USF Libraries
USF Digital Collections

Evaluating positive behavior support plan implementation in the home environment of young children with challenging behavior

MISSING IMAGE

Material Information

Title:
Evaluating positive behavior support plan implementation in the home environment of young children with challenging behavior
Physical Description:
Book
Language:
English
Creator:
Duda, Michelle A
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Component analysis
Early childhood
Family
Maintenance
Dissertations, Academic -- Special Education -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: In recent years, a central focus of the field of early intervention/early childhood special education has been to investigate ways to effectively support young children with challenging behavior and their families (Center for Evidence-Based Practice: Young Children with Challenging Behavior, 2003; DEC, 1999). Positive behavior support (PBS) is one of the most promising evidence-based practices for young children with challenging behavior and their families. The central purposes of PBS are to both help people develop and engage in socially desirable behaviors and to help minimize patterns of socially stigmatizing responding (Koegel, Koegel, and Dunlap, 1996). Research documenting the utility and applicability of PBS with preschool-aged populations remain scarce, particularly within natural environments (e.g., Blair, Umbreit, and Eck, 2000; Duda, Dunlap, Fox, Clarke, and Lentini, 2004; Moes and Frea, 2000).Several gaps in the research remain, including studies incorporating natural intervention agents, natural settings, and studies measuring technical aspects of behavior change (e.g., maintenance). Though studies of maintenance may be difficult to execute, they may provide researchers with a greater understanding of which factors in the change process are most critical to successful implementation, as well as to enhance the goodness of fit between specific plan components and the ecology in which implementation occurs (Albin, Lucyshyn, Horner, and Flannery, 1996). The purpose of this research study was to first assess the relationship of support plan components to behavior change, and then systematically fade the functional components, reducing the plan to naturalistic strategies that may be easy for the family to use over time.Results indicated each of the three child participants consistently maintained low levels of challenging behavior and high levels of engagement within each routine, despite the fact that clear functional relationships among individual intervention components were not attained. Procedural fidelity data indicated that intervention components were both implemented by the mother on a consistent basis and were easily adapted into natural family routines over time.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Michelle A. Duda.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 135 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001709508
oclc - 68623078
usfldc doi - E14-SFE0001297
usfldc handle - e14.1297
System ID:
SFS0025618:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Evaluating Positive Behavior Support Plan Implementation In The Home Environment Of Young Children With Ch allenging Behavior by Michelle A. Duda A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Special Education College of Education University of South Florida Co-Major Professor: Lise Fox, Ph.D. Co-Major Professor: Al bert Duchnowski, Ph.D. Glen Dunlap, Ph.D. Ann Cranston-Gingras, Ph.D. Date of Approval: October 10, 2005 Keywords: component analysis, ear ly childhood, family, maintenance Copyright 2005 Michelle Duda

PAGE 2

i Table of Contents Abstract iv List of Tables vi List of Figures vii Chapter One. Introduction 1 Purpose 4 Research Questions 5 Definitions 5 Artificial Intervention Components 5 Challenging Behavior 6 Functional Relationship 6 Procedural Fidelity 6 Maintenance 7 Systematic Fading 7 Chapter Two. Literature Review 8 Positive Behavior Support (PBS) 8 Definition and Theoretical Framework 8 Empirical Support 10 Gaps in Implementation for Y oung Children With Challenging Behavior 11 Multicomponent Behavior Support Plans 12 Contextual Fit 15 Longitudinal Research 16 Programmatic Features of Maintenance 18 Component Analysis 21 Conclusion 23 Chapter Three. Methodology 25 Purpose 25 Participants and Setting 25 Pilot Research Study: Assessmen t and Support Plan Implementation 27 Measurement and Design 27 Reliability 28 Procedural Fidelity 29

PAGE 3

ii Procedures 30 Data Analysis 30 Summary of Results 31 Research Study: Support Pl an Component Evaluation 32 Purpose 32 Research Questions 32 Participants and Setting 32 Dependent and Independent Variables 33 Measurement and Design 35 Reliability 35 Procedural Fidelity 36 Design 36 Procedures 37 Intervention Component Reduction 37 Component Analysis 39 Streamlined Plan 40 Systematic Fading of Artificial Components 40 Natural Only 41 Unanticipated Procedural Variations 41 Chapter Four. Results 42 Intervention Component Reduction 42 Procedural Fidelity Criterion 44 Parent Rating Scale 45 Summary 47 Systematic Behavioral Observations 47 Data Analysis 48 Component Analysis 48 Composite Challenging Behavior 49 Pilot Study 49 Component Analysis 50 Natural Only Condition 52 Max 53 Pilot Study 53 Component Analysis 54 Natural Only Condition 55 Zak 55 Pilot Study 55 Component Analysis 56 Natural Only Condition 57 Emmy 57 Pilot Study 57 Component Analysis 58 Natural Only Condition 59 Summary of Natural Only Conditions 59

PAGE 4

iii Summary 61 Chapter Five. Discussion 63 Review of Research Questions 63 Research Questions 64 Research Question #1: Which elements of a multicomponent intervention that was effective in reducing levels of challenging behavior are functional in maintaining low levels of challenging behavior? 64 Assessment of Functional Relationships 65 Natural Only Condition 68 Setting Event Variables 69 Research Question #2: Given that some elements are demonstrated to be functional in maintaining low levels of challenging behavior, is it possibl e to use a systematic fading procedure so that selected el ements are no longer needed to maintain low levels of challenging behavior? 71 Limitations 73 Contributions to Research and Practice 75 Conclusions 80 References 82 Appendices 101 Appendix A: Figures 102 Appendix B: Intervention Component Reduction Data 112 Appendix C: Natural Only Implementation Patterns 121 About the Author End Page

PAGE 5

iv Evaluating Positive Behavior Support Plan Im plementation in the Home Environment of Young Children with Challenging Behavior Michelle Duda ABSTRACT In recent years, a central fo cus of the field of early intervention/early childhood special education has been to investigate ways to effec tively support young children with challenging behavior and thei r families (Center for Eviden ce-Based Practice: Young Children with Challenging Behavior, 2003; DEC, 1999). Positive behavior support (PBS) is one of the most promising evidence-based pract ices for young children with challenging behavior and their families. The central purposes of PBS are to both help people develop and engage in socially desira ble behaviors and to help minimize patterns of socially stigmatizing responding (K oegel, Koegel, & Dunlap, 1996). Research documenting the utility and appl icability of PBS with preschool-aged populations remain scarce, particularly within natural environments (e.g., Blair, Umbreit, & Eck, 2000; Duda, Dunlap, Fox, Clarke, & Le ntini, 2004; Moes & Frea, 2000). Several gaps in the research remain, including studi es incorporating natural intervention agents, natural settings, and studies measuring tec hnical aspects of be havior change (e.g., maintenance). Though studies of maintenance may be difficult to execute, they may provide researchers with a great er understanding of which fa ctors in the change process are most critical to successful implementation, as well as to enhance the “goodness of fit”

PAGE 6

v between specific plan components and the ecology in which implementation occurs (Albin, Lucyshyn, Horner, & Flannery, 1996). The purpose of this research study was to first assess the re lationship of support plan components to behavior change, and then systemat ically fade the functional components, reducing the plan to naturalistic st rategies that may be ea sy for the family to use over time. Results indi cated each of the three child participants consistently maintained low levels of chal lenging behavior and high levels of engagement within each routine, despite the f act that clear functional relations hips among individual intervention components were not attaine d. Procedural fidelity data indicated that intervention components were both implemented by the moth er on a consistent ba sis and were easily adapted into natural family routines over time.

PAGE 7

vi List of Tables Table 1 Mean IOA Coefficients by R outine and Child: Pilot Study 28 Table 2 Operational Definitions and Dependent Variables for the Research Study 34 Table 3 Mean IOA Coefficients by Routine and Child: Research Study 36 Table 4 Intervention Components Included in the Component Analysis Procedures 51 Table 5 Changes in Variability a nd Level of Challenging Behavior and Engagement for Max: Pilot Study 53 Table 6 Changes in Variability a nd Level of Challenging Behavior and Engagement for Zak: Pilot Study 56 Table 7 Changes in Variability a nd Level of Challenging Behavior and Engagement for Emmy: Pilot Study 58 Table 8 Changes in Level of Co mposite Challenging Behavior Across Conditions and Routines 61 Table 9 Intervention Component Reduction Data: Clean Up 113 Table 10 Intervention Component Reduction Data: Twin Play 115 Table 11 Intervention Component Reduction Data: All Play 117 Table 12 Intervention Component Reduction Data: Dinner 119 Table 13 Natural Only Implem entation Patterns: Clean Up 122 Table 14 Natural Only Implemen tation Patterns: Twin Play 123 Table 15 Natural Only Implem entation Patterns: All Play 124 Table 16 Natural Only Implem entation Patterns: Dinner 125

PAGE 8

vii List of Figures Figure 1 Composite Percentage of In tervals with Challenging Behavior 103 Figure 2 Composite Percentage of In tervals with Challenging Behavior and Engagement: Baseline a nd Intervention—Clean Up and Twin Play 104 Figure 3 Composite Percentage of In tervals with Challenging Behavior and Engagement: Baseline and Intervention—Dinner and All Play 105 Figure 4 Composite Percentage of In tervals with Challenging Behavior: Component Analysis—Cl ean-Up and Twin Play 106 Figure 5 Composite Percentage of In tervals with Challenging Behavior: Component Analysis—Di nner and All Play 107 Figure 6 Component Analysis Individual Percentage of Intervals with Challenging Behavior: Clean-Up 108 Figure 7 Component Analysis Individual Percentage of Intervals with Challenging Behavior: Twin Play 109 Figure 8 Component Analysis Individual Percentage of Intervals with Challenging Behavior: All Play 110 Figure 9 Component Analysis Individual Percentage of Intervals with Challenging Behavior: Dinner 111

PAGE 9

1 Chapter One Introduction In recent years, a central fo cus of the field of early intervention/early childhood special education (EI/ECSE) has been to in vestigate ways to ef fectively support young children with challenging be havior and their families (Center for Evidence-Based Practice: Young Children with Challenging Behavior, n.d.; Center on the Social and Emotional Foundations for Early Learning, n.d.; DEC, 1999). Cha llenging behavior refers to “any repeated pattern of behavior or perc eption of behavior th at interferes with or is at risk of interfering with optimal learning or engageme nt in pro-social interactions with peers and adults” (Sm ith & Fox, 2003). While typical ly developing young children (i.e., under six years of age) often demonstrate such beha vior (e.g., tantrums, aggression, dropping to the floor, excessive crying), concerns usually reso lve as their repertoire of social and communication skills in crease. In contrast, childr en with challenging behavior continue to demonstrate these behaviors over time, using their be havior as a primary means of communication. As patterns of challenging beha vior intensify, opportunities for meaningful social interac tion and/or learning ar e lost, resulting in a host of negative child and family outcomes, su ch as eligibility for special education services, family stress, community isolation, and psychiat ric diagnosis/treatment (Campbell, 1994; Huffman, Mehlinger, Kerivan, Cava naugh, Lippitt, & Moyo, 2001; Keenan & Wakschlag, 2000; McEvoy & Reichle, 1995; Pierce, Ewing, & Campbell, 1999).

PAGE 10

2 Despite the variety of av ailable intervention proced ures to support young children with challenging behavior and their families available evidence supporting their efficacy varies drastically. Positiv e behavior support (PBS) is among the most promising evidence-based practices for young children with challenging behavior and their families. An empirically-supported model of problem so lving designed to enhance the capacities and skills of individuals and their families (Carr, Horner, Turnbull, Marquis, McLaughlin, McAtee, Smith, Ry an, Ruef, Doolabh, & Braddo ck, 1999; Horner, Dunlap, Koegel, Carr, Sailor, Anderson, Albin, & O’ Neill, 1990; Koegel, Koegel, & Dunlap, 1996), the central purposes of PBS are to both help people deve lop and engage in socially desirable behaviors and to he lp minimize patterns of so cially stigmatizing responding (Koegel, Koegel, & Dunlap, 1996). Research documenting the utility and ap plicability of PBS with preschool-aged populations is in its relative infancy. At present, studi es of preschoolers conducted within natural environments are relative ly scarce (e.g., Blair, Umbreit, & Eck, 2000; Duda, Dunlap, Fox, Lentini, & Clarke, 2004; Frea, Arnold, & Vittimberga, 2001; Schepis, Ownbey, Parsons, & Reid, 2000). Several gaps in the research remain, including studies incorp orating natural intervention agents (e.g., parents, siblings), natural settings (e.g., home, daycare/ preschool, community), and studies measuring technical aspects of behavior change, such as treatment integrity, maintenance, or generalization. Studies are needed to docume nt the extent by which support plans can be implemented by parents, child care providers, and teachers with a degree of fidelity and consistency required to ensure meani ngful outcomes over time.

PAGE 11

3 In addition to investigations of its tec hnical aspects, research ers continue to look toward ways to make behavioral interv ention more manageable for parents and community providers. A majo r goal has been to not onl y understand how to develop effective behavior support pl ans, but also to discover the means by which effective behavior plans may be maintained by careg ivers over time. Despite the apparent simplicity of such questions, challenging behaviors displayed by young children are often complex and may occur for eith er multiple reasons or for entirely different reasons in different contexts (e.g., setti ngs, routines, individuals, time of day). As complexity increases and the number of intervention com ponents expand, the relationship of each component to the desired outcome is often unkno wn. As a result, pa rents and caregivers are often asked to implemen t multi-component s upport plans over a prolonged period of time without truly knowing which aspects of the plan are mo st effective and likely to sustain meaningful be havior change. Following this rationale, one of the most challenging issues facing researchers, practitioners, and families alik e is maintenance. Maintenance is defined as “a stimulus control relationship that is st able or consistent across time” (Horner & Billingsley, 1988). At present, the existing literature base lacks studies that investigate critical features of support plan durability—researchers may wish to consider evaluating behavior support plans to learn why the process it self or which indivi dual components were most critical to achieving durable and meaningf ul outcomes. Likewise, the field also lacks studies of acquisition—studies often do not use an ad equate degree of experimental control allowing one to demonstrate clear functional relationships between the independent variable and changes in behavior (R usch & Kazdin, 1981). Though studies of

PAGE 12

4 maintenance and acquisition ma y be challenging and difficul t to execute, they may provide researchers with a gr eater understanding of which components are most essential to affect change or which factors in the ch ange process itself ar e most critical to successful implementation of the PBS model. Likewise, such studies may also help discover ways to streamline support plans, thereby enhancing both the practicality and durability of support plans, as well as the “goodness of fit” existing between specific features of a support plan and the ecology in which an intervention is implemented (Albin, Lucyshyn, Horner, & Flannery, 1996). Purpose This study was an extension of a recently conducted pilot re search study. The pilot study, which was presented by Fox, Clarke, and Duda (2005), provided an examination of the effect of multi-com ponent behavior support plans o n the challenging behavior of thre e young children within family routines. Documenting behavior change using a concurrent multip le baseline design across routines, this researcher and her colleagues implemented and measured the outcomes of four behavior support plans across four routines within the family’s hom e environment. As the family moved closer toward implementing and mainta ining their use of th e four plans on their own to support the engagement and reduced challenging behavi or of the children, it was unknown exactly which components were related to behavior change. Consequently, the purpose of this study was to first assess th e relationship of suppor t plan components to levels of problem behavior, and then systematically fade the functional components, reducing the plan to naturalistic strategies that may be easy for th e family to use over time.

PAGE 13

5 Research Questions 1. Which elements of a multi-c omponent intervention that was effective in reducing levels of challenging behavior are functiona l in maintaining low le vels of challenging behavior? 2. Given that some elements are demonstrated to be functional in maintaining low levels of challenging behavior, is it possible to use a systematic fading procedure so that selected elements are no longer needed to maintain the low levels of challenging behavior? Definitions For the purpose of this st udy, the following definitions were used. As a measure of consistency, selected definitions coinci ded with those specified by the Center for Evidence-Based Practice: Young Children w ith Challenging Behavior (2003) whenever possible. Artificial Intervention Components “Artificial” intervention components we re defined as components within a multicomponent positive behavior support plan that were introduced by this researcher during the pilot study and judged by the natural intervention ag ent as being cumbersome, requiring special materials, or difficult to implemen t in a variety of natural settings. Examples include social storie s, self-monitoring materials, and antecedent modifications (e.g., predetermined seating arra ngements, using music to cue the beginning of a routine).

PAGE 14

6 Challenging Behavior Smith & Fox (2003) defined “challenging be havior” as “any repeated pattern of behavior, or perception of behavior, that interferes with or is at risk of interfering with optimal learning or engagement in prosocial in teractions with peers and adults” (p. 6). Functional Relationship According to Sulzer-Azaroff and Mayer (1991 ), the term “functional relationship” refers to “a lawful relationship between values to two variables.” The authors noted that, “a dependent and independent variable are c onsidered to be functionally related if the behavior changes systematically with changes in value of the independent variable” (p. 590). Natural Intervention Components “Natural” intervention components were de fined as components that were either used by the natural interventi on agent prior to the initiati on of the pilot study or were judged by the natural intervention agent to be easily adaptable to natu ral family routines. Examples include specific praise, verbal cues, and choice making. Procedural Fidelity Procedural fidelity was defined as “the extent to which the independent variable is implemented” (LeLaurin & Wolery, 1992). Used to provide an estimate of the quality of an intervention’s implementa tion over time, the term “procedural fidelity” may be considered equivalent to other terms such as “treatment integrity,” “procedural reliability,” or “fidelity of treatment.”

PAGE 15

7 Maintenance Maintenance was defined as “a stimulus c ontrol relationship that is stable or consistent across time” (Hor ner & Billingsley, 1988). Systematic Fading According to Sulzer-Azaroff and Mayer ( 1991), the term fading referred to “the systematic, gradual removal of usually artifici al or intrusive prompt s, or discriminative stimuli such as direct ions, imitative prompts, physical guidance, and other cues.” The authors noted that the system atic fading procedure is used to “foster independence from supplemental prompts, and/or to shift control over to the stimuli design ated to evoke the response” (p. 590).

PAGE 16

8 Chapter Two Literature Review The purpose of this literatu re review was to present a nd discuss relevant research findings that provide a rationale for this research study. Toward such a goal, research literature have been presented and discus sed pertaining to positive behavior support (PBS) and specific gaps in its implemen tation for young children with challenging behavior. Specific topics we re discussed relative to bot h PBS and gaps in the PBS literature base for young children with challengi ng behavior including: 1) definition of PBS and theoretical framework; 2) empi rical support for PBS; 3) multi-component behavior support plans; 4) c ontextual fit; 5) longitudinal research; and 6) programmatic features of maintenance. Positive Behavior Support (PBS) Definition and Theoretical Framework Traditionally, challenging behavior has been addressed through the implementation of aversive appr oaches in which the behavior is responded to in a manner intended to reduce the future occurrence of challenging behavior (Horner et al., 1990). This type of consequence for disruptive beha vior is equivalent to what is commonly referred to as “punishment,” with attention directed toward disruptive behavior rather than providing the skill buildi ng opportunities for appropriate replacement behaviors. In response to such concerns, positive behavior support (PBS) emerged as an empirically-

PAGE 17

9 supported model of problem solving designed to enhance the capaci ties and skills of individuals and their families (Horner et al., 1990; Koegel Koegel, & Dunlap, 1996; Carr et al., 2002). PBS comprises a set of comprehensive inte rvention strategies custom designed to both help people de velop and engage in socially desirable behaviors and to help minimize patterns of soci ally stigmatizing behavior (Koegel, Koegel, & Dunlap, 1996). Descended from psychology and applied behavior analysis, PBS “represents an evolution toward a new appl ied science that 1) views consumers of research as collaborative partners; 2) values ecological, social, and internal va lidity; 3) seeks to promote lifestyle change; and 4) views social systems as units of analysis and intervention” (Carr et al., 1999; Turnbull, Turnbull, Wehmey er, & Park, 2003). Merging a conceptual framework for und erstanding the functional relati onships associated with a child’s behavior with a goal of achieving meaningful a nd durable lifestyle change, research indicates that PBS may be the best treatment approach toward the enhancement of a child and family’s qual ity of life (Singer, Goldberg-Hamblin, Peckham-Hardin, Barry, & Santarelli, 2002; Turnbull et al ., 2003). The PBS process incorporates the use of functional assessment to help lead key stakeholders to understand the purpose or function of challenging behavior, and the development of support strategies for prev enting challenging behavi or and teaching new skills (Fox et al., 2003; Fox, D unlap, & Cushing, 2002). Worki ng toward goals identified by the child’s parents and car egivers, PBS utilizes scie ntifically-endorsed research practices to help minimize ch allenging behavior a nd to teach new ski lls that are both more positive and socially ap propriate (Koegel, Koegel, & Dunlap, 1996). Support plans developed using the PBS model are individu alized and collaboratively designed to

PAGE 18

10 promote a functional understandi ng of the child’s behavior, prevent or minimize future occurrences of challenging behavior, enha nce critical skills (e.g., communication, language, socialization), and ultimately, im prove lifestyles and qua lity of life. Empirical Support A growing body of research has accumulate d documenting the efficacy of PBS as an empirically-supported practice (Carr, Ho rner, Turnbull, Marquis, McLaughlin, McAtee, Smith, Ryan, Ruef, Doolabh, & Bra ddock, 1999; Horner, D unlap, Koegel, Carr, Sailor, Anderson, Albin, & O’Ne ill, 1990; Koegel, Koegel, & Dunlap, 1996; Lucyshyn, Dunlap, & Albin, 2002). Initiall y, the vast majority of work in this area has either focused on conceptual issues (e.g., Carr et al., 1999; Dunlap & Fox, 1996; Horner et al., 1990; Weigle, 1997) or case st udies illustrating individu al-level support (e.g., Vaughn, Dunlap, Fox, Clarke, & Bucy, 1997; Clarke, Worcester, Dunlap, Murray, & Bradley Klug 2002; Dunlap et al., 1993; Vaughn, Clarke, & Dunlap, 1997; Dunlap, Foster-Johnson, & Robbins, 1990). Since then, PBS has expanded across envi ronments, populations, age ranges, and levels of prevention (i.e., tertiary to primary). Empirical demonstrations of the utility and applicability of PBS research and practice can now be found for children with a variety of medical/developmental di sabilities and challeng ing behaviors relative to: 1) a wide array of natural, complex community environments, including homes, general and special education classrooms, lib raries, churches, banks, restaurants, and retail stores (e.g., Carr et al ., 1999; Clarke et al., 2002; D unlap et al., 1995; Kern & Dunlap, 1999; Lucyshyn, Albin, & Nixon, 1997; Vaughn et al., 1997); and 2) primary, secondary, and tertiary levels of preventi on (Sugai et al., 2000; Lewis & Sugai, 1999;

PAGE 19

11 Sugai & Horner, 1994, 1999; Todd, Horner, Suga i, & Sprague, 1999; Sugai, Sprague, Horner, & Walker, 2000; Turnbull et al., 2002; Walker et al., 1996). Gaps in the PBS Literature Base for Young Children wi th Challenging Behavior Despite its rapid growth, several gaps in the PBS literature base remain. As reported by Carr and his colleag ues (1999), the field is work ing toward addressing five primary research gaps: 1) in creasing implementation of lifest yle change interventions; 2) measuring stimulus and respon se generalization; 3) condu cting research studies in applied family settings and contexts; 4) measuring the outcomes of multi-component stimulusand reinforcement-ba sed intervention plans as th ey are implemented within typical community settings; and 5) exploring ways to efficien tly modify environments as a means of preventing occurrences of challengi ng behavior (p. 75). In addition to these areas of interest, there has been a nationally recognized effort to at tend to the needs of preschoolers with challenging behavior and their families (Center for Evidence-Based Practice: Young Children with Challenging Behavior, 2003; Center on the Social and Emotional Foundations for Early Learni ng, 2003; DEC, 1999). From a prevention standpoint, children with challenging behavior who receive services and supports in their preschool years may acquire cr itical social and communica tion skills that serve as a foundation for long-term growth and devel opment (Bricker, 1992; Dunlap & Fox, 1996; National Research Council & Institute of Medi cine, 2000). Given that services designed to enhance a child’s progression of social and communication skill acquisition have the potential to minimize or prev ent subsequent delays late r in life, populations of young children with challenging behavi or represent an opportunity fo r the field of PBS to make a lasting and meaningful contribution to th e well-being of children and families.

PAGE 20

12 In support of such a goal a growing number of applie d research studies have demonstrated the applicability of PBS as a means of supporting young children with challenging behavior within hom e and community preschool sett ings (e.g., Duda et al., 2004; Frea, Arnold, & Vittimberg a, 2001; Lawry, Danko, & Stra in, 1999; Walker, Stiller, & Golly, 1999). Despite such progress, much work remains, particularly with regard to addressing gaps identified by Ca rr and his colleagues within natural preschool contexts (e.g., measuring outcomes of multi-compone nt stimulusand reinforcement-based intervention plans, measuremen t of maintenance outcomes). Multi-Component Behavior Support Plans One of the core characte ristics of PBS entails th e use of multi-component behavior support plans incorporating stimul usand reinforcement-based strategies (Lucyshyn, Horner, Dunlap, Albin, & Ben, 2002) Multi-component behavior support plans are explicitly designed to prevent and teach; children are taught age-appropriate social and communication skills, natural in tervention agents (e.g., parents, teachers, siblings) are taught to implem ent effective support strategies and natural environments are redesigned to prevent future occurrences of challenging behavior (i.e., environmental modification). When used together, multi-component behavior support plans incorporating both antecedentand consequence-based interven tion strategies help make challenging behavior functionally irreleva nt and less effective than using the functionally-equivalent, age-ap propriate skills taught in replacement (Favell & Reid, 1988; Horner et al., 1990; O’Neill et al., 1997). Relative to young children w ith challenging behavior, multi-component behavior support plans poses a challeng e for at least three reasons First, multi-component

PAGE 21

13 behavior support plans limit one’s ability to determine the impact of a specific intervention strategy. Unless one is able to assess changes in dependent measures when individual intervention com ponents are systematically in troduced and w ithdrawn, the efficacy of individual intervention com ponents cannot be measured when a multicomponent support plan is implemented. Although both this concern and replicability are frequently cited limitations a ssociated with PBS research st udies (e.g., Kern et al., 1994; Dunlap, White, Vera, Wilson, & Panacek, 1996; Moes & Frea, 2000), it is equally important to note that a centra l contribution of the PBS literature has been to validate the assertion that multiple interv entions may be an optimal me ans of achieving meaningful and durable behavior change over time, as well as an optimal practice for use in achieving contextual fit with family life (Horner et al., 1990; Lucyshyn, Horner, Dunlap, Albin, & Ben, 2002; National In stitutes of Health, 1990). A second limitation related to what is known about multi-component behavior support plans pertains to lim ited case illustrations. Though multiple examples of multicomponent behavior support pl ans can be found within th e PBS literature base (e.g., Clarke et al., 2002; Dunlap, Foster-Johnson, Clarke, Kern, & Childs, 1995; Dunlap et al., 1996; Ervin, Kern, Clarke, DuPaul, Dunlap, & Friman, 2000; Kern, Childs, Dunlap, Clarke, & Falk, 1994; Vaughn et al., 1997), relatively few ca n be found that demonstrate the ease of multi-component plans within natural environments or with natural intervention agents for young ch ildren with challenging behavi or (e.g., Blair, Umbreit, & Bos, 1999; Dunlap & Fox, 1999; Galensky, Mi ltenberger, Stricker, & Garlinghouse, 2001; Moes & Frea, 2000). C onsequently, it is reasonable to question the relative

PAGE 22

14 efficacy of multi-component behavior support plan s and/or degree of contextual fit with young children with challenging behavior. Finally, multi-component behavior s upport plans complicate implementation efforts. Given the fact that precise imple mentation of the independent variable (i.e., treatment integrity, Wolery, 1994) is of paramount importance to researchers, the degree of implementation precision may be logica lly jeopardized when one adds multiple intervention components. Th e implementation of a single component by an intervention agent is much easier than remembering an arra y of strategies. Additional concerns about the implementation of a “package” of components relates to the lack of inference that can be made about the value of any single component in the package or the need for all components to be used across stages of learning (e.g., initial acquisition, fluency, generalization). These reasons are in addi tion to the complexity of having multiple components, particularly in situations wh en: 1) multiple comp onents are sometimes needed only at the beginning of an intervention to promot e skill acquisi tion; and 2) intervention components vary in their relative ease of implem entation. Such concerns are only magnified within an appl ied research context, wher e multi-component behavior support plans are implemented by natural intervention agents within natural environments (e.g., home, preschool/daycare, co mmunity). In addition, researchers rarely take into account issues pertaining to “goodness of fit,” such as th e natural intervention agent’s perspective of an inte rvention component’s relative impor tance or the amount of effort and inconvenience associated with its implementation (Albin et al., 1996). Although the research literature on contextual fit (describ ed below) has contributed to increases in treatment fidelity (e.g., Albi n et al., 1996; Harrower, Fox, Dunlap, & Kincaid, 1999), it is

PAGE 23

15 clear that additional research in this area is needed, par ticularly with respect to case illustrations of multi-component behavior support plan implementation for young children with challenging behavior by na tural intervention agents within natural environments. Contextual Fit Contextual fit refers to the congruence existing between sp ecific features and components of behavior s upport plans and the ec ological and interpersonal variables relating to individuals and environments (Albin, Lucyshyn, Horner, & Flannery, 1996; Bailey, Simeonsson, Winton, Huntington, Comf ort, Isbell, O’Donnell, & Helm, 1986; Harrower et al., 1999; Lucyshyn, Dunlap, & Albin, 2002). Contex tual fit is a term used to describe the degree to which behavi or support plans consider and accommodate variables associated with the individual targeted for suppor t (e.g., specific strengths and challenges, values, goals, and beliefs), variab les associated with in dividuals responsible for plan implementation (e.g., speci fic skills or strengths, valu es, goals, and beliefs), and environmental factors (e.g., speci fic features of an environmen t, available resources). It has been proposed that when behavior s upport plans possess a high degree of contextual fit, they are more likely to be: 1) implemented with accur acy and precision; 2) applied across natural contexts; 3) impl emented over a prolonged peri od of time; and 4) rated as being effective and useful (Albin et al ., 1996; Lucyshyn, Horner, Dunlap, Albin, & Ben, 2002). At present, studies of the use of positi ve behavior support with preschoolers who have challenging behavior conducted within natural environments co mprise only a small proportion of the PBS research literature (e.g., Duda et al., 2004; Dunlap & Fox, 1999;

PAGE 24

16 Frea, Arnold, & Vittimberga, 2001; Schepi s, Ownbey, Parsons, & Reid, 2000; Blair, Umbreit, & Eck, 2000). Despite the recent a bundance of PBS resear ch and practice and recent emphasis on young children and their families, relatively few studies have reported demonstrations of maintenance of PBS inte rventions for preschoolers within natural environments (e.g., Hancock & Kaiser, 2002; Moes & Frea, 2000; Schreibman, Whalen, & Stahmer, 2000; Wert & Neiswo rth, 2003) and with natural in tervention agents (Baker, 2000; Barry & Singer, 2001; Frea, Arnold, & V ittimberga, 2001). Likewise, the need exists to explore the extent to which com ponents of corresponding behavior support plans are implemented with fidelity (e.g., identifying goals, coll ecting information, developing hypotheses, designing and implementing s upport plans, monitoring and evaluating interventions over time, maintenance, genera lization; Fox, Dunlap, Hemmeter, Joseph, & Strain, 2003; Koegel, Koegel, & Dunlap, 1996). Studies are needed to document the extent by which such support plans can be im plemented by parents, child care providers, and teachers with a degree of fidelity and consistency requ ired to ensure meaningful outcomes for young children with challeng ing behavior within their natural environments. Consequently, it is reasonable to conclude that demonstrations for preschoolers within natural environments (e.g., home, preschool/daycare, community) and with natural intervention agents (e.g., pare nts, teachers, siblings) are not only needed, but may serve an instrumental role in working toward th e articulation of recommended practices in facilitating effective intervention for parents and community providers alike. Longitudinal Research Another important way by which the exis ting literature base in PBS may be enhanced is through longitudina l research (Albin, Dunlap, & Lucyshyn, 2002; Carr et al.,

PAGE 25

17 1999; Dunlap, Clarke, & Steiner, 1999; Horner et al., 1990). One of the most convincing ways in which to demonstrate the durability of an intervention, longitudinal research is defined as “a type of investigation that involves describing changes in a sample’s characteristics over a specified period of time” (Gall, Borg, & Gall, 1996). Though uncommon, research documenting the sustainability and relati ve effect of intervention components paired with clearly defined in tervention procedures have the potential to serve as seminal research studies for the fi eld. Such studies have been recommended by several experts in the fiel d, most notably Carr and his colleagues (1999) and Albin, Dunlap, and Lucyshyn (2002). Families are looking for long-term solutions to their child’s challenging behavior (Carr et al., 1999). More speci fically, Carr and his colleagues (1999) note: “consumers tend to be concerned about problem behavior over long periods of time,” and that “the database reveals a substantial gap between the needs of consumers for long-term demonstrations of efficacy and the interests of researchers who follow i ndividuals for short periods of time, most typically for less than six mont hs and in no case fo r more than two years” (p. 76). Articulating their need and rationale for co llaborative research with families, Albin, Dunlap, and Lucyshyn (2002) str ongly support such claims. To gether, the authors cited a professional “obligation to ex tend research on PBS and to fu rther establish the external, social, and ecological validity of re search outcomes on PBS” (p. 375).

PAGE 26

18 Programmatic Features of Maintenance In addition to the aforementioned gaps in the PBS literature, there is even less information available regarding programmatic features of interven tion studies, such as maintenance or generalization (Horner, Dunlap, & Koegel, 1988; Stokes & Baer, 1977; Stokes & Osnes, 1989). Favell and Reid ( 1988) noted that the definitions of both maintenance and generalization relate to the improvement of target behaviors under conditions of reduced or di scontinued treatment (p. 185). Maintenance is generally defined as “how well the interv ention effects last over time ” (Carr, Levin, McConnachie, Carlson, Kemp, & Smith, 1994) or “the durabil ity of target behaviors under natural environmental conditions” (Sulze r-Azaroff & Mayer, 1991). Some researchers have assessed maintena nce with either a return to baseline conditions with continued measurement of dependent variables in the absence of independent variable implementation (e.g., Baker, 2000; Buggey, Toombs, Gardener, & Cervetti, 1999; Garfinkle & Schwartz, 2002; Koegel, Harrower, & Koegel, 1999), whereas others have used a “follow-up” condition to as sess changes in dependent variables with continued implementation of the independent variab le over time either with or without modifications in contingencies (e.g., Arme ndariz & Umbreit, 1999; Barry & Singer, 2001; Hancock & Kaiser, 2002; Hupp & Reitman, 2000; Koegel, Symon, & Koegel, 2002). In either ci rcumstance, the purpo se of “maintenance” or “follow-up” conditions are to fade from th e contrived context of clinical treatment toward a more natural context for th e child and family. Discrepancies in the way that maintenan ce conditions are included in research studies and embedded within behavior support plans may be linked to the complexity of

PAGE 27

19 the concept itself. Despite its apparent simplicity, maintena nce is far more complex than it may appear. Maintenance occurs as a fu nction of a stimulus control relationship between an intervention and a target behavi or across time (Horner & Billingsley, 1988). In other words, maintenance exists when there is a consistent patter n of behavior change when an intervention is applied. However, maintenance is a dynamic construct; it is influenced by both the stability or consistency of a stimulus control relationship, and the ever-changing context in which the target beha vior is observed (e .g., the child’s natural environment, intervention agents, reinforcem ent contingencies, variables influencing skill acquisition, the pass age of time). As a means of achieving a better grasp of its complexity, rese archers have studied the interaction between generalization, main tenance, and skill acquisition variables relative to changes in stimulus conditions, re sponse requirements, a nd reinforcer values (Dunlap, Horner, Carr, Sailor, Turnbull, Koegel, & Koegel, 1998; Horner & Billingsley, 1988). These researchers have argued that maintenan ce of specific replacement behaviors and skills (i.e., alte rnative and desired be haviors) are affected by a combination of instructional, antecedent, and consequence variables, including: 1. Selection of efficient and effective alternative behaviors to teach; 2. Teaching alternative behaviors to high fluency/accuracy criteria; 3. Teaching alternative behaviors as general case skills; 4. Avoiding presentation of setting ev ents and discriminative stimuli for challenging behavior; 5. Continued presentation of discriminative stimuli fo r alternative behavior; 6. Ensuring regular opportunity to perform alternative behavior;

PAGE 28

20 7. Providing reinforcement for desired behavior; 8. Providing reinforcement for a lternative behavior; and 9. Extinguishing challenging behavior; (Dunlap et al., 1998; Horn er & Billingsley, 1988). Researchers suggest that cons ideration of these nine vari ables contribute to both the design of interventions and the design of the environment in which the child is expected to perform desired and alternative behaviors. However, the ultimate durability of socially appropriate target behaviors also depends on macro-level ecological variables operating on the individual’s environmen t, including family and provider support systems, the community, state and local agen cies, and political/cultural va lues (Bronfenbrenner, 1986; Dunlap et al., 1998; Dunlap & Plienis, 1988; Horner & Billingsley, 1988. Together, this conceptualization suggests that maintenance is far more complex and dynamic than it may initially appear, particular ly in circumstances when desi red and alternative behaviors are measured well beyond th e initial impl ementation of intervention procedures Though few studies prioritize the measurement of mainte nance, its inclusion is critically important, as it allo ws one to assess the utility a nd efficacy of an intervention after its initial implementation and demons tration (i.e., the inte rvention phase of a research study; Dunlap, Horner, Carr, Sail or, Turnbull, Koegel, & Koegel, 1998; Horner & Billingsley, 1988; Horner, Dunlap, & Koeg el, 1988). Carr and hi s colleagues (1999) conducted a systematic review of the PBS literature, repo rting that while a relatively small proportion of the studies using PBS measured long-te rm maintenance effects (i.e., 5 months beyond intervention), tw o-thirds of those reporting sh ort-term outcomes (i.e., less than 5 months beyond intervention demonstrat ed success (relative to a 90% reduction

PAGE 29

21 criterion). None of the studies reviewed measured maintena nce for follow-up periods of 25 months or more (p. 48). Given such findings, it is r easonable to conclude that research studies incorporating the maintenance of target behaviors may not only help extend the longevity of support plans for preschooler s, but also to measure, document, and strongly support th e utility of PBS technology for pr eschool-aged children and their families. Component Analysis. Central to the issue of maintenance are the concepts of measurement and design (Bailey & Burch, 2002; Kazdin, 1982; Rusc h & Kazdin, 1981; Shirley, Iwata, Kahng, Mazaleski, & Lerman, 1997). Given the fact that maintenance allows one to assess efficacy of an interventi on, one must pay particul ar attention to the specific means by which dependent variables ar e measured in the ma intenance condition. In order to systematically assess changes in the dependent variable over time, a greater degree of experimental contro l must occur, which necessita tes the use of a design that allows one to analyze functiona l relationships between change s in variables. Such an issue is magnified further wh en one considers th e limitations associ ated with drawing conclusions about individual interven tions embedded within multi-component intervention packages (i.e., gr oups of interventions or s upports that are implemented either simultaneously or in succession). Under such c onditions, assessment of the independent variable is complicated by the f act that a functional relationship cannot be determined relative to a specific intervention component unless the im pact of the component is isolated and its strength demonstrated over time (e.g., Dunlap et al., 1996; Ervin et al., 2000; Kazdin, 1982).

PAGE 30

22 The existing literature base s in both PBS and EI/ECSE pr esently lack studies that investigate why the process itse lf or which individual components are most critical to achieving durable and meaningful outcomes over time. As a result, it becomes difficult to enhance the field without consideration of which compon ents are most essential to affect change or which factors in the change process itself ar e most critical to successful implementation of the model. Given such circumstances, it be comes necessary to consider the means to systemat ically assess the effects of th e individual components of a comprehensive intervention package (B ailey & Burch, 2002; Kazdin, 1982; Kern, Wacker, Mace, Falk, Dunl ap, & Kromrey, 1995). The research literatur e indicates that several single case research designs can be used when conducting component analyses. Indicated for the evaluation of multicomponent intervention packages, component analyses are typica lly conducted using designs that briefly withdraw the treatment after its eff ect has been established, or through the use of assessment probes (Kazdin, 1982; Tawn ey & Gast, 1984). In a discussion of designs used to examine transfer of traini ng and response maintenance, researchers have specified thr ee groups of designs: 1) probe designs; 2) withdrawal designs; and 3) between-group designs (K azdin, 1982; Rusch & Kazdin, 1981). Though each type of design has its own distinct ad vantages and disadvantages, Kazdin (1982) cited withdrawal designs offe r the researcher th e unique opportunity to assess changes in performance (i.e., changes in the depend ent variable) while specific intervention components are systematic ally excluded or includ ed from a multi-component intervention package (p. 213). Toward this goal, three variations of the withdrawal design exist: 1) the sequential-withdrawal; 2) the partial-withdrawal; 3) the combined

PAGE 31

23 sequential and partial-withdrawal design. According to Kazdin, the sequentialwithdrawal design entails “gradually withdr awing different components of a treatment package to see if behavior is maintained,” whereas the intervention is gradually withdrawn across different persons or baseli nes in partial-withdr awal designs (pp. 213215). Alternatively, both designs may be combined, thereby allowing the researcher to preview which components are most likely to be maintained before they are completely withdrawn from a multi-componen t intervention plan (Kazdin, 1982). Conclusion The purpose of this literatu re review was to synthesize the available research literature pertaining to this st udy. Research was discussed rela tive to PBS and gaps in the PBS literature base for young ch ildren with challenging behavi or. As a result of this literature review, several findi ngs appear noteworthy. The re search literature provides convincing demonstrati ons of the efficacy of PBS, an emerging field of behavioral science that has been applied successfully in a number of ca pacities (i.e., individual-, classroom-, and school-wide implementation; diverse settings, intervention agents, age ranges, and clinical populations ). As the field continues to grow, researchers have begun to investigate the imp act of PBS upon young children with challenging behavior and their families. Despite encouragi ng results, the increasing preval ence and widespread impact of challenging behavior upon both the family and service system provides a strong rationale for the continued ap plication of PBS technology fo r this population of children (Campbell, 1994; Division for Early Childhood, 1999; Powell, Fi xsen, & Dunlap, 2003). Important gaps remain as the field contin ues to strive toward desired quality of life outcomes for young children with challenging behavior and their families. Relatively

PAGE 32

24 few research studies using PBS with young child ren with challenging be havior have been conducted in natural environments or with na tural intervention agents. Likewise, experts in the field have called for a dditional longitudinal research studies (e.g., Carr et al., 1999; Horner et al., 1990; Albin, Dunlap, & Lucy shyn, 2002), assessment of technical aspects of applied research studies (e .g., treatment integrity or mainte nance; Favell & Reid, 1988; Kazdin, 1982; Wolery, 1994), or measured outcomes asso ciated with individual components of multi-component behavior support plans. As researchers in both PBS and EI/ECSE strive to enhance the accountability and quality of their research (Bailey, McWilliam, Darkes, Hebbeler, Simeonss on, Spiker, & Wagner 1998; Carta, 2002; Guralnick, 2000), each of the abov e issues deserve careful cons ideration, especially if the intent of the research is to obtain a convincing demonstratio n of specific components of multi-component PBS plans for young children with challenging be havior and their families.

PAGE 33

25 Chapter Three Methodology Purpose This study constituted an extension of a recently conducted pilot research study. This chapter has been organized into three major sections: 1) participants and setting; 2) description of the pilot re search study (i.e., purpose, dependent and independent variables, measurement and design, procedures, data analysis, results) ; and 3) description of this study (i.e., purpose, research questions, dependent and independent variables, measurement and design, data analysis, lim itations, contributions to research and practice). Participants and Setting The participants in both the pilot and this research study were a mother, her 5 year-old daughter Emmy, a nd fraternal twin 3 year-o ld sons Max and Zak (all pseudonyms). The family’s home environmen t was selected as the setting for both studies. With regard to presenting concer ns, the family was initially concerned with Max, who had a history of failur e to thrive, feeding difficul ties, and expressive language delays (i.e., delays with bot h expressive language and artic ulation of speech sounds). After receiving a developmental evaluation through the local early intervention program, Max was determined eligible for language a nd behavior support se rvices. Max reportedly learned to demonstrate a wide variety of challenging behaviors (e.g., tantrums, hitting,

PAGE 34

26 biting, throwing toys, excessive crying, difficulties with turn taking, noncompliance, elopement). Despite parent reports that Max was a lovi ng, affectionate, and curious young child, Max’s behavior had consisten tly disrupted his family’s functioning, particularly during playtime and dinner routines. This researcher initially established cont act with Max and his family in response to his need for behavior s upport services. After initial meetings, interviews, and behavioral observations, this researcher observed that Max’s two siblings also demonstrated challenging behavi or. Max’s fraternal twin br other, Zak, was described as a very bright and inquisi tive child who has a well-developed interest in vehicles. In spite of these strengths, Zak’s moth er reported that he frequen tly demonstrated challenging behavior, in the form of hitting, food dumping, spitting, biting, noncompliance, throwing toys, excessive crying, difficulties with turn taking, and elopement. In addition, it was also apparent that Max a nd Zak’s older sister, Emmy, consistently demonstrated challenging behavior. Accordi ng to both her mother and this researcher’s observation, Emmy was an artistic child w ho enjoys her role as a leader for her two younger brothers (e.g., helping mother with household chores). However, it also ap peared that Emmy’s leadership skills had proven to be a challe nge at times, as she had been observed modeling and encouraging noncompliant behavi or and inappropriate language for her brothers. Prior to initiating the pilo t study, parental informed consent was obtained as a means of ensuring both permissi on and a degree of commitmen t to the completion of the study. Approval from the univers ity institutional review bo ard (IRB) was obtained as a means of ensuring the safety and confidenti ality of the entire family. Consent was

PAGE 35

27 pursued relative to particip ation, confidentiality, and th e use of videotaping and photography. This process was re peated prior to the initiation of this research study, with informed consent to precisely ma tch the procedures of the study. Pilot Research Study: Assessm ent and Support Plan Implementation The purpose of the p ilot research study was to deve lop an effective assessmentbased behavior support plan that could be implemented by the pa rent in family routines to reduce problem behavior and promote child engagement or indepe ndence in targeted family routines. The depend ent variables in this study we re engagement and challenging behavior, both of which were op erationally defined in the same manner as in this research study (Table 1). Challenging behavior was me asured relative to both individual children (i.e., Max, Zak, and Emmy) and as an overall composite of all three children (i.e., reflecting the mother’s percep tion of whether or not cha llenging behavior was present during the session). Th e independent variable in this study was the implementation of an assessment-based individual-level positive be havior support plan (Horner et al., 1990; Koegel, Koegel, & Dunlap, 1996). Measurement and Design Systematic behavioral observations were used to measure changes in the dependent variables over time. Rates of challenging behavior, composite challenging behavior, and engagement were scored by tr ained observers via vi deotape using a 10second continuous interval recording system. Data were collected relative to operational definitions and expressed as the percentage of intervals of a dependent variable’s occurrence. Each session was independen tly videotaped by this researcher and subsequently scored by three tr ained observers for occurrence of each dependent variable.

PAGE 36

28 The two observers watched the videotape simultaneously a nd independently scored the occurrence/nonoccurrence of each dependent variable using the interval recording system. Reliability Interobserver agreement was scored for occurrence, non-occurrence, and total IOA for each operationally defi ned dependent variable. Interobserver agreement (IOA) was scored for occurren ce, nonoccurrence, a nd total IOA for both engagement and total challenging behavior per child. Reliability was assessed on at least 33% of all videotaped sessions Total IOA scores (e.g., means, ranges) were calculated for each routine and child par ticipant with no less than a 87% mean Total IOA score obtained between data collectors. Composit e IOA data indicated that reliability was achieved at a level of 93% (range = 89100%) for clean up, 96% (range = 90-100) for twin play, 97% (range = 93-100%), and 95% (89-100%) for dinner. Reliability coefficients for each i ndividual child are list ed below in Table 1. Table 1 Mean IOA Coefficients By Rou tine and Child: Pilot Study Routine Child Variable Total IOA (Range) Total Challenging 97% (89-100%) Max Total Engagement 97% (83-100%) Total Challenging 88% (92-100%) Clean Up Zak Total Engagement 97% (89-100%) Total Challenging 96% (90-100%) Max Total Engagement 96% (91-100%) Total Challenging 97% (93-100%) Twin Play Zak Total Engagement 97% (95-100%) Total Challenging 98% (93-100%) Max Total Engagement 98% (92-100%) Total Challenging 97% (94-100%) Zak Total Engagement 97% (93-100%) Total Challenging 97% (93-100%) All Play Emmy Total Engagement 96% (86-100%) (Table Continues)

PAGE 37

29 Table 1 (Continued) Total Challenging 97% (91-100%) Max Total Engagement 95% (81-99%) Total Challenging 94% (89-100%) Zak Total Engagement 94% (84-100%) Total Challenging 95% (90-98%) Dinner Emmy Total Engagement 95% (86-99%) With regard to design, data were eval uated using a concurrent multiple baseline design across four routines (T awney & Gast, 1984). This de sign was used to both ensure consistent evaluation of changes in the depe ndent variable (i.e., composite challenging behavior) and an adequate degree of experimental control both within and across conditions. Procedural Fidelity Procedural fidelity data we re collected across routines and conditions by trained data collectors as a means of assessing the degree to which intervention components were implemented wi th integrity. Data were obtained on the specific intervention components that coul d be directly observed during sessions. Employing the same videotapes used to record the dependent variable s, observers scored whether components were implemented as speci fied in the support pl an corresponding to the specific routine. Observers used a checklist of eac h component from the support plan for the specific ro utine, scoring whether each indivi dual component was observed during the session (i.e., yes or no). After obtain ing an implementation fi delity score for each component per individual sessi ons, an average score of com ponent fidelity was calculated for each component across all intervention sessions These data were later used to inform decision-making during the inte rvention component reduction process of the subsequent research study.

PAGE 38

30 Procedures Once the family’s concerns were iden tified, this resear cher and colleagues conducted the pilot research study in the family’s home envi ronment. Grounded in the family’s goals and obtained func tional assessment data (O’N eill, Horner, Albin, Sprague, Storey, & Newton, 1997; Repp & Horner, 1999) this researcher facilitated the collaborative development of individuali zed, comprehensive behavior support plans targeting four routin es: clean up, “twin play” (i.e., pl aytime with Max and Zak), “all play” (i.e., playtime with Emmy, Max, and Za k), and dinner. Afte r brief periods of coaching, each support plan was implemented by the children’s moth er until stable rates of behavior change were obtained (i.e., change s in each of the three operationally-defined dependent variables). Interven tion components were customized to fit within the context of each routine, each consisting of a co mbination of prevention strategies (e.g., antecedent modifications, choice making, clear expectations), parent responses, and skillbuilding interventions (e.g., compliance with expectations, play skills, teaching rules, self-monitoring, leading activi ties for younger brothers). Data Analysis Visual analyses served as the primar y means of analyzing changes to the dependent measures across conditions (i.e., base line, intervention). Data were graphed in order to determine changes in trend and le vel across conditions (K azdin, 1982). Visual analyses of trend considered changes in di rection both within and between conditions. Changes in level were assess ed through visual inspection of the magnitude of each dependent variable. With regard to procedural fidelity, data analys is were expressed as

PAGE 39

31 mean percentage of sessions with fidelity of implementation, with IOA estimates to be calculated across components and routines fo r a minimum of 33% of all sessions. Summary of Results Visual analyses of the obtained data indicated the presence of several noteworthy patterns. In general, data indicated that rates of challenging behavior consistently decreased during the interven tion condition across routines per child, while rates of engagement increased during in tervention across routines per child. Visual inspection of both dependent measures revealed that changes occurred rela tive to level and trend. Relative to composite challenging be havior, data indicat ed that both the percentage of observed interv als and trend decreas ed during intervention across routines. Data obtained following a brie f break in the data indicate a continuation of the same patterns (i.e., lower levels of composite challenging be havior, decreasing trend, less variability). With respect to data collected relative to in dividual children, rates of challenging behavior were cons istently scored at lower le vels for Max, Zak, and Emmy during the intervention c ondition across all routines. Visual analyses indicated that rates of challenging behavior were consistently lower, less va riable, and recorded in a decreasing trend across interv ention conditions. Finally, rates of engagement were consistently scored at a higher level fo r Max, Zak, and Emmy during the intervention condition across each rou tine. While visual inspection re vealed inconsistent patterns of direction during baseline conditions, analyses of trend indi cated that rates of engagement either maintained a flat or increasing trend across intervention conditions. Scores were also more tightly dispersed during the intervention conditio ns for each of the children,

PAGE 40

32 suggesting that rates of engagement were mo re consistent and le ss variable during each intervention condition. Research Study: Support Plan Component Evaluation Purpose In this research study, expe rimental procedures were used to systematically identify the functional com ponents of the multicomponent behavior support plan used by the parent and then fade the functional components of the pl an that were “artificial” or identified by the parent as burdensome for continued impl ementation. The goal was for the family to fade components of their behavior support plans so that the pl ans consisted of natural strategies that were easier to maintain within ever yday routines and settings. Toward this end, the fo llowing research questions were articulated: Research Questions 1. Which elements of a multi-c omponent intervention that was effective in reducing levels of challenging behavior are functiona l in maintaining low le vels of challenging behavior? 2. Given that some elements are demonstrated to be functional in maintaining low levels of challenging behavior, is it possible to use a systematic fading procedure so that selected elements are no longer needed to maintain the low levels of challenging behavior? Participants and Setting Max, Zak, Emmy, and their moth er participated in this research study. The entire study was conducted within the family’s home environment. In addition, parental informed consent was obtained prior to the commencement of this study as a means of

PAGE 41

33 ensuring both permission and a degree of commitment to the comp letion of the study. Approval from the university institutional re view board (IRB) was obt ained as a means of ensuring the safety and confid entiality of the entire famil y. Consents were pursued relative to participation, conf identiality, and the use of vi deotaping and photography. Dependent and Independent Variables The dependent variables in this research study were the same as those included in the pilot research study (i.e., challenging behavior, composit e challenging behavior, and engagement; Table 2). Each session was independently videotaped by this researcher and subsequently scored by three tr ained observers for occurrence of each dependent variable. Challenging behavior and engagement was meas ured relative to both individual children (i.e., Max, Zak, and Emmy) and as an overa ll composite of all three children’s challenging behavior. From the mother’s perspective, challenging behavior was scored if it occurred during an interval regardless of which child demonstrated the behavior. The same method of coding composite challenging behavior was used as in the pilot research study. Occurrences of com posite challenging behavior were scored whenever challenging behavior was observed with any child within the 10-second interval.

PAGE 42

34 Table 2 Operational Definitions of Dependen t Variables for the Research Study Behavior Recording Instructions Challenging Behavior Score challenging behavior in activity if child exhibits any of the following for the majority of the interval: Inappropriate Verbalizations : Score any occurrence in interval of screaming (voice tone louder than normal), crying, whining [e.g., high pitched begging, complaining, or acting as if crying without tears (i.e., fake crying)]. Inappropriate Social Interactions : Score any occurrence in interval of verbal resistance (e.g., verbal utterances that sugges t resistance to the situation) such as “No!,” or “Stop!” If the child expressed desire to do something else in a negative tone but was not indicating resistance to the current situ ation, the behavior may be marked as an “Inappropriate Vocalization.” This behavior should also be scored for any occurrence of social interaction directed toward another th at is considered bossy (e.g., “Gimme that!,” or “Shut up!”), mocking, or berating another person (e.g., imitating mother’s verbal instruction, arguing with mother). Include statements made to siblings with a connotation to chide them into appropriate behavior (e.g., egging on, instigating statements), that may result in being reprimanded). Aggression : Score any occurrence in interval of child attempting or following through with hitting, kicking, biting, wrestling, or attempting to pick up another person. Also score if child destroys another’s property (e.g., knocks down others’ block castle currently playing with, grabs another’s toy, physical “tug of war,” or struggle with another over object. Continue marking “Aggression” for each interval involved with struggle until behavior terminates. Include property destruction or attempt to deface or destroy others’ toys or materials. Out of Area : Score any occurrence in interval of child leaving assigned area (e.g., leaves dinner table before finished with food, runs out of play area to get mother). Inappropriate Use of Materials : Score any occurrence in interval of behavior in which materials are used in a manner that is inappropriate or not what object was intended for (e.g., spitting out food, throwing toys, standing on dinner chair, jumping off table, slamming doors). If materials are used in completing aggression, mark both categories. Noncompliance : Score any occurrence in interval of child failure or refusal to follow instructions or directives for 5 or more seconds (e.g., Mother instructs, “Let’s clean up,” “Take your plate to the garbage can,” or if child runs away or continues playing). Engagement Score engagement in activity if child is appropriately following sequence of activity for the majority of the interval. Engagement may still be scored if challenging behavior is recorded. If child exhibits challenging behavior throughout entire 10-second interval, do not score as engaged.

PAGE 43

35 The independent variable in this study wa s the implementation of an assessmentbased individual-level positive behavior support plan across four rout ines (Horner et al., 1990; Koegel, Koegel, & Dunlap, 1996). The independ ent variable was selected because the intention of this research study was to develop and implement four efficient, contextually-fitting positive behavior s upport plans for implementation by natural intervention agents within naturally-occ urring family routines over time. Measurement and Design Systematic behavioral observations of rates of dependent measures (i.e., challenging behavior, composit e challenging behavi or, engagement) were conducted and scored via videotape using a 10-second continuou s interval recording tool. Data collection corresponded to oper ational definitions and were expressed as the percentage of intervals of composite challe nging behavior, as well as perc entage of sessions in which a single intervention component was implemented as specifi ed in the ro utine-specific support plan. Reliability. Interobserver agreement was scored for occurrence, non-occurrence, and total IOA for each operationally defi ned dependent variable. Interobserver agreement (IOA) was scored for occurren ce, nonoccurrence, a nd total IOA for both engagement and total challenging behavior per child. Reliability was assessed on at least 33% of all videotaped sessions Total IOA scores (e.g., means, ranges) were calculated for each routine and child par ticipant with no less than a 88% mean Total IOA score obtained between data collectors. Thes e data are presented below in Table 3.

PAGE 44

36 Table 3 Mean IOA Coefficients By Routin e and Child: Re search Study Routine Child Variable Total IOA (Range) Total Challenging 98% (90-100%) Max Total Engagement 95% (76-100%) Total Challenging 98% (91-100%) Clean Up Zak Total Engagement 98% (91-100%) Total Challenging 98% (98-100%) Max Total Engagement 97% (95-99%) Total Challenging 99% (95-100%) Twin Play Zak Total Engagement 100% (100-100%) Total Challenging 97% (95-99%) Max Total Engagement 97% (91-100%) Total Challenging 99% (97-99%) Zak Total Engagement 99% (98-100%) Total Challenging 98% (97-100%) All Play Emmy Total Engagement 99% (97-100%) Total Challenging 99% (98-99%) Max Total Engagement 96% (95-98%) Total Challenging 96% (93-99%) Zak Total Engagement 95% (89-97%) Total Challenging 94% (93-95%) Dinner Emmy Total Engagement 95% (92-97%) Procedural Fidelity. In order to assess the degr ee to which intervention components were implemented with integrity, procedural fidelity data were collected during the natural only condition for each rou tine. Data were obtained relative to components that could be observed comple tely during sessions (i.e., intervention components that were clearly observabl e on videotape without interference or obstruction). Data were expressed as the m ean percentage of completed steps. Design. With regard to design, da ta were evaluated using a sequential withdrawal design. Using this design, individual com ponents were systematically withdrawn and represented in a non-random fashion in order to determine whether changes in dependent measures have maintained (Kazdin, 1982; Rusch & Kazdin, 1981). Often embedded within the context of withdraw al or multiple baseline designs, the sequential withdrawal

PAGE 45

37 design allows one to compare th e strength of the behavior sup port plan’s stimulus control relationship over the dependent measure as it is systematically faded to the use of naturalistic strategies (i.e., to identify the specific plan components that have a functional relationship to the dependent variable). Procedures Specific procedures used within this st udy are described below relative to each condition. Additional sections are discus sed relative to unanticipated procedural variations to the res earch study as it was or iginally proposed. Intervention Component Reduction The first set of proc edures were selected to facilitate the systematic reduction of the total number of in dividual intervention components into as efficient and durable a s upport plan as possible pe r routine. In order to complete this task, two sets of data were used to determ ine which individual artificial components would be included within the next phase of the research study (i.e., component analysis): procedur al fidelity data from the p ilot research study and parent rating scales completed by the mother for each routine. Data from th e former were used to determine which components were implemen ted by the mother on a consistent basis, whereas the latter data set were used to determine both the mother’s perceptions of intervention component efficacy and preference for long-term use. Guided by a set of decision rules for each data set, this researcher used these da ta to reduce the total number of individual intervention co mponents per routine in as objective and systematic a manner as possible. With regard to the procedural fidelity data this researcher used procedural fidelity data from the pilot resear ch study to determine the de gree of which individual

PAGE 46

38 intervention components were implemented as specified in their corresponding behavior support plan. Scoring each vide otaped session relative to th e presence or absence of individual intervention compone nts, trained observers obtain ed estimates of procedural fidelity for each intervention component per r outine. Given the fact that procedural fidelity data were presente d as the mean percentage of completed steps for each intervention component across intervention se ssions, it became necessary to develop a decision rule specifying whic h specific components would be includ ed and excluded from the component analysis. Consequently, a decision rule was created, sp ecifying that components with procedural fidelity estimates equal to or less than 50% would be assessed further to determine whether or not they would be included within the research study. The condition for exclusion involved visual analysis of the corresponding composite challenging behavi or graph; composite challe nging behavior data were visually analyzed by this res earcher on sessions in which the intervention component in question (i.e., the intervention component with a procedural fidelity estimate equal to or less than 50%) was both implemented and omitted. The second data set used to determin e which individual components would be included within the next phase of the research study were parent rating scales completed by the mother for each routin e. The mother was asked to complete a rating scale corresponding to procedural fidelity checkli sts for each routine. Following this procedure, the mother was presented with a thr ee-point Likert rating s cale to indicate: 1) whether she perceived there to be a relationship between a certain component and the child’s challenging behavior; and 2) whether the componen t was something she can see herself using six months into the future. Data were used to determine which individual

PAGE 47

39 components were considered candidates for subsequent compone nt analysis (i.e., components considered to be “artificial”) and which were considered “natural” to her family context (i.e., components that would st ay constant and remain in place during this analysis as specified in their corr esponding behavior support plans). Upon acquisition of both sources of data, this researcher made decisions to reduce specific intervention components. Though both sources of data were used, decisions of whether or not to include or omit specific components were weighted with procedural fidelity data (i.e., the procedural fidelity d ecision rule applied firs t, followed by parent responses reported via the rati ng scales). On occasions wh en an intervention component with low procedural fidelity data (i.e., equa l to or less than 50 %) had been rated by the mother as being either effective (i.e., de monstrating a relationship with her child’s challenging behavior) or desired for long-term use, this resear cher presented both sets of data pertaining to the specific component to the mother fo r her decision to include or exclude from the next phase of the research study (A ppendix B). Component Analysis The first step of the component analysis was to systematically test each component labeled “ar tificial.” Each component was assigned a letter and sequentially withdrawn, re-present ed, then withdrawn again in a “mini” reversal fashion. The sequen ce of the presentation of eac h component was determined by the mother’s rating of what specific steps she perceived as the most necessary to maintain low levels of challenging beha vior. In other words, it was the mother’s opinion that the removal of the specified com ponent from the behavior suppo rt plan would result in increased challenging behavior. Those percei ved as “most necessary” to keep levels of challenging behavior low were manipulated first, followed by the component with the

PAGE 48

40 second-highest “most necessary” ranking, a nd so forth. This step was repeated concurrently across routines until every “a rtificial” component was tested. Natural intervention strategies (i.e., th e “natural” components) stayed constant and remained in place during this analysis as specified in thei r corresponding behavi or support plans. Likewise, coaching was not provi ded to the parent during th is phase, though presentation of data and plan review occurred at the end of each “minireversal.” Data obtained using these procedures allowe d this researcher to further reduce the total number of intervention components per routine (see “Data An alysis” section for specific data analysis procedures used to determine the presence of a functional relationship). Based upon changes in level of composite challengi ng behavior, individual components that demonstrated change were ta gged for systematic fa ding in a subsequent phase of the study (i.e., the co mponent analysis was used to filter out which “artificial” components will remain in the streamlined plan, as de termined by changes in the dependent variable during systematic stimulus control manipulation). Streamlined Plan The purpose of th is condition was to combine the remaining “artificial” components that had demonstrated a functional relations hip to the dependent variable (i.e., composite ch allenging behavior) during the co mponent analysis with the “natural” components. Systematic Fading of Artificial Components In the next condition, each “artificial” component included was to be placed on a thinning schedule until components were no longer necessary to maintain levels of challenging behavi or (as indicated by stability of visual analyses). Natural in tervention strategies (i.e., the “natural” components) were to remain constant and in place du ring as specified in their

PAGE 49

41 corresponding behavior support plans. Prior to this c ondition, a single “artificial” component was scheduled for withdrawal, pr oviding an additional manipulation designed to assess the strength of the stimulus control relationship. Natural Only The final condition of this study entailed m onitoring the “natural” components over a brief period of time. In contrast to the intervention phas e of the pilot study, implementation of this c ondition did not entail coachi ng; the mother was asked to implement the behavior support plan indepe ndently without any fo rm of coaching or assistance. The parent did not receive a ny instruction regarding which components to implement. The purpose of this condition was to dem onstrate the ef ficacy of the “natural” support plan components within each routine. Data were collected to assess the plan over a brief period of time for the family’s even tual long-term use. Unanticipated Procedural Variations Although the aforementioned procedures were initially articulated for this research study, changes were made due to unanticipated outcomes. Upon completion of the component analys is, this researcher intended to implement streamlined plans for each routine, and then systematically fade those plans before moving into the natural only cond ition. However, as the research study progressed, it became apparent that the behavior support plans originally developed for the family were already streamlined in their cu rrent state. Conseque ntly, the streamlined plan and systematic fading procedures original ly articulated for this research study were no longer necessary for inclusion within the current research study.

PAGE 50

42 Chapter Four Results The results of this investigation have been presented in this chapter. Data pertaining to the in tervention component redu ction process (i.e., patt erns of reduction and parent ratings) have been pres ented by routine (i.e., clean up, twin play, all play, dinner), whereas changes in the dependent measures across the intervention condition of the pilot study, and both the component analysis and na tural only condition of the current study have been presented in rela tion to each participant. Intervention Component Reduction The first step of the proced ures for this research st udy entailed systematically reducing the total number of in dividual interventi on components into as efficient and durable a support plan as possibl e per routine (Tables 9-12; A ppendix B). The purpose of the intervention component redu ction procedures was to both reduce the total number of individual intervention compone nts per routine in as object ive and systematic a manner as possible, and to make each behavior suppor t plan easier for the parents to implement overt time. In order to comple te this task, two sets of data were used to determine which individual components would be included within the next phase of the research study (i.e., component analysis): procedural fidelity data fro m the pilot research study and parent rating scales completed by the mother for each routine. Data from th e former were used to determine which components were implemented by the moth er on a consistent

PAGE 51

43 basis, whereas the latter data set were used to determin e both the intervention agent’s perceptions of intervention co mponent efficacy and preference for long -term use. Using procedural fidelity data from the pilot st udy, a decision rule was cr eated exclusively for “artificial” intervention components (i.e., inte rvention components that are not typically incorporated within family r outines, as determin ed by the primary intervention agent of this research study). The deci sion rule specified that compone nts with procedural fidelity estimates equal to or less than 50% would be tagged for further assessment to determine possible inclusion with in the component analysis (i.e ., individual components were visually analyzed to determine whether their presence or absence appeared to influence changes in dependent measures). The second set of data enta iled the use of parent rati ng scales completed by the natural intervention agent for the current res earch study. The mother was presented with a three-point Likert rating s cale to indicate: 1) whether she perceived there to be a relationship between a certain component and the child’s challengi ng behavior; and 2) whether the component was something she can see herself using six months into the future. Data were used to determine wh ich individual components were considered candidates for subsequent component analys is (i.e., components considered to be “artificial”) and which were considered “natur al” to her family context (i.e., components that would stay constant and remain in place during this an alysis as specified in their corresponding behavior support plans). In the event that an intervention component with low procedural fidelity data (i.e., equal to or less than 50%) had been rated by the mother as being either effective (i.e., demonstrating a relationship with her child’s challenging behavior) or desired for long-

PAGE 52

44 term use, the mother was give n the opportunity to decide wh ether to include or exclude the specific intervention com ponent from the next phase of the research study. Procedural Fidelity Criterion Procedural fidelity data from the pilot study were used to complete the first step of this task. Looking at obtained procedur al fidelity estimates for each intervention component per routine, this re searcher identified several co mponents that were discarded using visual analyses (i.e., specific intervention component s whose procedural fidelity estimates were equal to or le ss than 50% were assessed relative to changes in dependent measures when the component was both implem ented and omitted). For example, data obtained during the all play rout ine indicated that pr ocedural fidelity was consistently low when the mother reviewed the rules pertai ning to Emmy’s social story and provided Emmy with access to the “rule lis t.” The procedural fidelity estimate for this intervention component was 33%. Conversely, th e mother consistently select ed a third toy set to help create a theme during the twin pl ay routine. Procedural fi delity for this intervention component was 100%. Applying the decision rule to these examples, the intervention component for the all play routine was discar ded, whereas the one us ed during twin play was retained. An analysis of decisions ma de using the procedural fi delity criterion yielded a number of distinct patterns across routines. With regard to the clean up routine, procedural fidelity was consis tently low (i.e., 50% or lower) for procedures requiring the mother to provide transition cu es. The same pattern was noted during the all play and dinner routines. Similarly, pr ocedural fidelity data indicated that the mother was less consistent implementing procedures requiring specific praise (i.e., twin play, all play,

PAGE 53

45 dinner). Finally, data indica ted that the mother implemented artificial components such as social stories and self-m onitoring procedures with less accuracy (e.g., reading all play social story to Emmy prior to all play rou tine, providing choice menu for Emmy if she matches with mother and has over 80% appropriate behavior). Parent Rating Scale The second step of the re duction process entailed fu rther reducing intervention components relative to parent perceptions regarding the utility of each support plan. The mother was asked to indicate whether indi vidual components for each routine were perceived to be directly related to her child ren’s behavior and to indicate whether or not she’d like to continue implementing the co mponent in the future (e.g., six months from now). Data were obtained fo r both questions using a threepoint Likert-type rating scale completed by the mother rela tive to each routine (Appendix B). With respect to the former question, scores were obtained relative to whet her the mother perceived a relationship existed between the component a nd her children’s beha vior (i.e., I do not think there is a relationship, felt unsure, and I think there is definite relationship). Across each routine, several trends were observed. First, th e mother tended to perceive that the majority of intervention components were related to her children’s behavior. Across each routine, the mother re ported that the majority of intervention components were related to the children’s be havior. Second, the moth er tended to report that intervention components tied to either setting clear expect ations and providing specific praise were more closely related to her children’s behavior. Again, this pattern was evident across routines.

PAGE 54

46 Finally, the mother reported that inte rvention components tied to antecedent modification were inconsistently tied to her children’s behavior. Within this general category of intervention compon ents, the mother consistently reported that components tied to environmental manipulations were more related to changes in her children’s behavior than those related to choice making. For exampl e, intervention components such as playing the Dragon Tales song dur ing clean up, having dinner completely prepared and on the table before dinner, as we ll as using a consiste nt seating arrangement during dinner were each rated as being tied to the child ren’s behavior. In contrast, intervention components tied to choice maki ng were estimated to be less related (i.e., earning an “unsure” rating). Ex amples of this pattern includ e choice of food items during dinner, access to additional toy sets during tw in play and all pla y, and choice of a preferred reinforcer during clean up. With regard to the latter question, the moth er was asked to estimate the degree to which she’d like to continue implementing individual inte rvention components in the future (e.g., six months from now). A review of parent ratings to th e specific intervention components the mother would like to continue in the future yielded similar trends as that of the first question. As noted with the pr evious question, the moth er’s responses held consistent across routines. Likewise, the majority of intervention components in each routine earned the most positive response (e.g., close relationship between an intervention component and my children’s behavior, woul d like to continue implementing a specific intervention component in the future). In addition, th e mother reported that both intervention components that entailed the use of materials and those used to provide choice were less preferred for subsequent inte rvention in the future. Examples of these

PAGE 55

47 types of less-preferred components included additional toy sets during twin play and all play, self-monitoring materials and the choi ce menu used during dinner and all play, choice of song or character during clean up, and choice of preferred reinforcer during clean up. However, it is stil l important to note that some “artificial” intervention components (e.g., choice of char acter during clean up) whose pr ocedural fidelity data fell below the 50% decision rule were included within the component analysis per the parent’s interest in incorpor ating the specific component w ithin natural routines over time. Summary. The purpose of the intervention component reduction procedures was to systematically reduce the total number of individual in tervention components into as efficient and durable a support pl an as possible per routine. A two-step process was used to accomplish this task (i.e., a predetermine d procedural fidelity criterion and parent ratings of each component’s utility). Several interventi on components were discarded using the procedural fidelity criterion, many of which were also rated to be discarded using the parent rating scale. However, several intervention components whose procedural fidelity coefficients suggested discontinuation (e.g., pr ior to the routine, dinner was completely prepared and put on the table) were retained with positive parent ratings, thereby resulting in a series of behavi or support plans that were both as streamlined and contextually fi tting as possible. These beha vior support plans were then marked for inclusion with in the subsequent compone nt analysis condition. Systematic Behavioral Observations In this section, time series data have be en presented and descri bed relative to both composite challenging behavi or and data obtained for each individual child (i.e.,

PAGE 56

48 challenging behavior, engagement). In order to adequately assess changes in dependent measures over time, data from the inte rvention condition of the pilot study, the component analysis, and the natural only condition have be en presented. Systematic behavioral observations were used to obtain estimates of composite challenging behavior across routines (Figures 2 through 5; Append ix A) and both chal lenging behavior and engagement for each individual child particip ant across routines and conditions (Figures 6 through 9; Appendix A). Data have been first presented relative to composite challenging behavior estimates, followed by challenging beha vior and engagement data obtained for each of the three in dividual child participants. In the final portion of this section, procedural fidelity data collected during the natu ral only condition have been presented. Patterns of plan implementati on are discussed relativ e to the remaining components in the natural only condition for each routine. Data Analysis Visual analyses served as the prim ary means of analyzing changes to the dependent measures. Data were graphed in order to determine changes in the trend and level of all three dependent variables acro ss conditions (Kazdin, 1982) Visual analyses of trend considered change s in direction both within and between phases, whereas changes in level were assessed through vi sual inspection of the magnitude of each dependent variable. Data obtained during th e current research study were also compared relative to the pilot study in order to pr ovide a context for analyzing changes in dependent measures over time. Component Analysis Due to the fact that the component analysis condition entailed three dependent variable observa tions across component groupings (i.e.,

PAGE 57

49 withdrawal, reintroducti on, withdrawal), data were assess ed by visual inspection of level changes across individual data points. Decisions to keep or cut each component during the component analysis condit ion were determined by the ma gnitude of change in level of the dependent variable as determined by visual inspection and an analysis of any setting events that may have impacted the data. A functional relationship was determined by either the magnitude of the difference between the level of the dependent variable when it was reintroduced versus when it wa s withdrawn. Data that demonstrated an appreciable difference between phases would therefore ha ve been judged to reflect a functional relationship. If the level of challenging be havior and enga gement did not fluctuate on sessions that were being targeted for the specific component, it was determined that a functional relationship could not be dem onstrated by this researcher. Composite Challenging Behavior Pilot Study. Across each routine of the pilot study (see Appendix A), visual analyses comparing baseline a nd intervention data consistent ly indicated that rates of composite challenging behavior demonstrated a decreasing tr end during the intervention condition. Relative to observat ions of trend, visual analyses of baseline data collected during the clean up routine main tained an increasing trend, a downward trend during twin play, and a slightly downward trend during the all play and dinner routines. Composite challenging behavior du ring baseline was highest in the cl ean up and twin play routines, and lowest during the dinner and all play routines. Data collected across both routines a nd conditions indicated that rates of composite challenging be havior consistently demonstrated both a lower level and lesser degree of variability during the intervention cond ition. Data indicated that rates of

PAGE 58

50 composite challenging behavi or across routines consis tently dropped during the intervention conditions. Across routines, mean levels of composite challenging behavior dropped a minimum of 43% and a maximum of 71%. With regard to variability, visual analyses indicated that compos ite challenging behavior during the clean up and twin play routines was more variable during the interv ention conditions. The opposite pattern was observed during the all play and dinner routines. Though the data obtained during the latter routines provide a more convincing demonstration of be havior change (relative to changes in variability), it is important to note that the increase d variability observed during the clean up and twin play conditions appeared to be a function of the sharp changes in level that were obtained. Together an assessment of changes in trend, level, and variability suggest that the children’s ra te of composite chal lenging behavior was much lower as the result of intervention implementation. Component Analysis. Specific intervention component s subjected to component analysis procedures are located in Table 4. A minimum of two intervention components were manipulated per routine, in addition to conditions m easuring changes in dependent measures in which both compon ents were included (i.e., “typ ical”) or omitted (i.e., the “NN” or “RN” condition).

PAGE 59

51 Table 4 Intervention Components In cluded in the Component Analysis Procedures. Clean-Up Twin Play All Play Dinner AB Typical AB Typical AB Typical AB Typical A Music A Full toy sets (Boys pick 2) A Full toy sets (Everyone picks 1) A Self monitoring B Dragon characters B Praise B Praise B Seating arrangement NN No music and No dragon characters RN Reduced toy sets No praise RN Reduced toy sets No praise NN No self monitoring No change in seating Across routines, visual an alyses of systematic behavi oral observation data across all four routines indicated that levels of composite challenging behavior were consistently lower during the component analys is conditions than during th e baseline condition of the pilot study (Figures 1 and 2). In contrast, rates of occurrence maintained a pattern similar to that of the pilot study inte rvention condition. For exampl e, mean rates of composite challenging behavior within the cleanup routine ranged fro m 16-34% during the conditions AB (typical), c ondition B (dragon characters), and condition NN (no music and no dragon characters). Similar results we re obtained during the other three routines and conditions, there by indicating that the children demonstrated similar rates of challenging behavior since th e initiation of the interven tion condition approximately a year before. The only exception to this pattern of low composite challenging behavior was observed in condition A of the clean up routine, which entail ed the manipulation of music (mean = 60%, range = 9-95%). Rates of composite challe nging behavior more than doubled that of the p ilot study intervention condition on the two occasions in which the

PAGE 60

52 component was manipulated, th ereby providing tentative sup port for the presence of a functional relationship between the inclusion/exclusion of music within the clean up routine and occurrences of composite challenging behavior. However, it is important to note that the data obtained dur ing this condition were in fluenced by a single child participant (i.e., Zak), whose behavior himself appeared to have been influenced by setting event variables (e.g., the presence of pr eferred objects, reactions to changes of any kind within the routine). As a result of these variables (described subsequently in Chapter 5), it is reasonable to conclude that visual analys es suggesting th e presence of a functional relationship between variables are presently in adequate and do not confirm such a relationship. Natural Only Condition. Within the natural only cond ition for each routine, the mother was given the opportunity to both im plement or omit any sp ecific intervention procedures she wished (Tables 13-16; A ppendix C). Implementing each behavior support plan without coaching or prior prep aration, the mother was encouraged to implement the specific components she felt neces sary to help her children maintain lower rates of challenging behavior and higher rates of engagement over time. Examples of specific intervention components implemented over the course of th e three natural only condition sessions included: 1 specific praise and cl early stated expectations (clean up); 2) choice of toy sets and play themes (twin play); 3) sel f-monitoring and expectations to ask Maggie for help (all play); and 4) an tecedent modifications such as seating arrangement and sitting with children fo r entire duration of mealtime (dinner). Visual analyses indicated that rate s of composite cha llenging behavior demonstrated a more variable and slightly increasing trend during the clean up and twin

PAGE 61

53 play routines, whereas a much less variab le and slight downward trend was observed during the all play and dinne r routines. Rates of com posite challeng ing behavior occurred at a lower level than that of the baseline conditions and more closely approximated intervention conditions of th e pilot study. The following mean rates of composite challenging behavior were obtained: clean up (m ean = 39%); twin play (mean = 25%); all play (mean = 13%%) ; and dinner (mean = 15%). Relative to pilot study comparisons, mean composite challenging behavi or data obtained in each of the natural only conditions approximated both the pilot study interventi on condition and condition AB of the component analyses (i .e., the typical condition). Max Pilot Study. With regard to Max’s behavior dur ing the pilot study, visual analyses indicated both challenging behavior and en gagement changed as a function of the implementation of behavior support plans across routines. With respect to changes in the variability and level of both dependent measures, data are presented below in Table 5. Table 5 Changes in Variability and Level of Chal lenging Behavior and Engagement for Max: Pilot Study Challenging Behavior Engagement Routine Baseline Mean (Range) Intervention Mean (Range) Baseline Mean (Range) Intervention Mean (Range) Clean Up 83% (73-100%) 8% (0-45%) 7% (0-20%) 94% (67-100%) Twin Play 47% (18-70%) 11% (0-27%) 51% (17-86%) 89% (72-100%) All Play 25% (13-65%) 7% (0-17%) 77% (39-88%) 95% (87-100%) Dinner 17% (6-24%) 13% (3-25%) 84% (64-94%) 90% (84-98%) Comparing data between both the baseline and interven tion conditions, rates of challenging behavior maintained a consistently downward trend, moderate degree of variability, and a lower level during the interv ention condition of th e clean up routine.

PAGE 62

54 Only small exceptions were noted across the other three ro utines. Challe nging behavior was more variable during the baseline condition in the twin play routine, but was less variable during the interventi on condition. The same pattern s were noted during the all play and dinner routines. Relative to ra tes of engagement du ring the pilot study, Max was consistently more engaged during each intervention condition. Specifically, data indicated that rates of engage ment during the clean up routin e were consistently upward in trend, less variable, and higher in level dur ing the intervention condition. This pattern was also evident within th e other three routines. Component Analysis During the component anal ysis, both Max’s challenging behavior and engagement maintained a similar degree of consistency as demonstrated during the intervention condition of the pilot study. In light of the fact that three individual intervention components were mani pulated for each condition (in addition to measuring behavior during the “typical” implementation of the behavior support plan), Max’s behavior only fluctuated slightly when praise was manipulated during the clean up routine (i.e., visual analyses indicated that chal lenging behavior increased and engagement decreased). Specifically, mean rates of challenging behavior within the clean-up routine ranged from 4-17% during conditions AB (mean = 7%), condition B (mean = 17%), and condition NN (mean = 4%). In contrast, mean ra tes of engagement ranged from 81-98% during the same rou tine across conditions AB (mean = 92%), condition B (mean = 81%), and condition NN (mean = 98 %). Similar results were obtained during twin play (e.g., means rangi ng from 12-23%), all play (e.g., means ranging from 19-200%), and dinner (e .g., means ranging from 17-35%).

PAGE 63

55 Although the initial child targ eted for support, Max’s beha vior changed very little in response to individual inte rvention component manipulation s. In other words, Max’s challenging behavior and enga gement did not appear to demonstrate a functional relationship to specific intervention compon ents. In contrast, however, both Max’s challenging behavior and engagement varied on occasions when he obtained less sleep than typical. On occasions when he obtaine d less sleep than typi cal, Max’s behavior the following day was described as both more challenging and less engaged than on occasions when he obtained his typical amount of sleep. Natural Only Condition Upon completion of the component analysis condition for each routine, Max’s behavior was briefl y assessed through the implementation of a natural only condition. Measur ing both rates of challengi ng behavior and engagement across three data points, Max’s rate of challenging behavior and engagement approximated the levels, variability, and tre nds consistently obse rved during the pilot study intervention and component analysis conditions. Zak Pilot Study. With respect to Zak’s behavi or during the pilot study, data consistently indicated his challenging beha vior and engagement also changed as a function of the implementation of behavior support plans across each of the four routines. Data are presented below in Table 6, descri bing the level and vari ability of both Zak’s rate of challenging behavi or and engagement over th e span of the pilot study.

PAGE 64

56 Table 6 Changes in Variability and Level of Chal lenging Behavior and Engagement for Zak: Pilot Study Challenging Behavior Engagement Routine Baseline Mean (Range) Intervention Mean (Range) Baseline Mean (Range) Intervention Mean (Range) Clean Up 57% (23-84%) 20% (0-92%) 37% (16-69%) 81% (80-100%) Twin Play 47% (27-56%) 18% (0-28%) 56% (46-73%) 92% (74-100%) All Play 25% (13-65%) 7% (0-17%) 77% (39-88%) 95% (87-100%) Dinner 29% (8-69%) 12% (0-24%) 84% (64-94%) 90% (84-98%) Comparing data between both th e baseline and intervention c onditions of the clean up routine, rates of challengi ng behavior maintained a cons istently downward trend, less variability, and a lower level dur ing the intervention condition. Similar trends were noted during the other three rou tines (i.e., twin play, all play, di nner). In relation to rates of engagement measured during th e pilot study, Zak was consiste ntly more engaged during each intervention condition. Specifically, data in dicated that rates of engagement for Zak during all four routines follo wed a consistently upward tr end, less variability, and a higher level during the intervention condition. Component Analysis During the component anal ysis, both Zak’s challenging behavior and engagement maintained a similar degree of consistency as demonstrated during the intervention condition of the pilot study. Mean ra tes of challenging behavior ranged from 19-85% during clean up, 3-25% dur ing twin play, 6-9% during all play, and 7-40% during the dinner routine. Likewise, mean rates of engagement ranged from 7483% during clean up, 87-100% dur ing twin play, 93% during all play, and 65-91% during dinner. Although Zak’s rates of challenging behavior and engagement across routines appeared similar to that of the pilot study interven tion condition, this researcher observed an additional pattern to Zak’s behavior in response to component analysis procedures.

PAGE 65

57 However, it is important to note that this pattern appear ed to be tied less to the manipulation of a specific intervention component than the presence of a change to the typical routine itself. For example, Zak demonstrated mo re challenging behavior on occasions when his preferred toys were not present during a specific routine (e.g., 85% challenging behavior and 20% engagement during condition A) In light of the fact that three individual intervention components were manipulated for each condition (in addition to measuring behavi or during the “typi cal” implementation of the behavior support plan), Zak’s behavior appeared to change more in response to changes in materials typically included within the routine than a specific component manipulated during the routine itsel f. This pattern was observed during the pilot study’s clean up routine across different inte rvention component manipulati ons (e.g., reduced number of toys, praise, self-monitoring) and routines (e.g., twin pla y, all play), thereby lending support to the notion that the concern with rout ine-specific materials may have influenced Zak’s response to task demands. Natural Only Condition Upon completion of the component analysis condition for each routine, Zak’s behavior was briefl y assessed through the implementation of a natural only condition. Measur ing both rates of challengi ng behavior and engagement across three data points, Zak’s rate of challenging behavi or and engagement approximated the levels, variability, and tre nds consistently obse rved during the pilot study intervention and component analysis conditions. Emmy Pilot Study. With respect to Emmy’s behavi or during the pilot study, data consistently indicated both challenging beha vior and engagement changed as a function

PAGE 66

58 of the implementation of be havior support plans across r outines. Comparing data between both the baseline and intervention conditions, rate s of challenging behavior maintained a consistently downward trend, le ss variability, and a lo wer level during the intervention condition of the al l play and dinner routines. Data reflecting the changes in both variability and level are presented below for each routin e and condition in Table 7. Table 7 Changes in Variability and Level of Chal lenging Behavior and Engagement for Emmy: Pilot Study Challenging Behavior Engagement Routine Baseline Mean (Range) Intervention Mean (Range) Baseline Mean (Range) Intervention Mean (Range) All Play 28% (11-36%) 6% (0-12%) 79% (71-97%) 97% (90-100%) Dinner 40% (22-66%) 8% (0-22%) 65% (50-78%) 93% (75-100%) Assessing the degree of change between c onditions and routines, data indicate that Emmy’s greatest degree of behavior change oc curred during the dinn er routine. Though Emmy’s rates of challenging behavior a nd engagement consistently improved (i.e., challenging behavior decreased, while engagement increased) relative to both routines and dependent measures, she demonstrated a greater degree of challenging behavior reduction over time than an increase in her rate of ta sk engagement. Component Analysis During the component analys is, both Emmy’s challenging behavior and engagement maintained a similar degree of consistency as demonstrated during the intervention condition of the pilot study. In light of the fact that three individual intervention components were mani pulated for each condition (in addition to measuring behavior during the “typical” implementation of the behavior support plan), Emmy’s behavior fluctuated very little (i.e., challenging behavior decreased, engagement increased). Specifically, mean rates of challenging behavior and engagement within the

PAGE 67

59 all play routine ranged from 7% and 96-97% during both conditions AB and condition B, respectively. With regard to the dinner routine, Emmy’s mean rates of challenging behavior and engagement during conditions CB, B, a nd A were 2-21% and 90-95%, respectively. Consequentl y, Emmy’s challenging behavi or and engagement did not appear to demonstrate a functional relati onship to specific inte rvention components. Natural Only Condition Upon completion of the component analysis condition for each routine, Emmy’s behavior was briefl y assessed through th e implementation of a natural only condition. Measur ing both rates of challengi ng behavior and engagement across three data points, E mmy’s rate of challenging behavior and engagement approximated the levels, variability, and tre nds consistently obse rved during the pilot study intervention and component analysis c onditions. Though Emmy exhibited a greater rate of challenging beha vior in the dinner ro utine during this condition, her mean rate of challenging behavior remains 26% lower than it was duri ng the pilot study baseline condition. Summary of Natural Only Conditions With regard to the specific components se lected for inclusion into the natural only support plans, several patterns were obser ved (Appendix C). Th e mother expressed interest in implementing a combination of antecedentand consequence-based interventions in order to both prevent the occurrence of chal lenging behavior and to teach her children prosocial skills. In reference to the former go al, one of the mother’s main priorities was to maintain a se nse of structure, en suring that expectations were clearly stated and understood (e.g., indi cating to the boys when it was time to clean up, telling the boys to play with th eir sister while she was in the k itchen during all play, stating that

PAGE 68

60 playtime was “all done” at th e end of twin play). Gi ven that both intervention components had been implemen ted in the previous conditio ns of the pilot and current research studies, these components were in cluded within each routine. Additional antecedent modifications were also retain ed for use within the dinner routine (e.g., seating arrangement, sitting with ch ildren for duration of meal). With respect to the latter goa l (i.e., to teach prosocial sk ills), the mother continued to deliver specific praise in each of the na tural only support plans. Procedural fidelity data indicated that the mother had learned to effectively deliver specific praise to her children and had demonstrated th e ability to consistently deliver specific praise to each of her three children for follow ing directions during the all play and dinner routines. Similarly, the mother also de monstrated an understanding of choice and preference, electing to provide choices of toy sets during twin play and all play, and to present the children with preferred reinfor cers whenever they successfully completed their routines. Though each routine necessitates subtle variations in procedure (e.g., the type of specific praise or language used in delivering clear expectations), each of these components were both preferred by the mother a nd implemented with precision acr oss previous conditions. Finally, it is important to acknowledge patterns in de pendent measures obtained during the natural only condition w ithin the context of both th e pilot and current research studies. Data collected nearly a year after initiating servi ces with the family indicated that rates of composite challe nging behavior rema in much lower than at its original baseline state (Table 8).

PAGE 69

61 Table 8 Changes in Level of Composite Challengi ng Behavior Across Conditions and Routines Routine Baseline Condition Mean Intervention Condition Mean Natural Only Mean Clean Up 96% 25% 39% Twin Play 69% 22% 25% All Play 54% 11% 13% Dinner 64% 21% 15% Across both the pilot and current research st udies, composite challe nging behavior data also indicate that rates of challenging behavi or have remained at levels similar to the intervention condition of the pilot study. Give n the fact that the intervention condition concluded approximately six months ago, th is finding supports th e durability of the children’s resulting behavi or change over time. Summary In accordance with data collection procedures artic ulated in the methodology, data were collected in order to obtain estimates of group a nd individual target behaviors (i.e., composite challenging be havior, challenging behavior, engagement). Behavior support plans for each routine were syst ematically reduced and implemented independently by the primary intervention agents (i.e., mother, older sister) within their natural routines. Across routines, levels of composite challe nging behavior and challenging behavior equated to or were less than the highest level obtained during intervention conditions, and le vels of engagement were higher than levels obtained during the intervention conditions of the pilot stu dy. These findings maintained across both time and child participants.

PAGE 70

62 Despite such encouraging findings, it is important to note that the component analyses conducted within each of the four rou tines did not result in the identification of a functional relationship betw een a single intervention component and changes in dependent measures. Although this aspect of the research study was not originally hypothesized, explanations of its occurrence are presented and disc ussed in the next chapter.

PAGE 71

63 Chapter Five Discussion Review of Research Questions This study entailed the demonstration of behavior change resulting from the implementation of four multi-component posit ive behavior support (PBS) plans. Each plan was derived from functional assessment s conducted within the family’s natural environment. Behavior support plans we re developed in collaboration with and implemented by natural interven tion agents, teaching age-appr opriate replacement skills. Efforts were made to ensure a high degree of contextual fit relative to both the ease of implementation and procedural fidelity over time. The results of this rese arch study indicated that th e children consistently maintained low rates of chal lenging behavior and high rates of engagement within each routine over time. In addition, procedural fidelity data indicated that intervention components were implemented as the parent ha d intended on a consis tent basis and that the plans were easily adapted into natural family routines. To further discuss these outcomes within the context of early inte rvention/early childhood special education (EI/ECSE), it is necessary to review the research questions ad dressed in the study. Specific outcomes are linked to individual questions as they are presented. Findings are discussed relative to eac h question, followed by a disc ussion of limitations and contributions to research and practice.

PAGE 72

64 Research Questions Research Question #1: Which elements of a multi-component intervention that was effective in reducing levels of challenging behavior are functional in maintaining low levels of challenging behavior? The purpose of the first re search question was to dete rmine the speci fic functional relationships between individual interv ention components of four multi-component behavior support plans. Chan ges in dependent measures we re assessed over time (i.e., between the pilot and current studies), as well as both within and between child participants and routines. Resu lts of the pilot stu dy indicated that rate s of both composite and individual challenging beha vior decreased steadily and maintained both lower levels and less variability as a func tion of the implementation of each behavior support plan in each routine. Conversely, changes in levels of engagement consiste ntly increased for all children across all four routines Such patterns of depend ent variable occurrence were also observed in the current study. With respect to the current study, this researcher initiall y sought to investigate the specific functional relationships of indivi dual components within a multi-component behavior support plan implemented within family routines, assess rates of dependent measures when nonfunctional components were removed, and th e resulting plan systematically faded over time. Using reduct ion procedures designed to systematically test intervention components, this researcher and the mother collaborated to determine which components were the most natural a nd implemented with th e highest degree of fidelity in order to attempt to understand long-term use, the mother also rated each behavior support plan component Following this procedure, component analyses were

PAGE 73

65 conducted within each routine in order to dete rmine the effectiveness of each “artificial” intervention component. Assessment of Functional Relationships. It was initially hypothesized that differences in the degree of functional relationship among “artificial” components would be evident, thereby necessi tating the remaining steps of the proposed methodological procedures in order to arrive at the most natural, contextua lly fitting support plan possible (i.e., to create a “streamlined plan” that consisted of inte rvention components demonstrating the greatest degr ee of functional relationship, and to systematically fade the remaining “artificial” components). Contrary to hypotheses stated by this researcher, the results obtained from the component analysis did not identify a set of “artificial” compone nts that clearly demonstrated an individual functional relationship. Data obtained across routines and conditions provided inconclusi ve evidence supporting stronger functional relationships than those obtained from the implementati on of each behavior support plan in its “typical” state. Though effort s were made to investig ate the impact of specific intervention components, it was not possible to detect a fu nctional relationship since challenging behavior was consistently low. Consequently, the findings of the current research study indicated that the implementation of each PBS behavior support pl an in its entirety was associated with durable behavior change over time rather than for individual in tervention components within each routine (i.e., a functional relationship was observed with the multicomponent behavior support pl ans rather than in dividual intervention components). More precisely, however, is the fact that these behavior supp ort plans appeared to have

PAGE 74

66 demonstrated a high degree of contextual fit (Albin et al., 1996; Lucyshyn et al., 2002). Following suit with previously reported findi ngs, each of the four behavior support plans was: 1) implemented with accuracy and preci sion; 2) applied across natural contexts; 3) implemented over a prolonged period of time; an d 4) rated as being effective and useful (Albin et al., 1996; Lucyshyn, Horner, Dunlap, Albin, & Ben, 2002). Though additional studies are needed within this area, the findings obtained in this research study, serve as a case illustration of contextu ally-fitting positive behavior support plan implementation within a family’s home environment. In an effort to explain the cause of this phenomenon, one mu st reconsider this finding relative to those docum ented within the PBS literature base. Previous studies documenting the implementation of PBS beha vior support plans for children and families describe the use of multi-com ponent behavior support plans as a limitation (Dunlap et al., 1996; Kern et al., 1994). In the past, this crit icism has been rendered primarily due to the fact that one cannot determin e which specific intervention co mponents were most closely related to changes in dependent measures (e.g., Carr, Horner, & Tu rnbull, 1999; Dunlap et al., 1996). Given th e fact that PBS is grounded in the science of ap plied behavior analysis, this limitation has been reported in the research literature with some regularity. While the premise of th is limitation makes sens e from the standpoint of replication and scientific rigor it is also important to cons ider the possibility that the same experimental limi tation may not be a limi tation at all within th e context of applied research. Each of the behavior support pl ans implemented in this research study were developed using PBS technology and included intervention co mponents that were both scientifically endorsed a nd reflective of evidence-bas ed practices. Given such

PAGE 75

67 characteristics, one may argue th at each behavior support plan had an adequate degree of technical adequacy. Consequently, one must also question whether the limitation associated with multi-component behavior supp ort plans is more of a theoretical than an applied research issue. A growing body of research ap pears to support this notion, asserting that multi-component behavior support plans are need ed in order to adequately program for lifestyle changes, ecological adjustments, and pr oactive strategies designed to promote stronger interpersonal relations hips and access to preferred activities (e.g., Horner et al., 1990; Lucyshyn et al ., 1997; Lucyshyn et al., 2002). Similarly, it is reasonable to questio n whether the findings obtained in this research study provide support to the notion that PBS multi-component behavior support plans have a greater likelihood of including intervention component s that are either “natural” and/or easy enough for natural interv ention agents to implement over time with a high degree of fidelity. Changes in depe ndent variables over tim e (i.e., across both the pilot and current studies) were due to the implementation of a series of contextuallyfitting, multi-component PBS be havior support plans rather th an for specific intervention components demonstrating a fu nctional relationship in the maintenance of both rates of challenging behavior and engage ment. Consequently, is was evident that the original behavior support plans implemented in both th e pilot and current research studies were already streamlined and had b ecome “natural” to the family (i.e., it was not necessary to create a more streamlined plan or to systematically fade “artificial” components). Following this argument, it does not appear that one may adequa tely answer this research question solely with data obtained in the current research study.

PAGE 76

68 Natural Only Condition The purpose of the natu ral only condition was to demonstrate the efficacy of the “natural” suppo rt plan components within each routine. Data were collected to assess the plan over a brief period of time for the family’s eventual long-term use. After receiving comprehensiv e behavior support for over a year, the mother began to implement each behavior s upport plan on an inde pendent basis. The culminating product of the in tervention component reduction process, each support plan was streamlined to include t hose intervention components that were determined to be both most preferred and most consistently implemented (as determined by both parent ratings and procedural fidelity data). With regard to the specific components se lected for inclusion into the natural only support plans, the mother expressed interest in learning how to prev ent her children from demonstrating further occurrences of challenging behavior, as well as identifying ways to teach them prosocial skills. In addition to achieving these goals through participation in the implementation of antecedentand cons equence-based interventions within each routine, the mother became bot h more familiar and proficient in identifying the specific triggers associated w ith each child’s challenging behavior Over the duration of the pilot and current research studies, the mother had learned how to consistently prevent her children from demonstrating challenging beha vior in favor of so cially appropriate alternatives. The mother lear ned a number of effective stra tegies, each of which she had become comfortable using within her natura l daily routines. A lthough positive, this phenomenon deserves acknowledgment, as the mother had become so fluent implementing intervention compon ents over the course of th e current research study that she no longer relied upon back up strategies (e.g., using physical guidance to help a child

PAGE 77

69 pick up a toy; requiri ng the children to remain at the table for 5 minutes before). On occasions when she observed challenging beha vior during the natural only conditions, the mother routinely implemented the strategies with which she was most comfortable and proficient. Although brief in its duration due to the time of the y ear in which the study was conducted (i.e., end of summer), data collect ed during the natural only conditions are encouraging. Following an extended period of intervention implementation and coaching during the pilot study and particip ation in the component analys is in the current research study, the mother independently implemented each support plan for three days. Aside from observing setting events that likely infl uenced the children’s behavior during the first day of the natural only c ondition (described in the section below), da ta indicated that levels of each dependent measure (e.g., co mposite challenging be havior, individual challenging behavior, individu al engagement) approximated levels previously obtained during the component analyses. Setting Event Variables. Analyzing the obtained resu lts further, it is also necessary to acknowledge the potential role of setting events that may have influenced the children’s behavior. This researcher compared changes in Zak’s behavior relative to changes across component manipulations o ccurring during both th e clean up and twin play routines. While conduc ting these comparis ons, it became evid ent via videotape observations that there was a relationship between Zak’s be havior across both the twin play and clean up routines. On several occasions where Zak experienced difficulty during the twin play routine, the same pattern would tend to occur in the clean up routine (which immediately followed). These comparisons indicated that there was yet another

PAGE 78

70 factor influencing Zak’s behavior: access to preferred toys (e.g., cars and trucks) during the twin play routine. On such occasions wh en cars and trucks were present, Zak’s rate of challenging behavior was consistently lowe r and his rate of enga gement higher than occasions when preferred toys were absent (e.g., cars and trucks). As a result of these two patterns of observations, it became evident to both the moth er and this researcher that Zak’s behavior was influenced by both the pres ence of cars and trucks and whether or not there was a change in his daily routine. C onsequently, it appeared more likely to this researcher that any changes in dependent measures occurring dur ing the conditions in which this pattern was observed are more likely attributable to the pattern itself rather than implying a functional relationship between a speci fic intervention component. In addition to observing unexpected change s in Zak’s behavior, it also appeared that Max’s behavior was influenced by fatigue. On occasions when their mother reported illnesses for any of the children, instances in which the children did not receive an adequate amount of sleep the night before, or occasions when the children did not fall asleep during naptime, Max c onsistently exhibited higher rates of challenging behavior and lower rates of engagement On days in which this pattern was observed (e.g., the first day of the natural only condition), vi deotape observations indicated that Max’s behavior followed a si milar pattern across morning and afternoon routines. In addition to these setting event variab les, it became increasingly apparent as data collection progressed that the children were sensitive to changes in routine, and on occasions when such changes were observed, ra tes of challenging behavior tended to be higher. Examples of such changes include interruptions observed during the clean up and twin play routines (e.g., repairmen, telephone calls) and instances when Emmy was home

PAGE 79

71 from school. While these unanticipated variable s were associated with higher levels of composite and individual challenging behavior, data indicated that Zak was particularly sensitive to the influence of such changes. For example, during the clean up routine, Zak became increasingly frustrated (as evidenced by elevated rate s of individual challenging behavior and decreased rates of engagement) when changes were made (e.g., inclusion or exclusion of music). The sa me pattern was also observed upon return from the family’s summer vacation. Research Question #2: Given t hat some elements are demons trated to be functional in maintaining low levels of challe nging behavior, is it possible to use a systematic fading procedure so that selected elements are no longer needed to maintain low levels of challenging behavior? The second research question addresse d the utility of systematic fading procedures used as a means of maintaining low levels of challenging behavior. Changes in dependent variables over time (i.e., across both the pilot and current studies) were due to the implementation of multi-component PBS behavior support plans rather than for specific intervention compone nts demonstrating a functi onal relationship in the maintenance of both rates of challenging behavior and enga gement. Upon completion of the component analysis, this re searcher intended to implemen t streamlined plans for each routine, and then systemati cally fade those plans before moving into the natural only condition. However, as the research study progressed, it became apparent that the behavior support plans originally develope d for the family had became sufficiently streamlined over time. In comparison between the interv ention conditions of the pilot study and natural only conditions of this resear ch study, the mother was able to use fewer

PAGE 80

72 intervention components to help her children reduce their rates of challenging behavior and to increase their rates of engagement. As the moth er became more proficient implementing each support plan, she reported that she had also learned when to use each intervention component (i.e., she understood how specific in tervention components were chosen to fit the function of her children’s beha vior). As a result of this understanding, it had become easier for the moth er to both implement each beha vior support plan and to also choose which individual components she in tended to continue to implement over time. Consequently, is was ev ident that the original beha vior support plans implemented in both the pilot and current re search studies had become “nat ural” to the family (i.e., it was not necessary to create a more streamlined plan or to systematically fade “artificial” components). Following this argument, it does not appear that one may adequately answer this research question solely with data obtained in the current research study. In light of such findings and tentativ e conclusions, it is also important to acknowledge an alternative expl anation for these outcomes. While it appears accurate to report that changes in dependent variable s were functionally related to the multicomponent behavior support pl ans themselves, it is also possible that changes in dependent measures within each natural only condition were due to the implementation of a multi-component behavior support plan that was m odified during the component analysis’ component reduction process. While separate conditions were originally proposed to create a streamlined plan (i.e., the streamlined plan and systematic fading of artificial component conditions), it is important to acknowledge the fact that the reduction procedures in the component an alysis may have served the same purpose (i.e., to create more efficient and streamlin ed plans). Although this explanation warrants further

PAGE 81

73 research, data from the current research study indicate similar patterns of trend, level, and variability during each natural only condition to those obtained during the pilot study. Consequently, if one were to s ubscribe to this explanation, it is possible to ar gue that: 1) a systematic fading procedure was used in th e current research study (i.e., the component reduction procedures); 2) visu al analyses supported similar patterns during the pilot study intervention and current study’s natural only conditions; and as a result; 3) rates of dependent variable occurrence were functi onally maintained by streamlined multicomponent behavior support plans. Limitations It is important to acknowled ge the limitations inherent to this research study. The first limitation pertained to external validity. Given the fact that the participants of both the pilot and current research studies were fro m an individual family of five, it is not possible to assume that the results of this study are directly repl icable with another family, other family members (e.g., father) or within another context (e.g., school or community). Generalizing results to other ch ildren, regardless of age, culture, gender, socioeconomic status, or diagnosis s hould be made with caution as well. The second limitation is associated with meas urement. It is po ssible that a degree of observer drift may have ex isted as a result of system atically coding behavioral observations over time. Howeve r, it is equally important to recognize that efforts to minimize these untoward effects were made, through both periodic review of operational definitions and interobserver agreement (via obs erver training prior to data collection and measurement of IOA per condition).

PAGE 82

74 With regard to the component analysis two limitations deserve consideration. The first pertained to the br ief withdrawal of intervention components in order to determine stimulus control relationships. There was a clear rationale for using component analysis to he lp the family fade the plan of be havior support so that it will be easy for them to maintain with in everyday routines and set tings. However, the mother may have experienced some degree of discomfo rt when asked to te mporarily withdraw a preferred intervention component. It is also important to note that each component was withdrawn for no more than two sessions, a nd the intervention co mponents were in no way designed to prevent accident or injury. Therefore, the brie f withdrawal did not appear to cause a measurable degree of stress or risk for the family other than a temporary change in routine. The second limitation related to component analysis is that one could argue the sequential withdrawal design was not n ecessarily required to assess response maintenance. As Rusch and Kazdin (1981) note d, “it is quite possible that behavior in a study may be maintained with a complete with drawal (i.e., a complete withdrawal of all components following acquisition ” (p. 134). Consequently, this researcher intended to manipulate at least one ar tificial component that ha d demonstrated a functional relationship to the dependent va riable during the component an alysis prior to initiating the “natural only” condition. A manipulation subsequent to the component analysis and “streamlined plan” condition was considered as a means of enhanc ing the rigor of the study and further demonstrate the strength of the stimulus co ntrol relationship associated with the specific intervention component (t hereby demonstrating the strength of the artificial component’s stimul us control relationship and ju stifying the selection of the

PAGE 83

75 sequential withdrawal design). Given the fact that the outcomes of the current research study did not demonstrate a functional rela tionship between changes in dependent measures and specific individu al intervention components, th e intended manipulation did not occur. Consequently, one could still argue that the sequ ential withdrawal design was not needed to assess response maintenance, but perhaps would have served more use in the creation of each be havior support plan. Finally, limitations existed re lative to the developmenta l maturation of the three children. For example, as th e study progressed, it was appare nt that Max and Zak were using expressive language in a more effici ent manner (i.e., use of grammar, syntax, length of utterance, ar ticulation that is easier to unders tand). While it is possible that changes in the boys’ language development co uld be attributable to the implementation of the independent variables in both the pilot and current re search studies, the opposite is equally possible (i.e., changes in thei r development over time influenced the implementation and measurement of the independent variable). Contributions to Research and Practice The current research study offers several contributions to research and practice. The pilot study provided a case example of the application of PBS with a sibling set of preschool-aged children and thei r parent. Though research is growing in this area, the unique features of this study ma y inspire future research (e.g. interventions designed to support fraternal twins and an older sibli ng, measurement of social validation and procedural fidelity). Given the fact that the current res earch study was a continuation of an ongoing research study (i.e., the pilot study), perh aps the most valuab le aspect of the

PAGE 84

76 research study is the fact th at it documented the utility of longitudinal, family-centered support consistently provid ed for over a year. Likewise, the current research study also demonstrated the maintenance of four multi-component PBS behavior support plans implemented within natural family routines. Though few studies prioritize maintena nce, its inclusion is critically important, as it allows one to assess th e utility and efficacy of an intervention after its initial implementation and demonstration (i.e., the intervention phase of a research study; Dunlap, Horner, Carr, Sailor, Turnbull, Koeg el, & Koegel, 1998; Ho rner & Billingsley, 1988; Horner, Dunlap, & Koegel, 1988). Su ch a contribution has been previously endorsed by Carr and his colleagues (1990) who argued that research studies incorporating the mainte nance of target behaviors may not only help extend the longevity of behavior support plan s, but also to document and str ongly support the overall utility of PBS technology for childre n and their families. Similarly, the current resear ch study offered an experime ntal demonstration of the relationship of individu al components to challenging behavi or. In the majority of studies on PBS, multi-component plans are developed that may include components that are perceived to have a relationshi p to reduce challenging behavior without data that affirms a functional relationshi p. While the current research study failed to offer conclusive evidence demonstrating a functional re lationship between sp ecific intervention components and dependent measures, the study may serve as a methodological case example of such an attempt. Though the findings were not anticipated, the procedures stated prior to the execution of this research study may de serve consideration in future research efforts, particularly those asse ssing the efficacy of individual intervention

PAGE 85

77 components embedded within functional assessment-based intervention plans for children and families (i.e., instances when PBS is not used). This is pa rticularly true given the fact that the efficacy of each behavior support plan was initially determined and then evaluated for durability. Give n that the majority of com ponent analyses are conducted prior to implementation, this feature may serv e to promote future re search, as the field continues to study ways by whic h both implementation of th e independent variable may be enhanced and investigation of specifi c features of the PB S model are conducted. Another methodological contribution associat ed with this research study pertains to the articulation of an eff ective intervention component re duction process. Using the procedures stated and implemented in the current research stu dy, it was possible to identify and eliminate specific interven tion components that were unnecessary to maintaining behavior change (i.e ., those with procedural fidelity coefficients less than or equal to 50 percent, those reported by the na tural intervention agent to be unnecessary and non-preferred for long-term implementation). Additional research in this area may help refine such procedures for use in both applied research and practice. The current research study also offers usef ul contributions to fu ture practice. The first contribution involves the use of natural intervention agents. Both the mother and older sister served as natura l intervention agents in the st udy. The mother served as the natural intervention agent for each of the four routines, while Emmy assumed such a role during the “all play” and “dinner” routines. In this regard, the current research study may serve as a useful case example for practitioners interested in facil itating the PBS process with children and families within home e nvironments, as well as conducting research entailing a high degree of collaborati on with natural intervention agents.

PAGE 86

78 Similarly, this research st udy also reflected an attempt to consider challenging behavior from the parent’s perspective. In addition to measuring challenging behavior and engagement relative to operational definitio ns, it was evident to this researcher that challenging behaviors demonstr ated by three children (on many cases simultaneously) created an appreciable amount of stress for the parent. In this regard, the parent communicated that she often experienced the st ress of her children’s behavior together (i.e., she perceived each child’s behavior to be challenging on instances in which at least one sibling demonstrated challe nging behavior, despite the fact that an individual child may have demonstrated prosoc ial behavior at the same tim e). As a result of these reflections, this researcher a ttempted to measure challenging behavior as a composite of the three children in addition to measur ing challenging behavi or and engagement individually for each child. Though each family system is different, the means by which challenging behavior was measured in this study may be a useful tool for further understanding family stress resulting from young children with challe nging behavior. The findings of the resear ch study also underscore th e value and importance of parent-child interaction. Thr ough participation in the resear ch, the mother was able to successfully implement comprehensive beha vior support plans designed to teach prosocial skills to her children within th e family’s natural environment (both with and without the assistance of her oldest child as an additional intervention agent). The mother learned a combination of new skills/st rategies, as well as the specific decision rules associated with when a nd where to use each st rategy. Given the fact that the mother demonstrated the acquisition and implementation of these sk ills over a prolonged period of time within her family’s natural environmen t, it is reasonable to assume that she may

PAGE 87

79 be more likely to independently prevent futu re occurrences of chal lenging behavior and to continue to teach prosocial sk ills to her children in the futu re. Within the larger system of care, this finding may serve as a rationale supporting the provision of indi vidual-level positive behavior support for young children and their families. Though further research in this area is necessary, the findings of both the pilot and current re search studies offer encouraging outcomes that support the use of PBS as an alternative to existing community-based methods of treatment (e.g., parent traini ng groups, parenting workshops, outpatient assessment and interventi on). In this regard, PBS methodology may be particularly useful when designing pr ograms for groups of parents and caregivers in need of strengthening the qu ality of their parent-child in teraction (e.g., foster parents, parents charged with abuse and neglect). An additional contribution pertains to family-centered pract ices. The current research study was intended to be as family-centered and collaborative as possible, thereby providing a potentially useful case example of fa mily-centered support practices (e.g., fostering collaboration, id entifying the family’s vision and goals, teaching the PBS process, designing streamlined behavior support plans directly linked to family goals). Similarly, the current research study offers a means by whic h to provide assessment and intervention related to a parent ’s perspective of his/her family’s stress. In addition to collaboratively developing and monitoring th e implementation of PBS behavior support plans in natural family contex ts, efforts were made to l earn more about implementation from the parent’s perspective (i.e., using a parent rating scale customized specifically to address parent perceptions of each intervention component’s utility and preference for implementation over time, provi ding frequent opportunities fo r feedback). Not only does

PAGE 88

80 such a step make practical se nse from the standpoint of s upport plan implementation, but it also communicates to the pare nt that a priority is placed upon the degree to which the plan is a good fit within fa mily preferences and natural r outines. Consequently, one might also find that such information may be useful in further re fining existing surveys designed to assess goodness of fit in future applied research and practice with children and families (Albin et al., 1996). Conclusions Research documenting the utility and appl icability of PBS with preschool-aged populations is in its infancy. Though studies of preschooler s conducted within natural environments are being reported with grea ter frequency, few incorporate a combination of natural intervention agents, natural setti ngs, and the measurement of technical aspects of behavior change (e.g., ma intenance). Though studies of maintenance may be difficult to execute, they may provide researchers with a greater understanding of which factors in the change process are most critical to succes sful implementation, as well as to enhance the “goodness of fit” betw een specific plan component s and the ecology in which implementation occurs (Albin, Lucy shyn, Horner, & Flannery, 1996). The purpose of this research was to fi rst assess the relationship of support plan components to rates of behavi or change, and then systema tically fade the functional components, reducing the plan to naturalistic st rategies that may be easy for the family to use over time. The results of this research study indicated that each of the three child participants consistently main tained low rates of challengi ng behavior and high rates of engagement within each routine. In additi on, procedural fidelity data indicated that intervention components were implemented as by the natural intervention agent (i.e., the

PAGE 89

81 mother) on a consistent basis and that the plans were easily adapted into natural family routines. Though clear functional rela tionships among individual intervention components were not attaine d, the current resear ch study offers tent ative support for the acknowledgement of multi-component PBS be havior support plans as an optimal intervention modality for young children with challenging be havior and their families within natural family routines. Demonstra ting a functional rela tionship between four multi-component PBS behavior support plans ov er time (as a relative strength than a perceived weakness), the curr ent research study demonstr ated both family-centered practices, as well as a means for measur ing maintenance and functional component relationships.

PAGE 90

82 References Albin, R. W., Luchysyn, J. M., Horner, R. H., & Flannery, K. B. ( 1996). Contextual fit for behavioral support plans. In L. K. Ko egel, R. L. Koegel, & G. Dunlap (Eds.). Positive Behavioral Support: Including people with di fficult behavior in the community (pp. 81-98), Baltimore, MD: Pa ul H. Brookes Publishing Co. Albin, R. W., Dunlap, G., & Lucyshyn, J. M. (2002). Collaborative research with families of positive behavior support. In J. M. Lucyshyn, G. Dunlap, & R. W. Albin (Eds.), Families and positive behavior su pport: Addressing problem behavior in family contexts, (pp. 373-390). Baltimore: Paul H. Brookes. Annie E. Casey Foundation (n.d.). Kids Count 2004 Data Book Online http://www.aecf.org /kidscount/databook/ Accessed January 03, 2005. Armendariz, F., & Umbreit, J. (1999). Us ing active responding to reduce disruptive behavior in a general education classroom. Journal of Positive Behavior Interventions, 1 (3), 152-158. Bailey, D. B., McWilliam, R. A., Darkes, L. A., Hebbeler, K., Simeonsson, R. J., Spiker, D., et al. (1998). Family outcomes in early intervention: A framework for program evaluation and efficacy research. Exceptional Children, 64 (3), 313-328.

PAGE 91

83 Bailey, D. B., Simeonsson, R. J., Winton, P., H untington, G. S., Comfort, M., Isbell, P., O’Donnell, K., & Helm, J. M. (1986). Family-focused intervention: A functional model for planning, implemen ting, and evaluating individual family services in early intervention. Journal of the Divisi on for Early Childhood, 10 156-171. Bailey, J. S., & Burch, M. R. (2002). Research Methods in Ap plied Behavior Analysis Thousand Oaks, CA: Sa ge Publications, Inc. Baker, M. J. (2000). Incor porating the thematic ritualistic behaviors of children with autism into games: Increasing social play interactions with siblings. Journal of Positive Behavior Interventions, 2 (2), 66-84. Barry, L. M., & Singer, G. H. (2001). A fam ily in crisis: Replac ing the aggressive behavior of a child with autis m toward an infant sibling. Journal of Positive Behavior Interventions, 3 (1), 28-38. Blair, K. S., Umbreit, J., & Bos, C. S. (1999). Using functional assessment and children’s preferences to im prove the behavior of young children with behavioral disorders. Behavioral Disorders, 24 (2), 151-166. Blair, K. S., Umbreit, J., & Eck, S. (2000). Analysis of multiple va riables related to a young child’s aggressive behavior. Journal of Positive Be havior Interventions, 2 (1), 33-39. Bredekamp, S. (1991). Developmentally appropriate practice in early childhood programs serving children birth through 8 Washington, D. C.: National Association for the Educa tion of Young Children.

PAGE 92

84 Bricker, D. (1992). The chan ging nature of communication a nd language intervention. In S. F. Warren & J. E. Reichle (Eds.). Causes and effects in communication and language intervention. Communication and language intervention series, Vol. 1. (pp. 361-375). Baltimore, MD: Paul H. Brookes Publishing. Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research perspectives. Developmental Psychology, 22 (6), 723-742. Buggey, T., Toombs, K., Gard ener, P., & Cervetti, M. (1999). Training responding behaviors in students with autis m: Using videotaped self-modeling. Journal of Positive Behavior Interventions, 1 (4), 205-214. Campbell, S. B. (1994). Behavior problems in preschool children: A review of recent research. Journal of Child Clin ical Psychology, 36 (1), 113-149. Carr, E. G., Dunlap, G., Horner, R. H., Koeg el, R. L., Turnbull, A. P., Sailor, W., Anderson, J. L., Albin, R. W., Koegel, L. K., & Fox, L. (2002). Positive behavior support: Evolution of an applied science. Journal of Positive Behavior Interventions, 4 (1), 4-16, 20. Carr, E. G., Horner, R. H., Turnbull, A. P ., Marquis, J., Magito-McLaughlin, D., McAtee, M., Smith, C. E., Anderson-Ryan, K. A., Ruef, M. B., Doola bh, A., & Braddock, D. (1999). Positive behavior support for people with developmental disabilities: A research synthesis Washington, D.C.: Americ an Association on Mental Retardation. Carr, E.G., Levin, L., McConnac hie, G., Carlson, J. I., Kemp, D. C., & Smith, C. E. 1994). Communication-based intervention for pr oblem behavior: A user’s guide for producing behavior change Baltimore, MD: Paul H. Brookes Publishing Co.

PAGE 93

85 Carta, J. J. (2002). An early childhood special education res earch agenda in a culture of accountability for results. Journal of Early Intervention, 25 (2), 102-104. Carta, J. J., Schwartz, I. S., Atwater, J. B., & McConnell, S. R. (1991). Developmentally Appropriate Practice: Ap praising its usefulness for young children with disabilities. Topics in Early Chil dhood Special Education,11 (1), 1-20. Center for Evidence-Based Practice: Young Children with Challenging Behavior (n.d.). www.challengingbehavior.org Accessed January 3, 2005. Center on the Social and Emotional Foundations for Ea rly Learning. (n.d.). http://csefel.uiuc.edu/ Accessed January 3, 2005. Clark, M. L., Cunningham L. J., & Cunningham, C. E. (1989). Improving the social behavior of siblings of autistic children using a gr oup problem solving approach. Child and Family Be havior Therapy, 11 (1), 19-33. Clarke, S., Dunlap, G., Foster -Johnson, Childs, K. E., Wilson, D., White, R., & Vera, A. (1995). Improving the conduct of child ren with behavioral disorders by incorporating student interests into curricular activities. Behavioral Disorders, 20 (4), 221-237. Clarke, S., Worcester, J. A., Dunlap, G., Mu rray, M., & Bradley-Klug, K. (2002). Using multiple measures to evaluate positiv e behavior support: A case example. Journal of Positive Beha vior Interventions, 4 (3), 131-146. Costello-Ingham, J. & Riley, G. (1998). Gu idelines for document ation of treatment efficacy for young children who stutter. Journal of Speech, Language, and Hearing Research, 41 (4), 753-770.

PAGE 94

86 Division for Early Childhood. (1999). DEC concept paper on the identification of and intervention with challenging behavior Washington, D.C.: Author. Doss, L. S., & Reichle, J. (1991). Repl acing excess behavior with an initial communicative repertoire. In J. Reichle, J. York & J. Sigafoos (Eds.), Implementing augmentative and alternativ e communication: Strategies for learners with severe disabilities (pp. 256-265). Baltimore: Paul H. Brookes. Duchnowski, A. J. (2000). Improving family support: An agenda for the next decade. Journal of Positive Beha vior Interventions, 2 (2), 117-118. Duda, M., Dunlap, G., Fox, L., Lentini, R., & Clarke, S. (2004). An experimental evaluation of positive behavior suppor t in a community pr eschool program. Topics in Early Childhoo d Special Education, 24 (3), 143-155. Dunlap, G., Clarke, S., & Steine r, M. (1999). Intervention research in behavioral and developmental disabilities: 1980 to 1997. Journal of Positive Behavior Interventions, 1 (3), 170-180. Dunlap, G., Foster-Johnson, L., Clarke, S., Kern, L., & Childs, K. E. (1995). Modifying activities to produce func tional outcomes: Effects on the problem behaviors of children with disabilities. Journal of the A ssociation for Pers ons with Severe Handicaps, 20 (4), 248-258. Dunlap, G., & Fox, L. (1999). A demonstrat ion of behavioral s upport for young children with autism. Journal of Positive Beha vior Interventions, 1 (2), 77-87.

PAGE 95

87 Dunlap, G., & Fox, L. (1996). Early interv ention and serious behavior problems: A comprehensive approach. In L. K. Koegel R. L. Koegel, & G. Dunlap (Eds.) Positive behavioral support: Including people with difficult behavior in the community (pp. 31-50). Baltimore, MD: Paul H. Brookes Publishing Co. Dunlap, G., Horner, R. H., Carr, E. G., Sailor, W., Turnbull, A., Koegel, R. L., & Koegel, L. (1998). Research and Rehabilitation Training Center on Positive Behavioral Support (RRTC-PBS) Washington, DC: National Institute on Disability and Rehabilitation Research/Office of Speci al Education Programs (OSEP), U.S. Department of Education. Dunlap, G., Kern, L., dePerczel M., Clarke, S., Wilson, D., Ch ilds, K. E., White, R., & Falk, G. D. (1993). Functional analysis of classroom variable s for students with emotional and behavi oral disorders. Behavioral Disorders, 18 (4), 275-291. Dunlap, G., Kern-Dunlap, L., Clar ke, S., Robbins, F. R. (19 91). Functional assessment, curricular revision, and se vere behavior problems. Journal of Applied Behavior Analysis, 24 (2), 387-397. Dunlap, G., & Plienis, A. J. (1988). Generalization and ma intenance of unsupervised responding via remote contingencies. In R. J. Horner, G. Dunl ap, & R. L. Koegel (Eds.), Generalization and Maintenance: Life-style changes in applied settings, (pp. 121-142). Baltimore, MD: Paul H. Brookes Publishing Co., Inc. Dunlap, G., White, R., Vera, A ., Wilson, D., & Panacek, L. (1 996). The effects of multicomponent, assessment-based curricular modifications on the classroom behavior of children with emotional and behavioral disorders. Journal of Behavioral Education, 6 (4), 481-500.

PAGE 96

88 Ervin, R. A., Kern, L., Clarke, S., DuPaul, G., J., Dunlap, G., & Friman, P. C. (2000). Evaluating assessment-based intervention strategies for children with ADHD and comorbid disorders within the natural classroom context. Behavioral Disorders, 25 (4), 344-358. Favell, J. E., & Reid, D. J. (1988). Ge neralizing and maintaining improvement in problem behavior. In R. J. Horner, G. Dunlap, & R. L. Koegel (Eds.), Generalization and Maintenance: Life-s tyle changes in applied settings, (pp. 171-196). Baltimore, MD: Paul H. Brookes Publishing Co., Inc. Feinberg, M. E., Neiderhiser, J. M., Simmens, S., Reiss, D ., & Hetherington, E. M. (2000). Sibling comparison of differential parental trea tment in adolescence: Gender, self-esteem, and emotionality as mediators of the parenting-adjustment association. Child Development, 71 (6), 1611-1628. Floyd, F. J., Harter, K. S., & Costigan, C. L. (2004). Family problem-solving with children who have mental retardation. American Journal on Mental Retardation, 109 (6), 507-524. Fox, L., Clarke, S., & Duda, M. (2005). PBS for children and fam ilies in the Part C system: An early intervention model Presentation conducted at the meeting of the Second International Conference on Positive Behavior Support, Tampa, Florida. Fox, L., Dunlap, G., & Cushing, L. (2002). Early intervention, positive behavior support, and transition to school. Journal of Emotional and Behavioral Disorders, 10 (3), 149-157.

PAGE 97

89 Fox, L., Dunlap, G., Hemmeter, M. L., Joseph, G. E., & Strain, P. S. (2003). The teaching pyramid: A mode l for supporting so cial competence and preventing challenging behavior in young children. Young Children, 58 (4), 48-52. Fox, L., Dunlap, G., & Philbric k, L. A. (1997). Providing i ndividual supports to young children with autism and their families. Journal of Early Intervention, 21 (1), 114. Frea, W. D., Arnold, C. L., & Vi ttimberga, G. L. (2001). A demonstration of the effects of augmentative communication on the extr eme aggressive behavior of a child with autism within an in tegrated preschool setting. Journal of Positive Behavior Interventions, 3 194-198. Gable, R. A., Hendrickson, J. M., & Van Acker, R. (2001). Maintain ing the integrity of FBA-based interventions in schools. Education and Treatment of Children, 24 (3), 248-260. Galensky, T. L., Miltenberger, R. G., Stricker J. M., & Garlinghouse, M. A. (2001). Functional assessment and the treatment of mealtime behavior problems. Journal of Positive Behavior Interventions, 3 211-224. Gall, M. D., Borg, W. R., & Gall, J. P. (1996). Educational research : An introduction White Plains, NY: Longman Publishers. Garfinkle, A. N., & Schwartz, I. S. (2002). P eer imitation: Increasi ng social interactions in children with autism a nd other developmental disabili ties in inclusive preschool classrooms. Topics in Early Childho od Special Education, 22 (1), 26-38.

PAGE 98

90 Gersten, R., Baker, S., & Lloyd, J. W. (2000) Designing high-quality research in special education: Grou p experimental design. Journal of Special Education, 34 (1), 2-18. Gresham, F. M. (1989). Assessment of treat ment integrity in school consultation and prereferral intervention. School Psychology Review, 18 27-50. Gresham, F. M., Gansle, K. A., & Noell, G. H. (1993). Treatment integrity in applied behavior analysis with children. Journal of Applied Be havior Analysis, 26 (2), 257-263. Guralnick, M. J. (2000). An agenda for change in early childhood inclusion Journal of Early Intervention, 23 (4), 213-222. Halle, J. (1998). Fidelity: A crucial question in translat ing research to practice. Journal of Early Intervention, 21 (4), 294-296. Hancock, T. B., & Kaiser, A. P. (2002). Th e effects of trainer-imp lemented enhanced milieu teaching on the social communication of children with autism. Topics in Early Childhood Special Education, 22 (1), 39-54. Harrower, J. K., Fox, L., Dunlap, G., & Kincai d, D. (1999). Functi onal assessment and comprehensive early intervention. Exceptionality, 8 189-204. Hieneman, M., Nolan, M., Presley, J., DeTuro, L., Gayler, W., & Dunl ap, G. (1999). Facilitator’s Guide, Po sitive Behavioral Support Tallahassee, FL: Florida Department of Education, Bureau of Instructional Support and Community Services.

PAGE 99

91 Horner, R. H., & Billingsley, F. F. (1988). The effect of compet ing behavior on the generalization and maintenance of adaptive be havior in applied se ttings. In R. J. Horner, G. Dunlap, & R. L. Koegel (Eds.), Generalization and Maintenance: Life-style changes in applied settings, (pp. 197-220). Baltimore, MD: Paul H. Brookes Publishing Co., Inc. Horner, R. J., Dunlap, G., & Koegel, R. L. (1988). Generalization and Maintenance: Life-style changes in applied settings Baltimore, MD: Paul H. Brookes Publishing Co., Inc. Horner, R. H., Dunlap, G., Koegel, R. L., Carr, E. G., Sailor, W., An derson, J., Albin, R. W., & O’Neill, R. E. (1990). Toward a technology of “nonaversive” behavioral support. Journal of the Association for Persons with Severe Handicaps, 15 (3), 125-132. Horner, R. H., Sugai, G., Todd, A. W., & Le wis-Palmer, T. (2000). Elements of behavior support plans: A technical brief. Exceptionality, 8 (3), 205-215. Huffman, L. C., Mehlinger, S. L., Kerivan, A. S., Cavanaugh, D. A., Lippitt, J., & Moyo, O. (2001). Off to a Good Start: Research on the risk factors for early school problems and selected fede ral policies affecting child ren's social and emotional development and their readiness for school Bethesda, MD: National Institute of Mental Health. Hupp, S. D., & Reitman, D. (2000). Parent-assi sted modification of pi votal social skills for a child diagnosed with P DD: A clinical replication. Journal of Positive Behavior Interventions, 2 (3), 183-187.

PAGE 100

92 Ingoldsby, E. M., Shaw, D. S., Owens, E. B., & Winslow, E. B. ( 1999). A longitudinal study of interparental conflict, emot ional and behavioral reactivity, and preschoolers’ adjustment probl ems among low-income families. Journal of Abnormal Child Psychology, 27 (5), 343-356. Johnson, L. J., Kilgo, J., Cook, M. J., Hammitte, D. J. Beauchamp, K., & Finn, D. (1992). The Skills Needed by Ea rly Intervention Administra tors/Supervisors: A Study Across Six States. Journal of Early Intervention, 16 (2), 136-145. Kazdin, A. E. (1982). Single-Case Research Designs: Methods for clinical and applied settings New York, NY: Oxford University Press. Keenan, K., & Wakschlag, L. S. (2000). More than the terrible tw os: The nature and severity of behavior problems in clinic-referred preschool children. Journal of Abnormal Child Psychology,28 (1), 33-46. Kern, L., Childs, K. E., Dunla p, G., Clarke, S., & Falk, G. D. (1994). Using assessmentbased curricular intervention to improve the classroom behavior of a student with emotional and behavioral challenges. Journal of Applied Behavior Analysis, 27 (1), 7-19. Kern, L., Wacker, D. P., Mace, F. C., Falk, G. D., Dunlap, G., & Kromrey, J. (1995). Improving the peer interactions of st udents with emotional and behavioral disorders through self-evaluation proc edures: A component analysis and group application. Journal of Applied Be havior Analysis, 28 (1), 47-59. Koegel, L. K., Harrower, J. K., & Koegel, R. L. (1999). Suppor t for children with developmental disabilities in full inclusion classroom s through self-management. Journal of Positive Beha vior Interventions, 1 (1), 26-34.

PAGE 101

93 Koegel, L. K., Koegel, R. L ., & Dunlap, G. (1996). Positive Behavioral Support: Including people with diffic ult behavior in the community Baltimore, MD: Paul H. Brookes Publishing Co. Koegel, R. L., Symon, J., B., & Koegel, L. K. (2002). Parent education for families of children with autism living in geographically distant areas. Journal of Positive Behavior Interventions, 4 (2), 88-103. Landy, S. (2002). Pathways to Compet ence: Encouraging Healthy Social and Emotional Development in Young Children Baltimore, MD: Paul H. Brookes Publishing Co. Lawry, J., Danko, C. D., & Strain, P. S. ( 1999). Examining the role of the classroom environment in the prevention of prob lem behaviors. In S. Sandall & M. Ostrosky (Eds.). Young exceptional children: Practical ideas for addressing challenging behaviors Young Exceptional Children Monograph Series No. 2 ., (p.p. 49-61). Denver, CO: Division for Early Childhood of the Council for Exceptional Children. LeLaurin, K. & Wolery, M. ( 1992). Research standards in ea rly intervention: Defining, describing, and measuring th e independent variable. Journal of Early Intervention, 16 275-287. Lewis, T. J., & Sugai, G. (1999). Effective beha vior support: A syst ems approach to proactive school-wide management. Focus on Exceptional Children, 31 (6), 1-24. Lucyshyn, J. M., Albin, R. W., & Nixon, C. D. (1997). Embedding comprehensive behavioral support in family ecology: An experiment al, single-case analysis. Journal of Consulting and Clinical Psychology, 65 241-251.

PAGE 102

94 Lucyshyn, J. M., Dunlap, G., & Albin, R. W. (2002). Families and Positive Behavior Support: Addressing Problem Be havior in Family Contexts Baltimore, MD: Paul H. Brookes Publishing Co. Lucyshyn, J. M., Horner, R. H., Dunlap, G., Al bin, R. W., & Ben, K. R. (2002). Positive Behavior Support with Families. In J. M. Lucyshyn, G. Dunlap, & R. W. Albin (Eds.). Families and Positive Be havior Support: Addres sing Problem Behavior in Family Contexts (pp. 3-43). Baltimore, MD: Paul H. Brookes Publishing Co. McEvoy, M., & Reichle, J. (1995). Developing and evaluating a model of inservice and technical assistance to prevent chal lenging behavior in preschoolers University of Minnesota, College of Education, Institute on Comm unity Integration, Center for Early Education and Development. McGillicuddy-De Lisi, A. V. (1993). Sibli ng interactions and children’s communicative competency. Journal of Applied Devel opmental Psychology, 14 (3), 365-383. Moes, D. R., & Frea, W. D. (2000). Usi ng family context to inform intervention planning for the treatment of a child with autism. Journal of Positive Behavior Interventions, 2 (1), 40-46. National Association for the Education of Young Child ren (NAEYC). (1991). Accreditation criteria and procedures of the National Academy of Early Education Programs Washington, D.C.: Author. National Institutes of Health. (1990). Treatment of destructive behaviors in persons with developmental disabilities (cons ensus development statement). Journal of Autism and Developmental Disorders, 12 403-429.

PAGE 103

95 National Research Council and In stitute of Medicine. (2000). From neurons to neighborhoods: The science of early childhood development Committee on Integrating the Science of Early Childhood Development. J. P. Shonkoff & D. A. Phillips (Eds.). Board on Childre n, Youth, and Families, Commission on Behavioral and Social Sc iences and Education. Wa shington, D.C.: National Academy Press. Nielsen, S., Olive, M., Donovan, A., & McEvoy, M. (1998). Challengi ng behaviors in your classroom? Don’ t react—teach instead! Young Exceptional Children, 2 (1), 2-10. O’Neill, R. E., Horner, R. H., Albin, R. W., Sprague, J. R., Storey, K., & Newton, J. S. (1997). Functional assessment and program development for problem behavior Pacific Grove, CA: Br ooks/Cole Publishing Co. OSEP Technical Assistance Center on Pos itive Behavioral Interventions and Supports (n.d.). www.pbis.org Accessed January 11, 2005. Park, J., Turnbull, A. P., & Tur nbull, H. R. (2002). Impacts of poverty on quality of life in families of children with disabilities. Exceptional Children, 68 (2), 151-170. Pierce, E. W., Ewing, L. J., & Campbell, S. B. (1999). Diagno stic status and symptomatic behavior of hard-to-manage preschool children in middle childhood and early adolescence. Journal of Clinical Child Psychology, 28 (1), 44-57. Powell, D., Fixsen, F., & Dunlap, G. (2003). Pathways to servi ce utilization: A synthesis of evidence relevant to young children with challenging behavior Tampa, FL: University of South Florida, Center for Evidence-Based Practice: Young Children with Challenging Behavior.

PAGE 104

96 Ram, A., & Ross, H. S. (2001). Problemsolving, contention, and struggle: How siblings resolve a conflict of interests. Child Development, 72 (6), 1710-1722. Repp, A.C. & Horner, R.H. (Eds). (1999). Functional analysis of problem behavior: From effective assessmen t to effective support Belmont, CA: Wadsworth Publishing. Rusch, F. R., & Kazdin, A. E. (1981). Towa rd a methodology of w ithdrawal designs for the assessment of response maintenance. Journal of Applied Behavior Analysis, 14 (2), 131-140. Sandall, S., & Ostrosky, M. (1999). Young exceptional childre n monograph series: Practical ideas for addres sing challenging behaviors Washington, D. C.: Division for Early Childhood of the Council for Exceptiona l Children. Schepis, M. M., Ownbey, J. B., Parsons, M. B., & Reid, D. H. (2000). Training support staff for teaching young children with disa bilities in an in clusive preschool setting. Journal of Positive Beha vior Interventions, 2 (3), 170-178. Schreibman, L., Whalen, C., & Stahmer, A. C. (2000). The use of video priming to reduce disruptive transition behavi or in children with autism. Journal of Positive Behavior Interventions, 2 (1), 3-11. Shirley, M., Iwata, B., Kahng, S., Mazaleski, J ., & Lerman, D. (1997). Does functional communication training compete with ongoing contingencies of reinforcement? An analysis during response acquisition and maintenance. Journal of Applied Behavior Analysis, 30(1) 93-104.

PAGE 105

97 Singer, G. H., Goldberg-Hamblin, S. E; Peck ham-Hardin, K. D.; Barry, L.; & Santarelli, G. E. (2002). Toward a synthesis of family support practices and positive behavior support. In J.M. Lucyshy n, G. Dunlap, & R. W. Albin (Eds.). Families and Positive Beha vior Support: Addressi ng Problem Behavior in Family Contexts (pp. 155-183). Baltimore, MD: Paul H. Brookes Publishing Co. Smith, B. J., & Fox, L. (2003). Systems of service delivery: A synthesis of evidence relevant to young children at risk of or who have challenging behavior Tampa, FL: Center for Evidence -based Practice: Young children with challenging behavior. Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10 (2), 349-367. Stokes, T.F., & Osnes, P. G. (1989). An operant pursuit of generalization. Behavior Therapy, 20 337-355. Sugai, G., Horner, R. H., Dunlap, G., Hienema n, M., Lewis, T., J., Nelson, C. M., Scott, T., Liaupsin, C., Sailor, W ., Turnbull, A. P., Turnbull III, H. R., Wickham, D., Wilcox, B., & Ruef, M. (2000). A pplying positive behavior support and functional behavioral assessment in schools. Journal of Positive Behavior Interventions, 2 (3), 131-143. Sugai, G., Sprague, J. R., Ho rner, R. H., & Walker, H. M. (2000). Preventing school violence: The use of offi ce discipline referrals to a ssess and monitor school-wide discipline interventions. Journal of Emoti onal and Behavioral Disorders, 8 (2), 94-101.

PAGE 106

98 Sulzer-Azaroff, B., & Mayer, G. R. (1991). Behavior analysis for lasting change Forth Worth, TX: Harcourt Brace. The No Child Left Behind Act of 2001 (n.d.). www.nclb.org/ Accessed January 3, 2005. Tawney, J. W. & Gast, D. L. (1984). Single Subject Research in Special Education Columbus, OH: Charles E. Merrill Publishing Co. Todd, A. W., & Horner, R. H. (2002). School-wide Information System Users Manual (2002). Eugene, OR: Educational a nd Community Supports, University of Oregon. Trivette, C. M. (2003). In fluence of caregiver responsiv eness on the development of young children with or at-risk fo r developmental disabilities. Bridges, 1 (3). Turnbull, III, H. R., Turnbull, A. P., Wehmeyer, M. L., & Park, J. (2003). A quality of life framework for special education outcomes. Remedial and Special Education, 24 (2), 67-74. University of Florida College of Medicine Perinatal Data/Research Center. (2000). Early Intervention Program (EIP) Center 01-Tampa: Graphics & Trends (Report #1). Gainesville, FL: University of Fl orida College of Medicine, Perinatal Data/Research Center. U.S. Department of Education, Office of the Secr etary, Office of Public Affairs. (2003). No Child Left Behind : A parent’s guide Washington, D.C.: Author. U.S. Department of Education, Office of Sp ecial Education and Reha bilitative Services (2002). A New Era: Revitaliz ing Special Education for Children and Their Families Washington, D.C.: Author.

PAGE 107

99 U.S. Department of Health and Human Servi ces, Office of the Surg eon General (2001). Youth Violence: A Report of the Surgeon General Washington, D.C.: Author. Vaughn, B. J., Dunlap, G., Fox, L., Clarke, S., & Bucy, M. (1997). Parent-professional partnership in behavioral support: A cas e study of community-based intervention. Journal of The Association for Persons with Severe Handicaps, 22 (4), 186-197. Vaughn, B. J., Clarke, S., & Dunlap, G. ( 1997). Assessment-based intervention for severe behavior problems in a natural family context. Journal of Applied Behavior Analysis, 30 (4), 713-716. Walker, H. M., Horner, R. H., Sugai, G., Bu llis, M., Sprague, J. R., Bricker, D., & Kaufman, M. J. (1996). Integrated appr oaches to preventing antisocial behavior patterns among school-age children and youth. J ournal of Emotional and Behavioral Disorders, 4 (4), 194-209. Walker, H. M., Stiller, B., & Golly, A. (1999) First steps to success: A collaborative home-school intervention for pr eventing antisocial behavior at the point of school entry. In S. Sandall & M. Ostrosky (Eds.). Young exceptional children: Practical ideas for addressi ng challenging behaviors Young Exceptional Children Monograph Series No. 2 Denver, CO: Divisi on for Early Childhood of the Council for Excep tional Children, 41-48. Webster-Stratton, C. (1999). How to promote children’ s social and emotional competence London: Paul Chapman. Weigle, K. L. (1997). Positive behavior s upport as a model for promoting educational inclusion. Journal of the Association for Persons with Severe Handicaps, 22 (1), 36-48.

PAGE 108

100 Wolery, M. (1994). Proc edural fidelity: A reminder of its functions. Journal of Behavioral Education, 4 (4), 381-386. Yeaton, W., & Sechrest, L. (1981 ). Critical dimensions in the choice and maintenance of successful treatments: Stre ngth, integrity, and effectiveness. Journal of Consulting and Clinic al Psychology, 49 156-167.

PAGE 109

101 Appendices

PAGE 110

102 Appendix A: Figures

PAGE 111

103 Figure 1. Composite Percentage of Intervals with Challenging Behavior

PAGE 112

104 Figure 2. Composite Percentage of In tervals with Challenging Behavior and Engagement: Baseline and Interv ention—Clean Up and Twin Play Composite Percentage of Intervals with Challenging Behavior Clean Up and Twin Play Max Zak

PAGE 113

105 Figure 3. Composite Percentage of In tervals with Challenging Behavior and Engagement: Baseline and Inte rvention—All Play and Dinner Composite Percentage of Intervals with Challengin g Behavior and Engagement All Play and Dinner Max Zak Emm y

PAGE 114

106 Figure 4. Composite Percentage of In tervals with Challenging Behavior: Component Analysis—Clean Up and Twin Play

PAGE 115

107 Figure 5. Component Analysis Indivi dual Percentage of Intervals with Challenging Behavior: Clean Up

PAGE 116

108 Figure 6. Component Analysis Indivi dual Percentage of Intervals with Challenging Behavior: Twin Play

PAGE 117

109 Figure 7. Composite Percentage of In tervals with Challenging Behavior: Component Analysis—All Play and Dinner 0

PAGE 118

110 Figure 8. Component Analysis Indivi dual Percentage of Intervals with Challenging Behavior: All Play

PAGE 119

111 Figure 9. Component Analysis Indivi dual Percentage of Intervals with Challenging Behavior: Dinner

PAGE 120

112 Appendix B: Intervention Component Reduction Data

PAGE 121

113 Table 9. Intervention Component Reduction Data: Clean Up Intervention Components Fidelity Score Parent Rating (Relation) Parent Rating (Use) Decision Remaining Components 1. Give 4 min. warning face to face with child (show photo). 28% Max 44% Zak 3 3 N/A N/A Cut 2. Give 1 min. warning face to face with child (show photo). 33% Max 44% Zak 3 3 N/A N/A Keep Give 1 min. warning face to face with child (show photo). 3. “All done” with previous activity clearly stated. 6% 3 3 Keep “All done” with previous activity clearly stated. 4. “Time to clean up” clearly stated. 94% 3 3 Keep “Time to clean up” clearly stated. 5. Provided child with opportunity to choose song/character. 76% 2 2 Keep Provided child with opportunity to choose song/character. 6. Played music to indicate beginning of activity. 100% 3 3 Keep Played music to indicate beginning of activity. 7. Counted number of toys children put in box. 6% Max 22% Zak 1 1 1 1 Cut 8. Provided praise for picking up toys. 88% Max 94% Zak 3 3 3 3 Keep Provided praise for picking up toys. 9. Celebrated goal at end of activity (i.e., done with clean up). 94% 3 3 Keep Celebrated goal at end of activity (i.e., done with clean up). 10. Gave verbal cue that clean up is over (i.e., all done). 94% 3 3 Keep Gave verbal cue that clean up is over (i.e., all done). 11. Provided a reinforcer (reward) for completing routine. 100% Max 100% Zak 3 3 2 2 Keep Provided a reinforcer (reward) for completing routine. (Table Continues)

PAGE 122

114 (Table Continued) 12. Provided choice of a preferred reinforcer (e.g., letting boys choose a badge/song). 100% Max 100% Zak 2 2 2 2 Keep Provided choice of a preferred reinforcer (e.g., letting boys choose a badge/song). 13. Provided verbal cue and photo of next activity. 43% Max 60% Zak 3 3 3 3 Keep Provided verbal cue and photo of next activity. Notes: Relation to Behavior: 1 = I do not think there is a relationship; 2 = Unsure; 3 = I think there is a definite relationship; N/A = No response. Long-Term Use: 1 = I’d really like to drop it; 2 = I’d like to drop it if possible; 3 = I can see myself using it; N/A = No re sponse. = Included wi thin component analysis.

PAGE 123

115 Table 10. Intervention Component Reduction Data: Twin Play Intervention Components Fidelity Score Parent Rating (Relation) Parent Rating (Use) Decision Remaining Components 1. Gave both children a clear verbal cue of what is going to happen (e.g., “It’s playtime children! Let’s pick some toys”). 70% 3 3 Keep Gave both children a clear verbal cue of what is going to happen. 2. Set clear expectation (e.g., “You are going to play while Mommy is in the kitchen doing her work”.) 69% 3 3 Keep Set clear expectation 3. Each child was given a choice of 1 toy set (e.g., cars, blocks). 79% 3 3 Keep Each child was given a choice of 1 toy set 4. Adult selected a third toy set to help create play theme. 100% 2 2 Keep Adult selected a third toy set to help create play theme* 5. A 4th toy set was selected and put in family room. 100% 2 1 Keep A 4th toy set was selected and put in family room* 6. Verbal prompt given to take toys into the family room. 83% Max 82% Zak 1 1 1 1 Keep Verbal prompt given to take toys into the family room. 7. Praised children (each individually) for bringing out toys (e.g., “yeah, it’s playtime. That is good helping bringing out the toys”). 42% Max 27% Zak 1 1 3 3 Keep Praised children (each individually) for bringing out toys (e.g., “yeah, it’s playtime. That is good helping bringing out the toys”). 8. Selected theme and presented it to children (e.g., “you are going to play while Mommy is in the kitchen doing dishes”). 100% 3 3 Keep Selected theme and presented it to children. 9. Once each child picks up first toy, adult set clear expectation (e.g., “Mommy will be in the kitchen, you keep playing”). 100% 3 3 Keep Once each child picks up first toy, adult set clear expectation. (Table Continues)

PAGE 124

116 Table (Continued) 10. Provided praise while out of area every 5 min. (e.g., “You are playing so nicely together. I like how you are playing while Mommy is doing her work”). 79% 2 2 Keep Provided praise while out of area every 5 min* 11. After 15 min. have passed, mother praised children and gave a verbal warning that playtime is almost “all done.” 64% 3 3 Keep After 15 min. have passed, mother praised children and gave a verbal warning that playtime is almost “all done.” Notes: Relation to Behavior: 1 = I do not think there is a relationship; 2 = Unsure; 3 = I think there is a definite relationship; N/A = No response. Long-Term Use: 1 = I’d really like to drop it; 2 = I’d like to drop it if possible; 3 = I can see myself using it; N/A = No re sponse. = Included wi thin component analysis.

PAGE 125

117 Table 11. Intervention Component Reduction Data: All Play Intervention Components Fidelity Score Parent Rating (Relation) Parent Rating (Use) Decision Remaining Components 1. Prior to playtime, “All Play” social story was read to Emmy. 44% 3 2 Cut 2. Provided clear verbal cue of what is going to happen (e.g., “It’s playtime everybody! Let’s pick some toys”). 75% 3 3 Keep Provided clear verbal cue of what is going to happen 3. Children provided clear expectation (e.g., “Pick some toys”). 88% 3 3 Keep Children provided clear expectation 4. Each child was given a choice of 1 toy set (e.g., cars, blocks). 88% 3 3 Keep Each child was given choice of 1 toy set* 5. Mom selected additional toy sets that are preferred by all children. 80% 2 1 Keep Mom selected additional toy sets that are preferred by all children* 6. Verbal prompts were given to take toys in the family room. 71% Max/ Zak 67% Emmy 2 2 1 Keep Verbal prompts were given to take toys in the family room. 7. Praised children for bringing out toys. 29% 3 3 Cut 8. Mom reviewed rules with Emmy (i.e., “I can be a helper at playtime”) and Emmy had access to “rule list”. 33% 3 2 Cut 9. Mom or Emmy gave suggestions for toy play activity. 100% 3 3 Keep Mom or Emmy gave suggestions for toy play activity. 10. Once each child picked up the first set of toys, Mom told children that she is leaving the area and Emmy will help (verbal prompts). 100% 3 3 Keep Mom told children that she is leaving the area and Emmy will help (verbal prompts). (Table Continues)

PAGE 126

118 (Table Continued) 11. During playtime, Emmy was coached on prompting and/or teaching brothers how to use toys, as well as how to provide praise. 100% 3 3 Keep Emmy was coached on prompting and/or teaching brothers how to use toys, as well as how to provide praise. 12. Mom provides praise to children while remaining out of area every 5 min. 100% 3 2 Keep Mom provides praise to children while remaining out of area every 5 min* 13. Mom provided specific praise to Emmy for being a “helper.” 90% 3 3 Keep Mom provided specific praise to Emmy for being a “helper.” 14. After 20 min. have passed, praise children and giver a verbal warning that playtime is “almost all done.” 44% 3 3 Cut 15. At the end of play Mom asked Emmy if everyone followed the rules and asked how she played. 0% 3 3 Cut Notes: Relation to Behavior: 1 = I do not think there is a relationship; 2 = Unsure; 3 = I think there is a definite relationship; N/A = No response. Long-Term Use: 1 = I’d really like to drop it; 2 = I’d like to drop it if possible; 3 = I can see myself using it; N/A = No re sponse. = Included wi thin component analysis.

PAGE 127

119 Table 12. Intervention Compone nt Reduction Data: Dinner Intervention Components Fidelity Score Parent Rating (Relation) Parent Rating (Use) Decision Remaining Components 1. Provided Dinner social story to Emmy prior to dinner. 38% 3 2 Cut 2. Emmy was given opportunity to read her social story by herself or read it to her brothers in order to review dinner rules. 38% 2 1 Cut 3. Emmy used self-monitoring materials and choice menus. 50% 3 3 Cut 4. Emmy was given opportunity to help set the table and/or put food on table. 90% 3 3 Keep Emmy was given opportunity to help set the table and/or put food on table. 5. Prior to sitting down, dinner was completely prepared and on the table. 50% 3 3 Keep Prior to sitting down, dinner was completely prepared and on the table. 6. Seating arrangement was modified. 90% 3 3 Keep Seating arrangement* 7. Choice of 2 food items was provided to children (1 preferred, 1 backup). 67% 2 1 Keep Choice of 2 food items was provided to children (1 preferred, 1 backup). 8. If Emmy/children refused to eat after choice was given, Emmy / children were instructed that they must sit at table for 5 minutes. 25% 3 3 Cut 9. Mom sat with children for entire duration of mealtime. 80% 3 3 Keep Mom sat with children for entire duration of mealtime* 10. Followed child’s lead for dinner conversation. 100% 3 3 Keep Followed child’s lead for dinner conversation. 11. Praise was provided throughout the routine. 40% Max 10% Zak 3 3 Keep Praise was provided throughout the routine. (Table Continues)

PAGE 128

120 (Table Continued) 12. Specific praise was provided to Emmy for appropriate behavior and self-monitoring. 10% 3 3 Keep Specific praise was provided to Emmy for appropriate behavior and selfmonitoring* 13. Each child asked to be excused or mother gave permission to leave table once finished with dinner. 90% Max/E mmy 89% Zak 3 3 Keep Each child asked to be excused or mother gave permission to leave table once finished with dinner. 14. After dinner, mother matched self-monitoring items with Emmy. 57% 2 2 Keep After dinner, mother matched selfmonitoring items with Emmy* 15. Choice menu provided to Emmy for self-monitoring if she matched with Mom and had over 80% appropriate behavior. 20% 3 3 Keep Choice menu provided to Emmy for selfmonitoring if she matched with Mom and had over 80% appropriate behavior. 16. Mother sets up a video for the boys and immediately starts “Mom and Emmy” time. 33% 3 3 Keep Mother sets up a video for the boys and immediately starts “Mom and Emmy” time* Notes: Relation to Behavior: 1 = I do not think there is a relationship; 2 = Unsure; 3 = I think there is a definite relationship; N/A = No response. Long-Term Use: 1 = I’d really like to drop it; 2 = I’d like to drop it if possible; 3 = I can see myself using it; N/A = No re sponse. = Included wi thin component analysis.

PAGE 129

121 Appendix C: Natural Only Implementation Patterns

PAGE 130

122 Table 13. Natural Only Implementation Patterns: Clean Up Intervention Components Natural Only Day 1 Natural Only Day 2 Natural Only Day 3 1. Give 4 min. warning. 2. Give 1 min. warning. 2. Give 1 min. warning. 2. Give 1 min. warning. 2. Give 1 min. warning. 3. “All done” clearly stated. 4. “Time to clean up” clearly stated. 4. “Time to clean up” clearly stated. 4. “Time to clean up” clearly stated. 5. Provided opportunity to choose song/character. 6. Played music to indicate beginning of activity. 7. Counted # of toys children put in box. 7. Counted # of toys children put in box. 7. Counted # of toys children put in box. 8. Praised Max for picking up toys. 8. Praised Max for picking up toys. 8. Praised Zak for picking up toys. 8. Praised Zak for picking up toys. 8. Praised Zak for picking up toys. 8. Praised Zak for picking up toys. 9. Celebrated goal at end. 10. Gave verbal cue that clean up is over. 11. Provided reinforcer for completing routine. 11. Provided reinforcer for completing routine. 12. Provided choice of preferred reinforcer. 12. Provided choice of preferred reinforcer. 13. Provided verbal cue and photo of next activity.

PAGE 131

123 Table 14. Natural Only Implementation Patterns: Twin Play Intervention Components Natural Only Day 1 Natural Only Day 2 Natural Only Day 3 1. Gave both children a clear verbal cue of what is going to happen. 2. Set clear expectation. 2. Set clear expectation. 2. Set clear expectation. 3. Each child given choice of 1 toy set. 3. Each child given choice of 1 toy set. 3. Each child given choice of 1 toy set. 3. Each child given choice of 1 toy set. 4. Adult selected 3rd toy set to create theme. 4. Adult selected 3rd toy set to create theme. 4. Adult selected 3rd toy set to create theme. 4. Adult selected 3rd toy set to create theme. 5. 4th toy set selected and put in family room. 5. 4th toy set selected and put in family room. 6. Verbal prompt given to take toys into family room. 7. Praised children for bringing out toys. 8. Selected theme and presented it to children. 8. Selected theme and presented it to children. 8. Selected theme and presented it to children. 9. Once each child picks up first toy, adult set clear expectation. 9. Once each child picks up first toy, adult set clear expectation. 9. Once each child picks up first toy, adult set clear expectation. 9. Once each child picks up first toy, adult set clear expectation. 10. Provided praise while out of area every 5 min. 10. Provided praise while out of area every 5 min. 11. After 15 min. have passed, mother praised children and gave a verbal warning that playtime is almost “all done.” 11. After 15 min. have passed, mother praised children and gave a verbal warning that playtime is almost “all done.” 1. Gave both children a clear verbal cue of what is going to happen.

PAGE 132

124 Table 15. Natural Only Implementation Patterns: All Play Intervention Components Natural Only Day 1 Natural Only Day 2 Natural Only Day 3 1. Prior to playtime, social story read to Emmy. 2. Provided clear verbal cue of what is going to happen. 3. Children provided clear expectation. 4. Each child given choice of 1 toy set. 4. Each child given choice of 1 toy set. 4. Each child given choice of 1 toy set. 4. Each child given choice of 1 toy set. 5. Mom selected additional preferred toy sets. 5. Mom selected additional preferred toy sets. 5. Mom selected additional preferred toy sets. 5. Mom selected additional preferred toy sets. 6. Verbal prompts to take toys into family room. 7. Praised children for bringing out toys. 8. Mom reviewed rules with Emmy. 9. Mom/Emmy suggested toy play activity. 9. Mom/Emmy suggested toy play activity. 9. Mom/Emmy suggested toy play activity. 9. Mom/Emmy suggested toy play activity. 10. Mom told children she is leaving area and Emmy will help. 10. Mom told children she is leaving area and Emmy will help. 10. Mom told children she is leaving area and Emmy will help. 10. Mom told children she is leaving area and Emmy will help. 11. Emmy coached on prompting and/or teaching brothers how to use toys and provide praise. 11. Emmy coached on prompting and/or teaching brothers how to use toys and provide praise. 11. Emmy coached on prompting and/or teaching brothers how to use toys and provide praise. 12. Mom praised children while remaining out of area every 5 min. 12. Mom praised children while remaining out of area every 5 min. 12. Mom praised children while remaining out of area every 5 min. 12. Mom praised children while remaining out of area every 5 min. 13. Mom praised Emmy for being a “helper.” 13. Mom praised Emmy for being a “helper.” 13. Mom praised Emmy for being a “helper.” 14. After 20 min., praised children/gave verbal warning that playtime is “almost all done.” 14. After 20 min., praised children/gave verbal warning that playtime is “almost all done.” 14. After 20 min., praised children/gave verbal warning that playtime is “almost all done.” 14. After 20 min., praised children/gave verbal warning that playtime is “almost all done.” 15. At end, Mom reviewed rules with Emmy. 16. At end, Mom praised Emmy for being a good helper. 16. At end, Mom praised Emmy for being a good helper.

PAGE 133

125 Table 16. Natural Only Impl ementation Patterns: Dinner Intervention Components Natural Only Day 1 Natural Only Day 2 Natural Only Day 3 1. Provided social story to Emmy prior to dinner. 2. Emmy given opportunity to read social story. 3. Emmy used selfmonitoring materials/choice menus. 4. Emmy given opportunity to set the table and/or put food on table. 5. Prior to sitting down, dinner completely prepared and on the table. 5. Prior to sitting down, dinner completely prepared and on the table. 6. Seating arrangement modified. 6. Seating arrangement modified. 6. Seating arrangement modified. 6. Seating arrangement modified. 7. Choice of 2 food items provided to children (1 preferred, 1 backup). 7. Choice of 2 food items provided to children (1 preferred, 1 backup). 8. If refused to eat, Emmy /children were instructed they must sit at table for 5 minutes. 9. Mom sat with children for entire meal. 9. Mom sat with children for entire meal. 9. Mom sat with children for entire meal. 9. Mom sat with children for entire meal. 10. Followed child’s lead for dinner conversation. 10. Followed child’s lead for dinner conversation. 10. Followed child’s lead for dinner conversation. 10. Followed child’s lead for dinner conversation. 11. Praised Max throughout routine. 11. Praised Max throughout routine. 11. Praised Max throughout routine. 11. Praised Zak throughout routine. 11. Praised Zak throughout routine. 12. Praised Emmy for appropriate behavior and self-monitoring. 12. Praised Emmy for appropriate behavior and self-monitoring. 12. Praised Emmy for appropriate behavior and self-monitoring. 13. Max asked to be excused or mother gave permission to leave table once finished. 13. Max asked to be excused or mother gave permission to leave table once finished. 13. Max asked to be excused or mother gave permission to leave table once finished. 13. Max asked to be excused or mother gave permission to leave table once finished. 13. Zak asked to be excused or mother gave permission to leave table once finished. 13. Zak asked to be excused or mother gave permission to leave table once finished. 13. Zak asked to be excused or mother gave permission to leave table once finished. (Table Continues)

PAGE 134

126 (Table Continued) 13. Emmy asked to be excused or mother gave permission to leave table once finished. 13. Emmy asked to be excused or mother gave permission to leave table once finished. 13. Emmy asked to be excused or mother gave permission to leave table once finished. 13. Emmy asked to be excused or mother gave permission to leave table once finished. 14. After dinner, mother matched selfmonitoring items with Emmy. 15. Choice menu provided to Emmy for self-monitoring if she matched with Mom and had over 80% appropriate behavior. 16. Mother sets up a video for the boys and immediately starts “Mom and Emmy” time. 16. Mother sets up a video for the boys and immediately starts “Mom and Emmy” time.

PAGE 135

About the Author Michelle Duda received a Bachelor’s Degr ee in Psychology at the University of Western Ontario in London, Ontario, Ca nada in 1999 and a Diploma from the Behavioural Science Technology program (Intensive Track) at St. Lawrence College in Kingston, Ontario, Canada in 2000. Later that year, Michelle began graduate study at the University of South Florida. She was the fi rst international studen t to earn a master’s degree in Applied Behavior Analysis in 2002, enrolled in the Special Education doctoral program the same year, and beca me a doctoral candidate in 2004. Michelle is a Board Certified Behavior Analyst with extensive experience supporting young children with challenging be havior and their families. She has conducted applied research in the areas of positive behavi or support, appl ied behavior analysis, and early intervention/early chil dhood special education. Michelle is the author/co-author of several p eer-reviewed articles and has pr esented/facilitated training ranging from the local to international level.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001709508
003 fts
005 20060614112135.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 060510s2005 flua sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001297
035
(OCoLC)68623078
SFE0001297
040
FHM
c FHM
049
FHMM
090
LC3965 (Online)
1 100
Duda, Michelle A.
0 245
Evaluating positive behavior support plan implementation in the home environment of young children with challenging behavior
h [electronic resource] /
by Michelle A. Duda.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (Ph.D.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 135 pages.
Includes vita.
520
ABSTRACT: In recent years, a central focus of the field of early intervention/early childhood special education has been to investigate ways to effectively support young children with challenging behavior and their families (Center for Evidence-Based Practice: Young Children with Challenging Behavior, 2003; DEC, 1999). Positive behavior support (PBS) is one of the most promising evidence-based practices for young children with challenging behavior and their families. The central purposes of PBS are to both help people develop and engage in socially desirable behaviors and to help minimize patterns of socially stigmatizing responding (Koegel, Koegel, and Dunlap, 1996). Research documenting the utility and applicability of PBS with preschool-aged populations remain scarce, particularly within natural environments (e.g., Blair, Umbreit, and Eck, 2000; Duda, Dunlap, Fox, Clarke, and Lentini, 2004; Moes and Frea, 2000).Several gaps in the research remain, including studies incorporating natural intervention agents, natural settings, and studies measuring technical aspects of behavior change (e.g., maintenance). Though studies of maintenance may be difficult to execute, they may provide researchers with a greater understanding of which factors in the change process are most critical to successful implementation, as well as to enhance the goodness of fit between specific plan components and the ecology in which implementation occurs (Albin, Lucyshyn, Horner, and Flannery, 1996). The purpose of this research study was to first assess the relationship of support plan components to behavior change, and then systematically fade the functional components, reducing the plan to naturalistic strategies that may be easy for the family to use over time.Results indicated each of the three child participants consistently maintained low levels of challenging behavior and high levels of engagement within each routine, despite the fact that clear functional relationships among individual intervention components were not attained. Procedural fidelity data indicated that intervention components were both implemented by the mother on a consistent basis and were easily adapted into natural family routines over time.
590
Adviser: Albert Duchnowski, Ph.D.
Co-adviser: Lise Fox, Ph.D.
653
Component analysis.
Early childhood.
Family.
Maintenance.
690
Dissertations, Academic
z USF
x Special Education
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1297