp. 7-37). New York: The Haworth Press, Inc. McWilliam, R. A., Tocci, L., & Harbin, G. (1998). Family-centered services: Service providers' discourse and behavior. Topics in Early Childhood Special Education, 18(4), 206-221. Mirenda, P. (2003). Toward a fimctional augmentative and alternative commrmication for students with autism: Manual signs, graphic symbols, and voice output communication aids. Language, Speech, and Hearing Services in Schools, 34, 203- 21 6. National Joint Committee for the Communication Needs of Persons With Severe Disabilities. (2002). Access to communication services and supports: Concerns regarding the application of restrictive “eligibility" policies [Electronic version]. Communication Disorders Quarterly, 23, 145-153. Nugent, M. D., & Constantine, L. L. (1988). Marital paradigms: Compatibility, treatment, and outcome in marital therapy. Journal of Marital and Family Therapy, 14, 351- 369. 138 Parette, H. P., & Angelo, D. H. (1996). Augmentative and alternative communication impact on families: Trends and future directions. Journal of Special Education, 30(1), 77-98. Parette, H. P., & Angelo, D. H. (1998). The impact of assistive technology devices and services on children and families. In S. L. Judge & H. P. Parette (Eds), Assistive technology for young children with disabilities: A guide to family-centered services (pp. 146-183). Cambridge, MA: Brookline Books. Parette, P., VanBiervliet, A., & Hourcade, J. J. (2000). Family-centered decision making in assistive technology, Journal of Special Education Technology, 15. Retrieved January 22, 2001, from http://jset.unlv.edu/15.1/parette/first.himl Pate, S. M. (1994). The dijferential impact of paradigmatic alignment/misalignment on post-divorce adjustment in divorced single-mother families. Unpublished doctoral dissertation, University of Florida, Gainsville. Pegorraro, C. (1999). Paradigrnatic orientation in home and private schooling families. Unpublished master's thesis, Michigan State University, East Lansing. Phillips, B., & Zhao, H. (1993). Predictors of assistive technology abandonment. Assistive Technology, 5, 36-45. Romski, M. A., & Sevcik, R. A. (2003). Augmented input: Enhancing communication development. In J. C. Light & D. R. Beukelrnan & J. Reichle (Eds), Communicative competence for individuals who use AAC: From research to eflecfive practice (pp. 147-162). Baltimore: Brookes Pub. Co. Scherer, M. J. (1996). Living in the state of stuck: How technology impacts the lives of persons with disabilities (2nd ed.). Cambridge, MA: Brookline Books. Schlosser, R. W. (1999). Social validation of interventions in augmentative and alternative communication. Augmentative and Alternative Communication, 15, 234-247. Schlosser, R. W. (2003). The eflicacy of augmentative and alternative communication: Towards evidence-based practice. San Diego: Academic Press. 139 Schlosser, R. W., & Blischak, D. M. (2001). Is there a role for speech output in interventions for persons with autism? A review. Focus on Autism and Other Developmental Disabilities, 1 6(3), 1 70-178. Siegel, E. B., & Cress, C. J. (2002). Overview of the emergence of early AAC behaviors: Progression from communicative to symbolic skills. In J. Reichle & D. R. Beukelrnan & J. C. Light (Eds), Exemplary practices for beginning communicators: Implications for AAC (pp. 25-57). Baltimore: Paul H. Brookes. Sigafoos, J ., & Drasgow, E. (2001). Conditional use of aided and unaided AAC: A review and clinical case demonstration. Focus on Autism and Other Developmental Disabilities, 16(3), 152-161. Smith, M. M., & Grove, N. C. (2003). Assyrnetry in input and output for individuals who use AAC. In J. C. Light & D. R. Beukehnan & J. Reichle (Eds), Communicative competence for individuals who use AAC: From research to ejfective practice (pp. 163-195). Baltimore: Brookes Pub. Co. Stevens, J. (1999). Power analysis, Intermediate statistics: A modern approach (2nd ed., pp. 121-143). Mahwah, N.J.: Lawrence Erlbaum Associates Publishers. Stockrnan, I. J., Boult, J ., & Robinson, G. (2004). Multicultural issues in academic and clinical education: A cultural mosaic. The ASHA Leader 9(13), 6-7, 20-22. Sweeney, L. A. (1999). Moving forward with families: Perspectives on AAC research and practice. In F. T. Loncke & J. Clibbens & H. H. Arvidson & L. L. Lloyd (Eds), Augmentative and alternative communication: New directions in research and practice (pp. 231-254). London: Whurr. Touliatos, J ., Perlrnutter, B. F., & Straus, M. A. (1999). Handbook of family measurement techniques. Newbury Park, CA: Sage. US. Census Bureau. (2000). QT -P32. Income dstribution in 1999 of huseholds and fmilies: 2000 [Data file]. Retrieved July 11, 2004, from http://factfinder.census.gov/servlet/SAFFPeople?_sse=on 140 US. Census Bureau. (2001). The population profile of the United States: 2000 [Internet Release]. Retrieved July 12, 2004, from http://www.census.gov/population/www/pop-profile/profile2000.html US. Census Bureau. (2002). Table I : Educational attainment of the population 15 years and over, by age, sex, race and Hispanic origin: March 2002 [Data file] . Population Division. Retrieved July 11, 2004, from http://www.census.gov/population/socdemo/education/ppl-169/tab01 .xls US. Census Bureau. (2004). Table 3: Annual estimates of the population by sex, race and Hispanic or Latino origin for the United States: April 1, 2000 to July 1, 2003 WC- EST 2003-03) [Data file]. Population Division. Retrieved July 12, 2004, from http://eire.census.gov/popest/data/national/tables/NC-EST2003-03.xls VanBiervliet, A., & Parette, H. P. (1999). Families, cultures and AAC [CD-ROM]. Little Rock, AR: Southeast Missouri State University and University of Arkansas for Medical Sciences. Villarruel, F. A., Imig, D. R., & Kostelnik, M. J. (1995). Diverse families. In E. E. Garcia & B. M. McLaughlin (Eds), Meeting the challenge of linguistic and cultural diversity in early childhood education (V 01. 6, pp. 103-124). New York: Teachers College Press. Ward, M. (1997). Family paradigms and older-child adoption: A proposal‘tor matching parents' strengths to children's needs. Family Relations, 46, 257-262. Zangari, C., & Kangas, K. (1997). Intervention principles and procedures. In L. L. Lloyd & D. R. Fuller & H. H. Arvidson (Eds), Augmentative and alternative communication: A handbook of principles and practices (pp. 235-253). Boston: Allyn and Bacon. 141 ullllrllilllgllgimlint"