|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 TEACHER IMPLEMENTED TREATMENT PROBES: A CONSULTATIVE PROCEDURE FOR FU NCTIONAL ANALYSIS OF CLASSROOM BEHAVIOR By ELIZABETH LANE WEEKS MCKENNEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Elizabeth L. W. McKenney
3 To all who have supported me in this work; especially those who have nurtured my dreams throughout my lifetime, my succe ss is based on your belief
4 ACKNOWLEDGMENTS Thanks go out to the m embers of my comm ittee for your support and insight, without which none of this would have been possible. Special thanks to my chair, Nancy Waldron, for sharing so much of your knowledge with me over the years, and to Maureen Conroy for continuing to offer me your confidence and suppor t across the miles. I would also like to thank my data assistants, Allison Sullivan and Susa nne Long, whose flexibility and enthusiasm has made this experience so very pleasant. Many, ma ny thanks go out to a ll of the teachers and students who participated in this study, for their time, their willingness to try something new, and their honest feedback. Finally, the greatest thanks of all go to my family, near and far, for their unwavering support and understanding.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 ABSTRACT ...................................................................................................................... .............10 CHAP TER 1 INTRODUCTION AND LI TERATURE REVIEW .............................................................. 12 Introduction .................................................................................................................. ...........12 Literature Review ...................................................................................................................15 Targeted Populations .......................................................................................................15 Functional Behavioral Assessment .................................................................................. 18 Functional Analysis in the Schools ................................................................................. 26 Student Involvement in Functional Assessments ............................................................ 32 Teacher Implementation a nd Treatm ent Integrity ........................................................... 34 The Role of Consultation in Intervention ........................................................................39 Conclusion and Research Questions ....................................................................................... 47 2 METHOD ........................................................................................................................ .......50 Participants and Recruitment .................................................................................................. 50 Teachers ...................................................................................................................... .....50 Teacher A .................................................................................................................51 Teacher B .................................................................................................................51 Teacher C .................................................................................................................51 Students ...........................................................................................................................52 Student A ..................................................................................................................53 Student B ..................................................................................................................53 Student C ..................................................................................................................54 Settings ...................................................................................................................... ......54 Experimental Design ....................................................................................................... 54 Measurement Procedures ................................................................................................. 55 Interobserver Agreement ................................................................................................. 57 Baseline .................................................................................................................... 58 Training .................................................................................................................... 59 FA implementation ...................................................................................................59 Procedures .................................................................................................................... ...........60 Preassessment ................................................................................................................. .60 Teacher Training .............................................................................................................60 Peer Training ...................................................................................................................62
6 Functional Analysis Conditions ...................................................................................... 62 Escape from task demands ....................................................................................... 63 Teacher attention ......................................................................................................64 Peer attention ............................................................................................................64 Control ...................................................................................................................... 65 Procedures for Monitoring Consultation .........................................................................66 3 RESULTS ....................................................................................................................... ........69 Teacher Integrity ............................................................................................................. ........69 Functional Analysis of Appropriate Classroom Behavior ...................................................... 72 Multielement FA ............................................................................................................. 73 Student A ..................................................................................................................73 Student B ..................................................................................................................73 Student C ..................................................................................................................74 Disruption .................................................................................................................... ....74 Peer Attention Reversal Analyses ................................................................................... 75 Student A ..................................................................................................................76 Student B ..................................................................................................................77 Multielement Data During Sessions of 100% Procedural Integrity ................................ 78 Social Validity ........................................................................................................................79 Teacher A ........................................................................................................................79 Teacher B .........................................................................................................................80 Teacher C .........................................................................................................................81 Student Reactivity ...........................................................................................................82 Consultation Outcomes ...........................................................................................................83 Teacher A ........................................................................................................................84 Teacher B .........................................................................................................................87 Teacher C .........................................................................................................................89 4 DISCUSSION .................................................................................................................... ...106 Overview of Findings ...........................................................................................................106 Teachers Procedural Integrity ......................................................................................106 Students Behavior during FA ....................................................................................... 107 Consultation and Social Validity ................................................................................... 107 Interpretation of Findings .....................................................................................................108 Teacher/Student A .........................................................................................................108 Teacher/Student B .........................................................................................................110 Teacher/Student C .........................................................................................................112 Theoretical Implications of Findings .................................................................................... 113 Practical Implications of Findings ........................................................................................ 115 Limitations ................................................................................................................... .........118 Teachers Procedural Integrity ......................................................................................118 Students Behavior During FA ...................................................................................... 119 Implications for Future Research .......................................................................................... 121 Summary ....................................................................................................................... ........123
7 APPENDIX A TEACHER RECRUITMENT FLYER ................................................................................. 125 B INFORMED CONSENT TEACHER PARTICIPANTS ..................................................126 C TEACHER PARTICIPANT DEMOGRAPHIC FORM ...................................................... 128 D INITIAL STUDENT PARTICIPAN T IDENTIFICATION FORM .................................... 130 E INITIAL OBSERVATION FORM ...................................................................................... 131 F INFORMED CONSENT PARENTAL CONSENT FOR STUDENT PARTICIPANTS .................................................................................................................. 132 G PEER CONFEDERATE ID ENTIFICATION FORM ......................................................... 134 H INFORMED CONSENT PARENTAL C ONSENT FOR PEER CONFEDERATES ......135 I PROBLEM IDENTIFICATION INTERVIEW TEACHER .............................................137 J PROBLEM IDENTIFICATION INTERVIEW STUDENT ..............................................139 K FUNCTIONAL ANALYSIS DESCRIPTIONS FOR TEACHER TRAINING .................. 141 L OPERATIONAL DEFINITION FORM .............................................................................. 143 M FUNCTIONAL ANALYSIS DATA COLLECTION FORM ............................................. 145 N SOCIAL VALIDITY FORM ............................................................................................... 146 LIST OF REFERENCES .............................................................................................................148 BIOGRAPHICAL SKETCH .......................................................................................................156
8 LIST OF TABLES Table page 1-1 Functional analysis in general education classroom settings ............................................. 49 2-1 Criteria for evaluating teacher in teg rity during classroom-based FA ................................ 68 3-1 Student A FA results; Teacher A procedural integrity ....................................................101 3-2 Student B FA results; Teacher B procedural integrity ..................................................... 103 3-3 Student C FA results; Teacher C procedural integrity ..................................................... 104 3-4 Teachers responses to social validit y interview (1= not at all, 7 = very) ....................... 105
9 LIST OF FIGURES Figure page 3-1 Teachers procedural integrity during FA sessions ........................................................... 92 3-2 Student A: Responses per minut e raising hand across conditions ..................................... 93 3-3 Student A: Responses per minute appr opriate peer speech across conditions. ..................93 3-4 Student A: Responses per minut e disruption across conditions. ....................................... 94 3-5 Student B: Responses per minute raisi ng hand/appropriate te acher speech across conditions. ................................................................................................................... .......94 3-6 Student B: Responses per minute appr opriate peer speech across conditions. ..................95 3-7 Student B: Responses per minut e disruption across conditions. ........................................ 95 3-8 Student C: Responses per minut e raising hand across conditions. .................................... 96 3-9 Student C: Responses per minute appr opriate peer speech across conditions. ..................96 3-10 Student C: Responses per minut e disruption across conditions. ........................................ 97 3-12 Student A: Frequency of di sruption; within sessions an alysis of Control and Peer Attention ..................................................................................................................... .......98 3-13 Student B: Frequency of a ppropriate peer speech; within sessions analysis of Control and Peer Attention..............................................................................................................98 3-14 Student B: Frequency of di sruption; within sessions an alysis of Control and Peer Attention ..................................................................................................................... .......99 3-15 Student A: Raising hand during sessions of 100% procedural integrity. ..........................99 3-16 Student A: Appropriate p eer speech during sessions of 100% procedural integrity. ......100 3-17 Student A: Disruption during sessions of 100% procedural integrity. ............................ 100
10 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TEACHER IMPLEMENTED TREATMENT PROBES: A CONSULTATIVE PROCEDURE FOR FU NCTIONAL ANALYSIS OF CLASSROOM BEHAVIOR By Elizabeth Lane Weeks McKenney May 2010 Chair: Nancy Waldron Major: School Psychology This investigation extended the litera ture on classroom based assessments and interventions with disruptive behavior by training general education mi ddle school teachers to conduct functional analyses, measuring the integrity with which they implemented functional analysis procedures in their classrooms, and evaluating the e ffects of functional analysis on students rates of appropriate classroom behavior. Behavioral c onsultation procedures, including performance feedback, were used to support and maintain teachers levels of integrity throughout the assessment process. Results indicated that teachers were able to implement functional analysis conditions with high integrity during the training phase of assessment. However, during classroom based functional analyses, two teacher s demonstrated high procedural integrity and one teacher demonstrated variables rates of proc edural integrity. Functional analysis yielded consistent results regarding the cl assroom behavior of one student. Factors contributing to valid functional analysis results appeared to include high procedural integrity by the implementing teacher and limited student reactivity to the asse ssment process. Social acceptability findings indicate that all teachers found the procedure to be useful for the assessment of behavior demonstrated by disruptive student s, and that they valued the information obtained. However,
11 teachers raised concerns about the feasibility of using a highly time intensive process to reduce rates of disruption. Consultation was an effectiv e way to support teachers in the implementation of a classroom-based functional analysis; highly collaborative consultati on appeared to produce the best outcomes via both teacher integrity and st udent response to the functional analysis.
12 CHAPTER 1 INTRODUCTION AND LI TERATURE REVIEW Introduction Functional analyses have been shown to be an effective tool in evaluating environm ental contributors to aberrant behavior (Carr et al., 1999; Iwata, Dorsey, Slifer, & Bauman, 1982/1994). Functional analysis (F A) procedures provides a dire ct link between assessment and intervention that is both evidence-based and in concordance with the recent emphasis for school psychologists and other educators to conduct mean ingful and useful assessments of student behavior and skills (Witt, VanDerHeyden, & Gilb ertson, 2004). However, efforts to introduce FA technology into the classroom environment ha ve been limited in their scope and efficacy (Broussard & Northup, 1995; Wright-Gall o, Higbee, Reagon, & Davey, 2006). Classroom-based assessments of problem behavior, when conducted, have typically used various forms of functional be havioral assessment methodology, or FBA (Ervin et al., 2001; Murdock, ONeill, & Cunningham, 2005; Shrive r, Anderson, & Proctor, 2001). FBA is a descriptive assessment of the relations between st udent behavior and classroom events, typically involving the use of interviews, questionnaires, and/or observati ons to develop hypotheses about the potential function of an identified behavior (Murdock et al., 2005; Shrive r et al., 2001). This differs from FA, which is an experimental mani pulation of environmental events in order to measure the effects on individual behavior (Erv in et al., 2001). Both FBA and FA can be considered under the more gene ral term, functional assessment In a recent review of 11 empirical investigations of be havioral function in school settings, Kates-McElrath, Agnew, Axelrod, and Bloh (2007) concluded that the gen eral consensus seems to be that the term functional assessment is an overarching term that does not necessarily involve the experimental analysis phase owned by the term functional analysis, (p. 53). Further, Stichter and Conroy
13 (2005) describe functional assessment as, a set of assessment procedures used to identify variables that promote and maintain challengi ng behavior, (p. 19). Thus, while functional assessment refers to the process of collecting information about a behavior and its possible environmental contributors, th e terms functional behavioral assessment (FBA) and functional analysis (FA) provide greater spec ificity in describing the specific processes involved in such an assessment. For the purposes of this discussion, FBA will refer to the indirect and descriptive techniques (e.g., interviews, questionnaires, observ ation) that have been used to form hypotheses about behavior, while FA will refer more strictly to the direct experimental manipulation of environmental contingencies for a specific behavi or. However, it is worth noting that some authors recommend the inclusion of FA into a school-based FBA process (e.g., Asmus, Vollmer, & Borrero, 2002; Kamps, Wendland, & Culpepper, 2006) while others restrict use of the term FBA to refer only to descriptive methods (e.g., Mu rdock et al., 2005; Shriver et al., 2001). At present, it seems that there is only clarity as to the definition of func tional assessment being distinct from that of functional anal ysis (Kates-McElrat h et al., 2007). At the advent of a legislative mandate to conduct function-based assessments of problem behavior (Individuals with Disa bilities Education Act, 1997), a plethora of research began on how to best accomplish this goal in schools, mo st of which has focused on descriptive FBA methods. However, little research has focuse d on the use of FA methodology in the classroom setting (Moore et al., 2002; Witt, Noell, LaFleur, & Mortenson, 1997). There is also a dearth of research that has extended the use of FA methodology to typica lly developing students, students with mild to moderate disruptive behavior, or adolescents (Boyajian, DuPaul, Handler, Eckert, & McGoey, 2001; Broussard & Northup, 1997; Er vin, DuPaul, Kern, & Friman, 1998; Flood, Wilder, Flood, & Masuda, 2002; Jones, Drew, & Weber, 2000; Moore et al., 2002; Northup et
14 al., 1997; Vollmer & Northup, 1996). Of the resear ch that has been conducted with these populations, very little has been undertaken in the classroom context (Boyajian et al., 2001; Broussard & Northup, 1995; Ervin et al., 1998; Moore et al., 2002; Northup et al., 1997), and to date no such procedures appear to have been conducted with typically developing adolescents exhibiting problem behavior. Additi onally, further research is needed to investigate the ability of teachers to manipulate and examine classroom environmental factors influencing students behavior; most research in this area has been developed and conducted by researchers, with relatively little participation in design and/or analysis by classroom teachers (Ervin et al., 2001). Functional behavioral assessment and functional analysis procedures have emerged out of the Applied Behavior Analysis (ABA) literature and have become increasingly present in school psychology and other intervention literature ba ses. ABA approaches to consultation and intervention within school setti ngs are designed to increase teach ers awareness of the ways in which their behavior and the cla ssroom environment affect student behavior (Erchul & Martens, 2002). As school psychologists increase their ex pertise and role in consultation within the school context, the ability to demonstrate and prom ote effective intervention strategies may be of increasing importance. Thus, more information is needed about the ways in which functional procedures for describing classroom behavior can be used in a vari ety of settings, with a variety of students, and by a variety of pr ofessionals. The purpose of this study is to explore the role of functional analysis techniques in the classr oom setting with the contexts of teacher implementation and consultation in schools. The in tent of this investigation is to examine a teacher implemented functional analysis procedure that can be used in general education middle school classrooms.
15 Literature Review Targeted Populations Adolescence is a sens itive pe riod of development that is defined by the transition it encompasses between childhood and adulthood and is characterized by a number of changes in an individuals biology, c ognitive characteristics and abilities, and emotional and behavioral expression (Santrock, 2001). During this period, students are uniquely affected by the contexts in which they are placed, such that the environments of school, the family, and peer groups take on special importance. Normativ e adolescent development is thought to benefit from a complex interaction of all of these settings in order to produce successful outcomes. Similarly, problems in adolescence are perceived to be due to an interaction of indi vidual and contextual characteristics that adversely affect an adoles cent in some noticeable manner (Santrock, 2001). Thus, the period of adolescence merits special attention to the manner in which relevant developmental contexts (e.g., schools) affect ad olescent behavior and ove rall adjustment. While students who meet criteria for a disrup tive behavior disorder are not exclusively responsible for demonstrating problem behavior, they tend to demonstrate the most pervasive and disruptive behaviors within the cla ssroom context (Boucher, 1999). Students who demonstrate disruptive behavior in the clas sroom (e.g., shouting out, noncompliance, using materials inappropriately) are frequently identif ied by teachers and referred to outside personnel for additional assistance or services (Rusby, Taylor, & Foster, 2007), and students with highly disruptive behavior are most likely to be noticed and perceived as problematic by teachers (Reddy, 2001). The prevalence of disruptive beha vior disorders may range between one and 16 percent, depending on the disorder under study and the method of classification (DSM-IV TR, 2000). Disruptive behavior disorders include such classifications as Attention Deficit
16 Hyperactivity Disorder (ADHD) Oppositional Defiant Disorder (ODD), and Conduct Disorder (CD). According to special education classificati on criteria, some students with disruptive behavior disorders are likely to be placed in the general education classroom. While many students who are identified as meeting criteria for Severe Emotional Disturbance (SED) are placed in highly restrictive settings, such as hospital, residential, or homebound settings, approximately 16% receive services in general education classrooms (Reddy, 2001). However, determining the prevalence of disr uptive behavior in general education is impeded by a lack of congruence between the behavioral characteristics of disruptive behavior disorders and current definitions of SED, which do not always seem to adequately fit the behavior patterns of disruptive behavior disorders, especially if students are cons idered socially maladjusted (Boucher, 1999). For example, in a survey of 397 students classified with SED in San Diego, California, of which approximately 308 students were between the ages of 12 and 18, 65.6% met criteria for a disruptive behavior disorder (G arland et al., 2001). Thus, while there is considerable overlap in identification of ED a nd disruptive behavior disorders, there is not complete agreement. Furthermore, while prevalen ce rates of disruptive behavior disorders, such as ADHD, may be as high as 16% of the populat ion (DSM-IV TR), the prevalence of students receiving services for ED in public education se ttings is less than 2% of the total school population (U. S. Department of Education, 1998). These findings indicate that there are many students demonstrating disruptive behavior who are not placed in special education classrooms, but rather, are part of the general education pop ulation, in addition to students identified with SED that may be receiving services in a general education setting.
17 General education teachers may be ill-equippe d to handle disruptive behaviors, as few have received the training and resources necessa ry to meet the needs of students demonstrating problem behavior (Reddy, 2001). Furthermore, di sruptive behavior in the classroom setting is associated with poor social and academic outco mes, especially among students who demonstrate pervasive conduct problems in a ddition to symptoms of hyperactiv ity or inattention (Gresham, Lane, & Beebe-Frankenberger, 2005). Such negativ e outcomes are likely to be associated with decreased opportunities for academ ic instruction and appropriate social interaction due to engagement in inappropriate behavior and its se quelae (Gresham et al., 2005). For example, not only does disruptive behavior preclude the oppor tunity for academic engagement and/or appropriate social interaction, it is often likely to result in students being removed from the classroom for disciplinary purposes. While the strictly defined disorders discusse d above are not the exclusive focus of this investigation, the similarity of these profiles to the behaviors targeted by this investigation makes an appropriate comparison. For example, the defining characteristics of ADHD, ODD, and CD include behaviors that can be disruptive to a classroom environment (e.g., fidgeting, leaving ones seat, excessive talking, physic al or relational aggression, de struction of ma terials, arguing, defiance, annoying others) (DSM-IV TR). As pr eviously mentioned, children and adolescents who receive the diagnoses of ADHD, ODD, a nd CD often are placed in restrictive and institutional placements (Nickerson, 2003; Re ddy, 2001; Sholevar & Eichelman, 1998). Thus, adolescents who receive treatments sensitive to the occurrence of disruptive behaviors may avoid the potentially restrictive effects of stigmatizi ng classifications (e.g., poor academic and social outcomes). Additionally, the districts that serv e these students may avoid the costs of more
18 intensive supports necessary for serving student s with severe behavior disorders later in adolescence (Reddy, 2001). The lack of congruence between special e ducation and clinical diagnosis means of disruptive behavior, as well as the presence of students considered at-risk for developing behavior problems, has led some to argue that a population-based approach to describing and approaching school-based behavior problems is mo re sensitive to and effective in addressing students behavioral needs (Baker, Kamphaus, Ho rne, & Winsor, 2006). In such an approach, students who demonstrate mild to moderate disrup tive behavior can be served in the second tier of a three tier prevention and inte rvention model, in which the firs t tier represents well-executed behavior management for all students, the s econd tier represents se lected, or secondary, interventions for students at-risk for further difficulties, and tert iary interventions are put into place for students with severe and pervasive beha vioral excesses and deficits (Baker et al., 2006; Sugai & Horner, 2006). In such a system, student s needs are defined by behavior rather than eligibility for diagnosis or identification, and assessment of problem behavior becomes tantamount to providing function-based intervention services (Sugai & Horner, 2006). However, more information is needed about how contextual assessments are a part of and inform a tiered model of behavior management in schools. Functional Behavioral Assessment Functional evaluation methodolog ies in school settings cannot be investigated without considering the frequency with which functional behavioral assessment (FBA) techniques are used. While functional analysis methodology is c onsidered more rigorous for its ability to empirically demonstrate the environmental phe nomena maintaining a targeted behavior, descriptive FBAs are ofte n preferred for their ease of impl ementation and speed of execution.
19 Recently, investigators have beco me interested in the validity of the use of FBA, as well as the degree to which such procedures accurate ly identify the factors at work in determining classroom behavior. In an inves tigation of the use of functional assessment in the schools, Ervin et al. (2001) conducted a meta-analysis of 100 in tervention studies using functional assessment procedures, which may or may not have incl uded an experimental functional analysis component. A total of 278 students, with and wi thout disabilities, refe rred for a variety of behavioral and academic problems were included. The results highlighted several weaknesses in the extant literature on functional assessment pro cedures. Perhaps most surprisingly, at the time of publication, no students without disabilities ab ove the age of 12 were included in any article examined. Furthermore, less than 11% of the tota l number of students did not have a disability. Reported disabilities included severe developm ental and cognitive delays, language disorders, learning disabilities, and behavior al disorders. Intriguingly, Ervi n et al. reported that over 90% of participants were examined via an experimental phase that involved direct manipulation of antecedent variables, consequences, or both. However, the authors did not indicate how the determination of experimental manipulation was made. Students without disabilities were more like ly to receive consequence manipulations, while those with disabilities more often received antecedent mani pulations. Finally, one of the most important findings of this study demonstrat ed that, when working with students without disabilities, school personnel (teachers or pa raprofessionals) were unlikely to conduct the manipulations (10%) and more likel y to work collaboratively with an investigator (30%), or be involved in a study where the investigator alone conducted sessions (53%). Thus, teachers of students without identified disabilities may be receptive to the opportunity to participate in a collaborative process of interven tion with an agent acting in a c onsultative role. Similarly, this
20 greater propensity toward co llaboration provides potential co nsultants with an increased opportunity to provide indirect se rvices to students demonstrati ng problem behavior, potentially facilitating an eventual decrease in the degree to which teachers re ly on an investigator alone to provide intervention services. However, this fi nding also indicates th at teachers of students without disabilities (e.g., general education teachers) might need further training in order to be able to independently conduct empirical analyses of student behavior an d implement data-based interventions. The above findings regarding the degree to wh ich teachers participated as intervention agents were not consistent among teachers of studen ts with one or more disabilities, in which the investigator (52%) or the teacher (23%) was more likely to manipulate classroom conditions alone than to work collaboratively (12%). This effect may be due to the classroom settings in which students with disabilities are typically studied. In those set tings, teachers are likely to be trained in special education and familiar with experimental manipulation of environmental variables, perhaps making them more comfortable with inde pendently conducting empirical analyses. Furthermore, the smaller classes in which students with disabilities are often placed may make it easier for teachers to create a time and place to conduct such analyses, while teachers of students with disabilities may be more motivated to participate in functional assessment procedures when problem behaviors are severe. Ervin et al. (2001) also noted that few studies addressed the issue of developing functional procedures with teachers and instructing them in their use. This is perhaps especially true with regard to more complex functional analysis procedures. Finally, very few studies reported data as to the integrity with which teachers implemented assessment procedures.
21 Other investigations in which FBA procedures have been used in schools and classrooms lend support for their use. For example, Broussard and Northup (1995) found that the hypotheses generated by descriptive assessments we re validated via functi onal analysis for three elementary school boys. Similarly, Murdoc k, ONeill, and Cunningham (2005) found that teachers effectively identified hypotheses and were pleased with the outcomes of FBA for eight adolescent boys demonstrating disruptive behaviors. The particip ating students were all being served in special education under the classification of a behavior disorder, with the exception of one student who was identified as meeting criteria for a learni ng disability. Each student received some instruction in a resource room (b etween one and four cla ss periods per day) and the rest in a general education classroom. Hypothese s about students beha vior and its likely antecedents and reinforcing consequences were developed in a team meeting format with each students junior high school teachers. Similar hypotheses were also obtained during an interview with each participating student. Each student wa s then observed between one and three times in each of his/her classes over a three week peri od, resulting in a total of four to fourteen observations per student. Results were conceptu alized in terms of f our common functions of behavior, social attention, escape from task demands, access to a preferred tangible item or activity, or automatic reinforcement (self-stimul ation). The most comm only identified functions were escape from task demands (f ive students) and social attenti on (three students). However, it was not specified whether social attention incl uded teacher attention, pe er attention, or both. There was 64% agreement between teacher, student, and observation with regard to identified functions, a finding which the auth ors interpreted as indicative of consistency across each form of assessment. If this finding is indeed an in dication of the strength of FBA as an assessment tool, those results may be due in part to the co nsiderable amount of observation time spent in the
22 classroom. The authors state that total time in observation ranged from fi ve to fifteen hours per student. However, at the conclu sion of the study, th e teachers reported that the procedures did not have a significant impact on the way in which they managed student behavior and did not clearly state that the process was useful in identifying antecedents and consequences to student misbehavior. Given the amount of time spent in meeting with teachers, and interviewing and observing students, a more dramatic effect on teachers ability to respond to problem behavior might be expected. While this FBA proce dure demonstrates evid ence of having reliably identified the function(s) of each students pr oblem behavior, the high ratio of time spent in assessment before beginning intervention planning or development calls into question the feasibility of such a procedure in schools set tings without extensive ex ternal (e.g., researcher) support. Recently, efforts to improve FBA procedures have included investigations of the potential of calculating conditional probabilities of a particular form of reinforcement given a particular behavior (Eckert, Martens, & DiGennaro, 2005). For example, rather than summing the total reinforcers allocated for each beha vior observed and cal culating percentages, conditional probabilities provide information about the contingent relation between the occurrence or non-occurrence of an identified behavior and a cer tain consequence. According to the procedures employed by Eckert et al. (2005), observation periods are di vided into intervals; each time the behavior (B) and a certain cons equence (C) occurs with in the interval the probability is calculated as ( p[C/B]). During intervals in which the identified behavior does not occur, but the consequence of interest is delivered, the probability is calculated as ( p[C/no B] (Eckert et al., 2005). Conditional probabilities can also be used to determine powerful
23 antecedents by calculating the proba bility of the identified beha vior given the occurrence of a particular antecedent (Eckert et al., 2005). Other research has compared the conditional probability of reinforcement to the background probability of reinforcement, in which the probability of reinforcement is measured at random points during the observation sessi on. In such a comp arison, the background probability describes the likeli hood of reinforcement independent of the occurrence (or nonoccurrence) of the behavior, providing a general description of the overall likelihood of reinforcement (Vollmer, Borrero, Wright, Van Camp, & Lalli, 2001). Subsequent comparison of the conditional probability and the background probability yields a description of the contingency between the target be havior and identified form(s) of reinforcement, which can be either neutral (conditional probability equal to background probability), positive (conditional probability greater than background probability), or negative (conditional probability less than background probability; Vollmer et al., 2001). In this way, conditional probabilities more precisely describe the availability of certain re inforcers in the classroom environment. This effort is considered an improvement over other forms of FBA (which rely on percentages rather than probabilities), in that the data obtained from the assessmen t describes the likelihood of a particular consequence or anteced ent being predictive of a target behavior independent of the occurrence of the identified target behavior. FBA has also been recommended for use within a Positive Behavior Support (PBS) framework, in which FBA results are typically used to develop Behavi or Intervention Plans (BIP) to address students problem behavior. In such a system, FBA is considered a cornerstone of providing intervention to stude nts at a Tier III, or tertiar y, level of intervention. Blood and Neel (2007) investigated the exte nt to which FBAs were used to inform planning and instruction
24 for students identified with emo tional and/or behavioral disorder s being served in self-contained classrooms. They found that students were sli ghtly more likely to have a BIP if an FBA had been conducted. However, not all of the FB As that had been conducted referenced a hypothesized function of the students behavi or, obviating the opportunity to implement function-based interventions. Furthermore, teacher s had little to no knowledge of the contents of the BIP, if one existed, nor any specific behavior al goals written in conj unction with students FBAs or IEPs. These findings indicate a lack of congruence between FBAs and instructional planning and the necessity for future research re garding the application of functional assessments to school-based intervention. As indicated above, the utility of functiona l assessment in schools is undermined by a lack of shared understanding of the procedures it encompasses. Shriver, Anderson, and Proctor (2001) identified several criteria that should be used to examin e the procedural strength of a given functional assessment proce dure. These procedures include the content of the assessment, the process through which it is c onducted, the demonstr ation of empirically supported functional relations, the reliability and ge neralizability of the measure, and evidence of its external, treatment, and social validity. At the current time, such stringen t evaluation criteria are not in place to examine the quality of a chosen proced ure. While Shriver et al.s (2001) research indicates that a consensus is being reached among researcher s with regard to the ideal components of an FBA, it is yet unclear whet her practitioners are employing recommended best practices when implementing functional assessment in the schools. Relatedly, Weber, Killu, Derby, & Barretto (2005) examined the level of support that State Education Agencies (SEAs) give school districts and local e ducation agencies regarding proce dural best practices for carrying out an FBA. They found that the majority of states supply inform ation regarding and/or
25 stipulating the use of a defined target behavi or, direct observation an d/or an A-B-C format, assessment in the ecological context, interviews, record re views, hypothesis development, student interviews, and team meetings. Further, a smaller number of states supplied information regarding scatter plot data, using an FAO form and/or checklists, conducting analog experiments (FAs), or including a process for identifying e ffective reinforcers (p. 741). However, no information is available regarding how or to what extent local agencies used these guidelines to inform FBA practice (Web er et al., 2005). The lack of clarity surrounding functional a ssessment procedures represents a major concern in the successful iden tification of functional rela tions between the classroom environment and student behavior Murdock et al. ( 2005) indicated that th e teacher teams of middle school students did not always agree on the behaviors to target during FBA. For example, in the assessment of a student with ex ternalizing behavior, on e middle school teacher may select disruption as the target beha vior of concern, while another may choose noncompliance as the primary concern. This indicates that FBA procedures may be describing idiosyncratic phenomenon rather than consistent functional relations. Further, Thompson and Iwata (2007) point out that because of the ubiquit ous nature of attention as a form of responding to problematic behavior, attention may be over identified as a potentially maintaining consequence during descriptive assessments of problem behavior. Such an effect may be especially problematic when using conditional probabilities to describe possible consequent relations (Thompson & Iwata, 2007). At the very least, more research is needed in the assessment of functional relations, especially at the middle school level, so that those maintaining variables most relevant to students can be identified and addressed within the classroom context.
26 Functional Analysis in the Schools Functiona l analysis (FA) in the school setting traditionally occurs in one of two possible setting formats, analog or natural. Analog se ttings typically refer to a location outside the classroom where the student intera cts with researchers, teachers, or paraprofessionals under a variety of predetermined conditions to identify those factors maintaining the target behavior. Analog settings are useful because they provide excellent control over all environmental factors affecting the integrity of the evaluation proce ss. However, SterlingTurner, Robinson, and Wilczynski (2001) have identified several important limitations when using analog FA procedures. The ability of the analysis to va lidly describe behavior in outside settings is compromised by the artificial natu re of the experimental environm ent, which is often foreign to the student. It may be difficult to evoke the target behavior in c onditions different than those that exist in the classroom, where the target behavior s typically occur. Students behavior may be functionally related to peer or t eacher behavior that is unique to the classroom environment, or may be related to the history of available reinfo rcement that is signaled by a particular stimulus in the classroom setting. If the classroom envi ronment itself is a discriminative stimulus (SD) for the availability of a particular type of re inforcement (e.g., teacher attention), an evaluation conducted in an analog setting may not accurately describe the behavior of interest and the environmental events that maintain that behavi or. Additionally, Sterli ng-Turner et al. (2001) note that most empirical evalua tions of analog procedures have been conducted with students who are developmentally delayed, calling into question the utility of the findings of such evaluations for typically de veloping students or students with mild disabilities. Sterling-Turner et al.s (2001) overview summarizes the various types of functional analysis procedures that have been used in the classroom setting, including a traditional multielement design (Iwata et al., 1982/1994). Othe r procedures that have demonstrated strong
27 empirical support and treatment efficacy include contingency reversals an d brief FA procedures (Northup et al., 1991). Furthermore, all of th ese approaches can be implemented using a hypothesis-driven approach, wherein only those factors suspected of being functionally related to the problematic behavior are evaluated (Sterlin g-Turner, Robinson, & Wilczynski, 2001). This approach may be especially powerful when multiple factors are hypothesized to maintain the behavior. For example, a hypothesis-driven approach has been used to evaluate the combined effects of escape and attention in maintaining ta ntrums at school (Mueller, Sterling-Turner, & Moore, 2005). Such an approach helps to abbreviate the assessment process, making FA procedures more suitable for the classroom environment. When FA is conducted in a classroom setting, concerns exist about the degree to which the internal validity that exists during an analog assessment can be maintained in a less controlled environment. The greatest limitation of any classroom-based FA procedure is the degree to which experimental control over all s ources of reinforcement may be impaired by the presence of other students and stimuli in the clas sroom environment. In an analog setting, it is possible to control sources of reinforcement by limiting involvement to highly trained personnel and carefully planning social and tangible stimu li to be included. During a classroom-based FA, though protocols may be developed for meas uring and limiting unintended delivery of reinforcement, the presence of other students and unexpected classroom events makes such rigorous experimental control improbable. There are other limitations inherent to many of the classroom-based FA procedures that have been developed to date. For instance, brief FA procedures do not examine all possible functions of the identified behavi or. In a hypothesis-driven brief FA, if disruptive behavior is hypothesized to be maintained by teacher atten tion, contingent escape from academic demands
28 and peer attention are not examined. Results of such an FA may be limited by a lack of information about other reinforcers maintaini ng the behavior. Furthe rmore, many classroom procedures may more closely resemble structural analyses, in which the independent variables being manipulated are actually ante cedents to behavior, and not cons equences. Such an effect is often seen in evaluating escape or demand conditi ons, in which the number of demands given or the difficulty of academic work is increased in order to show a related increase in problem behavior. In a reversal design, empirical evidence for an escape function is provided when the demands are lessened in later sess ions and the behavior decreases. However, without contingent reinforcement of the problem behavior when it occurs (e.g., providing escape from the demands when the identified behavior occurs) such a design is not truly a functional analysis and does not identify the environmental variable maintaining the behavior. Contingent reinforcement of the target behavior, when it results in an increase of that behavior provides a demonstration of the functional relation between a par ticular form of reinforcement (e.g. escape) and the identified behavior. The methodological variations described a bove have developed in an attempt to circumvent many of the challenges associated with working in natural settings. However, what appears to be needed is an FA procedure that is flexible enough to meet the demands of naturalistic settings without sacrificing the empi rical rigor of analog proc edures. By carefully considering the environment in which an analysis is to take place before beginning an assessment procedure, FA of problem behavi or can successfully identify relevant reinforcers in classroom settings. Such a consideration should include the target behavior(s), the availability of reinforcement for that behavior, and the feasibility of manipulating classroom contextual factors. In order to meet the challenges of natural settin gs, successful FAs will be sensitive to the unique
29 needs of each classroom without sa crificing procedural integrity. For example, classroom-based functional analyses have traditionally focused on problem behavior, which may explain the limited extent to which FA has been applied in t hose settings, as teachers may be resistant to temporarily increasing rates of pr oblem behavior for the purposes of assessment. Investigations of the use of FA to determine the function of appropriate classroom behavior have not been conducted to date. However, manipulating the c onditions under which appropriate behavior is likely to occur may increase teachers willingness to participate in classroom-based FA procedures and make such manipulations more amenable to a general education classroom environment. Conducting functional analyses in natural se ttings may provide teach ers and other care providers with more relevant information con cerning students behavior As many problematic classroom behaviors may be functionally related to classroom phenomena (e.g., escape from task demands or obtaining teacher attent ion), it follows logically that the classroom itself is the best place to evaluate those conditions. Furthermor e, although fewer investigations have been conducted of the use of functional analysis with typically developing students, much of it has focused on developing useful classroom-based procedures in the general education classroom. These procedures are su mmarized in Table 1-1. In each of the studies presented in Table 1-1, a function for the identified problem behaviors was successfully identified. In seve ral cases, the identified function was used to implement interventions with the target student that resulted in decreas ed problem behavior. There are some interesting points to note in the procedures desc ribed above. Primarily, a full multielement design was only used when the researcher or personnel within the university research setting were responsible for conducting the functional analysis itself. In the three
30 studies in which classroom teachers participated, a hypothesis-driven approach was used. In the Ervin et al. (1998) study, the teacher compared an tecedent conditions that were thought to evoke the problem behavior to those that were not. These comparisons were based either on task difficulty (indicating a presumed escape function) or the availability of pe er attention (indicating a presumed peer attention func tion). These decisions were ma de on the basis of descriptive assessments, and the strong differences noted betw een all evaluated conditi ons appear to verify the selected hypotheses. Similarly, Newcom er and Lewis (2004) conducted antecedent manipulations of classroom factors hypothesized to predict problem behavior for three general education students. After c onducting extensive desc riptive assessments, they carried out alternating treatments comparisons of potential setting events for problem behavior. The factors selected for comparison were ability to choose de sired peers versus being assigned disliked peers (potential peer attenti on function), hard versus easy task (potential escape function), and high levels of noncontingent teacher attention compar ed to no teacher attent ion (potential teacher attention function), A different participating student received each of these manipulations. While both antecedent investigations produced moderate to strong results, without direct comparison across possible functions, remaining reinforcing contingencies for the targeted behaviors may be overlooked. In the Moore et al. (2002) study, an altern ating treatments design was used to compare the relative effects of escape from a difficult activity and teacher at tention for disruption (yelling). While the observed differences between those conditions across students indicates that yelling was more strongly reinforced by escape than by attention, without comparison to other possible conditions, findings derive d from these conditions remain re lative. Furthermore, as this study focused on teachers integrity rather than on student changes, data on student behavior is
31 presented as mean frequencies rather than graphically, making definitive interpretation of these findings difficult. Finally, in a study conducted by Kamps et al. (2006), hypotheses developed during descriptive assessments (observation a nd teacher interviews) were used to design conditions comparing potential sour ces of reinforcement for inappr opriate behavior to potential intervention conditions (e.g., delivering praise and points on a token economy system for appropriate behavior). Based on the resulting data, interventions were developed that resulted in a decrease in problem behavior. While these an alyses again appeared to lead to effective interventions, without a comp lete multielement design it is difficult to form definitive conclusions about the functions of the behaviors under study. Additionally, mixed measurement methods, overlapping sessions (in which two cond itions appear to have been conducted during the same session), and limited data in regard to some conditions further limit the results of this study. Methods of measurement were not consistent across the two particip ants (one is a brief FA; the other is an extended design), providing li ttle information as to the generalizability of these procedures for assessing the function of student behavior. Also of note in the above studies is the age of participating students. Only two studies have evaluated the efficacy of functional analysis in assessing problem be havior demonstrated by adolescents, one of which was not conducted in a general education setting. Thus, while there is a substantiated research tradition that points to the efficacy of usi ng functional analysis in natural settings when evaluating typically developing students and those with mild disabilities, more information is needed. Specifically, procedures that rely on both the experimental rigor of a multielement design and the naturalistic power of teacher implementation should be examined. Furthermore, greater emphasis is needed on the a pplication of FA with adolescent populations to better understand the strengths and w eaknesses of FA as an assessment tool for school personnel.
32 Research has demonstrated th e potential of functional analysis as a method of assessing problem behavior in the general education classroom. However, more research is needed to identify the most efficient ways in which to use FA in that setting. Furthermore, while brief FA and hypothesis-driven approaches are promising, more stringent ev aluation procedures that are equally timeand resource-efficient offer the most accurate possible identification of the environmental factors maintaining problematic cla ssroom behavior. In pursuit of that goal, any discussion of the use of functional analysis in the classroom setting must also consider research on teacher behavior and the nature of consultation. Student Involvement in Functional Assessments Som e research has investigated the potential benefits of more directly involving students in the functional assessment process. Most of this research has been aimed at increasing the validity of the assessment process by confirmi ng the environmental variables identified by caregivers as those most relevant to the stude nt being assessed. For example, Reed, Thomas, Sprague, and Horner (1997) investigated the us e of the Student Guided Functional Assessment Interview with 5th, 6th, and 8th grade students. Eight male and two female students participated, three of whom were identified with ADHD. Their responses were compared to teachers responses on the corresponding Functional Assessm ent Interview. Both students and teachers were questioned about predicto rs for problem behavior (e.g., difficult task, peer teasing), potential setting events, (e.g., lack of sleep), potential reinforcing consequences (e.g., teacher attention, escape from the task, pe er attention), and support strate gies and potential intervention techniques that might reduce the behavior. While students and teachers had high rates of agreement on predictors and reinforcing consequences for behavior, they demonstrated low agreement on setting events for problem behavior and potential interv ention strategies.
33 Research conducted by Wehmeyer, Baker, Blumberg, and Harrison (2004) also investigated the amount of agreement between teachers and students on a descriptive functional assessment interview. Teachers were intervie wed using the Functional Assessment Interview; while students were interviewe d using a tool designed by the au thors, titled the Person-Guided Functional Assessment (PGFA). Te n students in grades one six were interviewed, six of whom were identified with a behavioral disorder, tw o with mental retardation, two with an autism spectrum disorder, and one with a learning disabil ity (sic). The PGFA was developed to include person-first language regarding st udents problem behavior, and an analysis of common routines that might be setting events/antecedents to prob lem behavior. Similar to the research conducted by Reed et al. (1997), there wa s strong agreement between stude nts and participating school professionals on all elements of the interview with the exception of setting events, for which agreement was very low. This effect is hypothes ized to be related to students greater knowledge of outside factors that may be influencing their behavior. Murdock et al. furthered this strain of research in a 2005 stud y investigating the degree of agreement between teacher and student interview and classroom observations. Eight junior high school students aged twelve fift een years participated, as well as their classroom teachers. The researchers reported that teacher responses, student responses, and classroom observation resulted in concordant findings approximately 64% of the time. When results were discordant, teacher report and observation we re most likely to agree. While only problem behavior definition and potential function were assessed, the results rega rding potential function were consistent with previous findings, and strong ag reement was observed. However, students and teachers demonstrated low agreement on the id entified problem behaviors themselves, as
34 students reported a number of behaviors that took place outside the classroom (e.g., smoking, skipping class) as most problematic; this inform ation was presumably unavailable to teachers. The investigations described above represent a first effort to include student information regarding problem behavior and its potentially maintaining c onsequences in the functional assessment process. However, none of the a bove mentioned studies included an intervention effort tied to the functional assessment process, providing little informa tion as to the ultimate utility of student involvement in descriptive asse ssment processes. Furthermore, experimental confirmations of the hypotheses during descriptive assessment (interviews and observations) would further support the utility of student involvem ent. Nonetheless, this research does suggest that students in upper elementary and middle scho ol grades can contribut e reliable information regarding their classroom behavior that may be useful during functional assessment. Teacher Implementation and Treatment Integrity Initial rese arch into the po ssibility of using FA in classroom settings has yielded encouraging results, including findings related to teachers ability to implement FA procedures with integrity (e.g., Iwata et al., 2000; Moor e et al., 2002; Wallace, Doney, Mintz-Resudek, & Tarbox, 2004). An investigation conducted by Iwata et al. (2000) demonstrated that undergraduate students could be trained to conduct FA with integrity. Prio r to training, students were given a written description of FA procedur es taken from Iwata et al. (1982/1994), asked to read them, given an opportunity to study them, and then asked to conduct an FA of problem behavior with a research confeder ate acting as the target individual. Integrity to the procedures was slightly higher than expect ed but variable, ranging from 50% to 89.5%. Students then received training, which included additional writt en materials, oral discussion by a graduate student trainer, video modeling, and a brief quiz. Second phase sessions then began, during which students had written outlines available to them and received performance feedback after
35 every third session. During this pha se, participants demonstrated an increase in their overall adherence to the FA procedures to 90% or higher. Several important limit ations to this study include the selected participants and the c ontext of the FA settings. While upper level undergraduate psychology students were selected for their presumed similarity to individuals working in applied settings with a bachelors degree (e.g., teachers ), it is unclear that these two groups are indeed as similar as they are presumed to be in this research. The participating undergraduates had all taken one co urse in applied behavior analys is prior to their participation in this study, indicating that th eir understanding of and enthusiasm for the principles underlying FA procedures may be quite different than th at possessed by teachers (a supposition that may be supported by the relatively high level of integr ity observed during baseli ne sessions). A more appropriate selection may have been undergraduate teachers-in-tra ining or individuals already working in educational/therapeutic settings. Re latedly, the training and FA sessions having been conducted in an analog, university setting may limit the generalizability of these findings to work with teachers and other intervention agents. Similar findings were obtained in a follo w-up study conducted by Wallace et al. (2004) which indicated that one teacher and two school psychologists who particip ated in the study were able to implement FA procedures with high inte grity. Training took place in a workshop format, during which participants were exposed to a desc ription of FA and its purposes, video modeling, and role playing opportunities. Baseline sessions were conducted prior to the workshop, during which the only information available to the partic ipants was a brief review of the Iwata et al. (1982/1994) procedures. The workshop lasted three hours and 35 other educators also participated. At the conclusion of the workshop, the participants engaged in simulated sessions designed to assess the function of head-hitting (with a researcher c onfederate acting as the target
36 individual). Participants we re provided feedback about their performance only if they demonstrated integrity below 90% during an assessment session. Each of the participants demonstrated increases to acceptable levels of integrity during the training/feedback phase, and one participant successfully conducted an FA of head hitting (w ith strong procedural integrity) 12 weeks following the initial training workshop. Moore et al. (2002) examined the extent to wh ich three elementary te achers were able to carry out functional analyses of disruptive behavior in their cl assrooms. Two of the student participants were typically de veloping, while one had been id entified with specific learning disabilities. Teachers were eval uated for their ability to respon d correctly to behaviors issued during the analysis conditions. Two conditions were examined, at tention and demand/escape. Teachers were provided with pe rformance feedback throughout the experimental phase. Performance feedback included presenting the da ta on percentage correct teacher responses (PCTR) to each teacher, praising them for correct implementation where appropriate, and reviewing elements of the protocol that had not been implem ented correctly. The results demonstrated that all three of the teachers we re able to implement FA procedures with high integrity during in-class sessions (Moore et al., 2002). This resear ch has promising implications for conducting further classroom-based functional analyses of problematic behavior. As noted by the authors, this research did not include a control condition. Th e presence of a control condition more soundly establishes that observed rates of behavior are functionally related to the reinforcers being manipulated by providing a comparison context in which reinforcement is not delivered contingent on the occurrence of the ta rget behavior(s). Common control conditions might include a free play or an alone/igno re condition, during which any occurrence of problem behavior is not socially reinforced (Iwata et al., 1982/1994). A free play condition is
37 typically set up as an enriched environment, in which desirable items and attention are readily available, and task demands are absent. An alone or ignore condition is one in which there are no desirable items or attention available, nor are there task demands. Problem behavior demonstrated during either of these conditions is considered functionally independent of the social context in which it occu rs, and thus may be an indication of automatically reinforced behavior (e.g., self-stimulati on, Iwata et al., 1982/1994). School -based research involving a control condition will be an important step in developing in-class FA procedures that are as functionally sound as analog functional analys es or those conducted by highly trained practitioners. When using FA techniques for a classroom evaluation, teachers are often not involved in the assessment process. In f act, most research surrounding the integrity with which teachers implement behavior analytic principles is relate d to intervention implemen tation, rather than the integrity of analysis (Moore et al., 2002; Noe ll et al., 2000; Witt et al., 1997). However, the integrity with which teachers implement a set of classroom-based procedures, be they assessment or intervention related, is essential to the success of the given procedur e. Wood, Umbreit, Liaupsin, and Gresham (2007) examined the impact of teacher integrity on the improved classroom behavior of a student who freque ntly disrupted the classroom. The teacher demonstrated variable performance throughout th e intervention, which was based on differential delivery of teacher attention and extinction of escape. However, when teacher behavior was analyzed across intervals, it was found that the student was much more likely to be on-task during intervals in which the teacher was deliveri ng the appropriate contingencies with integrity. In contrast, when the teacher did not deliver the contingencies accurately, the students demonstrated correspondingly lower rates of ontask behavior (Wood et al., 2007). Thus, the
38 integrity with which teachers carry out complex classroom contingencies may have a large effect on student behavior. It has been found that, under certain condi tions, teachers can implement intervention procedures with good integrity. Witt et al. (1997 ) examined the degree to which four general education elementary teachers were able to implement an academic intervention for an individual student. They found that providing the teachers with performance feedback enhanced the integrity of intervention implementation, and al so seemed to strengthen the effects of the intervention for the students. Noell et al. (2005) reported sim ilar findings in a comparison of three strategies for maintaining treatment integrity to selected interventions following behavioral consultation concerning 45 elemen tary school students. They compared three conditions of follow-up to intervention implementation with in a behavioral cons ultation model: weekly follow-up, commitment emphasis, and performance feedback. The results indicated that the provision of performance feedback strengthened treatment integrity and child behavior above the effects of brief weekly interviews or a comm itment emphasis. They defined performance feedback as meeting weekly with the teacher, reviewing student progre ss, graphing intervention implementation, praising correct implementation, reviewing the importance of any missed steps, and problem-solving when appropriate. While te achers across conditions di d not perceive that they implemented the intervention with differing levels of integrity, bo th visual analysis (graphic) and statistical analysis (ANOVA) re vealed a significant effect of providing performance feedback on teachers integrity to the intervention procedures. Furthermore, statistical analysis also indicat ed a difference in student outcomes; the students whose teachers received performance feedback during intervention implementation showed the greatest gains.
39 In another recent study, Codding, Feinberg, Du nn, and Pace (2005) found that bimonthly observation and performance feedback were ad equate to increase and maintain treatment integrity of teachers implementing behavior s upport plans. This finding, though yet to be replicated, may indicate that daily or weekly follo w-up is not always require d to maintain strong treatment integrity. Interestingly, research on tr eatment integrity has also recently incorporated the use of negative reinforcemen t for correct implementation of intervention steps (DiGennaro, Martens, & McIntyre, 2005). Teachers who ach ieved 100% integrity to the intervention on a given day of observation received performance feedback via a written note, and were able to avoid a meeting the next day. Teachers who did not achieve 100% integrity attended a traditional performance feedback meeting the next day, including discussion, review, and practice of any missed steps (D iGennaro et al., 2005). A strong body of evidence indicat es that performance feedback is necessary to maintain teachers integrity of intervention implementatio n. The consistent findings that performance feedback improves the integrity of intervention implementation argue for the need for similar procedures during assessment and analysis. The strong outcomes of teacher intervention implementation, when given performance feedbac k, indicate that teach ers may be able to implement classroom-based functional analys es with similar types of support. The Role of Consultation in Intervention Intervention efforts within school setti ngs often involve an extended collaboration between an interven tion agent (e.g., a teacher) and an individual external to the classroom who brings to bear his/her knowledge of intervention techniques on th e process. This pattern is repeatedly observable in the empirical literature that has hitherto been presented to discuss various elements of functional assessment and intervention in schools, as evidenced by such research as that conducted by No ell et al. (2005). Classroom-based activities, such as FBA/FA
40 and intervention implementation, often occur as onl y part of a larger c onsultative process. However, research into the use of FBA and FA pr ocedures has rarely investigated the role of such procedures within a consultative framewor k (Schill, Kratochwill, & Elliott, 1998). As professionals seek to provide services to a variety of individuals with different needs, consultative services have become increasi ngly necessary. Many individuals serve as consultants in school settings, among them are sc hool psychologists, special education teachers, behavioral consultants, and behavior analysts. However, consultants do not necessarily possess all of the knowledge with which to effectively intervene with a student; teacher involvement is necessary to ensure that the intervention selected is appropriate given th e classroom context, the students individual needs, and the demands of the particular school or curriculum in which the teacher and student are functioning. The colla borative nature and shared knowledge that necessitates consultation is, in fact, characterist ic of its very nature (Bergan & Kratochwill, 1990), making consultation an appropriate avenue through which to explore factors affecting the use of FBA and FA procedur es in the classroom. Behavioral consultation has emerged as a vali d and time efficient way in which to assess and treat students academic, soci al, and behavioral challenges (Wilczynski, Mandal, & Fusilier, 2000), though it is certainly not the only way in which consultation is practiced in school settings. For the purposes of the current discussion, however, the BC paradigm (Bergan & Kratochwill, 1990) is emphasized for its theoretical alignment w ith the principles of Applied Behavior Analysis. As FBA and FA procedures have emerged out the same empirical bases, the use of BC provides the most parsimonious manne r through which to explore the unique impact of consultation on assessment processes. These techniques appear to be especially successful
41 with individuals displaying discrete and/or ex ternalizing behaviors (S hapiro & DuPaul, 1996; Wilkinson, 1997). BC emerged from efforts to extend the proced ures of applied behavior analysis to consultative interactions in a systemized and clearly defined manner (Bergan, 1977; Kratochwill & Bergan, 1978). Defined as a problem-solvi ng process, BC was originally developed to address the needs of both individu als and organizations engaged in serving clients (Kratochwill & Bergan, 1978). Over time, BC has taken on a more individual focus, emphasizing the ability of successful consultation to increase the knowledge and skills of consultees in order to effectively meet the needs of ma ny potential clients. Neverthele ss, the original structure of BC remains; it is a four-step problem-solving process including problem identification, problem analysis, plan implementation, and plan evaluation (Kratochwill & Bergan 1978, 1990). Behavioral consultation is delineated by Berg an and Kratochwill (1990) as an indirect problem-solving process in which a consultant inte racts with a consultee in order to identify and treat any problem that arises in the consultees interactions with a clie nt. Erchul and Martens (2002) point out that a behavioral approach to consultation assumes that both appropriate and problematic behaviors are maintained by the envi ronmental principles that surround them. In this way, behavioral consultation is as much an assessment of the environment in which problems occur as it is of the individual client. In an effort to systematically identify and tr eat those aspects of the clients behavior that are problematic to the consulte e, behavioral consultation pro ceeds sequentially through several steps. Simultaneously, the consultant is concerne d with consultee behavior and behavior change throughout the consultative process (Erchul & Martens, 2002). While the consultant guides the process, and may use strategies to influence consultee behavior, the consu ltee is considered an
42 active participant in all stages of the process. The consul tant is assumed to be responsible for listening and attending to suggestions and concerns that the consultee brings to their interactions, as well as to nonverbal indications of the consultees comfort with each phase of consultation. When problems arise, careful analysis and further planning is warranted to ensure that integrity of analysis and treatment are mainta ined (Bergan & Kratochwill, 1990). Schill et al. (1998) investigat ed the utility of FBA proce dures within a BC format. Working with eleven Head Start teachers, they provided consultative servic es to address problem behavior among preschool students within a BC framework. Intervention services were either provided within a FBA format or what the rese archers titled a technological approach, in which the intervention was not based on an assessment of the potential functions of problem behavior. Rather, the technologica l intervention was delivered as a standardized set of self-help materials targeting either externalizing or inte rnalizing behavior problems according the needs of the target child. The FBA group participated in the problem identific ation interview (PII), problem analysis interview (PAI, following a functional assessment observation), and the treatment evaluation interview ( TEI; Bergan & Kratochwill, 1 990). The technological group did not participate in problem analysis activities or the PAI, but was given the treatment package with an explanation of how to use it in a br ief meeting conducted af ter the PII and before intervention implementation. The results revealed moderate positive intervention effects for both groups, indicating that the provisi on of descriptive analyses of problem behavior did not significantly improve treatment outcomes over the administration of a non-function-based intervention. The researchers attributed their findings to the power of the comparisons made across groups. However, the in ability of a descri ptive functional assessment to improve intervention outcomes above those of a nonfunc tional intervention, presumably due to the
43 inability of descriptive FBAs to correctly identify the operant function in all cases, has previously been discussed. Indee d, the authors also noted that the lack of observed differences in statistical outcome may have been because the implemented interventions were highly similar across groups. Most students received interventions based on differential reinforcement, irrespective of whether a hypothesi s regarding the function of their behavior was developed. However, these findings also indicate the suppor tive role of consultation in obtaining positive outcomes, albeit moderate ones, for students with minor problem behaviors. Most importantly, this study provides an excellent illustration of the ways in which BC can support a functional assessment process. Using an indirect method of service delivery, such as consultation, enables practitioners and others to provide services to more students than is possible in direct service delivery models. Many consultation procedures may also produce changes in the behavior of the consultee, resulting in an additional preventative effect (Bergan & Kratochwill, 1990). However, there are a number of divergent opinions about how consultation optimally is conducted, in addition to several important limitations in its use. While thes e complex issues cannot all be covered in this format, a review of the issues most salient to the current investigation is presented. BC and the strict application of ABA principles to classrooms have been called into question for several reasons. Many have cited the difficulty that teachers often have in identifying and describing the specific aspects of a students behavior th at are problematic (e.g., Murdock et al., 2005). Teachers ar e often resistant to intervention and may not implement the procedures decided upon during pr oblem analysis and plan impleme ntation phases, especially if there is not systemic support for innovative inte rventions within the sc hool (Piersel & Gutkin, 1983). It may also be unrealistic to expect that teachers will be able to sustain the amount of
44 time that is required by individual interactions wi th students, especially in crowded classrooms or for teachers who are already taxed by the demands of managing their students (Erchul & Martens, 2002). However, the n eed for consultation with a focus on skill-building in consultees continues, as research reveals that teachers ma y not have the necessary knowledge and skill to intervene effectively with students who present unique teaching challenges within the classroom (e.g., specific learning disabilities, mild mental handicaps, behavior disorders) (Wilkinson, 1997; Wilson, Gutkin, Hagen, & Oats, 1998). Furthermore, debate continues concerning the optimal level of consultant control within the format of behavioral consultation. While some argue that there should be little to no use of consultant control, others argue that outcomes are enhan ced by a decreased emphasis on collaboration in favor of providing greater c onsultant-directed expe rtise and guidance (e.g., Graham, 1998). Research findings intended to clarify this issue have thus far been mixed, with empirical studies supporting both stances (Bro wn et al., 2001; Graham 1998; Gutkin, 1999; Kratochwill & Van Someren, 1985). Witt, Gresham, and Noell raised questions about the efficacy of behavioral consultation in their 1996 work, Whats behavioral about Be havioral Consultation? In particular, they criticized behavioral consultations dependence on consultee report and indirect measures of assessment rather than on direct, experimental manipulation. Witt et al. wondered if behavioral consultation was truly able to identify and tr eat the primary behavior of concern without empirical evidence to support teachers statements With the primary outcome of interest in behavioral consultation being behavior change in both the client and the consultee, these authors argued for a more empirical approach to applyi ng behavioral techniques to the classroom. Subsequently, they present an alternative beha vioral consultation model that uses functional
45 analysis rather than descriptive FBA techniques to verify the function of the identified problem behavior in order to develop in tervention plans. They further advocate for the use of more stringent behavioral training tech niques to enhance teachers abilities to implement intervention appropriately, and the continuation of such t echniques throughout the monitoring and feedback phases of intervention implementation. The author s are particularly critical of behavioral consultations tendency to heavily rely on verbal exchanges in order to produce consultee and client change; they present their model under th e assumption that behavi oral consultation would be strengthened by following more closely to the lin es of behavior analysis rather than less so. It should be noted that while their concerns regarding the problem identification, problem analysis, and intervention deve lopment processes remain largely unaddressed, recent research into the use of performance feedback to incr ease teachers intervention integrity appears to answer their call for teacher tr aining and feedback procedures based on behavior analytic principles. Zins and Erchul (2002) give a definition of c onsultation that is inte nded to provide school psychologists with a synthesis of wh at are currently considered bes t practices in the field. They define consultation as a method of providi ng preventively oriented psychological and educational services in which consultants and consultees form cooperative partnerships and engage in a reciprocal, systematic problem -solving process enhanced by eco-behavioral principles. The goal is to enha nce and empower consultee systems, thereby promoting students well-being and performance, (p. 626). Though this definition operates within the paradigm of eco-behavioral consultation, which extends beha vioral consultation by including an ecological and systems focus, it provides a useful framework for discussing what have hitherto been considered the strengths and weakne sses of BC. A critical review of these issues is helpful in
46 examining how BC can be improved by more rigorous analyses. For example, Zins and Erchul begin their definition of consulta tion by classifying it as a preve ntively oriented activity, and end by stating that the go al of consultation is to empowe r consultees to promote students overall functioning. The implication, as has been discussed in the literatu re, is that BC does not plan or encourage generalization of skills by the consultee to othe r clients with similar presenting problems. The need, then, is for BC and relate d intervention strategies to offer increasingly analytic and skill-based services to teachers and other consultees, such that positive outcomes are generalized to more than one client without repeating the process of consultation which, though time-efficient for consultants, may be time-intensive for consultees. Time is a central issue to any form of consultative service delivery. After all, consultation is presented to professionals as a time-efficient way in which to disseminate their knowledge and services to as large a group as possi ble. In contrast to traditional assessment models, in which one student it assessed at a tim e, consultation allows school psychologists to work with several teachers in order to maxi mize outcomes for as many students as the teacher chooses. Originally, BC in the schools identif ied teachers as the primary agents with the responsibility of conducting assessments and implementing interventions (Bergan, 1977). Over time, however, some psychologists and others have found themselves spending more time involved in consultation activities than was originally suggested. Subsequent research supported this phenomenon and found that many teachers appreciated and rated more highly those consultants who were more ac tive and participatory throughout the consultative process (Wilczynski et al., 2000). Witt et al. (1996) cited consultants increased us e of direct observation and involvement with clients as a way of in creasing the content validity of behavioral consultation by verifying teachers re ports of problem behavior and its
47 antecedents/consequences. By augmenting consultees time in this way, consultants have been able to expedite and improve outcomes achieved for clients and consultees. Conclusion and Research Questions Adolescents who engage in disruptive behavi ors face unique challenges in the classroom and often pose difficulty for teachers and other stude nts, which can result in their being placed in restrictive educational settings. Several empirically supported me thods are available to assess and intervene with problem behavior in the general education classr oom, including functional behavior assessment (FBA) and func tional analysis (FA). Research findings regarding the utility of FBA procedures are mixed. The primary limitation of FBA is its reliance on indirect assessment and observation without experimental manipulation to verify the impact of certain antecedent and/or consequent factors on student behavior. Functional analysis meets these demands, and has recently been extended into th e classroom setting. Howe ver, applications of FA are typically limited, and teachers have very ra rely been trained to conduct FA within the classroom setting. Furthermore, very little res earch has examined the use of FA techniques in assessing the classroom behaviors of typically developi ng adolescents. Research into the utility of FBA and FA ha s often been conducted within the framework of behavioral consultation (BC) a procedure that is widely used to develop function-based interventions for students demonstrating problem behavior in the classroom. Research concerning the use of FA in classroom settings is strengthened by using a BC format because it mirrors the way in which such intervention se rvices are typically provided in schools. Furthermore, the use of BC provides an opportun ity and a structure through which to address issues of teacher integrity. Procedural integrity has been found to be a vital component of effective intervention implementation, and, thus, s hould be no less important to the successful implementation of a sophisticated analysis proce dure, such as FA, in the classroom setting.
48 Initial research into this area indicates that teachers are able to conduct FA with strong integrity when they are provided with adequate traini ng and performance feedback regarding their progress. The purpose of this study was to explore the role of FA in a general education middle school classroom setting. Additionally, teache r integrity to assessm ent procedures was measured. The intent of this investiga tion was to examine the following questions: 1) Given training and performance feedback, at what level of integrity do general education middle school teachers implement FA procedures? 2) When FA of appropriate behavi or is conducted with integrity in middle school classrooms, are there clearly identified experimental f unctions of students appropriate classroom behavior?
49Table 1. Functional analysis in ge neral education classroom settings Authors Student(s) Age Problem Beha viors Teacher Conditions evaluated Broussard & Northup, 1995 3 boys, 1 identified with ADHD 6 8 years Disruption, off-task, out of seat, aggression, destruction, crying, non-compliance Researcher Hypothesis-driven contingency reversal: escape Broussard & Northup, 1997 4 boys, 2 identified with ADHD 7 9 years Excessive activity, disruption, destruction Researcher Multielement: teacher attention, peer attention, escape Northup et al., 1997 1 boy, identified with ADHD 8 years Disruption out of seat, playing with objects University summer program for students with ADHD, teacher background not noted Multielement: teacher attention, peer attention, escape Ervin, DuPaul, Kern, & Friman, 1998 2 boys, identified with ADHD and ODD Setting: not specified 13 14 years Off-task, disruption Classroom teacher Brief reversal of baseline and intervention conditions: escape and peer attention Jones, Drew, & Weber, 2000 1 boy, identified with ADHD 8 years Disruption out of seat, playing with objects Clinic-based summer academic program, teacher background not noted Multielement: teacher attention, peer attention, escape Boyajian et al., 2001 3 boys, at-risk for ADHD 4 5 years Aggression Researcher/consultant Multielement: free play, attention, tangible, escape Moore et al., 2002 3 boys, 1 identified with SLD 4th 5th grades Disruption Clas sroom teachers Alternating treatments: teacher attention, escape Newcomer, & Lewis, 2004 1 boy identified with OHI, 1 boy, 1 girl 9 11 Aggression, off-task Classroom teache r Alternating treatments comparison of functionand non-function-based interventions (e.g., activity choice/no choice, hard task/easy task, attention/no attention) Kamps, Wendland, & Culpepper, 2006 1 boy and 1 girl identified as at-risk for emotional/behavioral disorder 7 years Non-compliance, disruption, off-task Classroom teacher Participant 1: Hypothesis-driven multielement (teacher attention, escape) Participant 2: Brief hypothesis-driven multielement (teacher attention, ignore)
50 CHAPTER 2 METHOD Participants and Recruitment Participan ts included middle school teachers a nd students at a university-based research school. All recruitment procedures took pl ace between November 2007 and January 2008. Three teacher/student dyads were recruited for participation. Specific procedures used for obtaining teacher and student partic ipants are described below. Teachers Consent to conduct research was obtained from the m iddle school assi stant principal at a developmental research school. After consenting to al low the research proj ect to go forth, the assistant principal disseminated a flier describing the project (Appendix A) to three teachers he thought appropriate for participation. Following hi s initial discussion with those teachers, the principal investigator approached each of them and introduced the resear ch project in greater detail. Teachers were told that the PI was evaluating ways to examine and treat disruptive behaviors in the classroom, such as shouting out, out of seat, and noncompliant behaviors, through the assessment of appropriate behavior. E ach of the three teachers gave his/her verbal consent to participate, whereupon they were gi ven written consent form s to sign (Appendix B). Subsequently, teachers nominated a student in their classroom for participation among those students who met the criteria described below, and completed a demographic form describing their years of experience, previous experien ce working with students demonstrating problem behavior, degree and level of e ducation, certifications, etc. (Appe ndix C). During initial contact with the assistant principal, preference was indi cated for teachers who had been teaching one or more years, as first year teachers may be unable to meet the additional demands of participation in this research. However, the assistant prin cipal indicated greater c onsideration for teachers
51 who had indicated a need for additional support se rvices regarding disruptive students in their classrooms. Further demographic informa tion on each teacher is provided below. Teacher A Teacher A was a Hispanic fem ale teaching 6th grade science. She obtained a Ph.D. in pharmacology and was teaching with a temporary cer tificate in middle school science. The year during which she participated in this study was he r first as a full-time teacher. She reported no previous experience working with students with disruptive behavior or participating in assessments of problematic classroom behavior. Teacher B Teacher B was a Caucas ian female teaching 8th grade math. She obtained a Ph.D. in educational anthropology and a M.Ed. in math education, and had certifi cations in secondary math and womens and gender studies. The year during which she participated in this study was her first as a middle school te acher; she had previously taught at the high school level for 24 years. She indicated extensive experience teachi ng students with disruptive behavior, including students identified with emotional/behavioral disord ers who were included in general education; she did not have previous experience conducti ng an assessment of problematic classroom behavior. Teacher C Teacher C was a Caucas ian male teaching 7th grade social studies. He obtained a M.Ed. in elementary and secondary education, and wa s certified in middle school instruction and secondary social studies. Prior to the school year during which he participated in this study, he taught for one year at the high school level. He reported no previous experience working with students with disruptive behavior; he reported th at other school personnel had come in to the
52 classroom to observe students in order to fo rmulate IEPs and 504 plans regarding classroom behavior. Students Participating teachers completed a brief form (Appendix D) identifying students who demonstrated disruptive behavior in their classrooms, accompanied by a verbal explanation of its purpose. Teachers were encouraged to identify al l students in their cla sses exhibiting disruptive behavior and to fill out a form on each student Students identified by their teachers were selected for participation accord ing to the following criteria: 1) A clear description of disrup tive behavior according to th e information provided by the teacher. Preference was given to those students demonstrating low intensity, high frequency, discrete instances of disruptiv e behavior, as verified by observation. 2) Observation by the PI to independently verify the occurrence and rate of disruptive behavior (Appendix E). One to two 15-mi nute observations was conducted of each student. The specified disruptive behaviors had to have occurred at least twice during each observation for that student to be selected for participation. 3) A preference was given for students for whom evidence was available that disruptive behavior was persistent acro ss contexts, including school en vironments (e.g., math class, language arts instruction) and, if possible, acad emic year. (Provided via teacher report and file review of all par ticipating students.) 4) Parental consent to participate (Appendix F). After parental consent for student particip ants was obtained, participating teachers identified peers who met the appropriate criter ia for participation as peer confederates (Appendices G, H). Teachers were prompted to identify five peers who might meet criteria for participation based on being socially mature and cooperative with teacher in structions. Teachers rated each of the five students according to the criteria listed in Appendix G. Those students who teachers rated as having the highest number of desired characteristics were recruited to participate; preference was given to students who were friendly with the target student. If the parents of an identified peer or the peer did not consent/assent to participate, another peer was
53 selected from the list. Finally, students who met the above criteria and the teachers referring them for participation participated independe ntly in problem-ident ification interviews (Appendices I, J), the purpose of which was to form a consensus about the problem behavior and develop an operational definition of all targeted behaviors (e.g., problem behavior and targeted appropriate replacement behavi ors). Students were included in the problem identification process in order to increase the reliability of problem identification and the level of student involvement in the consultation a nd intervention process. However, in order to reduce reactivity to the researcher's presence as an observer du ring FA sessions, all student interviews were conducted by school psychology graduate students w ho were not present in the classroom during FA sessions. Student A Student A was a Caucasian m ale in the 6th grade. He was not identified for special education services regarding academics nor behavi or. A review of his school records revealed that his academic performance was average in most areas. Upon initially referring this student for participation in this study, Teacher A reported that he frequently spoke to peers and shouted out during class, had difficulty staying in his seat, and was, hype ractive. She reported that he demonstrated this behavior in all of his classes. Student B Student B was a Caucasian m ale in the 8th grade. He was not identified for special education services regarding academics nor behavi or. A review of his school records revealed that his academic performance was average in most areas. Upon referring this student for participation, Teacher B reported that he frequently shouted out during class, was off-task, and would loudly bang his hands, feet, pencil, and chair against ot her nearby objects.
54 Student C Student C was an African Am erican female in the 7th grade. She was not identified for special education services regarding academics nor behavior. A review of her school records revealed that her academic performance was av erage to above average. Upon referring this student for participation, Teacher C noted that she frequently shouted out during class and was often out of her seat. Settings All training, assessm ent, and feedback pro cedures occurred in the general education middle school classrooms to which each particip ating teacher was assigned. Participating students sat in a seat assigned to them, chosen fo r its proximity to the t eacher and potential to decrease the availability of peer attention. Peers were sometimes asked to move seats to ensure the accurate delivery of attention. No other m odifications to the classroom environment were made. Experimental Design A m ulti-element design was employed for the FA conditions (Iwata et al., 1982/1994). However, to address the primary research questio n and evaluate teachers ability to implement FA, data on teacher integrity before training (baseline), immediately following training, and during classroom implementation of FA conditions, is reported using a nonconcurrent multiple baseline across participants design (Watson & Workman, 1981). A nonc oncurrent design was preferred for this study because it allowed trai ning to occur immediat ely before FA session implementation for each participating teacher. Due to the extended time necessary to consult with each teacher during this investigation, cl assroom FA implementation could not occur with all participating teachers simulta neously. In order to avoid an extended period of time between training and FA implementation, it was necessary to conduct initial mock FA sessions and
55 training temporally close to the implementation of classroom FA sessions. Therefore, baseline sessions conducted with this group (s) could not be concurrent. Non-concurrent multiple baseline designs are recommended for use when concurrent baselines are not feasible, and are appropriate given that certain requirements are met. Requirements include that baselin es are of different lengths and that the implementation of intervention (in this case, traini ng followed by classroom FA implementation) is planned and not serendipitous (Harvey, May, & Kennedy, 2004). Ther efore, this design is appropriate for this study. During multiple baseline analyses, phase changes are made when data demonstrates a stable trend within each phase. In this study, there were three phases of analysis: (1) baseline prior to teacher training, (2) training sessions, and (3) FA session implementation. During the baseline phase, data were collected on each teache rs level of integrity du ring one, two, or three sessions of each condition. The decision to di scontinue baseline session s during analysis is typically made on the basis of the stability of baseline data, such that a reasonable expectation of future performance can be derived from an an alysis of baseline performance (Kazdin, 1982). However, in this study baseline performance was measured for a predetermined number of sessions, according to the procedures used by Iwat a et al. (2000) and Wallace et al. (2004). Measurement Procedures The dependent variable in this m anipulation was the amount of procedural integrity with which teachers implemented FA procedures duri ng the assessment of appropriate classroom behavior. Teachers procedural integrity was m easured as the percent of correct steps per FA session, as well as the number of times each t eacher delivered reinforcement noncontingent on student behavior (an incorrect re sponse). A secondary dependent variable was the variation in the rate of appropriate behavior by the target student according to the manipulation of FA conditions. Appropriate behaviors were individually identified for each student in consultation
56 with the participating teacher. Preference was gi ven to discrete behaviors that were plausible replacements for the identified problem beha vior. For example, raising ones hand was considered an adequate replacement behavior for a student who freque ntly shouts out during class. Each appropriate behavi or was operationally defined in c onsultation with the participating teacher and reviewed during training (Appendix L). Independent variables included FA procedural training and performance feedback (for teachers integrity) and the differential reinforcement of appropriate behavior provided by each FA condition (for students behavior). Data were collected on the identified appropriate behavior of the student, disruptive student behavior, and correct implementation of FA procedures, including accurate de livery of the contingencies. Data on the accurate delivery of contingencies included integrity measurement of behavior direct ed toward peer confederates during the peer attention condi tion (e.g., prompting the peer to respond, when necessary, when the target student initiat ed during peer attention conditions). While disruptive behavior was not observed in order to determine function, data on in appropriate behavior was collected in order to determine any commensurate effects on disrupt ion associated with each condition (e.g., disruptive behavior decreases as ap propriate behavior increases). Data were collected by the PI in a frequencyby-interval paper and pencil format. Each 15 minute session was divided into one-minute intervals, providing a more precise and accurate estimate of interobserver agreement (IOA) than is allowed by a frequency only method. When a target behavior occurred, the minute was note d, the behavior was reco rded according to predefined codes for each student, and teacher inte grity was evaluated according to responses relevant to the condition and the behavior. Criteria by which teacher integrity was evaluated during each condition are presented in Tabl e 2-1. Additionally, data on noncontingent
57 reinforcement were collected during each condition. During the control condition, noncontingent attention delivered by the teacher on a fixed-time schedule of one minute was recorded as correct (delivered at the appropriate time) or incorrect (not delivered or delivered at the wrong time). These data points, because th ey were evaluated using a frequency method, were totaled with the other data regarding teacher integrity and included in the total percentage of steps completed with integrity by the teacher During teacher attention, escape, and peer attention sessions, data on noncontingent reinfo rcement were taken usi ng a partial-interval recording system, such that the number of minutes during which the relevant form of noncontingent reinforcement was observed was totaled and divided by the session length (15 minutes), then multiplied by 100 to obtain a percen tage of intervals during which noncontingent reinforcement occurred (Kazdin, 1982). For exam ple, if during a peer attention session, peer attention was delivered noncontinge ntly at any time during 5 of the 15 total intervals, the rate of noncontingent reinforcement would be calculated as (5/15 = .33) 100 = 33% of intervals. Paper and pencil data were preferred to computerized methods of data collection (e.g., video analysis or handheld personal computing te chnology) to reduce the imp act of reactivity to the observers presence (Appendix M for data coll ection form). In the research school under study, paper and pencil data is frequently co llected by consultants obs erving the classroom, making this method more amenable to the school context and less likely to distract students and/or teachers during data collection. Interobserver Agreement IOA data were collected on a m inimum of 25% of sessions by a trai ned graduate student observer acting as a data assistan t to the project. During FA impl ementation, criteria for IOA was set such that sessions on which IOA was less than 70% were not to be us ed to make decisions about the stability of student behavior or teacher integrity. IOA was calculated using the
58 proportional method of IOA, in which the smaller number of recorded events was divided by the larger number and multiplied by 100 to obtain a percentage of IOA (Kazdin, 1982) within each interval. IOA percentages were then totale d and averaged across each session to yield a percentage of total IOA. IOA was calculated within intervals to provide a more precise estimate of agreement than is offered by dividing the sm aller number of events by the larger number across the entire observation period (Lannie & McCurdy, 2007). IOA data were taken and reported on student appropriate beha vior, student disruptive behavior and teacher integrity to FA procedures across sessions. Baseline During baseline s essions, IOA was obtained on 100% (4/4) of sessions with Teacher A, 50% (4/8) of sessions with Teacher B, and 42% (5/ 12) of sessions with Teach er C. Variability in percentages of sessions observed is an artifact of the necessity to observe at least one condition for each teacher during baseline. As Teacher A only participated in four baseline sessions (one of each condition), and each condition had to be observed for IOA, 100% of sessions were observed. IOA during baseline sessions with Teacher A averaged 82% and ranged from 72% to 92%. During baseline sessions with Teacher B, IOA averaged 78% and ranged from 65% to 90%. The lower IOA noted during these sessions was hypothesized to be related to the fact that these sessions had to be videotaped (for which Teacher B gave consent), owing to scheduling constraints which prevented a data assistant from being present to take IOA data in vivo. Videotaped IOA sessions provide d a limited view of the classroom, obstructing some mock student behaviors and/or making verbal utterances difficult to hear. Finally, IOA during baseline sessions with Teacher C averaged 92%, and ranged between 87% and 96%.
59 Across conditions, total IOA averaged as follo ws: IOA for Teacher Attention sessions was 79% (range 65% to 92%), for Escape sessions was 85% (range 69% to 96%), for Peer Attention sessions was 85% (range 72% to 92%), and for C ontrol sessions was 91% (ra nge 90% to 100%). Training During training sessions, IOA data were taken on 79% of sessions with Teacher A, 75% of sessions with Teacher B and 100% of sessions with Teacher C. IOA averaged 86% during training sessions with Teacher A, and ranged fr om 67% to 96%. During training with Teacher B, IOA averaged 86% and ranged from 73% to 98% IOA during Teacher Cs training averaged 89% and ranged from 81% to 100%. Across conditions, total IOA averaged as follows: IOA for Teacher Attention sessions was 96% (range 90% to 100%), for Esca pe sessions was 85% (range 74% to 94%), for Peer Attention sessions was 86% (range 73% to 92%), and for C ontrol sessions was 88% (ra nge 81% to 92%). FA implementation During FA implem entation, IOA was collected on 38% of session with Teacher A. IOA averaged 94%, ranging from 92% to 99%. During Teacher Bs FA implementation, IOA was collected on 33% of sessions. Average IOA wa s 89%, ranging from 79% to 100%. IOA was collected on 44% of Teacher Cs FA implement ation sessions. Average IOA was 93%, and ranged from 77% to 100%. One IOA observation of FA implementation by Teacher C yielded IOA less than 70%; consequently, data from th at session was not used to represent teacher integrity levels or Student Cs performance during that condition. Across conditions, total IOA averaged as follo ws: IOA for Teacher Attention sessions was 92% (range 79% to 100%), for Escape session s was 98% (range 96% to 100%), for Peer Attention sessions was 95% (rang e 92% to 99%), and for Control sessions was 83% (range 77% to 92%).
60 Procedures Preassessment Prior to data collection, a problem identifica tion interview (PII) was conducted to identify the target inappropriate behavior(s) and inform the appropriate behavior s to be assessed. All interview procedures were conducted according to the procedures identified by Bergan and Kratochwill (1990) for behavioral consulta tion, and adapted to the current study. See Appendices I and J for the adapted teacher a nd student forms of the PII, respectively. Teacher Training Training p rocedures for the recruited middle school teachers occurred after all teacher and student participants were identified and consent had been obtained from the participating teachers and parents of the target students. Prior to training, teachers received and were instructed to read an outline of each FA condition, based on the pr ocedures section of the Iwata et al. (1982/1994) study (Iwata et al., 2000), and adapted to the assessment of appropriate behavior (Appendix K). Teachers read only the description of the impending session to be conducted, and were given five minutes to read the descrip tion before conducting the first baseline session. At subsequent baseline sessions, some of which occurred several days after the first baseline sessions, teachers were given between 30 seconds and two minutes to review the condition description. After readi ng the relevant section (e.g., escap e, teacher attention, peer attention, control), each teacher was instructed to conduct a mock FA session, with a trained graduate data assistant or the PI acting as the target student and/ or peer confederate. A data assistant acting as a peer confed erate was included in baseline a nd training sessions only during peer attention and contro l conditions, due to time constraints. Participating graduate assistants acting as the peer confederate dur ing peer attention a nd control conditions were instructed to provide the teacher an opportunity to prompt at least once every session, but to otherwise
61 demonstrate reasonably good integrit y to the procedures. The i ndividual acting as the target student, who was typically the PI, was instructed to demonstrate at least three behaviors per minute, and to demonstrate each of the target appropriate and disruptive behaviors at least once per session. During mock sessions, a higher in cidence of each specific target appropriate behavior was intentionally demonstrated during th e condition in which th at behavior might be expected to occur more frequent ly (e.g., raising hand during teach er attention conditions). Teachers were observed by the pr incipal investigator during mock FA sessions and their integrity to the FA procedures was measured as described above. All mock FA sessions followed the procedures for actual sessions desc ribed above, with the ex ception of shortening mock FA sessions to three minutes to allow fo r limited time. Following the training described below, teachers integrity during mock FA sessions was again assessed, in order to measure the effect of training on teachers procedural integrity (Iwata et al., 2000; Wallace et al. 2004). Performance feedback was delivered following each session, during which teachers were recognized and praised for correct ly implemented steps, and co rrective feedback was provided for each missed step. Initial training occurred fo llowing the baseline mock FA sessions described above. Initial training procedures included: 1) A visual (e.g., PowerPoint) presentation desc ribing the research aims and procedures, including video segments of mock FA conditions modeled by the PI and trained graduate data assistants and brief oral quizzes about previously presented material (Iwata et al., 2000). 2) An overview of the operational definitions of the disruptive behavi ors exhibited by the identified student participants. 3) Discussion and development of operational defi nitions for student appropriate behaviors that were to be assessed during research implementation according to the classroom expectations of each teacher. Operational defi nitions of disruptive behavior also reflected the non-occurrence of appropriate behavior; in stances of disruptive behavior constituted non-examples of appropriate behavior.
62 4) Discussion and development of peer attenti on conditions for each classroom according to the classroom expectations of each teach er (peer training is described below). 5) Additional mock FA sessions until each teacher reached a mi nimum of 80% integrity to research procedures at least twice in each c ondition and with an incr easing or stable trend, with performance feedback from the PI following each session. 6) Scheduling and research timelines for each teacher. Peer Training Prior to im plementation of FA conditions, id entified peers met with the PI to review conditions to be conducted that da y. Peers were told that they were helping the target student, who sometimes has difficulty behaving appropria tely during class, a nd that they would sometimes act as a classroom buddy for that st udent when he/she needed help. Peers were instructed to recognize th e target appropriate beha vior, and appropriate fo rms of attention were reviewed for the peer to use in response. For ex ample, when the target st udent spoke to the peer appropriately about his/her work du ring peer attention conditions, the peer was instructed to help the target student, but to speak only about the assigned activity. During control conditions, the peer was given a timer and instructed to check in on the target student every minute, ensure that he/she understood the work, and provide any necessary help. Du ring escape and teacher attention conditions, the peer was in structed not to speak to the ta rget student at any time. An index card, similar to that provided to the target student, also described expectations for peer behavior during each condition, and was delivered with the target students card at the beginning of each session. Functional Analysis Conditions After operational definitions were created and teachers were trained to implem ent procedures with integrity (e.g., 80% integrity during training sessions; see Appendix L for operational definition form), functional analys is conditions were implemented and data
63 collection on student behavior began. Analyses of student behavior were conducted using a multi-element FA design. Sessions were 15 minutes in duration, and data collection on each session continued until a stable three point tre ndline was observed in teacher integrity and, ideally, in student behavior. Three experimental conditions including esca pe from task demands, teacher attention, and peer attention were conduc ted, in addition to a control condition. Each condition is described below. Escape from task demands This condition was designed to identify behavior that occurred as a function of avoidance of difficult academ ic material. Teachers began this session by placing a card on the desk of the participating student that stated; Were going to work on material that may be difficult. Please do your best work. If you need a break, please as k for one by asking me for a break, and Ill give you 30 seconds to yourself. They then led the class into an activity that was previously identified as challenging for the target student. Data collection began when the instruction card was handed to the target student. Each time the target student requested a break, the teacher indicated to the student that it was time for a break. Breaks consisted of sitting quietly, putting ones head down, or engaging with a stress ball, (Student B only, per the recommendation of Teacher B). At the conclusion of 30 seconds, th e teacher indicated that the break was over and that the student should return to work. If the student again requested a break, one would be provided. No contingencies were provided for instances of disruptive behavior, unless the behavior escalated to a level of severity that the teacher percei ved it must be addressed. During training, teachers were instructed that if such an escal ation occurred, they should address it in a manner consistent with exis ting classroom procedures.
64 Teacher attention This condition was designed to identify behavior that occurred as a function of attention from the teacher. Teachers began this session by placing a card on the desk of the participating student that read, Were going to work on some material youre familiar with. Please do your best work. If you need help, let me know by raising your hand and Ill come talk to you. They led the class into an activity that was modera tely challenging for the target student. Data collection began when the instruction card was handed to the target student. After explaining the activity to the class (if necessary), the teacher sat at his/her desk or engaged in some other activity that gave the appearance of not providing attention to th e class. Each time the target student engaged in the identifie d behavior, the teacher walked to his/her desk and began speaking with the student. Teachers spoke to the student as they normally would (e.g., Do you need help? What are you having trouble with?) and then provided any necessary help. Teachers spoke to the student for approximately 15 seconds. If the student engaged in the target behavior again, the teacher w ould remain with the student for another 15 seconds. No contingencies were provided for instances of di sruptive behavior, unless the behavior escalated to a level of severity that the teacher perceived it must be addressed, in which case the teacher had been previously prompted to respond to it according to existing cl assroom procedures. Peer attention This condition was designed to identify behavior that occurred as a function of attention from peers. Peer confederates were identified fo r inclusion based on strong social skills, ability to follow instructions, and an existing relationship with the target student (Appendix G). Prior to beginning sessions, peers were instructed to recognize and deliver attention for each occurrence of the identified behavior, as described below. At the start of each new condition, peers were privately reminded of the expectations for their behavior by the PI. At the beginning of each
65 session, peers also received an instruction card similar to that given to the target student reminding them to respond to the target students requests for help. Teachers began sessions by placing a card on the desk of the participating stud ent that read, Were going to work on some material youre familiar with. Please do your best work. Youre allowed to talk quietly during this activity, so if you need help from (peers name), you may ask them for it by speaking quietly and appropriately about the work. The teacher led the class into an activity that was moderately challenging for the target student. Data collec tion began when the instruction card was handed to the target student. Each time the target student engaged in the identified behavior, the identified peer spoke with the student for a pproximately 15 seconds about the assigned activity and offered any necessary help. If the student engaged in the target behavior again, the peer spoke with the student for another 15 seconds. If the peer did not respond to the target students identified behavior, the teacher was responsib le for reminding him/her to do so by tapping him/her on the shoulder or quietly saying, (Peer), I think (target student) needs your help. Teachers only provided this directive once per in stance of target student identified behavior. Similarly, if the peer was providi ng attention to the ta rget student independently of the target students behavior, the teacher prompted the peer to return to his/her wo rk. No contingencies were provided for instances of disruptive behavior unless the behavior es calated to a level of severity that the teacher perceive d it must be addressed. If an escalation of disruptive behavior impeded classroom functioning, the teacher was encouraged to respond to it according to existing classroom procedures. Control The control condition was designed to provi de a basis of comparison for the other experim ental conditions. In this condition, reinforcement was not contingent on instances of the identified appropriate behavior, and the students did not experience the antecedents making such
66 behavior likely to occur. Rather, teacher and peer attention was deli vered noncontingently, and antecedents presumed to be least aversive (e .g., an easy academic task) were present throughout the condition. In this condition, teachers bega n sessions by placing a card on the desk of the participating student that read, Were going to do some work youre all familiar with. Ill be coming around to talk to you as you may need help. (Peer) will also check in with you from time to time. Please work quietly and do your best work. Data collection began at the conclusion of oral instructions. No contingencies were provided for instances of disruptive behavior or any of the target appropriate behavior s, but teacher and peer attention were provided on a fixed time schedule of every one minute. Participating peer confederates were instructed to provide such attention prior to the session, and had an i ndex card indicating when and how to provide attention. If they did not provide attention at the requisite time, the teacher was responsible for reminding them to check in on the target studen t. During training, teac hers were instructed that if an escalation of disr uptive behavior impeded classroo m functioning, they should respond to it according to existing classroom procedures. Procedures for Monitoring Consultation Consultation was based on the steps of the BC paradigm during which teachers participated in problem identification, problem analysis, plan (FA) implementation, and plan evaluation (social validity) stages However, it was also necessary to monitor the consultation process with each teacher. To do so, the resear cher maintained a log of each consultative interaction. In the log, events were monitored a nd recorded, such as the completion of important consultative and research procedur es. Additionally, aspects of each consultative interaction that seemed potentially important were noted. Elemen ts of each consultation that were recorded included teachers statements about the target students behavior and/or its causes, relevant interaction factors, such as a teachers tone or message c ontent during a meeting, teachers
67 apparent reactions to the implementation of the FA in their classrooms, procedural or logistical hurdles to be overcome, or statements rela ted to the development of each consultative relationship.
68 Table 2-1. Criteria for ev aluating teacher integrity during classroom-based FA Teacher correctly responds to target appropriate behavior Teacher correctly responds to disruptive behavior Teacher correctly prompts peer (if necessary) Teacher correctly ends reinforcement interval Peer correctly responds Control Escape Peer Attention Teacher Attention
69 CHAPTER 3 RESULTS This inves tigation examined two variables rela ted to the implementation of FA in general education middle school classrooms. The first, the integrity w ith which teachers implemented FA procedures, was examined across three phases: baselin e, training, and classroom implementation. The second factor under study wa s students responses to the FA conditions. Specifically, targeted appropriate behaviors and identified problem behaviors were observed and recorded. The frequency with wh ich students engaged in appropria te behavior across conditions provided evidence for the presence of a reinforc ing function for that be havior. Finally, all outcomes were examined in the context of a consultation conducted with each participating teacher, according to the procedures of BC. The results of this inve stigation are presented below. Teacher Integrity Teachers integrity to the FA procedures duri ng each phase of the analysis is presented graphically. The first phase, baseline, represen ts the level of integrity with which teachers performed an abbreviated mock FA session with only a written description as a guide. The second phase, training, represents teachers integrity during practice sessions (also an abbreviated mock FA) after vi ewing training presentations about how to conduct a classroombased FA, including video modeling of each cond ition. Teachers also re ceived performance feedback following each session during this phase. The third phase represents the introduction of the FA to the classroom and the target st udent, during which teachers were responsible for carrying out the environmental manipulations peculiar to each c ondition. Performance feedback was delivered approximately every f our sessions during this phase. Figure 3-1 depicts each teachers integrity to the FA procedur es during each phase of the assessment. During the baseline phase, Teachers A and B demonstrated integrity ranging from
70 20% to 87% of steps completed correctly. Mean rates of integrity du ring the baseline phases were 57% for Teacher A and 55% for Teacher B. The teacher attention condition stood out during this phase of the analysis as one duri ng which both Teachers A and B were able to perform with high levels of integrity despite having no training in how to do so. Teacher Cs baseline data is notable for the change in magnitude observed between sessions 4 and 5. This change is signified in Fi gure 3-1 by a phase change separating that data into baseline (bL) and baseline prime (bL ) phases. Due to time constraints, Session 5 began one week after Sessions 1 through 4 were conducted. Prior to beginning mock FA sessions that day, Teacher C inquired as to whether he should do what he thought the cards were telling him to do, or whether he should do what he would normally do in his classroom. During initial baseline sessions, the integrity demonstrated by Teacher C averaged 40%. When instructed to follow the cards as best as he understood them, he began to demonstrate high levels of integrity in all conditions, ranging from 85% to 100%. Following training only one mock FA probe was conducted in each condition with Teacher C. This modification was made because of the high levels of integrity he demonstrated during the baseline prime phase. During the training probes, Teacher C demonstrated 100% integrity in each condition. The stability with which such high integrity was observed across the baseline prime and training phases argued agains t the need to conduct further observations of Teacher Cs integrity to the FA procedures. Training sessions continued for Teacher A until requirements for stability at 80% criteria or above were met. For Teacher A, three sessions of each condition were sufficient to demonstrate stability of performance, with the exception of the control condition. Five training sessions of the control condition were conducted to demonstrate stability of integrity above 80%.
71 Teacher Bs training data reflects acceptable leve ls of integrity for a ll conditions after two sessions of each. Due to limited time, Teacher B proceeded into the third phase of the analysis on the strength of increasing trend in three out of four conditions. Levels of integrity during FA implementati on were highest for Teacher C, followed by Teacher A, then Teacher B. Teachers A and C both demonstrated consistently high levels of integrity across conditions. Thes e teachers were most likely to de monstrate slightly lower levels of integrity during peer attention (Teacher A mean integrity = 91%; Teacher C = 96%) and control conditions (Teacher A mean integrity = 92 %; Teacher C = 96%). Similarly, Teacher B demonstrated consistently low performance th roughout the control condition (mean integrity = 52%). However, she demonstrated variable performance during p eer attention sessions (mean = 88%, range 64% 100%). Table 3-1 details the areas in which each te acher missed steps to be implemented in the classroom-based FA. All three teachers demonstrated difficulty monitoring peer performance and prompting peers regarding th eir behavior when it was necessa ry during peer attention and control conditions. Teacher A was most likely to demonstrate erro rs in peer prompting, followed by the delivery of noncontingent delivery of attention during the control condition. Teacher B was most likely to make errors in ignoring instan ces of disruption, ignoring instances of raising hand outside of the teacher attention condition at a schedule of once for every three instances of hand raising, prompting the peers, and deliver ing noncontingent attent ion during the control condition. Teacher C was most likely to make e rrors in prompting the peer and in delivering noncontingent attention duri ng the control condition. Finally, data were also taken on the percent of intervals during which the target student received the form of noncontingent attention during each of the three experimental conditions
72 that was contingently available during that condition. Those data are represented by the closed data points on Figure 3-1, and indicate that each teacher was able to maintain low levels of the relevant form of noncontingent reinforcement during experimental conditions. Levels of noncontingent reinforcement during experimental conditions averaged 13.67% of intervals for Teacher A, 9.78% for Teacher B, and 7.57% for Teacher C. Sessions during which the percentage of intervals in which noncontingent attention exceeded 20% were Session 28 for Teacher A (73%, Session 10 for Student A) and Session 19 for Teacher C (40%, Session 3 for Student C). Teachers A and B demonstrated improvements in procedural integrity following training, modeling, and performance feedback regarding how to implement an FA. Improvements in integrity demonstrated by Teacher C appeared to be related to information supplied by the researcher during the baseline phase, rather than the instigation of training. All three teachers demonstrated high levels of procedural integrity throughout the training phase. During classroom implementation, Teachers A and C maintain ed high procedural integrity. Teacher Bs procedural integrity was vari able throughout that phase. Functional Analysis of Appr opriate Classroom Behavior Each teach er was responsible for conducting a classroom based analysis of appropriate behavior for a student demonstrat ing disruptive behavior in his/he r classroom. Data on each of the student target behaviors is presented graphically, in which session number is graphed on the x-axis and rate per minute is graphed on the y-axis. Rate per minute was calculated for each session by dividing the total frequenc y of instances of the target be havior by the total duration of the session. Sessions were typical ly 15 minutes in duration; howeve r, several sessions had to be shortened due to a sudden or unexpected change in classroom activity. In the case that a session
73 was shortened, rate per minute data were prorated by dividing the total frequency of the target behavior by the actual lengt h of the session duration. Multielement FA Student A Student A did not request a break in any of the conditions, thus, that data is not presented. He dem onstrated elevated rate s of raising his hand, appropria te peer speech, and disruption during all FA conditions. Figure 3-2 represents St udent As rate per minute of hand raising in each of the four FA conditions. The overlap in data points, lack of magnitude change between conditions, and lack of stable trend in any of the conditions indicate th at the function of hand raising was undifferentiated in experimental cond itions during which hand raising was reinforced compared to those when it was not. Figure 3-3 depicts Student As rate per minute of appropriate peer speech in each of the FA cond itions. During and after Session 6, a change in magnitude can be seen between rates of appropr iate peer speech during the peer attention condition, during which that behavior was consequated with brief amounts of peer attention, and the other three conditions. This change in ma gnitude continues with an increasing trend, until Session 14, when the rate per minute of appropriate peer speech declined during the final peer attention condition. However, rate s of appropriate peer speech in the other conditions remained stable until the conclusion of the FA, and an elevated rate of appropriate peer speech was demonstrated in all sessions. Thus, it appears that appropriate peer sp eech was reinforced by peer attention, though the data indicate a possibl e decreasing trend. Student B Student B did not dem onstrate requests for a break from the academic task during any condition; therefore, data regard ing that target behavior are not presented. Data regarding Student Bs rates of raising hand/appropriate teacher speech a nd appropriate peer speech are
74 provided in Figures 3-5 and 3-6, re spectively. Both indicate that th e function of either behavior was not differentiated by this analysis. Furtherm ore, the greatest stability of rates of each of the observed behaviors is noted duri ng the control condition. This suggests that the maintaining reinforcer for Student B to engage in any of th e target behaviors may have been present during the control condition, but was not id entified in this analysis. Student C Student C did not dem onstrate requests for a break from the academic task during any condition; therefore, data regarding that target behavior are not presented. Furthermore, Student C demonstrated only one instance of raising her hand during the FA, during Session 3, a peer attention condition. Data regarding that target behavior are depict ed in Figure 3-8; however, the lack of variability in that da ta over the length of the FA re nders any visual analysis and interpretation of that target behavior impossible. The rate at which Student C engaged in appropria te peer speech is illustrated in Figure 3-9. The high rates of that target behavior (approxi mately 1 2 times per minute) during Sessions 3 and 7 provide evidence of the reinforcing efficac y of peer attention for appropriate behavior. However, the trend in the rate at which Student C demonstrated appropriate peer speech is decreasing, and she demonstrated only 0.13 instances per minute of that behavior during the final peer attention condition. Disruption Although disruption was not consequated during th e FA, descriptive data on the frequency of its occurrence was taken in order to observe any commensurate effects of the FA on the frequency of disruption. Student As rate per m inute of disrup tion is graphed in Figure 3-4. Student A demonstrated a very high rate of disruption in Session 10, an escape condition. However, he was most likely to demonstrate elevat ed rates of disruption in the peer attention and
75 control conditions. Data regarding Student Bs ra tes of disruptive behavior is provided in Figure 3-7. The rate of disruption does not appear to be differentiated by any of the conditions presented. However, declining rates in disruption toward the conclusion of the FA suggest a possible treatment effect of the FA procedure. Over the course of the FA, Student C appeared to become increasingly reactive to the observer(s) presence in the classroom. Duri ng the final FA session, an escape condition, she engaged in aggression toward the peer confeder ate each time that Teacher C was not looking in her direction. That elevated ra te of aggression is represented in the rate of disruption, presented in Figure 3-10. Though aggression was not original ly a component of the operational definition of disruption demonstrated by Student C, it was r ecorded as such during that session on the basis of its potential to di srupt other students (specifically, th e student upon whom the aggression was occurring) from the classroom activity. That session was discontinued early and the FA was terminated as a result. The reactivity demonstr ated by Student C poses a serious threat to the internal validity of this FA and appears to have had a commensurate effect on the data path of her observed rates of appropriate behavior during the pe er attention conditions. Peer Attention Reversal Analyses Because the multielem ent data derived from the FA of Student As and Student Bs rates of appropriate behavior indicated a possible peer attention function, addi tional graphic analyses were conducted to further examine the differences in appropriate behavior demonstrated during the peer attention condition in comparison to the other conditi ons. In particular, control conditions (during which average rates per minute of appropriate peer speech often overlapped with those demonstrated during the peer attention condition) were further examined using a reversal design treating small time segments of the FA session as the unit of analysis. Breaking each session down into smaller units of time also allowed for an examination of trend within
76 each session, which may provide additional evidence regarding either students response to the experimental contingencies as they came under the stimulus control provided by each (Vollmer, Iwata, Zarcone, Smith, & Mazaleski, 1993). For exam ple, a high rate of appropriate peer speech at the beginning of each control condition, accompan ied by a decreasing trend in that behavior over the course of the session w ould suggest that a students ra te of appropriate peer speech decreased as a result of an extin ction effect of an experimental condition in which that behavior was not reinforced by peer attention. Likewise, an accelerating rate of the same appropriate response during the peer attention condition woul d further support the reinforcing efficacy of peer attention for that behavior. Student A Figure 3-11 depicts Student As rate of appropriate peer sp ee ch across control and peer attention conditions. Each unit de picted on the x-axis represents three minutes of the session, in escalating order. The y-axis depicts the total fr equency of appropriate pe er speech. Data are depicted cumulatively; thus, the ra te of increase on the rise of th e graph depicts th e total units of appropriate peer speech demonstr ated during each time increment. For example, in Figure 3-11, when the graph depicting appropria te peer speech during Session 6 increases from seven to eight between minutes 12 and 15, this indicates that Student A demonstrated one instance of appropriate peer speech during th at time. Cumulative graphing of the data allows both a minuteby-minute analysis of its trend and a composite analysis of the overall frequency of target behavior demonstrated within each session. Thus, the total freque ncy of appropriate peer speech within each session is represen ted at the data point above Mi nute 15 on the x-axis, allowing a similar comparison between conditions to that provided by the multielement data presentation. This data analysis still sugge sts the possibility of decreasi ng trend across both conditions, but further suggests that ra tes of appropriate peer speech were more likely to increase within each
77 peer attention condition. In c ontrast, during the control conditi on, the frequency of appropriate peer speech increased throughout only one session (Session 3). The point at which Student A ceased issuing appropriate peer speech duri ng the remaining conditions was during the 10-12 minute interval in Session 5, the 6 9 minute interv als in Sessions 11 and 16. In fact, in both of those sessions the final instance of appropriate peer speech occurred duri ng minute 7, indicating that appropriate peer speech did not persist in to the second half of a condition during which it was not reinforced by peer attention. Multielement presentation of the rates of disruption suggest ed that rates of appropriate peer speech were consistently highest during the peer attention and control conditions. However, cumulative line graphs of the frequency of disr uption during those conditions suggest a different pattern. The frequency of disrup tion during control and peer attention conditions in a reversal design using cumulative line graphs is depicted in Figure 3-12. This analysis suggests an increasing trend in the rate of disruption during the c ontrol condition, while during peer attention conditions the frequency of disruption did not in crease after the 12 minute increment in Sessions 6 and 9, and did not increase beyond the 9 minute increment in Session 14. Additional comparisons are necessary to fully interpret any possible effect of the p eer attention condition on the frequency of disruptive behaviors, but these findings suggest that the reinforcing effects of peer attention on rates of appropriate peer spe ech during the peer atten tion conditions may have contributed to a concurrent decrease in the frequency of disruption. Student B Despite undifferentiated rate s of responding during the FA, elevated rates of responding were noted during some experimental conditions Specifically, Student B demonstrated a high rate of appropriate peer speech during Session 2, a peer attention condition. However, he also
78 demonstrated elevated rates of appropriate pe er speech during the control condition. Thus, a within session reversal analysis of those conditions was conducted fo r each of those behaviors. Figure 3-13 illustrates the comparison between the frequency of appropriate peer speech observed during the peer attenti on and control conditions. This comparison does not indicate that peer attention is an effec tive reinforcer for appropriate p eer speech; the delivery of that consequence appears to have had a punishing e ffect during those conditions. Rather, it appears that there may be a reinforcing effect of an unknown consequence delivered during the control condition, or that the antecedents present dur ing the control condition (e.g., noncontingent attention) produced higher rates of appropriate peer speech. Figure 3-14 de picts a comparison of the frequency of disruption between the peer atte ntion and control conditi ons. This contrast indicates that rates of disruption may be reduced during the peer attent ion condition; however, this conclusion is only partially supported given the lack of observed stability in the data. While these analyses provide some additional evidence regarding the effects of the control condition on Student Bs behavior, it shou ld be noted that a ll control and escape conditions, and one peer attention condition, were conducted at less than 80% integrity. Thus, any conclusions about Student Bs performance during the FA are extremely tenuous. Multielement Data During Session s of 100% Procedural Integrity Due to the high level of integrity noted dur ing many of the FA sessions conducted with Student A, and lingering questions regarding the effect of teacher s procedural integrity on FA outcomes, the multielement FA data obtained with Student A was further analyzed in the context of Teacher As procedural integrity. Specifica lly, those sessions during which integrity to all procedures was 100% were gra phed alone (Figure 3-15, 3-16, a nd 3-17). Though a trend can not be observed in the single data poi nts present from the peer atten tion and control conditions, these analyses may clarify the results obtained from the full multielement design. Specifically, the
79 results remain unclear for rates of raising hand, a nd suggest that high differ ences in magnitude in rates of appropriate peer speech and disruption remain between the peer attention and other conditions. Social Validity After the conclusion of data collection, a social validity interview was conducted to ascertain ad ditional information about each t eachers perceptions of the FA training and implementation processes. This interview also served as the plan evaluation component of the consultation, according to the procedures of BC. The interview, which can be seen in Appendix N, consisted of 12 Likert-style items, on which 7 indicated a high level of the descriptor being probed in each item, and 1 indicated a low level of that descriptor. Each teachers responses to the Likert items are desc ribed in Table 3-4. Teacher A Teacher A indicated that she was so mewhat comfortable with the study, but that it took some time to grasp the concept of the FA procedures. According to her report, this decreased her overall level of comfort because she thought it was important to understand the procedures on a conceptual level prior to using them. Teacher A noted some differences in her social validity ratings over the course of the FA and reported two numbers for those items, these numbers are presented in Table 3-4. Specifical ly, she reported a decrease in th e participating peers apparent comfort level between the begi nning and end of the assessment sessions, an increase in disruption to the classroom over time, and an incr ease in the amount of disruption it caused other students. At the end of the interview, Teacher A was as ked about new information and/or skills that she had gained as a result of participati ng in this study. She reported the following:
80 1) She had delivered contingent escape to increase time spent on task for a student (other than the target student). 2) She had thought about using cards similar to those used during the FA to describe behavioral contingencies to disruptive students. 3) She was considering individual preferences for types of reinforcement as an intervention strategy for students who disrupted the classroom. 4) She was thinking about how to use group contingencies for completed work. Additionally, while responding to this question Teach er A discussed with the researcher at length how to address a hypothesized function of teacher attention for another students disruptive behavior. She also reported that she had recently considered how similar ideas about reinforcement might be maintaining tantrum beha viors exhibited by her three-year-old niece. Teacher B Teacher Bs responses to the social validity in terview are presented in Table 3-4. Overall, she indicated that she was comf ortable with the study, that it wa s somewhat time intensive to participate, but that it was minima lly disruptive to the classroom and its routines. She stated that she would encourage othe r disruptive students to participate in a similar type of assessment in the future. However, she indicated that her wil lingness to participate in a similar assessment in the future would depend upon the timing of such a procedure. Specifically, she reported that she would rather such an assessment begin in the fa ll, so that scheduling of FA sessions would fit more naturally into the classroom context and would not necessita te changing classroom activities in order to acc ommodate time constraints in data collection. She indicated that the information obtained from the assessment was so mewhat helpful, its effects being attenuated by the fact that it was not very reasonable to implement the FA conditions. Her primary concern about implementation centered on the control conditio n; she stated that it was not realistic to provide attention to one student every minute. She gave the pro cedure a global rating of
81 indicating that she perc eived it to be benefici al, saying her primary reason for doing so was the decrease in Student Bs rates of disruption over the course of the assessment. While she was uncertain whether it was the conditions themselves or an artifact of being experimented on that affected his behavior, she reporte d that Student B seemed to appreciate and be grateful for the help with his behavior. When asked what skills or knowledge she obtai ned from participating in the assessment, Teacher B indicated the following: 1) She was considering using cards similar to those used during the FA to describe to students the behavioral expectations during a specific activity, and/or to encourage peers not to attend to problem behavior dem onstrated by certain students. 2) Providing noncontingent attention to a disrup tive student, by checking on his/her progress, was a successful strategy to decrease disruptive behavior, and in the future she might do so as often as once every three minutes. Teacher C A social validity in terview was conducted with Teacher C following the conclusion of FA sessions in the classroom. His responses to the social validity questionn aire can be viewed in Table 3-4. Overall, he reported that the process was not as di sruptive as he had originally expected it to be, and that the information obtained was somewhat useful. He clarified, however, that the information obtained from the FA would be more useful at the beginning of the school year, rather than at the end. Teacher C indicated that he might not be willing to participate in similar processes in the future, and that his reas on was that he perceived once was sufficient to teach him the FA procedures and attempt the implem entation of an FA in his classroom. He did indicate that he would be willi ng to have another student participate in the future. Finally, he rated the assessment procedure as a four on the se ven-point scale, stating that he was unsure how feasible the procedure was wit hout extensive consultative suppor t, and simultaneously concerned about the level of invasiveness to the classroom when that support was provided (e.g., when the
82 consultant was in the classroom taking data). When asked what skills or ideas he gained from the process, he cited an increased awareness of the importance of ignor ing disruptive behavior and practice using that skill. He stated that he had been using that principle with a variety of students other than th e target student. Student Reactivity Prior to the implem entation of FA conditions in the classroom setti ng, the researcher did not interact with any of the pa rticipating target students, in an attempt to reduce potential reactivity to the researchers pr esence. However, due to the demands of data collection (e.g., needing to sit near the target students), at so me point during the FA im plementation each student became aware that he/she was being observed. Fu rther, each student demonstrated some aspects of reactivity to the observations and FA conditio ns. Student A appeared to note that he was being observed during approximately the 3rd or 4th observation. During those observations, he was noted to turn frequently in his seat to look at the researcher. St udent A began complaining about the FA conditions during approximately the 11th or 12th FA session. From that time forward, he was noted to mutter under his breath to the peer conf ederate prior to and following FA sessions. While the exact content of thos e statements in unknown, he appeared to be complaining about the FA process. Toward th e end of the FA he frequently greeted the researcher upon entering the classroom. During FA conditions themselves, however, he did not engage in any behaviors dire ctly indicating reactivity. Student B demonstrated relatively little reactivity to the FA procedures. At the end of the first class period during which the FA was conduc ted, he returned the last sessions condition cards to Teacher B and asked, Do you need these for other kids like me who need help with their behavior? When Teacher B reported this to the researcher, she indicated that she thought he seemed grateful for what he perceived to be as sistance. This was despite the fact that when a
83 peer asked him about the cards during an early teacher attention conditi on, he responded, I dont know, some stupid card telling me that I should be quiet and raise my hand if I want help. Thus, it is unclear to what degr ee he might have been embarrassed about being singled out for an assessment of his classroom behavior. However, he did not engage in any attempts to hide his participation. Additionally, Student B did not app ear to fatigue of the FA conditions until later in the assessment. Session 11 was the first th at he was observed to ask Teacher B, upon presentation of the condition cards, H ow much longer do I have to do this? Student C did not appear to note that she was being observed until approximately the 7th condition. However, after noting that the resear cher was observing her, she became extremely reactive to the procedure. She frequently asked Teacher C when and why she was being observed, and when it would be discontinued. She was not reassured by Teacher Cs explanations that he was being observed also. Student Cs reactivity became most notable during Session 10, when she began to aggress toward the peer confederate each time that the teachers back was turned. As previously not ed, this session and the FA were subsequently discontinued. Consultation Outcomes Experim ental procedures, specifically, training and performance feedback, were embedded in the procedural context of behavioral consu ltation (BC). As described in Chapter 2, BC was used to unify the steps of the study and as a supportive context intended to maximize the integrity with which FA procedures were impl emented by each of the participating teachers. Following each consultativ e interaction, important events occu rring during that interaction were logged by the researcher. Thus, an analysis of each of the consultati ons and its outcomes is presented via a) the stages of BC and b) their relevance to several factors known to influence consultative interactions. Specifi cally, those latter factors incl uded: entrance into consultation,
84 teachers perceptions of problem behavior, goa ls for consultation, the degree of consultant (researcher) directedness during verbal interactions, time allocated to meetings with the consultant, reactions to performance feedback me etings, the content of those meetings, and the overall nature of the c onsultative relationship. Teacher A Consultation began first with Teacher A in November, 2007 and concluded in April, 2008. Consultation began first with Teacher A primar ily because she was the most enthusiastic and flexible of the three participating teachers w ith regard to the time commitment and procedural rigor involved in the research pr oject. Given the high number of procedural demands involved in this research, choosing the most enthusiastic par ticipant with whom to begin was conceptualized as a means of reducing the variab ility inherent to consultative procedures with individuals with differing backgrounds, classroom pr ocedures, and perceptions of consultation. Put another way, choosing the most cooperative te acher participant at the onset of consultation decreased the potential likelihood of the consultation itself introducing new pr ocedural challenges to its implementation in a general education middle school classroom. As a result, the researcher was better able to carefully monitor other variables having the potential to in fluence the procedural integrity and results of the FA. Thus, consul tation with Teacher A served as a model for consultation steps to be taken with Teachers B and C. Entrance into consultation with Teacher A bega n as a result of her being recommended to participate by the middle school assistant princi pal. Thus, the consultation was not entirely voluntary. However, Teacher A began to expre ss enthusiasm for the project during the first meeting, and indicated that she had many stude nts she thought would be appropriate for participation. She seemed to perceive the pr oject as an opportunity to improve a students problem behavior, but did not expr ess specific attributions about pr oblem behaviors that students
85 demonstrated. During the selection process, she was careful to choose a student participant who was not already receiving any form of behavioral intervention se rvices, further suggesting that she perceived the consultation as an oppor tunity to target student behavior. Consultation with Teacher A became increasingly collaborative following the problem identification stage. She readily understood a nd applied the procedures to her scheduled classroom activities, easily adap ting her lesson plans and instru ctional strate gies per the requirements of the given condition. This unde rstanding appeared to begin early on in the consultation. After watching the training videos, in which each condition was modeled for her, she stated, I can see how I can use this with other students, too! This level of understanding appeared to increase when the FA began in th e classroom. Teacher A typically informed the researcher of which conditions would be appropr iate given the scheduled day and time of FA sessions. Rather than the researcher having to inform her of which c onditions needed to be conducted, she often began scheduling meetings by saying, We need to do a peer attention, right? I think we can do that during ______ activity on Tuesda y. Furthermore, she was accurate in her monitoring of which conditions needed to be conducted, and the activities that she selected were almost always appropriate for the requisites of a given condition. As the consultation developed, Teacher As goa ls continued to center on helping Student A improve his behavior. Secondary goals appeared to include helping the re searcher complete the project and learning new strategies for managing classroom behavior. Fu rther, the consultation was equally composed of consultant and consulte e directed statements. While she was receptive to the researchers statements about sessions to be conducted and suggestions for data collection times, Teacher A remained forthright in asserting days and times that were not available for
86 conducting FA sessions, concerns ab out the FA, and in discussing any potential obsta cles to the FA process. During FA implementation, performance feedback meetings were conducted with Teacher A on a variable ratio schedule of four sessions (performance feedback was conducted after every session during training). Thus, while meetings t ypically took place after four FA sessions had been conducted, they occasionally were scheduled after three or five se ssions. The introduction of a variable ratio schedule wa s not intentional, but was related to the feasibility of planning performance feedback sessions (i.e., performance feedback could not be given during the middle of a class period during which two sessions were to be conducted). Teacher A was flexible to the scheduling of performance feedback meetings, and consistently scheduled a sufficient amount of time to conduct such meetings. Consistent with the procedures outlined in Chapter 2, performance feedback sessions included review of the graphic results of both the teacher integrity data and Student As performance during the FA, as well as review of steps implemented correctly, praise offered for correct responding to specific cha llenges that may have occurred during the sessions, and review of st eps implemented incorrect ly. Teacher A often began performance feedback meetings by no ting or inquiring about steps implemented incorrectly (i.e., I didnt get over to him (to provide NCTA) every time (every minute during the control condition).) Such statements we re confirmed, but immediately followed by noting the overall integrity of the se ssion and steps implemented corre ctly. If necessary, steps implemented incorrectly were revisited to engage in problem solving about how to handle such scenarios in the future (e.g., I ts okay if you dont make it to pr ovide NCTA at the minute mark if youre speaking to another student, just make sure you get there as soon as possible.)
87 Teacher A consistently used performance feedba ck sessions as an opportunity to ask questions about her performance, and how the FA was proceeding. The consultative relationship with Teacher A was highly collaborative. Further, the high level of collaboration that characterized this consultation was present from early on in its development. At every stage of the BC process, she appeared to be comfortable expressing her questions, concerns, and enthusiasm about the project to the researcher. Teacher B Consultation with Teacher B began in January, 2008 and concluded in May, 2008. Teacher B was chosen to participate second because it was perceived that it may take slightly longer to complete training sessi ons with her than with Teache r C. Also recommended to participate by the middle school assistant principal, Teacher B appeared to perceive that she had little choice in participating, a nd that the project was being thru st upon her. During the initial meeting, she spoke extensively of her concerns that the project would require large amounts of her time and the level of intrusion to her cl assroom. When asked about potential student participants and her experience with students with behavior problems, Teacher B expressed internal, fixed attributions (A thanasiou, Geil, Hazel, & Copela nd, 2002) about students problem behavior. For example, she referred to st udents who demonstrate di sruptive behavior as troublemakers and bad kids. Specific to Student B, she credited much of his disruptive behavior to his developmental hi story and internal problems w ith self-control. During the problem identification interview, she indicated that she thought Student B was compelled to disrupt the classroom when it became too quiet. Teacher B appeared to relate to the research er most strongly on the shared experience of completing a dissertation, which often seemed to be the primary reason that she gave her continued consent to participate in the research project. The other goa l present throughout the
88 research process appeared to be a desire to im prove Student Bs behavior Teacher B often made statements that might indicate a strong pref erence for consultant directedness during consultation, saying such things as, Whatever you want, dear, in response to inquiries from the researcher about scheduling and conditions to be conducted. However, her statements during performance feedback meetings indicated a contra ry desire to have greater directedness as a consultee. For example, Teacher B often expressed frustration with the control condition and the challenge inherent to supplying a ttention to Student B every minut e, and would make statements such as, You just have to understand, its not realistic to get over there every minute. No teacher can do that on a daily basis. Attempts by the researcher to differentiate research procedures and daily teaching pr actices were unsuccessful in a ssuaging her concerns about the demands of the FA. Teacher B rarely allowed scheduling of m eetings longer than five minutes; thus, performance feedback sessions were typically brie f. As with Teacher A, performance feedback sessions occurred once per every four sessions on a variable ratio schedule, for the same reasons noted above. Performance feedback sessions with Teacher B consisted of reviewing graphic analysis of her procedural integrity, Student Bs data, steps implemented correctly, and steps implemented incorrectly. Occasionally, problem solv ing steps to address an issue in inconsistent implementation were reviewed. Most attempts to do so were met with a statement such as, At least you got data, right? Further, Teacher B often began scheduling and performance feedback meetings with statements such as, W hat do you want? While she did not appear to be angry or bothered by these meetings, she appear ed at all times to be overwhelmed by various demands upon her time, of which this project was another.
89 The blunt quality of her statements, though not necessarily an obstacle to the consultative process itself, made analyzing the content of her statements and proceeding accordingly difficult. Throughout the consultation, it was unclear to what degree she was di ssatisfied with the procedures, and which specific procedures were most bothers ome to her. Throughout FA implementation in the classroom, Teacher Bs statements did not reflect an increased understanding of the FA procedures. During sche duling meetings, and again immediately before conducting FA sessions, she asked to be reminded of the relevant procedures for the imminent conditions. Overall, there was some element of collaboration in consultation with Teacher B. She shared concerns about the procedures; howev er, efforts to validate her concerns and move into problem-solving steps were largely unsucce ssful. The skills she re ported gaining during the social validity interview were the first that she reported to the researcher. Though she gave positive ratings to the assessment procedure at th e conclusion of consultation, her statements throughout consultation reflected a pervasive diss atisfaction with the de mands upon her time. Teacher C Consultation with Teacher C began in January, 2008 and concluded in May, 2008. Entrance into consultation with Teacher C was, as with all the participating teachers, centered on the procedural aspects of affirming participati on, attaining consent, a nd identifying and obtaining consent from student participants. However, this phase seemed to persist longest, and was never fully replaced by other patterns of interacti ons. During problem identification, Teacher C identified the class in which Student C was pl aced as hyper. He reported that the class consisted of many students who were receiving remedial instructi on for low reading and/or math test scores. He seemed to make fixed, intern al attributions that th ese students behavior (including that of Student C) was endemic to their academic performance. Throughout consultation, Teacher C was minimally communica tive regarding his perceptions of the FA.
90 Thus, it is difficult to infer his goa ls for the consultation itself. It seems likely that the presence of the researcher at the impetus of the administ ration was perceived as a potential threat or a strong form of corrective feedback regarding his classroom management strategies. Thus, his primary goal appeared to be avoiding punishment from administration should he not cooperate. Additionally, he showed some interest in improving the classroom behavior of the target student. Teacher C readily set aside time to hold perf ormance feedback meetings. However, he frequently forgot about them or was not in his classroom when they were scheduled to occur. Performance feedback sessions with Teacher C we re conducted on a fixed ratio schedule of once every three sessions. Procedures during performance feedback were the same as were used with Teachers A and B. Teacher C occasionally asked fo r clarification about correct steps to be taken when he had made an error in FA implementa tion. He did not make statements about his understanding of the procedures or any inten tion to apply them to other students in the classroom. As is evident in his baseline da ta, Teacher C readily grasped the procedural requirements of each FA condition. Thus, consul tation was not organized around the concept of supporting him in understanding and applying these concepts to his classroom. Rather, most consultation interactions focused on next step s to be completed (e.g., scheduling sessions, conditions to be conducted), almost to the excl usion of other variables that may have been impacting his or Student Cs perfor mance during the classroom FA. Though all interactions with Teacher C regarding the FA were entirely consultant directed, attempts to broaden the focus of those intera ctions to classroom ma nagement strategies, behavioral principles, or general issues related to the challenges of teachin g were met with little to no response. Throughout baseline, training, and classroom FA sessions, Teacher C was primarily a passive participant in the consultation process. When attempts were made to engage
91 him in collaborative problem-solving, such as th rough a discussion of which conditions might fit best with upcoming classroom activities, he wa s noted to make statements such as, I dont know, it doesnt matter; tell me what you want to do. Thus, a collaborative approach was never taken that was specific to the FA sessions, and he seemed to continue to perceive decisionmaking responsibility for the FA to be solely th at of the researcher. However, outside of discussions about the FA, Teacher C engaged the researcher in collegial discussions about the general behavior of his classr oom, difficulties in management or humorous accounts of student behavior. These discussions were almost opposite in nature to those concerning the FA, in that while Teacher C appeared to ascribe to the rese archer decision-making expertise regarding the FA, he did not solicit information about appro aching classroom behavior more generally during other discussions. Rather, discussions of the FA versus outside conversations seemed to take place in parallel to one another.
92 Escape Teacher Attn TA TA Peer Attention Peer Attention Peer Attention Escape Escape Control Control Control NCR PA NCR TA NCR E 0 10 20 30 40 50 60 70 80 90 100 1234567891011121314151617181920212223242526272829303132333435 0 10 20 30 40 50 60 70 80 90 100 Peer Attention Escape Control Teacher Attention TA Teacher Attention Peer Attention Peer Attention Escape Escape Control Control NCR PA NCR TA NCR E 0 10 20 30 40 50 60 70 80 90 100 12345678910111213141516171819202122232425272829 0 10 20 30 40 50 60 70 80 90 100 Peer Attention Escape Control TA Teacher Attention Peer Attention Esc Control NCR TA NCR E NCRPA 0 10 20 30 40 50 60 70 80 90 100 1234567891011121314151617181920212223242526 0 10 20 30 40 50 60 70 80 90 100 Figure 3-1. Teachers procedural integrity during FA sessions Percent steps correct bL Training Classroom Teacher A Teacher B Teacher C bL Session Number
93 Figure 3-2. Student A: Responses per minute raising hand across conditions Figure 3-3. Student A: Responses per minute appropriate peer speech across conditions. 0 0.1 0.2 0.3 0.4 1 2 3 4 5 67 891011121314 15 1617 Session Number Teacher Attention Peer Attention Escape Control 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1 2 3 4 5 67 891011121314 15 1617 Session Number Teacher Attention Peer Attention Escape ControlRate per minute Rate per minute
94 Figure 3-4. Student A: Responses per minute disr uption across conditions. Figure 3-5. Student B: Responses per minute raising hand/ap propriate teacher speech across conditions. 0 0.2 0.4 0.6 1 2 3 4 5 6 7 8 9101112 Session Number Teacher Attention Peer Attention Escape Control 0 0.5 1 1.5 2 2.5 1 2 3 45 6 7891011121314 151617 Teacher Attention Peer Attention Escape ControlRate per minute Rate per minute
95 Figure 3-6. Student B: Responses per minute appropriate pe er speech across conditions. Figure 3-7. Student B: Responses per minute disr uption across conditions. 0 0.1 0.2 0.3 0.4 1 2 3 4 5 6 7 8 9101112 Session Number Teacher Attention Peer Attention Escape Control 0 0.2 0.4 0.6 0.8 1 1.2 1 2 3 45 6 7 8 91011 12 Session Number Teacher Attention Peer Attention Escape ControlRate per minute Rate per minute
96 Figure 3-8. Student C: Responses per minute raising hand across conditions. Figure 3-9. Student C: Responses per minute appropriate peer speech across conditions. 0 0.2 0.4 0.6 1 2 3 4 5 6 78 9 10 Session Number Teacher Attention Peer Attention Escape Control 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 1 2 3 4 5 6 78 9 10 Session Number Teacher Attention Peer Attention Escape ControlRate per minute Rate per minute
97 Figure 3-10. Student C: Responses per minute disruption across conditions. Figure 3-11. Student A: Frequency of appropriate peer speec h; within sessions analysis of Control and Peer Attention Control Peer Attention 0 2 4 6 8 10 12 3 6 9 1215 3 6 9 1215 3 6 9 12 153691215369121536912153 69 1215 3 691215 3 5 6 9 11 13 14 16 Time Increments Session Number 0 0.5 1 1.5 2 2.5 1 2 3 4 5 6 7 8 9 10 Session Number Teacher Attention Peer Attention Escape ControlRate per minute Frequency within 3 min intervals
98 Figure 3-12. Student A: Frequency of disruption; within sess ions analysis of Control and Peer Attention Figure 3-13. Student B: Frequency of appropriate peer speec h; within sessions analysis of Control and Peer Attention Peer Attention Control 0 1 2 3 4 5 36912 15 36912 15 36912 15 369 1215 369 12 15 3 6 9 1215 2369 10 11 Time Increments Session Number Control Peer Attention 0 5 10 15 20 25 3 6 9 1215 3 6 9 1215 3 6 9 12 153691215369121536912153 69 1215 3 691215 3 5 6 9 11 13 14 16 Time Increments Session Number Frequency within 3 min intervals Frequency within 3 min intervals
99 Figure 3-14. Student B: Frequency of disruption; within sess ions analysis of Control and Peer Attention Figure 3-15. Student A: Rais ing hand during sessions of 100% procedural integrity. 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 12 6 7 811 12 15 17 Session Number Teacher Attention Peer Attention Escape Control Peer Attention Control 0 2 4 6 8 10 12 3 6 9 12 153 69 12 15 36912153 69 121536912 15 36 91215 2 3 6 9 10 11 Time Increments Session Number Frequency within 3 min intervals Responses per minute
100 Figure 3-16. Student A: Appropriate peer speech during sess ions of 100% procedural integrity. Figure 3-17. Student A: Di sruption during sessions of 100% procedural integrity. 0 0.1 0.2 0.3 0.4 0.5 0.6 1 2 6 7 8 11 12 15 17 Session Number Teacher Attention Peer Attention Escape Control 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1 2 6 7 8 11 12 15 17 Session Number Teacher Attention Peer Attention Escape ControlResponses per minute Responses per minute
101 Table 3-1. Student A FA results; Teacher A procedural integrity Participant Session Condition Rate of responding Teacher integrity Weaknesses in integrity A 1 TA RH .27/min APS .07/min Disr .07/min 100% n/a 2 Escape RH .2/min APS 0/min Disr 0/min 100% n/a 3 Control RH 0/min APS .33/min Disr .13/min 95% Noncontingent Teacher Attention (NCTA) 5 Control RH 0/min APS .6/min Disr 1.33/min 85% NCTA; Prompting peer to deliver NC attn (NCPA) 6 PA RH 0/min APS .53/min Disr 1.33/min 100% n/a 7 TA RH .07/min APS .33/min Disr .27/min 100% n/a 8 Escape RH .07/min APS 0/min Disr .07/min 100% n/a 9 PA RH .07/min APS .6/min Disr .6/min 85% Prompting peer to deliver contingent attn. 10 Escape RH .07/min APS .07/min Disr 2.33/min 92% Correctly end escape interval; prompting peer not to respond to APS 11 Control RH .13/min APS .2/min Disr .53/Min 100% n/a 12 Escape RH .13/min APS 0/min Disr .07/min 100% n/a 13 PA RH .27/min APS .67/min Disr .67/min 87% Prompting peer to deliver contingent attn.; ignoring RH 14 PA RH .07/min APS .33/min Disr .2/min 91% Prompting peer not to deliver NCPA 15 TA RH .07/min APS .07/min Disr .33/min 100% n/a
102 Table 3-1. Continued. Participant Session Condition Rate of responding Teacher integrity Weaknesses in integrity 16 Control RH 0/min APS .07/min 98% NCTA 17 TA RH. .31/min APS .08/min 100% n/a
103 Table 3-2. Student B FA results; Teacher B procedural integrity Participant Session Condition Rate of responding Teacher integrity Weaknesses in integrity B 1 TA RH 0/min APS 0/min Disr .27/min 100% n/a 2 PA RH .13/min APS .27/min Disr .27/min 64% Correctly prompting peer to ignore disruptions; ignoring RH 3 Control RH .13/min APS .13/min Disr .2/min 39% NCTA, prompting peer to deliver NCPA, to ignore APS 4 Escape RH .47/min APS .07/min Disr 1/min 79% Ignoring RH; prompting peer not to provide NCPA 5 TA RH .4/min APS .07/min Disr 1/min 93% Ignoring disruptive behavior 6 Control RH .33/min APS .07/min Disr .67/min 59% Ignoring RH; providing NCTA; prompting peer to provide NCPA, to ignore APS 7 Escape RH 0/min APS 0/min Disr .4/min 57% Ignoring disruptive behavior 8 TA RH .13/min APS 0/min Disr .47/min 100% n/a 9 PA RH 0/min APS .07/min Disr .4/min 100% n/a 10 Control RH .27/min APS .13/min Disr .33/min 58% Ignoring disruptive beh; delivering NCTA, prompting peer to deliver NCPA, to ignore APS 11 PA RH 0/min APS 0/min Disr .09/min 100% n/a 12 Escape RH .08/min APS 0/min Disr .08/min 67% Ignoring RH
104 Table 3-3. Student C FA results; Teacher C procedural integrity Participant Session Condition Rate of responding Teacher integrity Weaknesses in integrity C 1 Escape 0 in all conditions 100% n/a 2 TA 0 in all conditions 100% n/a 3 PA RH .07/min APS 1.93/min Disr 1/min 89% Prompting peer to respond to APS 4 Control 0 in all conditions 94% Providing NCTA 5 TA 0 in all conditions 100% n/a 7 PA RH 0/min APS .87/min Disr 1.73/min 98% Correctly prompting peer to ignore disruption 8 PA RH 0/min APS .13/min Disr .13/min 100% n/a 9 Control RH 0/min APS .47/min Disr .53/min 97% Correctly prompting peer to provide NCPA when scheduled 10 Escape RH 0/min APS 0/min Disr 2.06/min 100% n/a
105 Table 3-4. Teachers responses to social va lidity interview (1= not at all, 7 = very) Question Teacher A Teacher B Teacher C How comfortable were you with this study? 5 6 4 How time intensive was it for you to participate in the training sessions? 4 5 5 How comfortable were you with the training sessions? 6 6 6 How comfortable did the peer appear with the assessment (FA) sessions? 6 (beginning) 3 (end) 6 6 How time intensive was it for you to participate in the assessment (FA) sessions? 6 5 1 How disruptive was this assessment to your classroom? 4 (beginning) 6 (end) 2 4 How much did you have to change your classroom routine to allow the target student and peer to participate in the assessment procedures? 5 3 4 How disruptive was this assessment to the other peers in the classroom? 4 (beginning) 6 (end) 2 1 How enthusiastic would you be about another student participating in this assessment? 6 6 5 How willing would you be to participate in this assessment again? 5 4 3 How useful was the information obtained from this assessment to you and the target student? 7 5 5 Overall, how would you rate this assessment procedure? 6 6 4
106 CHAPTER 4 DISCUSSION Overview of Findings The results of this study indicate that functional analyses m ay be implemented with high integrity by middle school teachers working with students who dem onstrate disruptive behavior. Further, those analyses may yi eld useful information relevant to intervention planning for students who demonstrate persiste nt disruptive behavior in the classroom setting. However, there are limitations inherent to this research st udy and the application of its findings to general education middle school settings, both of whic h are discussed below. Finally, additional questions raised by this re search are also discussed. Teachers Procedural Integrity Two of the three particip ating teachers (Tea chers A and C) demonstrated consistent implementation of FA conditions during both the training and classroom implementation phases. Teacher B was able to demonstr ate acceptable levels of integr ity during the training phase, but did not maintain those levels during all cond itions of the classroom-based FA. Prior investigations of the ability of persons unfamiliar with the theoreti cal principles of FA to conduct such assessments (Iwata et al., 2000; Moore et al., 2002; Wallace et al., 2004) have not examined the ability of educators to carry out a multielement design within the classroom setting. This study did so by measuring teachers implementation of four conditi ons (three experimental, one control) in an extended multielement design. W ith the exception of Teacher C, the results are consistent with those obtained during previous re search in this area (Iwat a et al., 2000; Moore et al., 2002; Wallace et al., 2004), in that teachers did not demons trate acceptable levels of procedural integrity until after training and perf ormance feedback procedures were introduced. However, Teacher Bs data pattern is discorda nt with the results obtained by Wallace et al.
107 (2004), in that she did not conti nue to demonstrate high levels of integrity during the classroombased FA. The provision of performance feedba ck at an average of every four sessions was sufficient to maintain levels of integrity demonstrated by Teachers A and C, but did not appear to be sufficient for Teacher B. Continuing investigations of performance feedback schedules have suggested that there is individua l variability in the amount of pe rformance feedback necessary to sustain teachers integrity to a selected pro cedure (Witt et al., 2004; Codding et al., 2005). Students Behavior during FA Each studen ts response to the FA conditions differed significantly. Responding demonstrated by Student A demonstrated evidence of a possible reinforcing effect of contingent peer attention for appropriate be havior, and unexplained effects of peer attention on disruptive behavior. Though low levels of integrity make it impossible to draw conclusions from Student Bs FA results, rates of disruption suggest a possible treatment effect of the extinction procedures present during all FA conditions. Finally, Student Cs data is also unable to be interpreted due to threats to its internal validity, but early rate s of responding, demonstrated prior to the apparent interfering effect of severe reac tivity, indicate the possibility of a peer attention function. Throughout each FA, there appeared to be uncontrollable, intervening factors affecting the outcomes of the FA that were not specifi c to the FA manipulations themselves. Consultation and Social Validity Consultation with each teacher differed greatly, in ways that may have been reflected in the data generated by each teacher/student dyad. However, the use of consultation procedures appeared to support the implementa tion of FA within th e classroom. It greatly benefitted the collection of information relevant to each target student, especially during the initial planning stages. Furthermore, it provided a context through which to work with each teacher in order to support his/her procedural integr ity. As has been noted, perfor mance feedback was a successful
108 strategy in maintaining the procedural integrity demonstrated by two out of three participating teachers. Social validity data obtained from a ll three teachers indicate that they found the procedure to be useful, and would be especially useful at the beginning of the academic year. Teachers mentioned concerns about the benefit of the information obtaine d relative to the level of intrusiveness required to conduct an FA. However, they all stated that they would support similar future efforts to be taken with a student demonstrating disruptiv e classroom behavior. Across participants, social validity ratings indi cated that they found this assessment strategy to acceptable and relevant to the behaviors demonstrat ed by each target student. Student reactivity as an indicator of social validity suggests that students were uncomfortab le with the assessment procedures, particular the length of the FA and the presence of an observer. Interpretation of Findings There are several findings that m erit further explanation. First, te acher integrity was not consistent across all participants The lack of integrity throu ghout much of the classroom FA implemented by Teacher B requires additional consideration. Second, though FA findings provided additional evidence regarding the best means through which to intervene with each students behavior, a lack of clarity pervaded all of the FA da ta, which must be examined in greater detail. Finally, though general proce dural guidelines of BC were followed, each consultation differed considerably. The variations in consultative interact ions, and their possible effects on the results of this inve stigation, should be reviewed. Teacher/Student A W ith respect to the level of collaboration present duri ng consultation, patterns of interaction, and self-reported sa tisfaction with the research, c onsultation outcomes with Teacher A were positive. The highly collaborative natu re of the consultation be tween Teacher A and the researcher appeared to benefit other aspects of the assessment. First, Teacher As levels of
109 integrity were high throughout the evaluation. A dditionally, the FA data obtained with Student A provided strong evidence of the reinforcing effects of peer attention on hi s rates of appropriate classroom behavior. Of the three students assessed in this investigation, the FA results obtained with Student A are arguably the most clear. However, despite this clarity, there were some unexplained rates of responding during the FA. Mo st notably, Student A was more likely to demonstrate disruption during peer attention and control conditions. Prior to the FA, Teacher A hypothesized that disruptive behavior was maintain ed by peer attention; appropriate behavior was also maintained by peer atte ntion during the FA Thus, it is perhaps surprising that Student A should continue to engage in disruptive behavi or when a) peer attention was contingently available for appropriate behavior as in the peer attention cond ition, and b) peer attention was continuously and noncontingently available dur ing the control condition. One possible explanation for this effect may be that the presen ce of some form of peer attention, no matter the contingency, served as either an antecedent to disruption or a discriminative stimulus within the classroom environment. If peer attention functioned as a discriminative stimulus for the availability of more peer atte ntion, the implication would be that the environmental conditions established during the FA were not consistent and/or powerful enough to overcome the influence of Student As learning history. In this case, once peer attention was delivered once, contingently or not, it served as an indicator that more peer attention would be available for additional behaviors. Because disruption was most likely the most frequently reinforced response in Student As learning history, that behavior would be the most likely to occur under similar conditions. Indeed, this appears to be the case. Fi gure 3-11 indicates that while appropriate behavior came under the stimulus c ontrol of peer attention conditions and responded to the lack of contingent reinforcement during control conditions; in Figu re 3-12 the opposite is
110 true. Disruption during peer attention conditions was extingu ished over time; however, during control conditions disrupti on continued to occur at a high rate This finding suggests that some element(s) of the control condition was sim ilar enough to Student As previous classroom experiences to continue to indu ce high rates of disruption. As previously stated, the FA conducted with Student A was the only assessment in this investigation to produce consistent findings. Howe ver, there is, nevertheless, some variability within those findings. The impact of Teacher As integrity on the variability of findings is unknown. For example, the apparent reinforcing f unction of peer attention on appropriate peer speech is further evidenced in Figure 3-16, in wh ich the one peer attention session conducted at 100% integrity produced clearly elevated rates of appropriate peer speech during that condition. Similar findings are observed regarding disrupt ion in Figure 3-17. Thus, the impact of procedural integrity remains unknown. Given that those conditions that produced the highest rates of responding, peer attention and control, were also those in which procedural integrity was typically lower than 10 0%, it is perhaps worth noting that the outcomes of the FA conducted with Student A may have been different had pro cedural integrity been equal to 100% during all sessions. Teacher/Student B The lack of a strong consultative relations hip, especially early on in the assessm ent process, appears to have been detrimental to the outcomes obtained with this dyad. One way to improve this relationship might have been to mo dify the performance feedback procedures to involve more frequent contact and additional re hearsal opportunities. For example, performance feedback sessions could have been carried out following each class period in which FA sessions were conducted. Additional rehearsal would have involved similar types of role playing practice as was used during the baseline and training phases of the assessment. Conducting additional
111 meetings with more intensive feedback proced ures may have also served to strengthen the consultative relationship. Typically, during meet ings, Teacher B was noted to be forthright about her concerns and direct in her stat ements. Thus, providi ng her with additional opportunities to discuss issues of concern may have engendered greater trust and collaboration between her and the researcher. Other modifications to the performance feedback procedures that may have been helpful include providing written performance feedback and using negative reinforcement procedures (DiGennaro, Martens, & McIntyre, 2005). Using such a procedure, written feedback would have been provided to Teacher B after each day of running sessions. Additionally, a meeting would be held to review missed steps unless the proc edures were implemented with 100% integrity. This approach has been shown to successful in in creasing and maintaining teachers integrity to a chosen procedure, and may be more successful th an other forms of performance feedback alone (DiGennaro, Martens, & Kleinmann, 2007). This procedure may have been especially useful with Teacher B in communicating that her professional time was valued by the researcher throughout the consultation process, and in prov iding her with a permanent record of her performance to be referred back to prior to additional sessions. Because consultation with Teacher B was not successful in maintaining high levels of procedural integrity throughout the FA process, additional consideration should be given to the importance of establishing a collaborative relation ship with some consultees prior to conducting such an assessment. Indeed, the volume of the research literature concerning teachers assessment and intervention integr ity demonstrates the ubiquitous nature of concerns about teachers ability to modify classroom procedures in order to change student behavior. The findings of this study suggest that not only is it crucial for teachers to demonstrate strong
112 integrity in order to implement classroom-bas ed FAs, but that it may be the nature of consultation itself that inhibits performance during measures of classroom integrity. Perhaps, with some teachers, it is necessa ry to engage in an extended period of problem identification and goal-setting, in order to ensure that consultant and consultee have a shared sense of the problem and agreed upon procedures for how to assess and address it. In some cases, this may involve teachers collecting data about students and/or th eir own classroom behavior. For others, it may involve several more focused conversations about the problem beha vior and a critical examination of the classroom factors affecting it. In most cases, it is perhaps both strategies that would be effective. While this may appear to be a more time-intensive approach to problemsolving, greater time spent on understanding the problem and goals for consultation at its inception would almost certainly yield more effici ent assessment and inte rvention procedures, as well as require fewer modifications to the proc edure once it is implemented. Most importantly, however, spending additional time in developing co llaboration prior to classroom use of any procedure is likely to improve consultative outcomes, including the integrity with which procedures are implemented by teachers. Teacher/Student C As with Student B, the FA outcom es obtai ned with Student C we re inconclusive. However, the implementation of the FA with Teacher and Student C was unique. Two issues appeared to affect the outcomes of this assessment: the absence of collaboration between the researcher and Teacher C, and the severe reactivity demonstrated by Student C. The latter factor appears to have had the most direct effect on the student behaviors obser ved during the FA. While some reactivity was demonstrated by all thr ee student participants in this study, Student C was the only student participant whose reactivity a ppeared to affect her ra tes of both appropriate
113 and disruptive behaviors during the FA. There are several stra tegies that may have reduced reactivity to this procedure. The researcher did not interact with each of the target students in an effort to limit reactivity. However, each student demonstrated some degree of reactivity nonetheless. A better strategy might have been to make students awar e of the researchers presence and the FA conditions following the student PII. This woul d have provided each student an opportunity to ask questions about the process, address his/he r concerns, and gain a better sense of his/her enthusiasm (or lack thereof) for the research itsel f. Further, as student s inevitably became aware of the FA conditions and the fact that the researcher was observi ng them, it may have been most prudent to obtain target students assent to participate as we ll. Though there is debate about the degree to which children and adolescents can tr uly give informed assent, and the conditions under which it may be acceptable not to obtain assent (such as in the intended use of this FA procedure), there are alternative protocols that may be used to measure assent. These include allowing informed dissent, in which the student is informed of the research procedures and can withdraw participation, and caref ul monitoring by participating caretakers to determine whether a particular behavior might indicate dissent (Lewis, 2002) Certainly, in the case of Student C, the introduction of aggressive beha vior in Session 10 was taken as an indicator of dissent. However, more careful monitoring of her behavior prior to that session, or having presented her with the option of withdrawi ng participation, might have prevented the aggression she demonstrated toward the peer confederate. Theoretical Implications of Findings The prim ary goal of this research project was to determine whether general education middle school teachers are able to implement FA with high integrity within the classroom setting. The results indicate that while it is pos sible for teachers to l earn these procedures under
114 carefully controlled conditions, it may not always be successfully applied to the classroom itself. Further, the quality of the consultative re lationship through which teachers learn and are supported in conducting FA may im pact their procedural integr ity. However, it is unknown which and to what extent specific elements of consultation are needed in order to produce high integrity to FA procedures. Performance feedba ck was enough to sustain procedural integrity for all teachers during training, and fo r two teachers during the classroom FA. This finding is in accordance with previous investigat ions of performance feedback that also found that procedure was highly effective during consultation. Howeve r, it also indicates that there are some limitations to the use of performance feedback, pe rhaps the most important of which is that a certain form or frequency of delivery may not be universally applicable, even though certain consultations are similar in nature. Instead, it may be necessary to rely on problem identification statements and early consultative interactions in order to determine an optimal form and intensity of providing feedback. A secondary goal of this study was to dete rmine the degree to which FA conducted in classroom settings yielded meaningful results regarding students behavior. Modifications made to the FA procedure in order to make it amenab le to a middle school classroom setting highlight relevant conceptual issues. Appropriate behavior rather than disruptive behavior, was reinforced during the FA in order to circumvent the logistic al and acceptability issues that would be raised by temporarily increasing rates of disruption w ithin an instructional setting. However, this modification was made according to the assumpti on that differing topographies of behavior are maintained by the same reinforcing functions (Borrero and Vollmer, 2002; Frea and Hughes, 1997). This assumption has been experimentally examined in recent research with adolescents demonstrating severe problem behavior in an institutional setting, and support has been found for
115 the continuity of function across topographies of behavior (Perrin, Sellers, Badley, & Marcus, 2008; Sellers & Strickland, 2008). Thus, for Student A, the finding that appropriate peer speech was maintained by peer attention would argue that disruption is also maintained by peer attention. Indeed, this statement was supported by Teacher As statements. But elevated rates of disruption during peer attenti on conditions, as previously di scussed, indicate that other phenomena were exerting an unknown influence on disruption that was not measured in this investigation. Thus, this study provides some evidence in support of th e theoretical assumption that function is the same across different behaviors. However, the current findings also suggest that two or more variables may be operating to maintain sophist icated behavioral repertoires, especially in typically de veloping students, in ways that were not measured by this investigation. Practical Implications of Findings In nearly thirty years of use, FA has been conducted with m any indi viduals and the various settings in which they are served. The many wa ys in which it has been applied is a clear and convincing argument for the util ity of FA in determining be havioral function and planning intervention. However, this does not mean that FA is necessarily applicable to all contexts, individuals, or beha viors. In applying FA to general edu cation settings, one of the goals of this research was to evaluate how useful this pro cedure was to an environment in which it had previously not been introduced. In doing so, the utility of the outcomes must be critically analyzed in terms of its efficiency, requi red effort, and impact upon the instructional environment. In this study, the results of the FA were only conclusive for one student. Furthermore, the results of conducting FA with Students B and C highlight key issues that a ppeared to affect the FA outcomes. Specifically, those are that consultation must exist, be highly collaborative, and carefully monitored in order for teachers to im plement FA contingencie s within the classroom
116 context. Furthermore, students own reactivity to observation and e xperimental manipulation must be considered and accounted for. These factors indicate that, whil e FA may produce valid outcomes for some students, the time, effort, a nd skill involved remains extremely high. Taken together, this is a strong indica tion that school-based assessments of problem behavior should be carefully planned to achieve an optimal balance of these factors. While it is possible to implement FA in a classroom environment, it ma y not always be feasible or necessary. The demands upon teachers time and instructional re sources must be considered relative to a particular student's needs. In some cases, a descriptive FBA may be sufficient to develop intervention goals and effective strategies (Ervin et al., 1998). When stude nts behavior does not respond as expected to function-based interventions developed on the basis of descriptive observations, experimental observations may be merited. The goal of BC is not only to improve outcomes fo r the consultee and clie nt relative to the current problem, but to strength en the ability of the consultee to approach similar future problems (Bergan & Kratochwill, 1990). In this research, teac hers ability to handle future problems was only indirectly measured during the social validity interview by asking about skills they acquired through participation in the FA. Nevertheless, each teacher provided important insight into the potential of th ese procedures to impact his/her future classroom management. Perhaps unsurprisingly, Teacher A indicated the greatest number of skills attained, and reported that she had already been thinking about ways in which to apply FA principles to her classroom. Teacher B indicated some generalization of these skills, while Teacher C reported only that he had learned the importance of ignor ing disruption for some students. Each of the skills reported by the three participating teachers, however, was th eoretically consistent with the principles of behavior analysis, from which FA is derived. Thus, several importa nt observations can be made.
117 First, Teacher C demonstrated extremely high levels of procedural integrity, but did not subsequently indicate a thorough understanding of the application of FA principles to his classroom. On the contrary, Teacher B had variab le levels of procedural integrity within the classroom, but reported having acquired several new management strategies as a result of her participation. Teacher A demonstrated both high integrity and extensive self-reported generalization of theoretical knowle dge. Variables that might be impacting these results include the amount of time spent conducting the FA (Teach er A participated in the highest number of training and classroom implementation sessions; Teacher C participated in the least), years teaching (Teacher B may have augmented her understanding of FA principles, despite her low levels of integrity, by applying previous classroom experience), and, of course, the level of collaboration present during consultation. The high level of collaboration between Teacher A and the researcher undoubtedly resulted in the highest am ount of procedural knowledge transferred, despite the effort to standardize the transmission of knowledge through performance feedback sessions. The level of theoretical knowle dge teachers gained offers a potential alternative to the time-consuming process of conducting an extended multi-element FA. Thus, it is an important adjunct to an investigation that assesses the use of these procedures in general education classrooms. If teachers abilities to respond to problem behavior can be addressed through consultation that focuses on the acquisition of re levant classroom management skills, concerns related to the feasibility of conducting an FA may be circumve nted. Providing teachers with this knowledge may also remediate the noted limitations in implementing interventions derived from descriptive functional assessment techniques, including the possibilities of low intervention integrity or that the identified function is inco rrect (Ervin et al., 2001; Wood et al., 2007). If
118 teachers have a more general understanding of the founding principles of FA, such as behavioral function and schedules of reinforcement, their ability to accommodate function-based intervention is likely to be increased. Lastly, the present findings suggest important ramifications for practitioners acting as consultants in school settings. Not only shoul d consultants carefully consider the type of assessment to be conducted and the level of theo retical knowledge that might be beneficial to teachers, but they should also evaluate the consultation itself over the course of its development. Of course, careful monitoring is a hallmark of BC (Bergan & Kr atochwill, 1990). However, the interaction processes through which positive consulta tion outcomes are attained are as yet poorly understood (Graham, 1998). The findings of this study suggest that the level of collaboration remains an important indicator of the success of a given consultation, de spite the presence of highly technological procedures for sustaini ng intervention/assessment procedures (e.g., performance feedback). Thus, c onsultants should pay close atten tion the consultees response to each phase of the consultation, and be prepared to make modifications as necessary. Limitations Teachers Procedural Integrity One of the prim ary limitations inherent to th e study of teachers procedural integrity was introduced by the use of a multiple baseline design. While this design is preferable for investigations in which the independent va riable cannot be removed once it has been implemented (e.g., training in FA cannot be remove d), there may be difficulty in demonstrating internal validity of the experi mental manipulation. Experimental control is demonstrated when changes in the dependent variable only occur upo n the introduction of th e independent variable (Kazdin, 1982). Thus, repeated dem onstrations of this change provi de evidence for the effect of the independent variable in a multiple baselin e design. In this investigation, Teacher C
119 demonstrated a change in observed levels of procedural integrity during the baseline phase, prior to the introduction of training a nd performance feedback. Though this effect has previously been explained, it nevertheless poses a threat to th e internal validity of the results observed with Teacher C. Furthermore, the simultaneous impl ementation of training and performance feedback does not reveal which element is primarily respon sible for the change observed in the integrity demonstrated by Teachers A and B. It may be th at one of the two elements is the responsible variable, or that each has a specific potential impact upon teachers procedural integrity during FA. As has been discussed, the impact of cons ultation upon teacher integrity outcomes is unknown. There appears, however, to have been an effect of the consultative relationship on the integrity with which teachers implemented FA pr ocedures. Thus, the assessment of training and performance feedback in terms of its ability to sustain proced ural integrity in middle school classrooms was affected by an intervening vari able. Perhaps if the study had been conducted with teachers with whom a highly collaborati ve consultative relationship had already been established, the results w ould be clearer as to th e precise impact of training and performance feedback. Conversely, given that consultation interactio ns have the potential to interfere with prolonged classroom-based research, perhaps the implementation of training, and particularly performance feedback, could have been more sens itive to the preferences of each consultee. Students Behavior During FA It is unknow n to what degree all sources of reinforcement were controlled during FA conditions. Thus, observed eff ects on student behavior may have been compromised by the presence of other sources of reinforcement throughout the FA. While noncontingent attention was measured specific to the condition being implemented, other sources of noncontingent attention were not. For instance, there is no way to state with certainty that there was not peer
120 attention available during teacher attention conditions. While anecdotally it was noted that most sources of reinforcement appeared to be well-controlled, the lack of systematic observation of that phenomenon precludes any cert ainty. Another classroom factor that was not systematically observed was the presence/absence of other prefe rred peers. Because the specific behavior of peers other than the confederate was not obser ved, it is unknown to what degree specific behaviors demonstrated by specifi c peers might have influenced target students behavior. Given the high rate of responding observed during peer attention conditions, this may have had a strong impact on demonstrated rates of behavior. Other potential influences on target student behavior include classroom activity, distal antecedents and consequences for target behavi or, and the strength of each students previous contact with the learning environment. While efforts were made to control classroom activity by specifying the type of academic demands that should be present during each condition, ultimately teachers selected th e activity during which FA conditions would be conducted. As previously discussed, during peer attention and control conditions, peer speech was typically permitted throughout the classroom. Thus, activity type necessarily differed (it was an activity students could work on together), and the ove rall volume of the classroom was greater. Furthermore, the impact of events occurring prior to or after target behaviors, but outside of the observation period, is also unknown. Had the target students been more directly involved in the assessment process, they could have been interv iewed about related expe riences before and/or after the observation periods. For example, prior to running sessions, students could have completed a brief written interview about events o ccurring that day that ma y have affected their classroom behavior (e.g., disagree ment with a friend, being teased, bad grade, being sick, etc.). Occasional interviews might also have measured a ny distal consequences related to participation
121 in FA conditions, such as being teased by peers, not being able to intera ct with preferred peers during conditions, or any eff ects on work completion. This may have supplied information allowing the FA conditions to be modified in orde r to produce higher internal and social validity (as evaluated by the participating students). Finally, the results also indicated that students previous experiences in each classroom environm ent were not completely controlled for within the FA. This may have been most evident in ra tes of disruption, which was not fully eradicated even when appropriate behavior was maintained by reinforcers available within the FA. Implications for Future Research In relation to the poin ts described above, there are several areas of additional research indicated by this study. More knowledge is needed about the condi tions under which this procedure is most useful, including form and frequency of problem behavior, teacher characteristics, and the nature of the consulta tive relationship within which the procedure is supported. If this procedure is only useful for fr equent and persistent forms of disruption, other assessment protocols remain to be developed for infrequent, highly intens e classroom disruptions (e.g., severe aggression; Sterling-Turner et al., 2001). As teacher char acteristics, including consultative goals and perceptions of problem beha vior, appeared to impact the integrity with which teachers implemented the procedures, additional information is needed about the conditions under which teachers are likely to dem onstrate high integrity. Finally, there remains a dearth of research delineati ng the technology through which str ong consultative outcomes are obtained. While this investigation extends the literature by describing the intensity of consultation necessary to train teachers in a soph isticated assessment procedure, it remains to be seen whether high levels of collaboration can be obt ained in all consultations or whether that is necessary in order to obtain strong outcomes in all cases. Indeed, Teacher C demonstrated high procedural integrity in the absence of any collaborative relationship. However, despite
122 demonstrating slightly more collaboration duri ng consultation, Teacher B did not demonstrate high integrity in the classroom FA. Thus, additiona l inquiry is needed into what conditions merit that time is spent on developing high levels of collaboration, and when adherence to the procedures is sufficient to produce positive outcome s for target students. Finally, some estimate is needed of the relative contribution of bot h training and performance feedback to teachers sustained procedural integrity. If one procedure is adequate to improve and maintain integrity during FA, valuable time may be saved for both teachers and consultants in the implementation of FA. The FA of students behavior during this st udy also raised importan t questions regarding the nature of classroom behavior. First, the i ssue of how appropriate peer speech and disruption are related remains to be adequately explained. It may be that these behaviors have developed as a response class for students who demonstrate disruptive behavi or, which may be functionally defined as behaviors that can be used to access peer attention. If that is the case, then it may be harder to extinguish rates of disruption while si multaneously increasing a ppropriate peer speech than if each behavior operated independently of one another (Sajwaj, Tw ardosz, & Burke, 1972). More research into the relation of appropriate and inappropriate forms of seeking peer attention by general education adolescents is clearly merited, as this will undoubtedly remain a primary reinforcer for much of this population. Th e high rates of respondi ng observed during peer attention conditions also raise other questions. For example, when interviewed, all of the participating teachers indicated that they used or had tried reprimanding the target student in an effort to decrease disruption. Furthermore, such attempts were unsuccessful, indicating that they either had reinforcing effects upon the behavior or did not adequa tely address other sources of reinforcement. One related area for research w ould be to investigate wh ether the provision of
123 teacher attention contingent upon the demonstration of problem behavior increases the opportunity for students to receive peer attention. For example, it may be that during periods of teacher reprimands or conversations about problem behavior, di sruptive students also receive increased amounts of peer attention. This peer attention could take the form of teasing, laughing, or otherwise commenting on the target students behavior. An investigation of this phenomenon in a classroom setting might apply conditional pr obability technology to measure the contingent relationship of teacher reprimands and peer attention, thus providing a more inclusive measure of the multiple forms of reinforcement potentially maintaining problem behavior. Finally, the variety in rates of respondi ng across conditions demonstrated by Students A and B may also indicate that those students be havioral repertoires we re not adequately s upported by the artificial division of sources of reinforcem ent. Indeed, the maintenance of elevated rates of responding suggest that the control condition might be more representative of the classroom context, in which there are often many forms of noncontinge nt reinforcement, than other conditions examined in this investigation. Further, it is likely that a variety of sources of noncontingent reinforcement during that condition acted as di scriminative stimuli for multiple behavioral topographies maintained by all or several availa ble reinforcers. Thus, perhaps future FA research in middle school classrooms should in clude not only individual assessments of behavioral function, but conditions in which mu ltiple potential functions are combined (e.g., escape-to-attention; Mueller et al., 2005). Summary The results of this investigation indicate that FA is a useful and feasible procedure for identifying behavioral function in general educati on middle school settings. Further, using FA to assess desired appropriate behaviors led to increases in the observed rate of appropriate classroom behavior, and participating teachers vi ewed these procedures as acceptable for their
124 classrooms. However, FA and the consultativ e procedures necessary to sustain it in the classroom setting is extremely time intensive, and may not decrease rates of disruption demonstrated by participating students. Add itional consideration s hould be given to the consultative context in which these procedures mu st occur, including factor s likely to affect the integrity with which teachers implement FA. Re latedly, careful considera tion should be given to the appropriateness of this procedure, given th e identified problem behavior, student cooperation, teacher goals, and available resources. Additiona l research is needed to identify the conditions under which these procedures are most likely to be successful, as well as to the unique characteristics of middle school classrooms that li kely affect the rate of disruptive behavior demonstrated by students.
125 APPENDIX A TEACHER RECRUITMENT FLYER Attention Middle School Teachers: Are you interested in treating problem behaviors in your classroom with a focus on appropriate behavior? Would you like to be able to design and use interventions specific to your classroom? Would you like to work in consultation with another professional? If so, please contact Liz McKenney. I am working on a research project to develop a consultative process to assess and treat problematic behavior in middle school classrooms. And I would love to work with you! You can contact me at: email@example.com This research has been approved by the Institutional Review Board at University of Florida and by P.K. Yonge. Participation in this project is completely voluntary.
126 APPENDIX B INFORMED CONSENT TEACHER PARTICIPANTS Dear EducatorI am a graduate student at the University of Florida, under the supervision of Dr. Nancy Waldron, conducting research on ways to help teachers identify a nd intervene with problematic behavior in their classrooms. The purpose of the study is to develop an in-class procedure you and other teachers can use to identify and addres s factors that may help to support appropriate behavior for students who disrupt the classroom. The results of th is study have the potential to help you address problem behaviors you currently s ee in your classes, and they may help future teachers address problematic behavior in thei r classrooms. I am asking you to participate because you have been identified as an edu cator interested in expanding your teaching techniques for addressing behavioral problems. As part of this study, Id like to ask you to identify student s in your class who regularly demonstrate disruptive behaviors youd like to address. Students you nominate will then be asked for consent from their parents to particip ate. For each participa ting student, I will work with you to identify the problematic behavior s youd like to replace with an appropriate behavior. For example, you may indicate that disruption (talking wit hout raising ones hand, talking while the teacher is talking) is a probl ematic behavior for a student in your class. Together, we will set up several conditions throug h which to evaluate what may be maintaining, or reinforcing, each students behavior. In each of the different conditions, we will identify ways to provide reinforcers for the occurrence of a desired and appropriate replacement behavior. These reinforcers are not unlike those that natu rally occur in classroom s, and include your attention to the identified student, a peers at tention to the identified student, and allowing the identified student to escape from an academic ac tivity for a short period of time. The rate of each students behavior, or number of times they engage in the behavior during a certain amount of time, will be calculated. We will then compare behavior in each of the conditions to determine which condition best supports appropria te behavior. I understand that teachers face a variety of challenges and unforeseen dilemmas every day. Therefore, I will observe in your classroom as you are conducting these sessions, in order to determine how easily you can implement these conditions within your classroom. I will also provide and seek feedback about ways in which we can improve the analysis. This study has the potential to provide you with valuable information about how students behavior is maintained and can be altered. Wh en we are able to es tablish the function of students behaviors, we can work together to develop interventi ons to improve individual student behavior. There are no known ri sks to the participants. For the protection of your privacy, your name will be changed when the data is reported. Data generated by you about each student will be kept c onfidential to the extent provided by the law. You have the right to withdraw consent for your participation at any time without consequence. If you have any questions about this research protocol, please contact me at 246-1480, or my faculty supervisor, Dr. Waldron, at 392-0723, ext. 232. Questions or concerns about your rights
127 as a research participant may be directed to the UFIRB office, University of Florida, Box 112250, Gainesville, FL 32611, (352) 392-0433. Elizabeth L.W. McKenney, M.Ed. I have read the procedure described above. I voluntarily give my cons ent to participate in Elizabeth McKenneys study of teacher-led, in-class analyses of appropriate behavior. I have received a copy of this description. _______________________ ________ Participant signature Date
128 APPENDIX C TEACHER PARTICIPANT DEMOGRAPHIC FORM The following is a brief set of questions asking about your ex periences as a teacher. Completing this form provides me with information about yo ur previous experience that may be helpful in designing and describing this re search. Thank you for taking the time to complete this form accurately. Teachers Name: Degrees Earned: Additional Certifications Earned (if any): Years of Full-Time Teaching (after internship): Preschool Kindergarten 1st grade 2nd grade 3rd grade 4th grade 5th grade 6th grade 7th grade 8th grade 9th grade 10th grade 11th grade 12th grade Please describe any previous experience working with students with disruptive behavior and/or disabilities:
129 Have you ever participated in a classroom-based assessment of a students problematic behavior? If so, please describe this experience.
130 APPENDIX D INITIAL STUDENT PARTICIP ANT IDENTIFICATION FORM This research project is intended to investigate classroom factors that increase appropriate behavior in students with disruptive behavior. The following questions are intended to identify students with whom this research would be appropriate. Please fill out one form for each student in your class(es) who has exhibited disruptive behavior so far in the course of this academic year. Disruptive behavior should include behaviors that have persisted thus far this year and for which you might have sought additional support or ar e considering seeking such support. Your name __________________________ Grade level ___________ Students name __________________________ Period _______________ 1) Please describe each form of the disruptive be havior that the student exhibits (e.g., yelling out, talking to peers, etc.): ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________2) Does the behavior occur at any time during th e period more than another? Are there any triggers for the behavior? If so, please describe them. ________________________________________________________________________ ________________________________________________________________________ 3) Please describe ALL the contexts in whic h you have seen the behavior occur (e.g., math instruction, group work, lunchroom, ha llway, etc.). If th ere is more than one form of disruption, please list the contexts for each individual behavior. ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 4) To your knowledge, has this student demonstr ated these or similar behaviors during past school years? If so, please describe what you know of the be havior prior to this year. ________________________________________________________________________ ________________________________________________________________________
131 APPENDIX E INITIAL OBSERVATION FORM Student: Behavior 1 Antecedents & Consequences Behavior 2 Antecedents & Consequences Minute 0 1 Minute 1 2 Minute 2 3 Minute 3 4 Minute 4 5 Minute 5 6 Minute 6 7 Minute 7 8 Minute 8 9 Minute 9 -10 Minute 10 11 Minute 11 12 Minute 12 13 Minute 13 14 Minute 14 15 Notes (classroom activity, peer behavior, other):
132 APPENDIX F INFORMED CONSENT PARENTAL CONSENT FOR STUDENT PARTICIPANTS Dear Parent/GuardianI am a graduate student at the University of Florida, under the supervision of Dr. Nancy Waldron, conducting research on ways to help teachers identify a nd intervene with problematic behavior in their classrooms. The purpose of the study is to develop an in-class procedure teachers can use to identify and address factors th at may help support appropriate behavior in the classroom. The results of this study are intended to be individua lized to the needs of your child, and may also help future students. Id lik e to ask for your permission for your child to participate in this research. Your childs teacher will identify one or more pr oblematic behaviors theyd like to address. For example, teachers might indicate that disrupt ion (talking without raising ones hand, talking while the teacher is talking) is a problematic behavior for some students. The teacher will set up several conditions through which to evaluate what environmental factors maintain, or reinforce, an appropriate alternative behavi or. In each of the different conditions, the teacher will offer a different form of reinforcement for appropriate behavior. These will include the teachers attention, another students atten tion, or a brief break from a challenging class activity. The number of times your student demonstrates the behavior during a certain amount of time will be counted. Teachers will then compare behavior in each of the conditions to determine which condition best supports your students appropriate behavior. I will observe as teachers are conducting these studies, in orde r to determine how easily teacher s are able to administer the conditions and evaluate the behavior. This study has the potential to provide you and your childs teacher with useful information about problematic behaviors in the classroom a nd how to encourage appropriate behavior. In addition, information provided by this type of res earch can provide a direct link to interventions designed for your child. There are no known risks to your child. For the protection of your familys privacy, stud ents names will be changed when the data is reported. Data generated by the teachers about each student will be ke pt confidential to the extent provided by the law. Participation or nonparticipation in this study will not affect your childs grades or placement in any programs. You have the right to withdraw consent for your child's participation at an y time without consequence. If you have any questions about this research protocol, please contact me at 246-1480, or my faculty supervisor, Dr. Waldron, at 392-0723, ext. 232. Questions or concerns about your child's rights as a research participant ma y be directed to the UFIRB offi ce, University of Florida, Box 112250, Gainesville, FL 32611, (352) 392-0433. Elizabeth L.W. McKenney, M.Ed.
133 I have read the procedure described above. I voluntarily give my consent for my child, _________________, to participate in Elizabeth McKe nney's study of teacher-led, in-class analyses of problematic behavior. I have received a copy of this description. ____________________________ ___________ Parent / Guardian Date
134 APPENDIX G PEER CONFEDERATE IDENTIFICATION FORM Please identify 3 5 students whom you think would be appropriate for inclusion in this project to deliver attention to the target student cont ingent on the target a ppropriate behavior. 1) 4) 2) 5) 3) For each of these students, please cons ider the following yes/no questions: Student #: 1 2 3 4 5 Is this student mature enough to participate in this study? Does he/she have what you would consider strong social skills? Does he/she follow directions well? Is he/she friends with the target student?
135 APPENDIX H INFORMED CONSENT PARENTAL CONSENT FOR PEER CONFEDERATES Dear Parent/GuardianI am a graduate student at the University of Florida, under the supervision of Dr. Nancy Waldron, conducting research on ways to help teachers identify a nd intervene with problematic behavior in their classrooms. The purpose of the study is to develop an in-class procedure teachers can use to identify and address factors that may help to support appropriate classroom behavior. One of the teachers with whom I am working, _______________________, has identified your child as someone who consistently follows classroom instructions and demonstrates supportive social beha viors, who may be able to assi st us in our evaluation. With your permission, Id like to ask your child to help in this research project by acting as a peer confederate. The problem behaviors of identified students may involve such behaviors as disruption (talking out of turn or without raising one s hand) or out-of-seat behavior. One of the ways in which we will be evaluating student behavior will be th rough providing peer attention for targeted appropriate behavior. For example, when an id entified student raises his/her hand instead of shouting out, a peer confederate might ask, Do you need help? By providing attention to the identified student for appropriate behavior, we are able to determine if peer attention is a motivating factor in the behavior of the target student. This pr ovides teachers with a direct link to intervening with problem behavior for that individual student. Your child will be asked to provide attention to the identified student when he/she engages in one or more appropriate goal behaviors. The attention that your child provides will be brief, and will not negatively affect his/her academic e ngagement. Your childs teacher will carefully plan the evaluation sessions during class time so that no students ability to learn is negatively impacted. For the protection of your familys privacy, stud ents names will be changed when the data is reported. Data generated by the teachers about each student will be ke pt confidential to the extent provided by the law. Participation or non-p articipation in this study will not affect the students grades or placement in any programs. You have the right to withdraw consent for your childs participation at any time without conseque nce. There are no known risks to your child. If you have any questions about this research protocol, please contact me at 246-1480, or my faculty supervisor, Dr. Waldron, at 392-0723, ext. 232. Questions or concerns about your childs rights as a research participant ma y be directed to the UFIRB offi ce, University of Florida, Box 112250, Gainesville, FL 32611, (352) 392-0433. Elizabeth L.W. McKenney, M.Ed.
136 I have read the procedure described above. I voluntarily give my consent for my child, _________________, to participate in Elizabeth McKe nney's study of teacher-led, in-class analyses of problematic behavior. I have received a copy of this description. ____________________________ ___________ Parent / Guardian Date I have read the procedures above. I give my assent to participate in Ms. McKenneys study of classroom behavior. I have a re ceived a copy of this description. ____________________________ ___________ Student Signature Date
137 APPENDIX I PROBLEM IDENTIFICATION INTERVIEW TEACHER (Adapted from Bergan & Kratochwill, 1990) Teachers Name:___________________________ Date:___________________ Identified Student:_______________________ The purpose of this interview is to discu ss the problem behavior that _______________ exhibits and begin to identify when/why it happens and what other behavior might be more appropriate instead. Please feel free to stop me and ask que stions along the way. I want to make sure I understand whats happening in your classroom, so don t hesitate to clarify when necessary. 1) I believe youve told me that you re concerned about _________________, is that the behavior that youd like to discuss? What other behavior s, if any, are you concerned about? 2) Behavior specification (What does he/s he do when he/she is being disruptive?) a)Prioritize each behavior on a scale of 1 to 10 in terms of disruptiveness to classroom functioning. 3) Clarify that this is a problem behavior that occurs in the ph ysical setting of the classroom. 4) During which classroom activities, if any, is the problem behavior particularly likely to occur? 5) What typically happens immediately before the problem behavior occurs? 6) What do peers do/say immediately befo re the problem behavior occurs? 7) What typically happens immediately after the problem be havior occurs? a) What are you likely to do? b) What do peers typically do? 8) Summarize and validate sequential c onditions, record teachers responses. 9) Query each potential function of student behavior (teacher attention, peer attention, escape) while restating sequential conditions. (Example : Youve said that Johnny shouts out during independent work, and that you ty pically take him aside and talk to him about why doing so is inappropriate. Do you think that Johnny might s hout out to get your atte ntion? His peers attention? Away from the assignment?) 10) How often does this behavior occur during a typical class pe riod? Upper limit? Lower limit? 11) How long does it last (if releva nt)? Upper limit? Lower limit?
138 12) Summarize, validate, and confirm frequency and duration, record teachers responses. 13) What other behavior(s) could __ ___________________ use that would be appropriate for your classroom? Is this a behavior that he/she has demonstrated previously? 14) Students sometimes demonstrate problem behavi or because it is easier and/or faster than appropriate behavior. Is the sel ected appropriate behavior roughl y equal to the problem behavior in terms of ease and efficiency? Thanks for taking the time to discuss this wi th me. I think weve come to a good shared understanding of this students behavior and what appropriate beha vior he/she could use in its place. Our next steps are to go through a brief tr aining and then begin the assessment to see what conditions will help to elicit ap propriate behavior from this student. (Discuss any scheduling issues that need to be covered). Do you have any questions for me?
139 APPENDIX J PROBLEM IDENTIFICATION INTERVIEW STUDENT (Adapted from Bergan & Kratochwill, 1990) Students Name:___________________________ Date:___________________ Participating Teacher:_______________________ Subject:_________________ I want to talk to you a little bit today because some of your teachers are concerned about your behavior during class. Specifically, theyve said that you are sometimes disruptive to the classroom. Youre not in trouble, and your parents and teachers have given their permission for me to talk with you I just want to talk to you a littl e bit about what happens in class when youre disruptive, so that we can help your teacher to work with you on improving your behavior, is that okay? So were going to discuss this problem behavior and try to think about when/why it happens and what other behavior might be more appropriate instead. Please feel free to stop me and ask questions along the way. I want to make sure I understand whats happen ing in your classroom, so dont hesitate to clarify when necessary. 1) What do you typically do that is disruptive to the cl assroom? (If student answers vaguely or says, nothing, follow with, When your teachers talk to you about your behavior, what do they say is the problem?) 2) Is there anything else that you do during class th at may be problematic? 3) What specifically are you doing when you typically get into trouble? 4) Thinking about (identified subj ect area), are there certain activities or assignments that are hard or frustrating for you? What about when youre working by yourself? In groups? (Write down all of students responses for this prompt). 5) Is there anything that could happen right before you ______________ that might make you especially likely to do that? 6) Is there anything that your peers might do that might make you likely to _____________? 7) What typically happens right after you _________________? a) What does Mr./Ms. _________________ do? b) What do your peers typically do? 8) Summarize and validate sequential condi tions, record students responses. 9) Query each potential function of student behavior (teacher attention, peer attention, escape) while restating sequential conditions. (Example : Youve said that you usually stand up and walk around during independent work, and that your teacher yells at you a nd tells you to go back
140 to your seat. Meanwhile, your friends will talk to you if you talk to them, even though theyre working. Do you think that when you do this what you really want is to get Mr./Ms. _______________s attention? To be able to talk to your friends? To get out of having to do the assignment?) 10) How often does this behavior occur during a typical class pe riod? Upper limit? Lower limit? 11) How long does it last (if relevant )? Upper limit? Lower limit? 12) Summarize, validate, and confirm frequency and duration, record students responses. 13) What else could you do instead that would be appropriate for your classroom? Is this something that you know how to do? 14) Sometimes we do something were not supposed to because its easier and/or faster than something more appropriate. Is the appropriate behavior we ju st talked about roughly equal to ____________________ (identified problem behavior) in terms of how hard or easy it is to do? Thanks for taking the time to discuss this wi th me. I think Ive got a better unde rstanding of your classroom behavior and wh at might be helpful for you. Do you have any questions for me?
141 APPENDIX K FUNCTIONAL ANALYSIS DESCRIPT IONS F OR TEACHER TRAINING (Adapted from the methods section according to Iwata et al., 1982/1994) Teacher Attention. This condition is designed to identify behavior that occurs as a function of attention from the teacher. The teacher and st udent are in the classroom together during academic instruction. Teachers begin the sessio n by placing an instruction card on the desk of the target student which states, Were going to work on some material youre familiar with. Please do your best work. If you need help, le t me know by (insert targ et behavior) and Ill come talk to you. Having explained the activity to the class, the teacher sits down at his/her desk or engages in some other activity that gives the appearance of not providing attention to the class. Each time the target student engages in the identified behavior, the teacher will walk to his/her desk and begin speaking with the student. Teachers should speak to the student as they normally would (e.g., Do you need help? Wha t are you having trouble with?), and then provide any necessary help. Teachers will speak to the student for approximately 15 seconds using such statements before returning to their previous activity. If the student engages in the target behavior again, the teacher should remain with the student for another 15 seconds. No contingencies, including teacher attention, s hould be provided for instances of disruptive behavior. Peer Attention. This condition is designed to identify be havior that occurs as a function of attention from peers. At the start of each sess ion, peers should be privately reminded of the session condition by the teacher. Peers will be prompted to listen to teacher instructions in order to know when to begin the session. Teachers be gin the session by placing an instruction card on the desk of the target student which states, Were going to work on some material youre familiar with. Please do your best work. Youre allowed to talk quietly during this activity, so if you need help from a friend, you may ask (peer conf ederate) for it by (insert target behavior). Each time the target student engages in the identified behavior, the identified peer(s) will speak with the student. Peers will be instructed to speak to the target student as they normally would given the specific behavior, and speak to the stud ent for 15 seconds before returning to their own work. If the student engages in the target behavior again, the peer will speak with the student for another 15 seconds. If the peer does not respond to the target students identified behavior, the teacher should remind him/her to do so by tappi ng him/her on the shoulder or quietly saying, (Peers name), I think (target student) needs yo ur help. Teachers should only provide this directive once per instance of ta rget student identified behavior Similarly, if the peer is providing attention to the target student independently of the target students behavior (e.g., the target student hasnt asked for peer help), the t eacher should prompt the peer to return to his/her work. No contingencies, includ ing peer attention, should be pr ovided for instances of disruptive behavior by the target student. Escape from task demands. This condition is designed to id entify behavior that occurs as a function of avoidance of difficult academic materi al. Teachers will begin this session by placing an instruction card on the desk of the target student which states, Were going to work on material that may be difficult. Please do your be st work. If you need a short break, you may ask for one by (insert target behavior). I will let you know that you can take a break by (using an identified signal behavior). Each time that the target student engages in the identified
142 appropriate behavior, the teacher should indicate to the student that he/she can take a break. Breaks can consist of sitting quietly, putting ones head dow n, walking around the room, or sitting in the teachers office. At the conclusi on of one minute, the teach er indicates that the break is over and that the student should return to work. If the student again engages in the identified behavior, another break should be pr ovided. No contingenc ies, including breaks, should be provided for instances of disruptive behavior. Control condition. The control condition is designed to observe the occurrence of behavior under conditions in which reinforcement is not contingent on instances of the identified appropriate behavior, but the student does not ex perience the antecedents making such behavior likely to occur. Thus, teacher attention and peer attention are available and the presented task is not difficult. In this condition, teachers begin the session placing an instruction card on the desk of the target student that states, Were going to do some work youre all familiar with. Ill be coming around to talk to you as you may need help. (Peer confederate) w ill also check in with you from time to time. Please work quietly and do your best work. No contingencies will be provided for instances of disruptive behavior or the target appropriate be havior, but teacher and peer attention should be provided every one minut e. Participating peer confederates will be quietly instructed to provide such attention prio r to the session, and should be prompted to do so by the teacher if they are not pr oviding attention every minute.
143 APPENDIX L OPERATIONAL DEFINITION FORM Participating Teacher:________________________ Subject:_________________ Students Name:_______________________ Inappropriate Behavior 1: Operational definition: Examples: Non-examples: Inappropriate Behavior 2: Operational definition: Examples: Non-examples: Inappropriate Behavior 3: Operational definition: Examples: Non-examples: ----------------------------------------------------------------------------------------------------------Appropriate Behavior 1: Operational definition: Examples: Non-examples: Appropriate Behavior 2: Operational definition: Examples: Non-examples:
144 Appropriate Behavior 3: Operational definition: Examples: Non-examples:
145 APPENDIX M FUNCTIONAL ANALYSIS DATA COLLECTION FORM Teacher: Condition: Correctly responds to target appropriate behavior Correctly responds to disruptive behavior Correctly prompts peer (if necessary) Peer correctly responds Instances: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Beh: Min: Min: Min: Min: Min: Min: Min: Min: Min: Min: Teacher correctly delivered instruction card: Noncontin g ent Deliver y of Reinforcement
146 APPENDIX N SOCIAL VALIDITY FORM Date: ____ _________________ Teacher: ___________________ Target student: ________________ I am examining the acceptability of my research procedures, and would appreciate your honest feedback. Using the scales below, please comp lete each item by circling the number that best indicates how you feel about this study. Thank you for your assistance. 1. How comfortable were you with this study? Not at all Neutral Very Comfortable Comfortable 1 2 3 4 5 6 7 2. How time intensive was it for you to pa rticipate in the training sessions? Not at all Neutral Very Intensive Intensive 1 2 3 4 5 6 7 3. How comfortable were you w ith the training sessions? Not at all Neutral Very Comfortable Comfortable 1 2 3 4 5 6 7 4. How comfortable did the peer(s) ap pear with the assessment sessions? Not at all Neutral Very Comfortable Comfortable 1 2 3 4 5 6 7 5. How time intensive was it for you to pa rticipate in the assessment sessions? Not at all Neutral Very Intensive Intensive 1 2 3 4 5 6 7 6. How disruptive was this asse ssment to your classroom? Not at all Neutral Very Disruptive Disruptive
147 1 2 3 4 5 6 7 7. How much did you have to change your classr oom routine to allow the target student and peer(s) to participate in the assessment procedures? Not much Neutral Very At all Much 1 2 3 4 5 6 7 8. How disruptive was this assessment to the other peers in the classroom? Not at all Neutral Very Disruptive Disruptive 1 2 3 4 5 6 7 9. How willing would you be to allow another stud ent to participate in this assessment? Not at all Neutral Very Willing Willing 1 2 3 4 5 6 7 10. How willing would you be to particip ate in this assessment again? Not at all Neutral Very Willing Willing 1 2 3 4 5 6 7 11. How useful was the information obtained fr om this assessment to you and the target student? Not at all Neutral Very Useful Useful 1 2 3 4 5 6 7 12. Overall, how would you rate this assessment procedure? Not at all Neutral Very Beneficial Beneficial 1 2 3 4 5 6 7 Comments:
148 LIST OF REFERENCES American Psychiatric Association: Diagnostic and statistical manual of mental disorders, Fourth Edition, Text Revision. Washington, DC, American Psychiatric Association, 2000. Asmus, J. A., Vollmer, T. R., & Borrero, J. C. (2002). Functional behavioral assessment: A school-based model. Education and Treatment of Children, 25, 67 90. Athanasiou, M. S., Geil, M., Hazel, C. E., & Copeland, E. P. (2002). A look inside school-based consultation: A qualitative study of the beliefs and practices of school psychologists and teachers. School Psychology Quarterly, 17, 258 298. Baker, J. A., Kamphaus, R. W., Horne, A. M., & Winsor, A. P. (2006). Evidence for population-based perspectives on childrens be havioral adjustment and needs for service delivery in schools. School Psychology Review, 35, 31 46. Bergan, J. R. (1977). Behavioral consultation. Columbus, OH: Charles E. Merrill Publishing Co. Bergan, J. R., & Kratochwill, T. R. (1990). Behavioral consultation and therapy. New York: Plenum Press. Borrero, J. C., and Vollmer, T. R. (2002). An application of the matching law for severe problem behavior. Journal of Applied Behavior Analysis, 35, 13 27. Boyajian, A. E., DuPaul, G. J., Wartel Handler M., Eckert, T. L., & McGoey, K. E. (2001). The use of classroom-based brief f unctional analyses with preschoolers at-risk for Attention Deficit Hyperactivity Disorder. School Psychology Review, 30, 278 293. Broussard, C., & Northup, J. (1995). An appro ach to functional assess ment and analysis of disruptive behavior in re gular education classrooms. School Psychology Quarterly, 10, 151 164. Broussard, C., & Northup, J. (1997). The use of functional analysis to develop peer interventions for disruptiv e classroom behavior. School Psychology Quarterly, 12, 65 76. Brown, D., Pryzwansky, W. B., & Schulte, A. C. (2001). Psychological consultation: Introduction to theory and practice (5th ed.). Boston: Allyn & Bacon. Boucher, C. R. (1999). Students in discord: Adolescents with Emotional and Behavioral Disorders. Westport, CT: Greenwood Press.
149 Carr, E. G., Horner, R. H., Turnbull, A. P., Marquis, J. G., Magito McLaughlin, D., McAtee, M. L., Smith, C. E., Anderson Rya n, K., Ruef, M. B., Doolabh, A., & Braddock, D. (1999). Positive Behavioral Support for people with developmental disabilities: A research synthesis. Washington, D.C.: American Association on Mental Retardation. Codding, R. S., Feinberg, A. B., Dunn, E. K., & Pace, G. M. (2005). Effects of immediate performance feedback on implem entation of behavior support plans. Journal of Applied Behavior Analysis, 38, 205 220. Cone, J. D. (1997). Invited essay: Issues in f unctional analysis in behavioral assessment. Behaviour Research and Therapy, 35, 259 275. DiGennaro, F. D., Martens, B. K., & McInty re, L. L. (2005). Increasing treatment integrity through negative reinforcement: Effects on teacher and student behavior. School Psychology Review, 34, 220 231. DiGennaro, F. D., Martens, B. K., & Klei nmann, A. E. (2007). A comparison of performance feedback procedures on teacher s' treatment implementation integrity and students' inappropriate behavior in special education classrooms. Journal of Applied Behavior Analysis, 40, 447 461. Eckert, T. L., Martens, B. K., & DiGennaro, F. D. (2005). Describing antecedentbehavior-consequence relations using conditional pr obabilities and the general operant contingency space: A preliminary investigation. School Psychology Review, 34, 520 528. Erchul, W. P. (1999). Two steps forward, one step back: Collabora tion in school-based consultation. Journal of School Psychology, 37, 191 203. Erchul, W. P., & Martens, B. K. (2002). School consultation: C onceptual and empirical bases of practice (2nd ed.). New York: Kluwer Academic/Plenum Press. Ervin, R. A., DuPaul, G. J., Kern, L., & Fr iman, P. C. (1998). Classroom-based functional and adjunctive assessments: Proact ive approaches to intervention selection for adolescents with Attention Defi cit Hyperactivity Disorder. Journal of Applied Behavior Analysis, 31, 65 78. Ervin, R. A., Radford, P. M., Bertsch, K., Pipe r, A. L., Ehrhardt, K. E., & Poling, A. (2001). A descriptive analysis and critique of the empirical literature on school-based functional assessment. School Psychology Review, 30, 193 210. Flood, W. A., Wilder, D. A., Flood, A. L., & Masuda, A. (2002). Peer-mediated reinforcement plus prompting as treatment for off-task behavior in children with Attention Deficit Hyperactivity Disorder. Journal of Applied Behavior Analysis, 35, 199 204.
150 Frea, W. D., & Hughes, C. (1997). Func tional analysis and tr eatment of socialcommunicative behavior of adolescents with developmental disabilities. Journal of Applied Behavior Analysis, 30, 701 704. Graham, D. S. (1998). Cons ultant effectiveness and tr eatment acceptability: An examination of consultee requests and consultant responses. School Psychology Quarterly, 13, 155 168. Gresham, F. M., Lane, K. L., & Beebe-Frankenberger, M. (2005). Predictors of hyperactive-impulsive-inattention and conduct problems: A comparative follow-back investigation. Psychology in the Schools, 42, 721 742. Gutkin, T. B. (1999). Collaborative versus directive/prescriptive/expert school-based consultation: Reviewing and re solving a false dichotomy. Journal of School Psychology, 37, 161 190. Harvey, M. T., May, M. E., & Kennedy, C. H. (2004). Nonconcurrent multiple baseline designs and the evaluation of educational systems. Journal of Behavioral Education, 13, 267 276. Individuals with Disabiliti es Education Act, 20 U.S.C. 33 (1997). Iwata, B., Dorsey, M., Slifer, K., Bauman, K., & Richman, K. (1982). Toward a functional analysis of self-injury. Analysis and Intervention in Developmental Disabilities, 2, 3-20. (Reprinted in Journal of Applied Behavior Analysis, 1994, 27, 196209). Iwata, B. A., Wallace, M. D., Kahng, S., Lindberg, J. S., Roscoe, E. M., Conners, J., et al. (2000). Skills acquisition in the implemen tation of functional analysis methodology. Journal of Applied Behavior Analysis, 33, 181 194. Jones, K. M., Drew, H. A., & Weber, N. L. (2000). Noncontingent peer attention as treatment for disruptive classroom behavior. Journal of Applied Behavior Analysis, 33, 343 346. Kamps, D., Wendland, M., & Culpepper, M. (2006). Active teacher participation in functional behavior assessment for students with emotional and behavioral disorders risks in general education classrooms. Behavioural Disorders, 31, 128 146. Kates-McElrath, K., Agnew, M., Axelrod, S., & Bloh, C. L. (2007). Identification of behavioral function in public school s and a clarification of terms. Behavioral Interventions, 22,47 56. Kazdin, A. E. (1982). Single-Case Research Designs: Me thods for Clinical and Applied Settings. New York, NY: Oxford University Press.
151 Kratochwill, T. R., & Bergan, J. R. (1978) Evaluating programs in applied settings through behavioral consultation. Journal of School Psychology, 16, 375 386. Kratochwill, T. R., & Van Someren, K R. (1985). Barriers to treatment success in behavioral consultation: Current limitations and future directions. Journal of School Psychology, 23, 225 239. Lannie, A. L., & McCurdy, B. L. (2007). Preventing disruptive beha vior in the urban classroom: Effects of the Good Behavior Game on student and teacher behavior. Education and Treatment of Children, 30 85 98. Lewis, A. (2002). Accessing, through research interviews, the views of children with difficulties in learning. Support for Learning, 17, 110 116. McComas, J., Hoch, H., Paone, D., & El-R oy, D. (2000). Escape behavior during academic tasks: A preliminary analysis of idiosyncratic establishing operations. Journal of Applied Behavior Analysis, 33, 479 493. Moore, J. W., Edwards, R. P., Sterling-Turner, H. E., Riley, J., DuBard, M., & McGeorge, A. (2002). Teacher acquisit ion of functional analysis methodology. Journal of Applied Behavior Analysis, 35, 73 77. Mueller, M. M., Sterling-Turner, H. E., & M oore, J. W. (2005). Towards developing a classroom-based functional analysis condition to assess escape-to-atten tion as a variable maintaining problem behavior. School Psychology Review, 34, 425 431. Murdock, S. G., ONeill, R. E., & Cunningham, E. (2005). A comparison of results and acceptability of functional behavioral asse ssment procedures with a group of middle school students with Emotional/B ehavioral Disorders (E/BD). Journal of Behavioral Education, 14, 5 18. Newcomer, L. L., & Lewis, T. J. (2004) Functional behavioral assessment: An investigation of assessment relia bility and effectiveness of function-based interventions. Journal of Emotional and Behavioral Disorders, 12, 168 181. Nickerson, A. B. (2003). Identifying skills and behaviors for successful inclusion of students with emotional or behavioral disorders. Behavioral Disorders, 28, 401 409. Noell, G. H., Witt, J. C., LaFleur, L. H., Mo rtenson, B. P., Ranier, D. D., & LaVelle, J. (2000). Increasing intervention implemen tation in general education following consultation: A comparison of two follow-up strategies. Journal of Applied Behavior Analysis, 33, 271 284.
152 Noell, G. H., Witt, J. C., Slider, N. J., C onnell, J. E., Gatti, S. L., Williams, K. L., Koenig, J. L., Resatar, J. L., & Duhon, G. J. (2005). Treatment implementation following behavioral consultati on in schools: A comparison of three follow-up strategies. School Psychology Review, 34, 87 106. Northup, J., Jones, K., Broussard, C., DiGi ovanni, G., Herring, M., Fusilier, I., & Hanchey, A. (1997). A preliminary analys is of interactive effects between common classroom contingencies and methylphenidate. Journal of Applied Behavior Analysis, 30, 121 125. Northup, J., Wacker, D. P., Sasso, G., Steege, M., Cigrand, K., Cook, J., & DeRaad, A. (1991). A brief functional analysis of aggressi ve and alternating beha vior in an outclinic setting. Journal of Applied Behavior Analysis, 24, 509 522. Perrin, C. J., Sellers, J. A., Badley, B. M ., & Marcus, A. (2008, May). Comparison of findings across functional analysis and concurre nt operant assessments. In Sellers, J. A. (Chair), Extensions of experimental anal ysis to assess appropriate behavior of adolescents. Symposium presen ted at the annual meeting of the Association for Behavior Analysis, Chicago, IL. Piersel, W. C., & Gutkin, T. B. (1983). Resistance to school-based consultation: A behavioral analysis of the problem. Psychology in the Schools, 20, 311 320. Reddy, L. A. (2001). Serious emotional dist urbance in children and adolescents: Current status and future directions. Behavior Therapy, 32, 667 691. Reed, S., Thomas, E., Sprague, J. R., & Horn er, R. H. (1997). The student-guided functional assessment interview: An analys is of student and teacher agreement. Journal of Behavioral Education, 7, 33 49. Sajwaj, T., Twardosz, S., & Burke, M. (1972). Side effects of extinction protocols in a remedial preschool. Journal of Applied Behavior Analysis, 5 163 175. Santrock, J. W. (2001). Adolescence. Boston: McGraw-Hill. Schill, M. T., Kratochwill, T. R., & Elliott, S. N. (1998). Functional assessment in behavioral consultation: A treatment utility study. School Psychology Quarterly, 13, 116 140. Sellers, J. A., & Strickland, C. C. (2008, Ma y). Functional analysis of appropriate behavior with adolescents. In Sellers, J. A. (Chair), Extensions of experimental analysis to assess appropriate beha vior of adolescents. Symposium presented at the annual meeting of the Association for Behavior Analysis, Chicago, IL.
153 Shapiro, E. S., & DuPaul, G. J. (1996). A school-based consultation program for service delivery to middle school students with A ttention Deficit/Hyperactivity Disorder. Journal of Emotional and Behavioral Disorders, 4, 73 81. Shriver, M. D., Anderson, C. M., & Proctor, B. (2001). Evaluating the validity of functional behavior assessment. School Psychology Review, 30, 180 192. Sholevar, E., & Eichelman, B. S. (1998). Adolescent aggressive behavior: From laboratory to clinic. In Shwartzberg, A. Z. (Ed.), The adolescent in turmoil (pp. 109-123). Westport, CT: Praeger Publishers. Sterling-Turner, H. E., Robinson, S. L., & Wilczynski, S. M. (2001). Functional assessment of distracting and disruptiv e behaviors in the school setting. School Psychology Review, 30, 211 226. Stichter, J. P., & Conroy, M. A. (2005). Using structural analysis in natural settings: A responsive functional assessment strategy. Journal of Behavioral Education, 14, 19 34. Sugai, G., & Horner, R. R. (2006). A prom ising approach for expanding and sustaining School-Wide Positive Behavior Support. School Psychology Review, 35, 245 259. Taylor, J., & Miller, M. (1997). When timeout works some of the time: The importance of treatment integrity and functional assessment. School Psychology Quarterly, 12, 4 22. U. S. Department of Education. (1998). Twentieth Annual Report to Congress on the Implementation of Individuals with Disabilities Education Act. Vollmer, T. R., Borrero, J.C., Wright, C.S., Van Camp, C., & Lalli, J. S. (2001). Identifying possible contingencies during de scriptive analyses of severe behavior disorders. Journal of Applied Behavior Analysis, 34, 269 287. Vollmer, T. R., & Northup, J. (1996). Some implications of functional analysis for school psychology. School Psychology Quarterly, 11, 76 92. Vollmer, T. R., & Northup, J. (1997). Applied Behavior Analysis and School Psychology: An introducti on to the mini-series. School Psychology Quarterly, 12, 1 3. Vollmer, T. R., Marcus, B. A., Ringdahl, J. E ., & Roane, H. S. (1995). Progressing from brief assessments to extended experimental analyses in the evaluation of aberrant behavior. Journal of Applied Behavior Analysis, 28, 561 576. Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazaleski, J. L. (1993). Within-session patterns of self-injury as indicators of behavioral function. Research in Developmental Disabilities, 14, 479 492.
154 Wallace, M. D., Doney, J. K., Mintz-Resudek, C. M., & Tarbox, R. S. F. (2004). Training educators to impleme nt functional analyses. Journal of Applied Behavior Analysis, 37, 89 92. Watson, P. J., & Workman, E. A. (1981). The non-concurrent multiple baseline acrossindividuals design: An extension of th e traditional multiple baseline design. Journal of Behavior Therapy and Experimental Psychiatry, 12, 257 159. Weber, K. P., Killu, K., Derby, K. M., & Barre tto, A. (2005). The status of Functional Behavioral Assessment (FBA): Adherence to standard practice in FBA methodology. Psychology in the Schools, 42, 737 744. Wehmeyer, M. L., Baker, D. J., Blumberg, R., & Harrison, R. (2004). Selfdetermination and student involvement in functional assessment. Journal of Positive Behavior Interventions, 6, 29 35. Weissenburger, J. W., Fine, M. J., & Poggio, J. P. (1982). The relationship of selected consultant/teacher characterist ics to consultation outcomes. Journal of School Psychology, 20, 263 270. Wilczynski, S. M., Mandal, R. L., & Fusilie r, I. (2000). Bridges and barriers in behavioral consultation. Psychology in the Schools, 37, 495 504. Wilkinson, L. A. (1997). School-based behavior al consultation: Delivering treatment for childrens externalizing beha vior in the classroom. Journal of Educational and Psychological Consultation, 8, 255 276. Wilson, C. P., Gutkin, T. B., Hagen, K. M., & Oats, R. G. (1998). General education teachers knowledge and self-reported use of classroom interventions for working with difficult-to-teach students: Implications fo r consultation, prerefe rral intervention, and inclusive services. School Psychology Quarterly, 13, 45 62. Witt, J. C., Gresham, F. M.., & Noell, G. H. (1996). Whats behavior al about Behavioral Consultation? Journal of Educational and Psychological Consultation, 7, 327 344. Witt, J. C., Noell, G. H., LaFleur, L. H., & Mortenson, B. P. (1997). Teacher use of interventions in general educat ion settings: Measurement and analysis of the independent variable. Journal of Applied Behavior Analysis, 30, 693 696. Witt, J. C., VanDerHeyden, A. M., & Gilberts on, D. (2004). Troubleshooting behavioral interventions: A systematic process for finding and eliminating problems. School Psychology Review, 33, 363-383. Wood, B. K., Umbreit, J., Liaupsin, C. J., & Gresham, F. M. (2007). A treatment integrity analysis of f unction-based intervention. Education and Treatment of Children, 30, 105 120.
155 Wright-Gallo, G. L., Higbee, T. S., Reagon, K. A., & Davey, B. J. (2006). Classroombased functional analysis and intervention for students with emotional/behavioral disorders. Education and Treatment of Children, 29, 421 436.
156 BIOGRAPHICAL SKETCH Elizabeth Lane W eeks McKenney was born in Miami, Florida, and grew up in the suburbs of Atlanta, Georgia from the time sh e was seven. Though primarily an only child, she was fortunate enough to live down the street from her half-sister, Janet. Elizabeth graduated from Centennial High School in 1999 and went on to attend Tulane University. While at Tulane, she majored in psychology, took minors in Span ish and dance, and became a member of Chi Omega Fraternity and the Newcomb Dance Compa ny. Elizabeth graduated summa cum laude in 2003 and was among one of the last classes to gr aduate from Newcomb College before it was disbanded in 2006 following Hurricane Katrina. In 2002, a position as a classroom teacher at a summer treatment program for children with ADHD taught Elizabeth the potential of effec tive behavioral interventions. She pursued her interest in those techniques into her graduate caree r at the University of Florida. During graduate school, Elizabeth had the good fortune to broade n her understanding of behavior problems and behavior analysis through grant work with chil dren with autism spectrum disorders under the direction of Dr. Maureen Conroy. At the same time, she learned the invaluable skills of effective consultation through practicum work at P. K. Yonge Developmental Research School under the supervision of Dr. Nancy Waldron. The influential leadership of both of E lizabeths advisors has become the cornerstone of this dissertation proj ect and of her future professional goals. In 2004, Elizabeth married her college sweethe art, Mark. Their daughter, Samantha, was born in 2007, and she is a daily source of joy in their otherwise hectic lives. She completed her internship and was conferred her Ph.D. in Ma y, 2010 and plans to remain involved in both K schools and university academic setti ngs as a school psychologist.