Training Parents to Enhance Social Skils in Children with Developmental Delays: A Component Analysis by Kely D. Schleismann A disertation submited to the Graduate Faculty of Auburn University in partial fulfilment of the requirements for the Degre of Doctor of Philosophy Auburn, Alabama August 2, 2014 Keywords: Autism Spectrum Disorders, Parent Training, Behavioral Skils Training Component Analysis, Parental Stres Copyright 2014 by Kely D. Schleismann Approved by Jennifer Gilis, Co-chair, Asociate Profesor, Binghamton University Sacha Pence, Co-chair, Asistant Profesor, Auburn University Elizabeth Brestan Knight, Asociate Profesor, Auburn University Chris Correia, Asociate Profesor, Auburn University ii Abstract Social skils deficits are a core feature of ASD (American Psychiatric Asociation [APA], 2000; Parsons & Mitchel, 2002), and they are related to a myriad of other social, developmental, and psychological chalenges (Rogers, 2000). As such, it is important to identify efective means of teaching social skils to children with ASD. Research suggests that naturalistic training techniques such as Pivotal Response Training (Koegel, O?Del, & Koegel, 1987; Stahmer, 1999) and techniques such as Integrated Play Groups that provide exposure to and prompted interaction with peers (Wolfberg & Schuler, 1993 & 1999) can enhance social skils in this population. However, generalization of these skils, or the ability to apply these skils in diferent contexts with diferent people, remains problematic (Stahmer, 1995). Training the adults who are with the child throughout his day to deliver social skils intervention would provide maximum exposure to naturalistic learning opportunities and may improve generalization. Indeed, researchers have suggested that training parents in other behavioral training techniques improves the generalization of the skils taught (Lovaas, Koegel, Simons, & Long, 1973; Schreibman & Koegel, 1996). Many studies have demonstrated succesful training of parents in a wide variety of behavioral interventions (e.g., Lafasakis & Sturmey, 2007; Kroeger & Sorenses, 2010; Wang, 2008) including naturalistic training techniques (e.g., Gilet & LeBlanc, 2006). The majority of these studies, however, have relied on the use of multicomponent training packages. Therefore the component or components responsible iii for bringing about the desired results remains unknown. In addition, given the range of outcomes observed in parent training literature, parent characteristics such as stres may impact outcome (Bagner & Graziano, 2012; Strauss et al., 2012). The primary aim of the current was to investigate how to efectively train parents to implement a social engagement procedure. Furthermore, it systematicaly analyzed the components of the training package to determine which components are responsible for behavior change and identify the most eficient method of training possible. A non- concurrent multiple baseline design was used to examine the efects of the training package. To further analyze the relative contribution of each component of the training package, each component was presented systematicaly, using an ABC design, until measures of participants? implementation of the procedure were stable. A secondary aim of the study was to consider how factors such as parent stres might be related to skil acquisition. Results indicated that al participants who participated through completion were able to implement the procedure with fidelity following training. Furthermore, results suggest that fedback is an efective and eficient method of training when presented alone and may acount for the majority of changed observed in parent behavior. Finaly, parent afect remained neutral or improved over the course of their participation. Results are discussed in terms of possible reasons for the observed changes. Clinical implications and future directions are also discussed. iv Table of Contents Abstract ...................................................................................................................................ii List of Tables .........................................................................................................................vi List of Figures .......................................................................................................................vii List of Abbreviations ............................................................................................................vii Introduction .............................................................................................................................1 Social Skils Interventions ............................................................................................5 Pivotal Response Training ..................................................................................8 Integrated Play Groups .....................................................................................10 Summary ..........................................................................................................12 Parent Training ..........................................................................................................12 The Ned for a Component Analysis ..........................................................................15 Factors that Impact Training Succes .........................................................................18 Current Study ............................................................................................................19 Method ..................................................................................................................................21 Participants ................................................................................................................21 Asesments .....................................................................................................21 Participant Characteristics ................................................................................24 Seting and Materials .................................................................................................29 Design and Procedure ................................................................................................30 v Parent Training Procedures ........................................................................................31 Dependent Variables ..................................................................................................33 Interobserver Agrement ............................................................................................35 Procedural Integrity ...................................................................................................36 Social Validity ...........................................................................................................37 Results ..................................................................................................................................39 Discussion .............................................................................................................................58 Components Analyzed ...............................................................................................59 Affect ........................................................................................................................62 Parent Acquisition, Afect, and Child Behavior: Puting it Together ...........................65 Limitations ................................................................................................................67 Strengths and Future Directions .................................................................................69 References .............................................................................................................................71 Footnote ................................................................................................................................89 vi List of Tables Table 1.1 ...............................................................................................................................24 Table 1.2 ...............................................................................................................................25 Table 2 ..................................................................................................................................34 Table 3 ..................................................................................................................................36 Table 4 ..................................................................................................................................40 vii List of Figures Figure 1 .................................................................................................................................40 Figure 2.1 ..............................................................................................................................42 Figure 2.2 ..............................................................................................................................43 Figure 3.1 ..............................................................................................................................55 Figure 3.2 ..............................................................................................................................56 vii List of Abreviations ABA Applied Behavior Analysis APA American Psychiatric Asociation AS Asperger Syndrome ASD Autism Spectrum Disorders BST Behavioral Skils Training DSM-5 Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition DTT Discrete Trial Training HFA High Functioning Autism ID Intelectual Disability IPG Integrated Play Groups MR Mental Retardation PRT Pivotal Response Training 1 Training Parents to Enhance Social Skils in Children with Developmental Delays: A Component Analysis Introduction Social skils can be defined as the verbal and nonverbal behaviors that lead to succesful or positive social interactions (Rao, Beidel, & Murray, 2008). The presence of a social skils repertoire provides opportunities to interact with others in a manner that is reinforcing to both parties and to adapt to diferent social contexts (DiSalvo & Oswald, 2002). For most people, social skils are acquired over the course of development through interactions with parents, teachers, siblings, and peers across a variety of contexts (e.g., school, sports, family activities, etc.). However, these skils are often lacking or limited in children with Autism Spectrum Disorders (ASD; Parsons & Mitchel, 2002; Rao et al., 2008). As characterized in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5; American Psychiatric Asociation [APA], 2013), the term ?Autism Spectrum Disorders? refers to a group of neurobiological disorders characterized by deficits in two main categories: a) social communication and interaction, and b) restricted interests and repetitive behaviors. Among the defining features of this spectrum of disorders are impairments in social skils (APA, 2013; Parsons & Mitchel, 2002; Rao et al., 2008). In fact, social skils deficits serve as a major source of impairment in this population regardles of language or cognitive ability (Carter, Davis, Klin, & Volkmar, 2005). Common chalenges faced by this population include, but are not limited to, 2 dificulty developing and maintaining age-appropriate peer relationships, poor use of nonverbal communication (e.g., eye contact, gestures, body posture, etc.), dificulty interpreting the social cues of others, problems understanding and expresing emotions, and a lack of reciprocity during social interchanges (Atwood, 2000; Weis & Haris, 2001; White, Koenig, & Scahil, 2007). ASD can be reliably diagnosed in children ages two to thre years old (Chawarska & Volkmar, 2005; Landa, 2008); however, some researchers have demonstrated evidence of impaired social skils in children within the first year of life (Maestro et al., 2002; Osterling & Dawson, 1994; Osterling, Dawson, & Munson, 2002; Werner, Dawson, Osterling, & Dinno, 2000). For example, Werner and colleagues examined home videotapes of 15 children at 8 to 10 months of age who were later diagnosed with an ASD and compared them to videos of 15 typicaly developing same-aged children. The children with ASD les frequently responded to their names and were les likely to look at the face of another person while smiling compared to typicaly developing children. In a similar study, Osterling and colleagues (2002) reviewed home videotapes of the first birthday parties of thre groups of children: those with an ASD, those with mental retardation, and typicaly developing children. Results revealed that children with an ASD showed les gesturing, orienting to name, looking at objects held by others, and looking at people when compared to typicaly developing children. Furthermore, children with an ASD exhibited les orienting to name and looking at others than children with mental retardation. Thus, some of the earliest signs of ASDs include deficient social engagement (e.g., eye contact, responding to name, etc.) and poor social-communicative behaviors (e.g., gesturing, requesting, etc.). 3 Not only are the social-communicative deficits a defining feature of ASD, but some have also considered them to be the most debilitating because of the myriad of chalenges with which they are asociated (Rogers, 2000). For example, social skils deficits often lead to social exclusion, ridicule, and rejection by the peer group (Church, Alisanski, & Amanullah, 2000; Howlin, 1997; Litle, 2001). Bauminger and Kasari (2000) conducted a study of friendship among children with autism. Despite a reported desire for more social interaction, children with autism often expresed experiencing poor social support and more lonelines. Locke, Ishijima, Kasari, and London (2010) found similar results among adolescents with high functioning autism (HFA). Specificaly, they asesed levels of lonelines, friendship quality, and the extent of the social networks of adolescents with HFA as compared to their typicaly developing peers. Results indicated that the adolescents with HFA reported higher levels of lonelines and poorer friendship quality. Furthermore, more adolescents in this group were isolated or peripheral with respect to their level of integration into the clasroom. Social skils deficits may also lead to problems with mood and anxiety later in life (Myles, Bock, & Simpson, 2001; Tantam, 2003). However, results supporting this asertion are mixed. Many researchers have documented the occurrence of comorbid mood and anxiety disorders in both children (e.g., Abdalah et al., 2011; Amr et al., 2012; Bryson, Corrigan, McDonald, & Holmes, 2008; Joshi et al., 2010; Leyfer et al., 2006; Matila et al., 2010) and adults with ASDs (e.g., Ghaziuddin & Zafar, 2008; Lugnegard, Halerback, & Gilberg, 2011; Ryden & Bejerot, 2008). Tantam (2000) posited that individuals with Asperger syndrome experience teasing and bullying, which may lead to increased frustration, low self-estem, and suspiciousnes of others. He goes on to 4 suggest that these experiences combined with an increased understanding of how one is perceived by others contributes to the co-occurrence of other psychological disorders (e.g., anxiety disorders) in individuals with Asperger syndrome (Tantam, 2000). Some researchers have documented a relationship betwen social functioning, negative peer relationships, and anxiety in both typicaly developing individuals (Ginsburg, La Greca, & Silverman, 1998; La Greca & Lopez, 1998) and individuals with an ASD (Belini, 2004), providing additional support for Tantam?s asertion. However, other researchers have failed to support this notion. Specificaly, Gren, Gilchrist, Burton, and Cox (2000) investigated the relationship betwen social functioning and later psychological functioning among male adolescents with Asperger syndrome as compared to adolescents with conduct disorder. They found no significant correlations betwen psychiatric symptoms and interpersonal dificulties for either group. Nevertheles, the adolescents with Asperger syndrome experienced more severe social dificulties than those with conduct disorder, thus reinforcing the notion that social skils impairments are a core behavioral deficits in ASD. Social skils deficits do not subside with age and maturity. Rather these dificulties persist into later childhood (Church et al., 2000) and adulthood (Rao et al., 2008). Matson, Dempsey, and LoVullo (2009) asesed the social skil functioning of 336 adults with intelectual disability. They found that the presence of an ASD diagnosis was asociated with greater levels of social impairment while characteristics such as gender, age, ethnicity, deafnes, or the co-occurrence of epilepsy were not asociated with any diferences in social skils. These results clearly demonstrate that the same distinguishing social impairments that are present in childhood remain throughout the lifespan. 5 In sum, social skils impairments are a defining set of behavioral deficits of individuals with an ASD. Given the pervasive and persistent nature of these deficits, in combination with the host of additional problems to which these deficits are related, it is critical to investigate methods of improving social skils for this population. Social Skils Interventions Historical atempts to teach social skils to children with ASD aimed to teach the building blocks of social skils in a structured seting and were based on the principles of operant conditioning (Frankel, Leary, & Kilman, 1987; Lovaas & Taubman, 1981; Parsons & Mitchel, 2002). These interventions were succesful at training specific behaviors such as making eye contact, emiting a vocalization in response to another person, increasing functional communication skils, and reducing problem behavior (Lovaas, 1987; Lovaas, Koegel, Simons, & Long, 1973; Lovaas & Taubman, 1981; Schreibman, 2000; White et al., 2007). However, the learning environment was structured such that naturaly occurring discriminative stimuli were replaced with contrived, trainer-driven trials (Frankel, et al., 1987). As a result, one major criticism of these interventions involves the lack of generalization to other people or contexts (Parsons & Mitchel, 2002; White et al., 2007). This is problematic because, while the child might be able to respond appropriately to an adult in a highly structured seting, he might continue to have dificulty interacting with pers and other people that he encounters in his daily life. In recent years the number of social skils intervention studies being conducted has increased dramaticaly (Matson, Matson, & Rivet, 2007; Reichow & Volkmar, 2010). In a review of studies investigating social skils interventions, Matson and colleagues 6 (2007) noted an increase in the number of studies over the 25-year span covered by their review (i.e., 1979 to 2006). They indicated that only five studies were published betwen 1979 and 1985 while 30 studies were published betwen 2001 and the time the manuscript was writen. Reichow and Volkmar (2010) described a continuation of this trend with 54 studies published betwen 2001 and 2007. A number of qualitative reviews of this literature have noted that children with an ASD respond positively to a wide variety of interventions (e.g., Cappadocia & Weis, 2011; DiSalvo & Oswald, 2002; Matson et al., 2007; Rao et al., 2008; Reichow & Volkmar, 2010; Rogers, 2000; Schreiber, 2011; White et al., 2007). For example, Rogers (2000) presented a review of social skils interventions for children with ASD, highlighting studies that have been efective at teaching social skils to individuals with ASD across thre age groups: preschool children, school-aged children, and adolescents. Among the efective interventions, she listed video-modeling, adult-directed approaches (e.g., visual-cuing, direct instruction, social stories, adult instruction in social skils games, social skils groups, and Pivotal Response Training [PRT]), and peer-directed approaches (e.g., peer mediated interventions, peer tutoring, and peer training in PRT). While the list of efective interventions is encouraging, there is stil work to be done to establish empirical support for interventions for children with ASD. Reichow and Volkmar (2010) sought to identify evidence-based practices for teaching social skils acording to the criteria set by Reichow, Volkmar, and Cicheti (2008). 1 They identified eight categories of social skils instruction (i.e., ABA, naturalistic interventions, parent training, peer training, social skils groups, visual, video modeling, and other). The only intervention types to met criteria to be clasified as empiricaly supported were social 7 skils groups and video modeling for school-aged children. None of the intervention types had enough empirical support to met clasification criteria for the preschool age group or the adolescents and adults. However, the authors noted that criteria were not applied to the thre most commonly used techniques (i.e., ABA, parent-training, and peer-training) because of the wide variety of study procedures. Furthermore, this research synthesis restricted the scope of the review by only including recent publications (i.e., those published from 2001 to 2008). Nevertheles, social skils intervention research remains crucial. Many researchers agre that peer involvement sems to be a central component to many succesful social skils intervention (Atwood, 2000; Kohler, Strain, Hoyson, & Jamieson, 1997; Odom & Strain, 1987; Rogers, 2000). Some of the available techniques include antecedent interventions designed to maximize interactions betwen children with autism and their peers (e.g., Integrated Play Groups), peer-instruction to teach peers how to initiate and reinforce interactions with children with autism (e.g., peer-networks, PRT), and initiation training for children with autism to change peer expectancies (DiSalvo & Oswald, 2002). The basic premise behind these approaches is that typicaly developing peers can serve as models for children with les advanced skils (McEvoy & Odom, 1987). While a number of studies have demonstrated the efectivenes of peer- mediated interventions (e.g., Pierce & Schreibman, 1997; Roeyers, 1996; Trembath, Balandin, Togher, & Stanclife, 2009; Wolfberg & Schuler, 1993), gaps in the literature with respect to interventions that result in adequate generalizability of acquired skils and socialy validity remain. Thus, additional research is necesary to addres these problems. The following section wil present two social skils intervention approaches ? Pivotal 8 Response Training and Integrated Play Groups ? that show promise with respect to the development of an efective social skils intervention that may lead to beter generalization of skils and more socially valid interventions. Pivotal Response Training. One approach that has explicitly aimed to increase the generalization of acquired skils is known as Pivotal Response Training (PRT) (Koegel, O?Del, & Koegel, 1987; Schreibman & Koegel, 1996; Stahmer, 1999). PRT is an intervention based in the principles of Applied Behavior Analysis (ABA). It employs a variety of techniques recommended by Stokes and Baer (1977) designed specificaly to promote generalization (LeBlanc, Esch, Sidener, & Firth, 2006). For example, training takes place in the seting in which the behavior is expected to occur, multiple change agents (i.e., teachers, parents, peers, etc.) are trained to deliver the intervention, and multiple exemplars are presented (LeBlanc et al., 2006). In this approach, pivotal behaviors are targeted. Pivotal behaviors are behaviors that, when trained, are expected to lead to change in a wide range of additional behaviors, (Koegel & Frea, 1993). Examples include motivation to respond to social stimuli, responding to multiple discriminative stimuli, and engaging in self-management strategies. Interventions are designed to increase child motivation by alowing the child to choose prefered contexts, such as aces to specific toys or engagement in specific activities. Target behaviors are then modeled and natural reinforcers are delivered contingent on imitation of or approximations to those behaviors. Finaly, maintenance tasks are interspersed with new target skils in order to alow continued high rates of reinforcement (Stahmer, 1999). Research has demonstrated that PRT can be succesful in teaching language skils (Koegel, O?Del, & Koegel, 1987), symbolic play (Stahmer, 1995), and socio-dramatic 9 play (Thorp, Stahmer, & Schreibman, 1995). There is also some evidence that PRT may be more efective at establishing generalization and maintenance than other forms of social skils training. For example, Lydon, Healy, and Leader (2011) compared video modeling and PRT for teaching play skils to children with ASD. Both interventions resulted in an increase in play behaviors. Interestingly, PRT resulted in greater gains in the generalization seting than did video modeling. Despite these promising results, one criticism of PRT is that skils often fail to generalize to other peers. Stahmer (1995) used PRT to increase rate, complexity, and creativity of symbolic play in seven children with ASD. The rate of positive social responses increased following the intervention, and these skils generalized to new toys, new setings, and other adults. However, participants? responses to peers did not increase. In order to addres this shortcoming, some researchers began teaching peers to deliver PRT directly. Such an approach would more closely approximate the context in which the behavior is expected ? one of the techniques to promote generalization recommended by Stokes and Baer (1977). One study to utilize this strategy was conducted by Pierce and Schreibman (1997). They trained peers to use PRT to increase maintenance of interactions and initiations in two children with ASD. A multiple baseline design across peer trainers was used, and the results were replicated in each of the two participants. The peers succesfully implemented PRT and facilitated increases in social engagement and initiations in the participants. After two or thre peers implemented the intervention with each child with ASD, advances in social engagement generalized to a new peer. 10 Integrated Play Groups. Another method of teaching social skils to children with autism is known as Integrated Play Groups (IPG). Wolfberg and Schuler (1993) developed this approach as a comprehensive model of play. This approach incorporates a transactional developmental approach (Prizant, Weatherby, & Rydel, 2000) within a sociocultural framework (Rogoff, 1990). In this model, children with social skils deficits (i.e., ?novice players?) participate in play groups with typicaly developing children (i.e., ?expert players?). An adult facilitator both encourages appropriate play and interaction in the novice players and fosters aceptance in the expert players (Wolfberg & Schuler, 1993). Progresively more competent forms of play are encouraged through various antecedent manipulations. First, intervention takes place in natural, integrated setings in which children with autism and similar developmental disabilities are presented with opportunities to interact with more socialy competent peers. Secondly, group members are selected such that there is a balance of age and developmental level. More specificaly, there are typicaly thre to five members in the group with a greater number of socialy competent children than children with autism. Children are fully imersed in play, rather than breaking down play into discrete subtasks. Furthermore, the physical arangement of the play space is designed to maximize participation and social interaction, and play materials are carefully selected such that they wil appeal to children at diferent developmental levels. A consistent routine is also established from the beginning to create a sense of predictability. Finaly, each child?s level of competence is asesed, and the amount of support and guidance is adapted to met each child?s needs. In addition to these antecedent manipulations, prompts and prompt fading are used as an 11 adult guides the child in participating in more and more complex forms of play then systematicaly decreases the amount of support provided. Several studies have shown this model to be succesful in increasing more complex and social forms of play, and these behaviors were maintained when adult prompts were withdrawn (Wolfberg & Schuler, 1993 & 1999; Yang, Wolfberg, Wu, & Hwu, 2003; Zercher, Hunt, Schuler, & Webster, 2001). Furthermore, the children?s acquired social interaction behaviors generalized to other setings and people acording to parent report (Wolfberg & Schuler, 1993; Yang et al. 2003). In one investigation, Wolfberg and Schuler (1993) evaluated the efectivenes of this approach for two children with autism and thre of their typicaly developing peers. A multiple probe design was used to ases changes in quality of play (i.e., no interaction, object manipulation, functional play, and symbolic/ pretend play) and social integration (i.e., isolate, orientation, paralel or proximity play, and play with a common focus or cooperative play). Results showed decreases in manipulation and gains in functional object use. There were also decreases in isolate play and increases in common focus and paralel/proximity play. Higher rates of appropriate play were not maintained when adult support was withdrawn at Probe 1, but were restored at Probe 2. In another study, Zercher, Hunt, Schuler, and Webster (2001) examined the efect of an integrated play group seting on social behaviors of 6-year-old twin boys with autism. The boys? sisters served as the expert players, and an adult trainer provided coaching in how to involve them in a variety of play themes. A multiple baseline design with thre conditions (i.e., baseline, play with adult coaching, and play without adult coaching) was used to evaluate the efects of the play group. Results showed an increase 12 in joint atention, symbolic play, and language skils. Furthermore, parents reported an increase in the twins? interactions with their peers, though no formal evaluation of generalization was conducted. Although these results are promising, additional research is necesary to determine the generalizability of the skils learned in IPGs. Specificaly, there are no objective measures of behavior in diferent contexts or with diferent people, so it is not possible to determine whether the behaviors truly generalized. Sumary. The interventions discussed above have demonstrated some succes with respect to acquisition, generalization, and maintenance of social skils. Common among these procedures is a naturalistic approach in which children with developmental disabilities are prompted to interact with their peers during play. While the inclusion of peers sems crucial to succesfully teaching social skils to children with developmental disabilities, mere exposure to peers is not sufficient (DiSalvo & Oswald, 2002; Rogers, 2000). The state of the literature appears to support the use of trained adults to mediate interaction betwen children with ASD and their pers or family members in order to bring about change in social skil development. Parent Training There are numerous reasons to train parents to implement social skils training for their children. First, in order to maximize naturalistic teaching opportunities so as to promote generalization and maintenance of skils, the adults who are with the child throughout his or her day should be trained to encourage appropriate interactions. Parents are in just such a position to offer the support needed. Researchers and practitioners have long recognized the benefits of parental involvement in behavioral therapies for children with ASD, noting improvements in generalization and maintenance of skils when parents 13 are trained to deliver behavior interventions (Koegel, Schreibman, Briten, Burke, & O?Neil, 1982; Lovaas, et al. 1973; Schreibman & Koegel, 1996). For example, Lovaas and colleagues (1973) provided intensive behavioral treatment to 13 children with autism. The children enrolled in an intensive behavioral intervention and were aranged into four groups based on the time period during which they were enrolled: (a) in the first group the parents were not involved in treatment, (b) in the second group the parents were trained to deliver the treatment procedures, (c) the parents of the children in the third and fourth groups received training and consultation services. Results showed improvement in the speech, play and social behaviors of al children as wel as decreases in self- stimulation and echolalia. Furthermore, the children whose parents received training continued to improve. Therefore, training parents to deliver social skils intervention may aid in the generalization of these skils. Furthermore, behavioral interventions have been criticized for lack of acesibility and afordability for the families who need them (Parsons & Mitchel, 2002; Rogers, 2000). Several authors have suggested that parent training may be more economical and alow more families to aces critical services (Schreibman & Koegel, 1996; Schultz, Schmidt, & Stichter, 2011). One training method that has received considerable atention in the literature is Behavioral Skils Training (BST). BST is a training package consisting of instruction, modeling, rehearsal, and fedback. It has been used to train paraprofesional staf and teachers to implement Discrete Trial Training (DT; Dib & Sturmey, 2007; Koegel, et al., 1977; LeBlanc, Riciardi, & Luiseli, 2005; Sarokoff & Sturmey, 2004, 2008), deliver mand training (Nigro-Bruzzi & Sturmey, 2010), and conduct preference asesments (Lavie & Sturmey, 2002). In addition, BST 14 has been used succesfully to train paraprofesional staf to conduct more naturalistic teaching procedures such as imbed teaching into everyday routines (Schepis, Reid, Ownbey, & Parsons, 2001), teach adaptive skils (Palmen, Didden, & Korzilius, 2010), and provide alternative and augmentative communication instruction (Wood, Luiseli, & Harchik, 2007). Although BST has been used extensively in the staf training literature, the explicit use of BST in the parent training literature is les common. In fact, only thre studies in the present review ere identified as stating the use of BST. Two of these studies trained parents to implement DT (Lafasakis & Sturmey, 2007; Ward-Horner & Sturmey, 2008). A third study, conducted by Stewart, Car, and LeBlanc (2007) evaluated the efectivenes of BST to train family members of a boy with Asperger?s disorder to implement a BST package targeting social skils. Many more studies have investigated training packages that include some form of the same components used in BST to train parents to implement DT (Crocket, Fleming, Doepke, & Stevens, 2007), teach imitation skils (Ingersoll & Gergans, 2007), and enhance joint atention (Rocha, Schreibman, & Stahmer, 2007). Therefore, BST and similar training packages appear to be efective in training both paraprofesional staf and parents to deliver a wide variety of interventions. In addition, there are numerous studies that used instruction, modeling, rehearsal, and fedback to teach parents how to implement naturalistic teaching strategies that target language and play skils (e.g., Coolican, Smith, & Bryson, 2010; Koegel, Symon, & Koegel, 2002; Reagon, & Higbee, 2009; Symon, 2005). For example, Gilet and LeBlanc (2006) taught parents to implement NLP through the use of didactic instruction, 15 modeling, rehearsal and both imediate and delayed fedback. Parents were able to implement NLP with fidelity following training. Furthermore, parents? implementation of the procedure resulted in increases in rates vocalizations and appropriate play. In a similar study, Coolican and colleagues (2010) used a brief training program to teach parents of children with autism to implement PRT. The training program took place over the course of thre 2-hour sesions and involved didactic instruction, modeling, rehearsal and fedback with more time spent on the later two components. Parents? fidelity scores increased following training, and the intervention resulted in increases in the children?s use of functional verbal utterances. In short, BST and similar training packages can also be used to teach parents to implement naturalistic procedures to increase language and social skils in their children. While a number of studies have demonstrated succes in training participants to master a set of target skils through the use of multicomponent training packages, the specific components that are necesary to bring about mastery of the trained material remain unknown. Additional research is necesary to identify the efective components in order to develop the most eficient and cost-efective training packages possible. The Ned for a Component Analysis As sen in the review presented above, there is ample research evidence for the efectivenes of training packages. Most involve some form and combination of didactic instruction, modeling, rehearsal, and fedback; however, it is unclear which of these components are necesary and/or sufficient to efectively teach the desired skils. Schultz, Schmidt, and Stichter (2011) suggested that a component analysis of training strategies 16 has important implications for practice. Specificaly, knowledge of the components necesary and sufficient for adequate training may lead to more economical packages. Few studies have conducted systematic analyses of the efectivenes of individual BST components; however, there is evidence that some components may be more efective than others. For example, several studies have shown that didactic training is insufficient when delivered alone (Feldman, Case, Rincover, Towns, & Betel, 1989; Sterling-Turner, Watson, & Moore, 2002; Sterling-Turner, Watson, Wildmon, Watkins, & Litle, 2001). Sterling-Turner and colleagues (2002) investigated the efectivenes of a training package administered to four teachers who sought consultation for the disruptive behavior of individual students. Consultation was provided in four phases. An initial consultation period was delivered to identify the problem and gather baseline data on target students? behavior. Teachers? implementation of individualized behavior plans was then asesed after didactic training was provided. Finaly, teachers? implementation of the plan was measured after modeling, role play, and fedback were provided. The percentage of participants? total treatment integrity was low following didactic instruction with participant averages ranging from 7 percent to 70 percent. After additional training was provided, however, these scores rose to a range of 81 percent and 97 percent. Although these studies demonstrate that didactic training should be used in conjunction with other training techniques to be efective, they do not provide any information regarding which of the other training components are necesary. Other studies have suggested that rehearsal and fedback are efective strategies for training individuals to deliver behavioral services (Bolton & Mayer, 2008; Jones, Wickstrom, & Friman, 1997; Mortenson & Wit, 1998; Noel et al., 1997; Parsons & 17 Reid, 1995; Schepis, Reid, Ownby, & Parsons, 2001; Shanley & Niec, 2010). For example, Kaminski, Vale, Filene, and Boyle (2008) provided support for the importance of rehearsal in a meta-analytic review of parent training programs. They found that rehearsal, particularly rehearsal with one?s own child, was reliably asociated with higher efect sizes. There was litle diference, however, betwen studies that included other BST components such as modeling and role play and those that did not. In addition, Leblanc, Riciardi, and Luiseli (2005) used an abbreviated performance fedback intervention to improve staf?s implementation of discrete trial instruction. Training consisted of a verbal review of the discrete trial instruction checklist and fedback on the implementation of each skil. During fedback, the trainer delivered praise for corectly implemented skils and clarification and verbal direction for skils that were not implemented with 100% acuracy. No modeling, role playing, or practicing correct performance of skils took place. Staf were able to deliver discrete trial instruction with fidelity folowing the intervention, and skils were maintained 11 weks after training. Another study conducted by Mueler, Piaza, Moore, and Keley (2003) suggested that other components may be just as efective as rehearsal and fedback. These researchers investigated the efectivenes of thre diferent training packages to teach parents to implement feding protocols. After establishing the efectivenes of a full training package consisting of writen protocols, verbal instruction, modeling, and rehearsal, the authors tested various combinations of these training components. One package included writen protocols, verbal instruction, and modeling; a second package consisted of writen protocols, verbal instruction, and rehearsal; and a third package included writen protocols and verbal instruction only. To acount for the efects of mere 18 exposure to the training material, verbal instruction was delivered twice in the last training package. Al thre training packages resulted in parents? implementation of a feding program with high treatment fidelity. The authors speculated that the mode of presentation may not be as important as simply presenting the material more than one time. In short, few studies have investigated the efectivenes of the individual components of BST. There is some evidence, however, that not al components are necesary to train individuals to deliver behavioral services with integrity. Therefore, additional research is necesary to identify the components responsible for observed changes in behavior. Factors that Impact Training Suces In examining the training literature, it is clear that packages vary in length and intensity, and some learners acquire skils more readily than others. Much of the variation in acquisition may be acounted for by the particular training components delivered, but there may be other factors, such as parent characteristics, at play. Researchers have investigated predictors of atrition with mixed results. For example, some have suggested that demographic variables such as low socioeconomic status or minority status may predict dropout (Fernandez & Eyeberg, 2009; Lavigne et al., 2010). Others have found that factors such as parent stres beter acount for atrition rates (Werba, Eyeberg, Boggs, & Algina, 2006) or les than optimal outcome (Bagner & Graziano, 2012; Strauss, et al. 2012). Bagner and Graziano (2012) have suggested that multiple factors may have a cumulative impact on the lack of succes in parent training. They examined the impact of factors such as socioeconomic status, maternal education, family structure, 19 minority status, maternal inteligence, and maternal distres on both atrition and outcome. Minority status and family structure predicted dropout, and maternal education was related to outcome. Furthermore, they found that risk of dropout increased dramaticaly with each additional risk factor (Bagner & Graziano, 2012). From their results, they concluded that it is important to regularly ases risk factors and atempt to ameliorate their efects so as to increase succes and decrease atrition. Parent stres is one such risk factor that has been given considerable atention in the parent training literature. Some studies have shown that high parent stres may impede parental treatment fidelity thereby negatively impacting child outcome (Bagner & Graziano, 2012; Strauss et al., 2012). However, other studies suggest that training can decrease parent stres (McConachie & Diggle, 2007, Sanders & Woolley, 2005; Tonge et al., 2006). For example, Ken, Couzens, Musprat, and Rodger (2010) demonstrated that parent training workshops and profesional support decreased parent stres and increased parent self-eficacy to a greater extent than parents receiving similar information via self- study DVD and activity sheet package. In sum, stres has been indicated as an important factor in parent training. Whether stres negatively impacts training outcome or training positively impacts stres, research suggests that it may be beneficial to monitor levels of parent stres during training. Curent Study The primary aim of the current study was to evaluate how to most efectively train parents to implement a behavioral social engagement procedure similar to PRT. In addition, a systematic analysis of the components of the training package was conducted to determine which components primarily acounted for the behavior change and to 20 identify the most eficient method of training possible. A secondary aim of this study was to consider parent afect, including observed parent stres, and how it might be related to skil acquisition. 21 Method Participants Parents of children with ASD were recruited from locations that serve children with developmental disabilities (e.g., speech and language treatment facilities, occupational therapists offices, pediatricians' offices, schools, the university clinic, other local clinicians). Participants al had a child with a developmental delay or social skils deficit betwen the ages of 2 and 7 years. A total of 10 caregiver/child dyads were recruited for the study. One dyad was unable to participate due to scheduling conflicts. One dyad completed baseline sesions but discontinued participation before training began due to dificulty traveling to the training site. A third dyad began training sesions, but was unable to continue for personal reasons, and they discontinued participation before any change in behavior was observed. These first thre participants wil not be discussed in the results section as there were not suficient data collected. A fourth dyad, began treatment and demonstrated improvement, but they discontinued participation for personal reasons before mastery criteria were met. Another dyad completed training, but ilnes and other family circumstances prevented them from returning for generalization and follow-up sesions. The remaining five participants completed training, generalization, and two to thre follow up sesions as scheduling permited. Assesments. Participants completed questionnaires and asesments designed to identify characteristics that may impact training. First, they completed a demographic questionnaire that requested information pertaining to the parent?s age, gender, level of 22 education, and the extent of any formal or informal training related to developmental disabilities. It also inquired about the child?s age, diagnosis, participation in prior social skils interventions, and any current services the child was receiving. Finaly, the parent was asked to identify the third party with whom the child would be interacting for generalization purposes and specify that person?s age and relationship to the child. Participants were also asked to complete the Parenting Stres Index ? Third Edition (PSI; Abidin, 1995). This is a norm-referenced asesment designed to identify stres in the parent-child relationship. It consists of 101 items and yields a total stres score as wel as scores on six child-related subscales and seven parent-related subscales. This measure has been widely used in the literature and has demonstrated adequate reliability and validity (Abidin, 1995). In order to gather information about the child participants? diagnoses, developmental level, social skils, and language abilities, a series of asesments were given. First, a measure of the child?s developmental level or inteligence was administered. One of two asesments was used depending on the child?s age and language ability. These asesments include the Mulen Scales of Early Learning (Mullen, 1995), or the Kaufman Brief Inteligence Test-Second Edition (KBIT-2; Kaufman & Kaufman, 2004). The Mullen is a standardized, norm-referenced asesment that measures a child?s skils in four domains (Visual Reception, Fine Motor, Receptive Language, and Expresive Language), and yields an Early Learning Composite score. The KBIT-2 measures verbal and nonverbal cognitive skils and provides an IQ composite score. Each the Mullen and the KBIT-2 demonstrate good psychometric properties (Kaufman & Kaufman, 2004; Mullen, 1995). 23 The Childhood Autism Rating Scale-2 nd Edition (CARS2; Scholper, Van Bourgondien, Welman, & Love, 2010) and the Pervasive Developmental Disorder- Behavior Inventory, Parent Rating Form (PD-BI PRF; Cohen & Sudhalter, 2005) were administered to confirm the diagnosis of the child participants and provide an estimate of symptom severity. The CARS2 is a 15-item behavior rating scale that is completed by an evaluator based on direct observation. It helps identify children with autism and provides a measure of symptom severity. The PDD-BI is a norm-referenced questionnaire that measures children?s functioning in communication, reciprocal social interaction, ritualistic activities, and learning skils. It is comprised of five composites scales (Aproach/Withdrawal problems, Receptive/Expresive Comunication abilities, Expresive Social Comunication abilities, Repetitive, Ritualistic, and Pragmatic Problems), as wel as an Autism Composite. Research has demonstrated sound psychometric properties for these two asesments (Cohen, 2003; Cohen, Schmidt- Lackner, Romanczyk, & Sudhalter, 2003; Scholper et al., 2010). Parents also completed the Vineland Adaptive Behavior Scales, Second Edition (VABS-I; Sparow, Cicheti, & Bala, 2005). This is a norm-referenced rating scale that provides measures of the child?s adaptive comunication, daily living, and socialization skils relative to same-age peers as wel as a measure of maladaptive behaviors. Research has demonstrated adequate psychometric properties of the VABS-II (Sparow et al., 2005). A preference asesment was also conducted with al child participants in order to identify activities in which they would readily engage as wel as potential reinforcers that 24 could be delivered contingent on appropriate behavior. The format of the asesment (e.g., fre operant, paired stimulus, multiple stimulus without replacement, etc.) was chosen acording to the needs of each child. Because of his advanced verbal repertoire, Child 2 was able to give a verbal report of possible prefered items and activities at the start of treatment. He then participated in a brief multiple stimulus without replacement at the start of each sesion. Participant characteristics. Descriptions of each participant dyad appear in the paragraphs below. Se Table 1.1 for a summary of demographic information and asesment results for adult participants, and Table 1.2 for child demographics and asesment results. Table 1.1 Adult Demographics and Assesment Results Participant Age Relationship to Child Ethnicity Education DD-related training PSI %ile Parent 1 32 Mother African American Bachelor's Degre none 90-95 Parent 2 52 Grandmother Caucasian Associate's Degre none 90-95 Parent 3 34 Mother African American Master's Degre Observation only 45-50 Parent 4 32 Mother Caucasian Some Colege none N/A Parent 5 32 Mother Serbo- Croatian Some Doctoral- level Training none 60-65 Parent 6 30 Mother African American Bachelor's Degre none 90-95 Parent 7 43 Mother Caucasian Some Master's- level Training Yes >99 Note: A sumary of the demographic information and asesment results for adult participants. 25 Table 1.2 Child Demographics and Asesment Results Participant Age Gender Diagnosis KBIT- 2 Mulen CARS2 PDD- BI VABS-II Adaptive VABS-II Maladaptive Child 1 7y 1m F Multiple Disabilities N/A N/A 55 N/A 45 Elevated Child 2 6y 5m M Autism 93 N/A 28 70 101 Clinicaly Significant Child 3 6y 5m M Autism 41 N/A 30 56 78 Elevated Child 4 5y 4m M Autism N/A Very Low 36.5 N/A N/A N/A Child 5 4y 10m F Agenesis of the Corpus Calosum 91 N/A 24 25 91 Average Child 6 4y 8m M PDD-NOS 65 N/A 31 40 111 Elevated Child 7 2y 5m M PDD-NOS N/A Very Low 39.5 N/A 69 N/A Note: A sumary of demographic information and asesment results for child participants. Scores on the KBIT-2 and VABS-II Adaptive Composite are standard scores with a mean of 10 and a standard deviation of 15. Descriptive categories are listed for the Mulen as participants scored to low to calculate a score for some or al of the domains. Scores on the CARS2 below 30 sugest "Minimal-to-No symptoms" of ASD; 30-36.5 suggest "Mild- to-Moderate symptoms" of ASD; and 37 and above sugest "Severe symptoms" of ASD. The notation "N/A" is used to indicate scores that were not calculated because the asesment was not administered, the parent did not return the measure, or there was not sufficient data to calculate score. Parent 1 was a 32-year-old, African American female who was maried with two children. She had earned a bachelor?s degre and worked as a Registered Respiratory Therapist and she reported no previous training relevant to developmental disabilities. Her total stres score on the PSI fel betwen the 90 th and 95 th percentiles. Her daughter, Child 1, was a 7-year, 1-month-old African American female with multiple disabilities including developmental delays, autism, and epilepsy, al diagnosed by a neurologist. She 26 was nonverbal and only emited nonfunctional groans and grunts. Intelectual testing could not be completed as she did not atend to asesment stimuli or respond to verbal instructions. Likewise, an Autism composite score on the PD-BI could not be calculated as sections related to expresive communication could not be scored. Child 1?s CARS2 score indicated that she exhibits severe symptoms of ASD. Her mother?s ratings on the VABS-I suggested that her adaptive behavior skils were low and her maladaptive behaviors were elevated. Parent 2 was a 52-year-old Caucasian woman who was maried. She and her husband serve as the legal guardians to their two grandchildren. Parent 2 earned an Asociate?s degre in acounting and two profesional certificates and she worked as a medical transcriptionist. Parent 2 reported no training relevant to developmental disabilities. Her score on the PSI indicated that her level of stres fel betwen the 90 th and 95 th percentiles. Her grandson, Child 2, was a 6-year, 5-month-old Caucasian male who was diagnosed with autism by his pediatrician. Child 2?s KBIT-2 score suggested that his intelectual functioning was within the average range. His CARS2-HF score revealed mild to moderate symptoms of ASD. His PD-BI Autism Composite score was high as compared to other children with ASD. His VABS-I scores indicated adequate adaptive living skils and a clinicaly significant level of maladaptive behaviors. Parent 3 was a 34-year-old, African American woman who was maried with thre children. Her highest level of education was a master?s degre, and she worked as a physical therapist. Parent 3 reported that she did not have any training related to developmental disabilities, but she had observed and actively participated in her son?s speech, occupational therapy, ABA, and special education services over the past thre 27 years. Her total stres score on the PSI fel betwen the 45 th and 50 th percentiles, suggesting that she experiences average levels of stres as compared to the general population. Her son, Child 3, was a 6-year, 5-month-old African American male who was diagnosed with autism by a developmental neurologist. His KBIT-2 scores revealed intelectual functioning in the lower extreme range. His CARS2 score indicated the presence of mild to moderate symptoms of ASD, and his PD-BI Autism Composite was typical of a child with ASD. Child 3?s VABS-I score indicated moderately low levels of adaptive behavior and elevated levels of maladaptive behavior. Parent 4 was a 32-year-old Caucasian woman who was separated and had two children. She had completed some college and was a stay-at-home mother. She had not received any training related to developmental disabilities. Her son, Child 4, was a 5- year, 4-month-old Caucasian male with autism. An Early Learning Composite score on the Mullen could not be calculated because his performance was too low, and his CARS2 score was at the high end of the mild to moderate range of symptoms of ASD. Despite numerous atempts during and after the study to have Parent 4 complete the asesment packet, she did not return a completed PSI, PD-BI, or VABS-I; therefore, scores for those asesments are not available. Parent 5 was a 32-year-old, female originaly from Serbia. She was maried with two children, and the family was bilingual in Serbo-Croatial and English. Parent 5 had earned a master?s degre in biochemical engineering, worked as a regulatory manager, and was atending school to earn a doctoral degre in biochemical engineering. She reported no training related to developmental disabilities and she was experiencing average levels of stres as indicated by her scores on the PSI faling betwen the 60 th and 28 65 th percentiles. Her daughter, Child 5, was a 4-year, 10-month, Serbian female who was diagnosed with agenesis of the corpus calosum. Her scores on the KBIT-2 indicated average intelectual functioning. Her CARS2 score revealed minimal symptoms of ASD, and her PD-BI Autism Composite score suggested fewer symptoms of ASD than is typicaly observed in children with the disorder. Although she did not have autism, Child 5 met criteria for the study in that she was diagnosed with a developmental delay, and she exhibited deficits in social interaction. Parent 6 was a 30-year-old, African American female. She was maried with thre children and was a stay-at-home mother. She had earned a bachelor?s degre in marketing and had no prior training specific to developmental disabilities. Parent 6?s scores on the PSI suggest that her level of stres fals betwen the 90 th and 95 th percentiles. Her son was a 4-year, 8 month-old male who was diagnosed with Pervasive Developmental Disorder ? Not Otherwise Specified by his pediatrician. His scores on the KBIT-2 revealed his level of intelectual functioning in the lower extreme. His CARS2 score indicated minimal symptoms of ASD, and his Autism Composite score on the PD-BI fel just within the low end of the range typical for children with ASD. Child 6?s adaptive skils were adequate, and he exhibited elevated levels of maladaptive behaviors acording to ratings on the VABS-II. Parent 7 was a 43-year-old Caucasian female. She was maried with one child. Parent 7 had earned a bachelor?s degre in rehabilitation services, and she had taken graduate level clases in ABA. She had previously worked in several diferent group homes for adolescents with developmental disabilities, but was a stay-at-home mother at the time of her participation in the study. Her scores on the PSI indicated levels of stres 29 that were above the 99 th percentile. Her son was a 2-year, 5-month-old male who had been diagnosed with Pervasive Developmental Disorder ? Not Otherwise Specified by a provider who specialized in diagnostic and consultation services for children with ASD. An Early Learning Composite score on the Mullen could not be calculated because his performance on the receptive and expresive language domains was too low to score. Similarly, an autism composite score on the PD-BI could not be calculated as sections related to expresive communication could not be scored, but Child 7?s CARS2 score suggested the presence of severe symptoms of ASD. His adaptive behavior composite score fel in the low range. Seting and Materials Sesions were video recorded with either a hand-held camera on a tripod or a built-in video monitoring system. Sesions took place in one of two locations. First, some participants came to the university clinic where sesions took place in a clinic playroom containing a child-sized table and chairs, a smal adult-sized table, and bookshelves with a variety of age-appropriate toys (e.g., Mr. Potato Head, blocks or Legos, a dollhouse with furniture and dolls, etc.). Alternatively, some participants atended sesions at a speech and hearing clinic where training took place in smal treatment rooms that contained an adult-sized desk and chairs, a child-sized table and chairs, and cabinets that housed treatment materials. The trainer aranged age-appropriate toys and games on the floor and on the child-sized table. There was at least one individual available to serve as a play partner during al sesions. An undergraduate research asistant served as the play partner during treatment sesions in order to control for extraneous variables that might impact training. Whenever possible, a typicaly developing peer was recruited to serve as 30 a play partner during baseline, generalization, and follow-up phases. Typicaly, the peer selected for participation was a sibling, another family member, or a family friend. Design and Procedure In order to demonstrate functional control of the training package, a non- concurrent multiple baseline design across participants was utilized. To further analyze the relative contribution of each component of the training package, each component was presented systematicaly, using an ABC design, until measures of participants? implementation of the procedure were stable. For the purpose of this study, stability was defined as at least thre sesions with data points within 10% of each other and a stable or decreasing trend. Participants reached mastery criterion when they implemented the procedure with 90% fidelity in thre consecutive sesions. In applied setings, components of BST are often delivered in two groups or phases. In the first phase, the trainer instructs and demonstrates the procedure to the trainee with didactic instruction and modeling. This phase often does not require the presence of the child. In the second phase, the traine is offered a chance to implement the procedure and receive more personalized instruction with rehearsal and fedback. Because components are often combined in this manner, the component analysis focused on these two combinations. Identifying the efectivenes of each of these combinations is clinicaly significant as it wil help determine whether children must be present for training to be efective or whether parents can be trained independently. Furthermore, within the later condition, fedback and rehearsal were further separated such that participants received fedback alone first followed by fedback plus rehearsal if necesary. 31 Evidence presented by Mueler, Piaza, Moore, and Keley (2003) suggests that multiple components may be equaly as efective when presented with didactic training, thus, it was important to acount for order efects. As such, a modified counterbalancing procedure was used. Participants were divided into two groups. The first group of participants received didactic instruction and modeling first followed by fedback, and the second group received the same components in reverse order. Fedback plus rehearsal was introduced following fedback alone for participants who did not mastery criteria and showed stable acuracy scores during fedback. Parent Training Procedures Baseline sesions. During baseline sesions, participants received a writen description of the procedure to review for ten minutes prior to the start of the sesion. They were then instructed to encourage the child and the play partner to play together following the procedure to the best of their ability. A variety of age-appropriate toys were available, but no specific instructions regarding the use of the toys was provided. Participants did not receive fedback on their eforts at facilitating play. Training sesions. Training sesions lasted no longer than 30 minutes. During each sesion, a training probe during which the participant conducted the procedure under baseline conditions took place during the first 10 minutes for data collection purposes. Training activities (i.e., didactic instruction, fedback, etc.) took place during the second portion of the sesion. Descriptions of the components are presented below. Didactic Instruction and Modeling. In these sesions, the trainer presented the rationale for the procedure, verbaly reviewed the step-by-step instructions, and provided examples. The trainer then modeled the procedure in a role-play with the parent playing 32 the part of his or her child. Finaly, the trainer answered any questions posed by the participant except those directly related to the participant?s own performance so as to avoid providing fedback. The participant was then instructed to conduct the procedure until stability criteria were met. Instructions were repeated as often as requested by the participant, but no additional fedback was provided. Fedback. During the first fedback sesion, the participant was instructed to conduct the procedure for approximately 15 minutes during which the trainer offered in vivo coaching and imediate fedback. After this 15-minute period the trainer provided fedback acording to the steps of efective fedback (Bolton & Mayer, 2008; Parsons & Reid, 1995; Schepis, Reid, Ownbey & Parsons, 2001). For subsequent fedback sesions, the participant conducted the procedure for the data collection period and received fedback on his or her performance after the data collection period ended. Fedback plus Rehearsal. These sesions were identical to fedback sesions with the addition of the parents? rehearsal of any incorrectly performed treatment components in a role play with the trainer playing the part of the child. Generalization. Probes to ases generalization of parents? implementation of the procedure in a diferent seting and/or with a diferent play partner were conducted. When available, a child peer (e.g., a sibling, neighbor, or friend of the participant) served as a play partner. The first generalization probe took place under baseline conditions. If the parent did not implement the procedure with 90% fidelity during this probe, he or she received the training package that was found to be efective during the component analysis in the new seting or with the new play partner. 33 Follow up sesions. Two to thre follow-up sesions were conducted under the same conditions as baseline betwen two and six weks after completion of training as time and scheduling alowed. Dependent Variables Dependent variables included a rating of parents? acuracy of implementing the social engagement procedure as wel as a rating of the parents? afect during interactions. These variables are described in more detail in the following sections. Parent Acuracy. Data were collected during 5- to 8-minute observation sesions, depending on the needs of the participant. Specificaly, higher functioning participants, or those children who could sustain atention in play activities for longer periods of time, participated in 8-minute sesions, while others (i.e., Dyad 1and Dyad 7) participated in 5-minute sesions. A 30-second, partial interval recording system was used to rate parents? performance of the following behaviors: 1.) Bring children within arm?s reach, 2.) Talk up items or activities, 3.) Verbaly prompt an appropriate interaction, 4.) Model an appropriate interaction, 5.) Physicaly prompt the interaction, 6.) Reinforce appropriate behavior (Se Table 2 for operational definitions). During each 30- second interval, each step of the social engagement procedure was coded as ?correct? if every instance of the behavior in the interval was correct, ?incorect? if one or more instances of the behavior was performed incorrectly, or ?no opportunity? if the parent did not have an opportunity to engage in the behavior during that interval. An overal acuracy score was calculated by dividing the number of correct intervals by the sum of the correct and incorrect 34 Table 2 Operational Definitions of Adult Behaviors Behavior Corect Incorrect No Opportunity Have 2 or more children in a group The adult prompts any unengaged child(ren) to join the group within 10 seconds of the noticing the unengaged child(ren) The child(ren) are not engaged for more than 10 seconds All child(ren) present are engaged with the group and no child(ren) are excluded from the group throughout the interval Bring children within arm?s reach The children must be within an arm?s reach of each other and/or the adult The child(ren) are further than an arm?s reach for more than 10 seconds with no atempt by the adult to bring the child(ren) back to the group All child(ren) are within an arm?s reach of each other and/or the adult throughout the interval Talking up items or activities Talking enthusiasticaly about the item or activity and encouraging children to get involved Siting quietly while the children play with no attempts to engage children in the activity All children are actively engaged in the activity Verbaly prompt an appropriate interaction Verbaly instructing one or more of the children in the group to perform an action or emit an appropriate verbalization during an interval No prompts ocur during the interval The children are actively engaging with each other independently and no prompt is necessary Model an appropriate interaction Wait 5 seconds after the verbal prompt then model an action or appropriate verbalization Models after the child has emited the behavior, does not wait 5 seconds after the verbal prompt, waits longer than 10 seconds after the verbal prompt, or does not model an appropriate interaction The child performed the behavior independently or folowing a verbal prompt Physicaly prompt the interaction Wait 5 seconds after the model prompt then physicaly prompt an action or appropriate verbalization Physicaly prompts after the child has emitted the behavior, does not wait 5 seconds after the model prompt, waits longer than 10 seconds after the model prompt, or does not physicaly prompt The child performed the behavior independently or folowing the verbal or model prompt Reinforce appropriate behavior Deliver a reinforcer (to be determined on an individual basis) immediately following any appropriate behavior Does not deliver a reinforcer within 5 seconds of the apropriate behavior, or delivers a reinforcer folowing an inapropriate behavior The child does not perform an appropriate behavior 35 intervals and multiplying by 100. To ensure that participants had an adequate number of opportunities to run the procedure, the research asistant serving as the play partner was instructed to disengage from the social interaction approximately once per minute. Data were graphed and visualy inspected. Afect Ratings. Each sesion, parents? level of happines, interest, and stres was rated based on a scale developed by Koegel, Symon, and Koegel (2002). Each of the thre components was rated on a 6-point Likert scale ranging from 0 to 5. For happines and interest, higher scores suggested more positive afect. Specificaly, a score of 0 or 1 indicated a negative interaction style (i.e., discontent, limited interaction); a score of 2 or 3 indicated a neutral interaction style (i.e., neither happy nor unhappy, a moderate number of interactions); and a score of 4 or 5 indicated a positive interaction style (i.e., smiles or laughs, frequent interaction). In the current study, stres was rated such that higher scores indicated higher levels of stres. For example, a score of 0 or 1 indicated the presence of few indicators of stres and a relaxed interaction style, a score of 2 or 3 suggested that the parent was neither stresed nor relaxed, and a score of 4 or 5 indicated that the parent was tense or frustrated. Interobserver Agrement Parent Acuracy. For data collection purposes, undergraduate research asistants were trained on scoring criteria to 90% agrement. During training probes, the trainer or a research asistant coded adult behavior in vivo for treatment decision-making purposes. Later, a second independent coder scored video-taped sesions to obtain interobserver agrement (IOA) for at least 30% of al sesions for each participant evenly distributed throughout baseline, treatment, follow up, and generalization phases. Point-by-point 36 agrement was calculated by dividing the number of agrements by the number of agrements plus disagrements and multiplying by 100 to obtain a percentage. An agrement was defined as both raters giving the same score for each behavior in each interval. The average IOA was 95.74% (range = 88.50-100.00) across al sesions. Afect Ratings. A second independent rater also provided afect ratings for at least 30% of al sesions for each participant. To calculate IOA, an agrement was defined as both observers? scores being within one point of each other. Percent agrement was calculated by dividing the number of agrements by the number of agrements plus disagrements and multiplying by 100. The average IOA was 94.59% (range = 67.00- 100.00) across al sesions. Procedural Integrity Procedural integrity was calculated for at least 30% of al sesions for each participant evenly distributed across al phases of the study. A series of behaviors for each training component was identified (se Table 3). Each behavior was coded as ?correct? if it was present and ?incorrect? if it was absent. A procedural integrity score was calculated by dividing the number of correct behaviors by the number of correct plus incorrect behaviors and multiplying by 100 to obtain a percentage. The average procedural integrity score was 98.43% (range = 60.00-100.00). Table 3 Trainer behaviors Didactics Cover al of the topics listed on the training shet Present the step-by-step instruction shet and verbaly review each step Provide an oportunity to ask questions and answer al questions presented by the participant Feedback (from Parsons & Reid, 1995; Schepis et al., 201) Positive or empathetic general statement about the teaching sesion Praise for identifying and creating oportunities to teach and performing teaching skils corectly 37 Identify teaching skils that may have ben performed incorectly Describe how to corectly perform those skils Provide an oportunity to ask questions about the fedback and answer any questions posed Offer a final positive or encouraging statement Modeling Demonstrate how to run the procedure with the children for 5 minutes Role play Play the part of the child Respond to the participants prompts both corectly and incorectly Engage in behaviors similar to those observed by the target child with whom the participant is working Social Validity Social validity, or the degre to which the community, individual, or family finds a measure or treatment aceptable, is an important component of any treatment evaluation (Schwartz & Baer, 1991). Therefore, participants were asked to evaluate the training package and the trained intervention via a questionnaire. Specificaly, participants were asked to rate the extent to which the training package was aceptable and whether they felt they learned how to more efectively promote social engagement as a result of the training package. In addition, they were asked to rate the aceptability and efectivenes of each of the components individualy. Finaly, participants were asked to rate the extent to which the child?s social skils improved following treatment. Participants ranked the extent to which they agred with each statement on a 5-point Likert scale in which 1 indicated that they agred ?very much,? while 5 indicated that they did not agre at al. Social validity questionnaires were completed by the five participants who completed al training and folow up sesions. Al five participants gave a rating of 1 or 2 for the aceptability and efectivenes of the treatment package as a whole as wel as each of its components. Thre of the five participants indicated feling beter able to teach 38 their child social skils by rating that item a 1 or 2. One participant felt somewhat able to teach social skils as indicated by a rating of 3, and one did not fel able to teach her child social skils as indicated by a rating of 4. Finaly, thre of the five participants indicated that they noticed improvements in their children?s social skils with ratings of 1 or 2, while two participants reported noticing litle to no improvement in social skils by the end of the study with ratings of 4 or 5. 39 Results Thre participants received fedback first, while four participants received didactics and modeling first. In order to examine overal trends in the data and compare efectivenes of each of the treatment conditions, the average number of sesions from the start of treatment to demonstrated mastery was calculated (Se Figure 1). Calculations were based on the first treatment sesion after baseline through the sesion in which the parent met mastery criteria, and these calculations do not include results for Parent 2 (in the fedback condition) given that she did not met mastery criteria before withdrawing from the study. Results indicate that parents who received fedback first met mastery criteria in an average of 4 sesions, while those parents who received didactics and modeling first met mastery criteria in an average of 15.9 sesions (Se Figure 1). In addition, the percent increase in average acuracy ratings from baseline in the treatment condition to each of the thre main treatment phases was calculated (Se Table 4). The average acuracy rating across sesions for parents who received didactics and modeling first (Parents 4, 5, 6, and 7) increased by 4%, 13%, 44%, and 35% from baseline to didactics and modeling respectively. Those same parents then made increases of 8%, 34%, 98%, and 46% from baseline to fedback. Finaly, the two parents who continued with fedback plus rehearsal made gains of 36% and 104% from baseline to fedback plus rehearsal. Parents who received fedback first (Parents 1, 2, & 3) increased by 144%, 106%, and 66% respectively. 40 In order to further examine the efect of each component of the training package on an individual level, acuracy scores were graphed and visualy inspected. Graphical displays of the results can be found in Figure 2.1 for parents who received fedback first and Figure 2.2 for those who received didactics and modeling first. Data for each parent- child dyad are described below. Figure 1. The average number of sesions from the first training sesion to the sesion in which mastery criteria were met for participants receiving feedback first and those receiving didactics and modeling first. Table 4 Percent Increase from Baseline to Training Phases Condition Participant Baseline to Didactics & Modeling Baseline to Feedback Baseline to Feedback plus Rehearsal Feedback Parent 1 N/A 144 N/A Parent 2 N/A 106 N/A 41 Parent 3 N/A 66 N/A Didactics and Modeling Parent 4 4 8 36 Parent 5 13 34 N/A Parent 6 44 98 N/A Parent 7 35 46 104 Note: Percent increase in average acuracy score from baseline in the treatment condition to the three main treatment phases. Participants Receiving Fedback First Dyad 1. Dyad 1 participated in thre baseline sesions, four fedback sesions, and one generalization sesion. A fre operant preference asesment revealed that Child 1 prefered to play with play food and bals. A paired stimulus preference asesment revealed Goldfish crackers to be the most prefered edible reward. Due to her severe deficits, targeted skils for Child 1 included basic interactions such as waving ?helo,? rolling a bal to her play partner, and handing a toy to her play partner. In baseline, Parent 1 scored an average of 40.13% correct (range = 36.40-43.90). Performance rose to over 90% corect after the first fedback sesion, and she met mastery criteria in the minimum number of sesions (i.e., three) required. One additional fedback sesion was conducted as it was necesary at the time to confirm the acuracy of the live coding before changing phases. Parent 1?s average acuracy score for these four sesions was 97.80% (range = 93.20-100.00) for a 144% increase from baseline to fedback. Because no typicaly developing peer was available for generalization, the generalization probe was conducted with a diferent research asistant in a slightly diferent seting (i.e., playing on the floor as opposed to playing at the table). Parent 1 performed 83.70% correct during the generalization probe. 42 Figure 2.1. Acuracy ratings of parents who received fedback first. Closed circles represent acuracy in the training condition (i.e., with a graduate research asistant as the play partner), and open circles represent accuracy in the generalization probes (i.e., with a peer as the play partner and/or in a diferent setting). Phases include baseline (BL), fedback (FB), and generalization (Gen) as well as folow up at 2, 4, and 6 weeks (Parent 3). 43 44 Figure 2.2. Acuracy ratings of parents who received didactics and modeling first. Closed circles represent accuracy in the training condition (i.e., with a graduate research assistant as the play partner), and open circles represent accuracy in the generalization probes (i.e., with a peer as the play partner and/or in a diferent seting). Phases include baseline (BL), didactics and modeling (D+M), fedback (FB), fedback plus rehearsal (FB+R), live modeling plus fedback (M+FB), coaching via bug in the ear plus modeling plus fedback (Bug+M+FB), and generalization (Gen) as wel as follow up at 2, 3, 4, and/or 6 weks. Because of Child 1?s multiple disabilities, it was evident from the start of her participation in the study that the social skils procedure would not be appropriate for her. Rather, Child 1 would require a more intensive, most-to-least prompting procedure in order to efect behavior change. The original intent was to train Parent 1 on the original procedure, then, when no progres was observed in Child 1, implement a more appropriate procedure. However, upon beginning baseline sesions with the new procedure, it became clear that Child 1?s program would require substantialy more supervision and oversight than could be offered within the constraints of the study. As such, a referal was made to a local behavior analyst, and Dyad 1 discontinued the study. Dyad 2. Dyad 2 was unable to complete the study due to family conflicts; therefore only a brief discussion of trends in the data wil be presented for this dyad. Dyad 2 completed five baseline sesions and five fedback sesions. Child 2 was able to verbaly expres his preference for activities. His preferences varied from sesion to sesion, but they often included activities with the dollhouse and dolls, puppets, cars, and action figures. Skils included giving compliments, offering a friend a turn, and acepting a friend?s choice of activity when it difered from his own. Parent 2 received an average acuracy score of 53.72% (range = 50.81-56.63) in the generalization condition and 34.38% (range = 20.38-49.31) in the treatment condition. Despite the variability in acuracy observed in baseline, treatment was initiated because the overal trend was decreasing. Parent 2?s average acuracy score 45 across the five fedback sesions rose to 70.75% (range = 48.75-84.88) for a 106% increase from baseline to fedback. Acuracy scores were highly variable at the beginning of treatment, but variability reduced in the last thre sesions with scores substantialy higher than those observed in baseline. Had Dyad 2 continued with treatment, fedback and rehearsal would have been delivered in the next sesion. Although Parent 2 did not remain in the study long enough to met mastery criteria, a treatment efect was observed with an increase in acuracy scores after fedback was provided. Dyad 3. Dyad 3 completed eight baseline sesions, including two generalization probes, eight fedback sesions, thre generalization sesions, and follow-up probes at two, four, and six weks post-training. An MSWO preference asesment revealed Child 3?s most prefered activities to be play-doh, play food, and, cars. His most prefered tangible rewards were a slinky, a toy fire truck, and Flarp. Child 3?s 4-year old, typicaly developing sister served as the play partner in the generalization seting. Targeted skils included sharing, turn taking, requesting, commenting on play activities, and giving compliments. During baseline, Parent 3 received acuracy scores averaging 54.60% (range = 59.88-49.31) in the generalization condition and 53.72% (range = 75.00-42.13) in the treatment condition. A decreasing trend was observed at the end of the baseline phase. Acuracy rose to 69.00% after the first fedback sesion then 96.00% after the second fedback sesion. Parent 3 met mastery criteria within the first four sesions after starting the fedback phase; however, an additional sesion was run before scores for the first four sesions were confirmed. There was a dip in performance during this fifth sesion, 46 with an acuracy score of 77.90%. Anecdotaly, the appointment was disrupted by the crying from an infant sibling in the next room imediately prior to and continuing part way through this sesion. As such, it is possible that distraction may acount for the dip in performance. Because of the decreasing trend observed, the fedback phase was continued. The fedback plus rehearsal phase was not introduced because the parent had mastered the procedure with fedback alone. Parent 3?s acuracy scores once again rose to over 90%, and she met mastery criteria within the next thre sesions. Parent 3 conducted the social engagement procedure during thre generalization probes with Child 3 and his typicaly developing sister. In these thre sesions, Parent 3?s acuracy score averaged 95.92% (range = 91.63-100.00). No fedback was necesary in the generalization condition. Acuracy scores remained high in both treatment and generalization conditions during 2-, 4-, and 6-wek follow up sesions with scores at an average of 94.72 (range = 88.75-100.00). During the last follow-up sesion, Parent 3 reported that she observed improved social skils in both Child 3 and his sister, who was socialy anxious. Specificaly, Parent 3 stated that Child 3 was making more independent comments during play and asking ?Wh- questions? more frequently than he did before implementation of the social engagement procedure. His sister was also reportedly making more comments and speaking more frequently to people outside of her imediate family. Five months after the family completed participation in the study, Parent 3 contacted the researcher to report that the procedure continued to be efective. She explained that the children had learned more efective ways to communicate with each other and those communication skils generalized to interactions with others. 47 Participants Receiving Didactics and Modeling First Dyad 4. Dyad 4 participated in four baseline sesions, including one generalization probe, thre didactic and modeling sesions, five fedback sesions, and five fedback and rehearsal sesions. An MSWO preference asesment revealed Child 4?s most prefered activities to be play-doh, puzzles, and Mr. Potato Head. His most prefered tangible rewards were a toy phone, and a spinning light, and his most prefered edible reward was Skitles. Child 3?s 6-year old, typicaly developing brother served as the play partner in the generalization seting. Targeted skils included turn taking, requesting, and sharing. Parent 4 obtained an average acuracy score of 57.59% (range = 55.63-59.38) during the baseline phase in the training condition with an acuracy score of 45.00% in the generalization probe. Didactic training resulted in litle improvement, with an average acuracy score of 59.90% (range = 57.81-63.25). Training continued with fedback, and Parent 4?s average acuracy score across the five fedback sesions was 62.18% (range = 52.81-71.19). With no substantial improvement in acuracy of implementation and the observation of a downward trend, fedback plus rehearsal was introduced. Across the five sesions in this phase, Parent 4?s average acuracy score was 78.08% (range = 67.88- 90.50). After the standard training, Parent 4?s acuracy scores had not met mastery criteria; therefore, additional training was offered. First, the trainer provided in vivo modeling for two sesions. In these sesions, the trainer modeled the procedure with Child 4 and the research asistant for 10 minutes while Parent 4 observed. This period of modeling was followed by additional didactics (i.e., explaining what was done during 48 modeling and why), as wel as continued fedback. In these two sesions, Parent 4 received acuracy scores of 79.63% and 66.69%. Finaly, a ?Bug-in-the-ear? device was used for nine sesions. During bug-in-the- ear sesions, the trainer provided in vivo coaching and imediate fedback through a portable bug-in-the-ear device. These sesions difered from standard fedback sesions in that the trainer provided step-by-step instruction imediately prior to the parent completing each step, in addition to imediate fedback. In this phase, Parent 3 received an average acuracy score of 91.19 (range = 80.8-96.63), and she reached mastery criteria after eight bug-in-the-ear sesions. Due to family stresors and ilnes, Dyad 4 was unable to return for generalization probes and follow-up sesions. However, anecdotal reports indicate that Parent 4 continued to use the procedure after training, and Child 4?s social communication was improving. Dyad 5. Dyad 5 completed six baseline sesions, including 3 generalization probes, four didactic and modeling sesions, six fedback sesions, one generalization sesion, and follow-up probes at two and four weks post training. An MSWO preference asesment revealed a train set and play-doh to be Child 5?s most prefered activities. Neither tangible nor edible rewards were used as verbal praise was sufficient to maintain target behaviors. Child 5?s 2-year old, typicaly developing sister served as the play partner in the generalization seting. Targeted skils including commenting on play activities, asking questions, turn taking, and sharing. Parent 5?s average acuracy score in baseline was 68.11% (range = 63.00-74.88) in the training seting and 42.90% (range = 39.00-45.13) in the generalization seting. 49 Scores did not improve substantialy after didactic and modeling sesions, with an average acuracy score of 76.91% (range = 70.50-84.50). After one fedback sesion, Parent 5?s acuracy score was 72.56%, then it rose to 96.25% after the second fedback sesion. The average acuracy score across al six fedback sesions was 90.98% (range = 72.56-100.00), and mastery criteria were met in the last thre sesions. Parent 5?s acuracy score remained high in the generalization condition at 95.43%, and no additional training was necesary. Across al six folow-up sesions, Parent 5 received an average acuracy score of 88.11% (range = 75.44-97.5), with scores over 90% in the last two probes. Dyad 6. Dyad 6 completed eight baseline sesions, including two generalization probes, four didactic and modeling sesions, ten fedback sesions, two generalization sesions, and follow-up probes at thre, five, and seven weks post-training. An MSWO preference asesment revealed play-doh, a magnetic drawing board, and coloring to be Child 6?s most prefered activities. Neither tangible nor edible rewards were used as verbal praise was sufficient to maintain target behaviors. Child 6?s 2-year old, typicaly developing sister served as the play partner in the generalization seting. Targeted skils included sharing, turn taking, requesting, and giving compliments. Average baseline acuracy scores for Parent 6 were 41.89% (range = 34.5-52.13) and 46.63% (range = 42.44-50.83) in the training and generalization setings respectively. Conditions necesitated the trainer also serve as the play partner for the first thre didactic sesions rather than conduct in vivo coding of parent behaviors as usual. An equipment malfunction during those sesions resulted in an inability to code behaviors via video. As such, there are no objective data for didactic sesions 1-3. However, the 50 trainer estimated acuracy scores for these sesions to be betwen 60% and 65% based on observations during play sesions. A fourth probe in the didactic condition was conducted to confirm Parent 6?s estimated scores while avoiding unnecesary practice efects. The score for this probe sesion was 55.94%. Because this score was below the estimated scores for the first thre didactic and modeling sesions, training continued with fedback. Parent 6?s acuracy scores were variable during the first seven fedback sesions and ranged from 60.50% to 92.69%. Because of the variability observed in these sesions, training continued with fedback, and Parent 6 met mastery criteria in fedback sesions 8-10. Her average acuracy score across al fedback sesions was 83.04% (range = 60.50-95.80). In the first generalization probe, Parent 6 received a score of 88.06%. Her acuracy score then improved to 97.69% after fedback was provided. In follow-up sesions, Parent 6?s acuracy score dropped to a range of 83.25% to 85.75%. However, after brief fedback after the first 6-wek follow-up probe, acuracy scores returned to 94.25% and 93.56% in the generalization and training conditions respectively. Upon completion of training, Parent 6 explained that she had been using the procedure with Child 6 and a same-aged, typicaly developing neighbor. She noted an increase in Child 6?s sharing with that child as wel as other peers, and she stated that she had observed him aking more frequent appropriate social comments during play. Dyad 7. Dyad 7 completed 12 baseline probes, including two probes in the generalization condition, four didactic and modeling sesions, four fedback sesions, seven fedback and rehearsal sesions, one generalization probe, and two sesions in each of 2-, 3-, and 4-wek follow-up visits. A fre operant preference asesment was 51 conducted because removal of stimuli elicited tantrums from Child 7. This preference asesment revealed cars, blocks, puzzles, and a Blue?s Clues computer to be Child 7?s most prefered activities. These activities provided opportunities for interaction, and aces to these activities was used as the reinforcer for appropriate social interactions. The typicaly developing 2-year-old son of a family friend served as the play partner in the generalization condition. Because he was an early learner and demonstrated problem behaviors (i.e., crying, flopping to the ground, and kicking) when his play was interupted, the initial focus of training was on building paralel play skils and increasing tolerance of the social overtures of others. Additional skils targeted included turn taking and requesting. Under baseline conditions, Parent 7 received an average acuracy score of 45.89% (range = 33.00-61.48) in the training condition and 39.45% (range = 33.00-45.9) in the generalization condition. Upon implementation of didactics and modeling, Parent 7?s average acuracy score rose to 61.85% (range 53.54-73.80). With stable scores in didactic sesions, fedback was implemented. After the first fedback sesion, Parent 7?s acuracy score dropped to 26.00%. As part of her fedback, Parent 7 had been instructed to prompt the play partner to interact with or give toys to Child 7 rather than the reverse as she had done previously. Consequently, Parent 7 turned her atention to prompting the play partner to engage in play behaviors but not behaviors that required a social interaction. For example, she would prompt the play partner to roll his car or stack his blocks, but she did not have Child 7 engage in these behaviors. As such, this score largely reflects the absence of prompted social interactions. After 52 receiving additional fedback, however, Parent 7?s acuracy score rose to 83.30% and remained stable in the next two sesions with scores at 79.40% and 80.00%. Because mastery criteria had not yet been met, fedback plus rehearsal was introduced. Parent 7 received an average acuracy score of 93.74% (range 80.80-100.00) across the seven sesions in this phase, meting mastery criteria in the last thre sesions. One generalization probe was conduced during which Parent 7 received an acuracy score of 86.70%. Fedback was provided in the generalization condition, but it was not possible to conduct additional generalization probes due to limited time availability of the typicaly developing peer. Follow-up sesions were conducted at two and four weks post training in the training seting and thre weks post training in the generalization seting. Parent 7?s average acuracy scores across al follow up sesions was 94.08% (range 86.60-100.00). After training, Parent 7 reported that progres towards appropriate social interactions was limited. However, she did note some improvement in requesting items both verbaly and through simple hand signs. Afect Ratings The first step in analyzing the afect rating scales was to calculate the correlation betwen each of the thre subscales. The Spearman rank order correlation coeficient was calculated to determine the strength and direction of the relationships betwen each of the thre subscales. Results revealed that happines, interest, and stres were highly correlated with each other. Specificaly, results indicated that as happines ratings increased, interest ratings also increased (? = 0.805, p ?. 01), and stres ratings decreased (? = -0.745, p ?. 01). Furthermore, as interest ratings increased, stres ratings decreased (? = -0.613, p ?. 01). Because of the high correlation betwen the two positive afect 53 ratings, namely happines and interest, only happines was examined in the current study. Stres was retained as a separate rating to provide a measure of negative afect. These data were graphed and visualy inspected. Graphical displays of the results can be found in Figure 3.1 for parents who received fedback first and Figure 3.2 for those who received didactics and modeling first. Data for each parent-child dyad are described below. Paterns of change in afect varied greatly from participant to participant. For some participants, litle to no change was evident over the course of treatment. For example, Parent 1 exhibited neutral ratings of both happines and stres throughout the duration of her participation in the study. Similarly, Parents 2, 3, and 5 demonstrated litle change in afect. Unlike Parent 1, however, these parents exhibited positive to neutral afect for the duration of their participation. More specificaly, Parent 2 received positive ratings of happines during baseline, which dropped slightly to neutral ratings once treatment began. Her ratings of stres, however, remained low over the course of her participant. Parent 3 received high happines ratings for the duration of the study, apart from thre sesions in the middle of her participation in which happines dropped to more neutral scores. Anecdotaly, these sesions correspond to those during which her infant child could be heard crying in the next room. Parent 3 similarly received low to neutral ratings of stres during her participation. Finaly, Parent 5 exhibited a similar patern to Parent 3 in that happines ratings were high, and stres ratings were low over the course of her participation. For other parents, changes in afect across phases of treatment were observed. Overal, these changes tended to be in the positive direction such that happines ratings 54 increased and stres ratings decreased by the end of participation, but specific paterns of change varied. For example, Parent 4 began with positive to neutral ratings of happines and low to neutral ratings of stres. During phases in which she was receiving fedback, happines decreased and stres increased. Anecdotaly, Child 4 exhibited problem behavior including crying, hiting, and atempting to elope during these sesions. When this parent began receiving more extensive support via a bug-in-the-ear device, her happines increased and stres decreased. Change in afect for Parent 6 was gradual with neutral happines and stres ratings in the beginning of her participation slowly, but steadily, fading to high happines and low stres ratings by the end of her participation in the study. Finaly, Parent 7 exhibited primarily high stres and low to neutral happines ratings during the baseline phase. Once treatment started, her happines ratings rose to neutral or positive, and these ratings ended in the positive range. Similarly, her stres ratings decreased to mostly low or neutral during treatment, and these ratings ended in the low range by the end of her participation. 55 Figure 3.1. Ratings of hapines and stres for parents who received fedback first. Diamonds represent happines and squares represent stres. Closed marks represent afect in the training condition (i.e., with a graduate research asistant as the play partner), and open marks represent happines in the generalization probes (i.e., with a per as the play partner and/or in a diferent setting). Phases include baseline (BL), feedback (FB), and generalization (Gen) as wel as folow up at 2, 4, and 6 weks (Parent 3). 56 Figure 3.2. Ratings of hapines and stres for parents who received didactics and modeling first. Diamonds represent hapines and squares represent stres. Closed marks represent afect in the training 57 condition (i.e., with a graduate research assistant as the play partner), and open marks represent happiness in the generalization probes (i.e., with a per as the play partner and/or in a different setting). Phases include baseline (BL), didactics and modeling (D+M), fedback (FB), fedback plus rehearsal (FB+R), live modeling plus fedback (M+FB), coaching via bug in the ear plus modeling plus fedback (Bug+M+FB), and generalization (Gen) as wel as folow up at 2, 3, 4, and/or 6 weeks. 58 Discusion The purpose of the current study was to examine the efectivenes of a behavioral skils training (BST) package for training parents to implement a behavioral social engagement procedure. It also aimed to identify the components of that training package that were responsible for behavioral change. These aims were tested through a multiple baseline design across participants to demonstrate functional control of the training package with a systematic presentation of individual components to examine their unique contributions to parent acquisition. A secondary aim of the study was to consider factors that might be related to parents? acquisition of the procedure. To this end, parent afect was coded during each sesion and paterns of change were examined. Overal, the study demonstrated that BST is efective for training parents to implement a behavioral social engagement procedure with fidelity. Al six parents who completed the training portion of the study met mastery criteria. The seventh parent, who discontinued participation prior to completing her training, demonstrated increased treatment fidelity following training despite not meting mastery criteria. This finding is consistent with previous literature demonstrating the efectivenes of BST to train a variety of teaching strategies aimed at increasing social interaction or play skils (Coolican et al., 2010; Gilet & LeBlanc, 2006; Koegel et al., 2002; Reagon, & Higbee, 2009; Symon, 2005). In addition, most parents who learned to implement the procedure with a trained research asistant were able to generalize those skils to implementing the procedure either with a diferent research asistant in a diferent seting (i.e., Parent 1) or 59 with typicaly developing peers (i.e., Parents 3, 5, 6, and 7). In this way, the current study extends the literature by teaching parents a procedure to encourage social interaction betwen their children with developmental delays and their peers. Components Analyzed The primary purpose of the current study was to conduct a component analysis of the training package as recommended by Schultz and colleagues (2011) in order to identify those components that most directly bring about behavior change, thus, leading to more economical training packages. To this end, components of traditional BST were presented systematicaly in order to beter ases the relative contribution of each component or combination of components. The analysis of these components is presented below. Didactic training was not analyzed separately because of the extensive evidence in the literature that didactic training alone is not sufficient (Feldman et al., 1989; Sterling-Turner et al., 2002; Sterling-Turner et al., 2001). As such, the first combination of components to be analyzed included didactic training and modeling in the form of role play. There has been very litle investigation of the relative efectivenes of modeling. In fact, only one study was identified as addresing this research question (Mueler et al., 2003), and it found that there was no diferences in the efectivenes of modeling, rehearsal, or repeated verbal instruction. The current study failed to support these findings. While participants who received didactics and modeling first demonstrated some slight improvements in their acuracy of implementation, the magnitude of change from baseline to training sesions was smal, and no participants approached mastery criteria after this training phase. A potential reason for this discrepancy may pertain to the 60 type of modeling utilized. Specificaly, modeling in the current study took place in the form of role play with the trainer playing the part of the adult and the parent playing the role of the child. Results may have difered if parents had observed the role play rather than participated in it. For example, two therapists ? one playing the part of the child and one playing the part of the parent ? could have modeled the procedure as was done by Muller and colleagues (2003). Nevertheles, the curent study provides valuable insight into the limited efectivenes of modeling in the form of role play with the parent. The next components to be examined were fedback and fedback plus rehearsal. Numerous studies have demonstrated that rehearsal and fedback, when presented as a brief training package, can result in skil acquisition (Bolton & Mayer, 2008; Jones et al., 1997; Leblanc et al., 2005; Mortenson & Wit, 1998; Noel et al., 1997; Parsons & Reid, 1995; Schepis et al., 2001; Shanley & Niec, 2010). The current study supported these findings in that the majority of parents met mastery criteria following the fedback or fedback plus rehearsal training conditions. Two of the thre participants who began their training with fedback (Parent 1 and Parent 3) met mastery criteria following fedback alone. Parent 2 also demonstrated marked improvement from baseline to the fedback phase of treatment despite not meting mastery criteria before withdrawing from the study. In addition, the magnitude of change from baseline to fedback was notable. Furthermore, by delivering components systematicaly, the curent study extends the literature base by demonstrating that fedback or fedback and rehearsal are more efective than didactics and modeling. In general, improvements made by participants who received didactics and modeling first were more gradual and took place over a greater number of sesions than those observed in participants who began training with 61 fedback. Two of the four participants who began with didactics and modeling (Parent 5 and Parent 6) met mastery criteria following fedback. In addition, although these parents made litle improvement from baseline to the didactics and modeling phase, the magnitude of change from baseline to the fedback phase was substantial. Despite strong evidence that fedback alone was efective for four participants, thre participants required additional training beyond fedback. One participant met criteria following fedback plus rehearsal (Parent 7), one participant required training beyond that which was originaly specified in the protocol (Parent 4), and the extent of training that one parent (Parent 2) would have needed cannot be determined. These results suggest that, as one might expect, some parents required more practice and higher levels of support than others. Future research should aim to develop means to identify such parents prior to the start of training so as to beter tailor training methods to met the needs of the parent. In sum, parents who received didactics and modeling first took an average of more than thre times as long as parents in the fedback condition to reach mastery criteria. A larger sample size and more balanced distribution of participants is necesary to strengthen the findings of the current study. However, when these results are considered with previous literature providing support for fedback as an isolated training component (Bolton & Mayer, 2008; Jones et al., 1997; Kaminski, Vale, Filene, and Boyle 2008; Mortenson & Wit, 1998; Noel et al., 1997; Parsons & Reid, 1995; Schepis et al., 2001; Shanley & Niec, 2010), they provide evidence that focusing training eforts on the provision of fedback is more efective and eficient than providing didactics and modeling. 62 Affect Previous research has suggested that parent stres may either decrease following parent training (Ken et al., 2010; McConachie & Diggle, 2007, Sanders & Woolley, 2005; Tonge et al., 2006) or negatively impact treatment outcome (Bagner & Graziano, 2012; Strauss et al., 2012). As such, the current study included ratings of parent afect in order to explore its relationship with parent acquisition of the trained procedure. Both positive afect (i.e., happines) and negative afect (i.e., stres) were examined. In general, results suggested that afect either stayed the same or improved over the course of the study. In other words, parents with moderate to high ratings of happines and moderate to low ratings of stres (Parents 1, 2, 3, and 5) retained those ratings over the course of the study. Conversely, the afect of parents with low levels of happines and high levels of stres (Parents 6, and 7) improved over the course of the study such that happines increased and stres decreased. A third patern of change was observed in Parent 4, who began with high ratings of happines and moderate to low ratings of stres, displayed more negative afect when training began, and returned to improved levels of afect once she reached mastery of the procedure. In sum, afect ratings suggest that the training package did not have a negative impact on parent afect, and these results corresponded to parent reports of social validity in that al parents rated the training package as aceptable. Because afect was not a primary target of the current study, factors that may have impacted these changes were not controlled; therefore, it is not possible to determine the cause of these changes. However, observed changes raise interesting questions with regard to potential causes. The discussion that folows wil first consider anecdotal 63 observations from the current study to formulate hypotheses for these changes then later examine evidence from the literature to further support these observations. In keeping with researchers who have found that parent stres decreases as a result of parent training, (Ken et al., 2010; McConachie & Diggle, 2007, Sanders & Woolley, 2005; Tonge et al., 2006), the first potential reason for the observed changes is that the behavioral intervention, or training package, afected not only parents ability to implement the social engagement procedure, but also impacted their afect. It was observed that changes in afect did sem to corespond to parent acquisition of the procedure. Results from Parents 6 and 7 demonstrate this point. Both of these parents demonstrated gradual improvement in procedural fidelity from baseline to post-treatment phases. Their changes in afect were also gradual, starting with moderate to low happines ratings and moderate to high stres ratings and ending with the reverse. Similarly, but in a slightly diferent manner, Parent 4?s changes also correspond to her mastery of the procedure. While she demonstrated positive afect during baseline and in the didactic and modeling phase of treatment, her afect shifted to lower levels of happines and higher levels of stres during the fedback phases, perhaps as she became more aware of her own inacuracy. Her afect shifted again, however, at the end of the study as she demonstrated mastery of the procedure. Finaly, several parents (Parents 1, 3, and 5) demonstrated litle change in afect despite marked improvements in acuracy of implementation. For parents 3 and 5, ceiling and floor efects, or poor sensitivity of the rating scale, might acount for this lack of change. The lack of change for Parent 1 suggests that perhaps another variable, such as child behavior, may be at play. This factor is discussed below. 64 As previously mentioned, it is possible that problem behaviors exhibited by the child, such as non-compliance, crying, and aggresion, may have played a role in changes in parent afect, or lack thereof. Although such problem behaviors were not tracked explicitly in the current study, anecdotal evidence suggests that these factors may wel have played a role. For example, Child 4 was observed to cry, hit and kick his mother and the adult play partner, and atempt to elope from the play area. These behaviors were not observed frequently in the beginning of the study, but they increased as his mother was instructed to prompt more interactions and folow through with her prompts. Again anecdotaly, Parent 4 was not always efective in her prompting, frequently and unnecesarily prolonging restriction to Child 4?s prefered items and activities, which often resulted in problem behavior. Towards the end of the study, Parent 4 received more direct support from the therapist and more efectively delivered reinforcement for appropriate behaviors, and Child 4 exhibited few problem behaviors. At this time, Parent 4?s afect returned to more positive levels. Similar child behavior problems were observed in Children 6 and 7. Child 6, for example, was non-compliant (i.e., did not respond to commands given by his mother), frequently eloped from the play area, and engaged in disruptions in the form of throwing toys and other materials. Child 7 cried, dropped to the floor, and pushed other people away when his play was interupted. The behaviors of both of these children semed to occur more frequently early in their participation, and they semed to decrease over time, possibly corresponding to positive changes in parent afect. Final support for the impact of problem behavior on parent afect come from the anecdotal observation of problem behaviors observed by the children whose parents did 65 not display changes in afect over time. Children 1 and 2, for example, both engaged in some problem behaviors. For Child 1, behaviors included elopement from the play area, mouthing objects, and negative vocalizations, and Child 2 exhibited behaviors such as verbal protests and non-compliance. Given their brief participation, there was litle opportunity for any significant change in problem behavior or parent afect to occur. Finaly, Children 3 and 5 engaged in very few problem behaviors during their participation. Both children were compliant and readily responded to their parents? prompts. One interesting observation is a drop in happines for Parent 3 during sesions 11-13. As previously mentioned, Parent 3?s infant daughter could be heard crying in the next room during these sesions, which may acount for this drop in afect. Parent Acquisition, Afect, and Child Behavior: Puting it Together As might be expected, some parents required training beyond fedback, and two of these parents had already received didactic training. These results suggest that there may be parent- and child-specific factors, as opposed to the form of training delivered, that may impact the extent of training that a parent wil require. Some of factors that sem to be relevant base on observations of the curent participants include the parent?s level of stres as wel as symptom severity, or extent of problem behavior of the child. Indeed, there is evidence in the literature that provides support for the impact of these factors. For example, there is evidence that high parent stres may impede parental treatment fidelity thereby negatively impacting child outcome (Bagner & Graziano, 2012; Strauss et al., 2012). Furthermore, studies have demonstrated that symptom severity and behavior problems in children with developmental delays are asociated with increased levels of parent stres, depresion (Ingersoll & Hambrick, 2011; Lecavalier, Leone, & 66 Wiltz, 2006). Finaly, behavior problems have been shown to be more predictive of levels of parent stres than are adaptive skils (Lecavalier, et al. 2006; Peters-Schefer, Didden, & Korzilius, 2012), cognitive and developmental delays (Baker, Blancher, Crnic, & Edelbrock, 2002; Baker, McIntyre, Blacher, Crnic, Edelbrock, & Low, 2003; Hering, Gray, Tafe, Tonge, Sweney, & Einfeld, 2006; Peters-Schefer et al., 2012), and diagnosis (Hering, Gray, Tafe, Tonge, Sweney, & Einfeld, 2006; Peters-Schefer et al., 2012). Given this evidence, it folows that parents whose children display high rates of problem behavior may experience higher levels of stres and may, therefore, require more training and support than other parents. Considering the evidence presented above, it is posible that examining factors such as parent stres and using that information to guide decisions about what training procedures should be utilized for each parent may ultimately lead to more efective and eficient training procedures. A prime example of how this approach may have been used in the current study can be sen with Parent 4. Specificaly, Parent 4 reported a significant number of stresors both with respect to her response to her son?s chalenges as wel as broader life stresors. Had these factors been considered earlier in training, higher levels of support may have been presented from the beginning thereby decreasing the overal time spent training and sparing the family from the stres related to inefective training. In sum, there are a number of factors that impact parent acquisition of the target behaviors. Examples include not only the methods by which material is delivered, but also factors such as parent stres and child problem behavior. Often these isues are overlooked in the behavioral literature. However, while atention to such factors may 67 sem to delay the start of addresing the main purpose of the training, taking time to addres them ay ultimately lead to more eficient training. This recommendation is based only on anecdotal evidence from the curent study and a select few studies that lend support to those observations. Future studies should examine the relationships betwen these factors more systematicaly. Limitations There are a number of limitations to this study. First is related to the number and distribution of participants. Only seven out of the ten participants recruited participated in the study and only thre of those participants received fedback first, making the groups unequal. Furthermore, one of the participants receiving fedback first withdrew from the study before demonstrating mastery of the procedure; therefore, it is not possible to determine how much additional training this participant would have needed to achieve mastery. Replication of the efectivenes of fedback or fedback plus rehearsal in a greater number of participants would strengthen the component analysis. Although this study was primarily completed to examine parent behavior, a limitation might be argued for the absence of a measure of child social initiations and responses. Without these data, the current study cannot verify that this parent- implemented treatment resulted in gains in appropriate social initiations and responses in the children. Unfortunately, in order to ensure parents were following the procedure and prompting appropriately, children were left with litle opportunity to independently initiate social interactions. Future studies might addres this problem by adding probes of child behavior with no adult prompting periodicaly over the course of the study. In addition, and perhaps more importantly, future studies should include criteria for fading 68 prompts such that children are ofered independent opportunities to engage with their peers. Despite this limitation, however, the study offers anecdotal evidence of the efectivenes of the parent-implemented intervention. For example, Parent 2 reported to the primary investigator that her son had begun to comment more frequently without prompting, and he began to ask various ?wh? questions in appropriate situations. Similarly, Child 4 was reported to have made gains in compliance and frequency of functional verbalizations by both his mother and the speech therapist that refered him to the study. Finaly, Parent 6 reported that her son demonstrated an improved ability to take turns and compromise with his peers. While these reports are not objective and may be skewed by expectation of treatment efects, they do provide some evidence that the treatment was beneficial and socialy valid. A third limitation pertains to the absence of objective data for child problem behavior. As previously discussed, problem behavior semed to play a major role in parents? ability to implement the procedure with fidelity based on anecdotal evidence. Including a continuous measure of such behaviors would alow for more firm conclusions as to their impact on the results of the current study. A final potential limitation pertains to the fact that current study quantified parent afect by using operational definitions found in Koegel, Symon, and Koegel (2002) rather than use repeated administrations of a psychometricaly sound asesment instrument. This rating scale alowed for the continuous measurement of overt behaviors asociated with parent afect while avoiding the monetary and temporal costs asociated with using some asesments. Ratings were operationaly defined to the extent possible, but it is 69 possible that parents would have reported levels of happines and/or stres that are much diferent from those estimated by an outside observer. Furthermore, the current study found that ratings of happines, interest, and stres were very highly correlated with one another, suggesting that these scales may not represent thre independent factors. Future studies might alow parents to rate their own afective experiences so as to acount and control for diferences betwen self-and observer-report. Strengths and Future Directions The current study provides a systematic investigation of the components used in a behavioral skils training package and provides evidence for fedback as the component responsible for a significant portion of the behavioral change that was observed in parents. These results wil be beneficial in designing and implementing parent training programs that are both efective and eficient. Such eficient packages may decrease cost and alow providers to deliver services to more parents than would otherwise be served. There are numerous other iterations of BST packages that can be investigated in this manner in pursuit of increasingly eficient training methods. For example, the current study used modeling in the form of role-play. It is possible, however, that diferent results would be observed if methods such as video modeling or in vivo modeling were used. It may also be beneficial to determine whether similar results could be obtained by providing fedback during role-play conditions, and whether those results would generalize to sesions with the children. Finaly, future studies might also investigate the application of this training model in a group format. In addition to providing insight into efective and eficient training strategies, the current study also incorporated a method to monitor parent afect. Although this aspect of 70 the current study was not the primary focus and the methods utilized do not alow for any direct statements regarding the connection betwen parent afect and skil acquisition, results do provide evidence that afect may be important to consider when training some parents. Future studies should more directly investigate this relationship as wel as examine whether steps taken to decrease stres can positively impact parents? ability to learn and implement skils. A final strength of the current study is related to participant demographics and the diversity of the sample. Many single subject studies in the literature related to parent training and social skils interventions for children with ASD include middle clas, Caucasian families. The current study included Caucasian, African American, and European families with a range of education and socioeconomic status. This is an important aspect to consider as it gives some evidence for the external validity of these findings. 71 References Abdalah, M. W., Greaves-Lord, K., Grove, J., Norgaard-Pedersen, B., Hougaard, D.M., & Mortensen, E. L. (2011). Psychiatric comorbidities in autism spectrum disorders: Findings from a Danish historic birth cohort. European Child & Adolescent Psychiatry, 20, 599-601. Abidin, R. R. (1995). Parenting Stres Index. Lutz, FL: Psychological Asesment Resources. American Psychiatric Asociation. (2000). Diagnostic and statistical manual of mental disorders (Fourth Edition ? Text Revision). Washington D.C.: American Psychiatric Asociation. Amr, M., Raddad, D., El-Mehesh, F., Bakr, A., Salam, K., & Amin, T. (2012). Comorbid psychiatric disorders in Arab children with autism spectrum disorders. Research in Autism Spectrum Disorders, 6, 240-248. Atwood, T. (2000). Strategies for improving the social integration of children with asperger syndrome. Autism, 4, 85-100. Bagner & Graziano, (2012). Bariers to succes in parent training for young children with developmental delay: The role of cumulative risk. Behavior Modification, 37, 356-377. Baker, B. L., Blacher, J., Crnic, K. A., & Edelbrock, C. (2002). Behavior problems and parenting stes in families of thres-year-old children with and without developmental delays. American Journal on Mental Retardation, 107, 433-444. 72 Baker, B. L., McIntyre, L. L., Blacher, K., Crnic, C. Edelbrock, C., & Low, C. (2003). Pre-school children with and without developmental delay: behavior problems and parenting stres over time. Journal of Intelectual Disability Researh, 47, 217- 230. Bauminger, N. & Kasari, C. (2000). Lonelines and friendship in high-functioning children with autism. Child Development, 71, 447-456. Belini, S. (2004). Social skil deficits and anxiety in high-functioning adolescents with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 19, 78-86. Bolton, J. & Mayer, M. D. (2008). Promoting the generalization of paraprofesional discrete trial teaching skils. Focus on Autism and Other Developmental Disabilities, 23, 103-111. Bryson, S. A., Corrigan, S. K., McDonald, T. P., & Holmes, C. (2008). Characteristics of children with autism spectrum disorders who received services through community mental health centers. Autism, 12, 65-82. Cappadocia, M. C. & Weis, J. A. (2011). Review of social skils training groups for youth with Asperger Syndrome and High Functioning Autism. Research in Autism Spectrum Disorders, 5, 70-78. Carter, A. S., Davis, N. O., Klin, A., & Volkmar, F. R. (2005). Social development in autism. In F. R. Volkmar, R. Paul, A. Klin, & D. Cohen (Eds.), Handbook of Autism and Pervasive Developmental Disorders: Vol. 1. Diagnosis, Development, Neurobiology, and Behavior. Hoboken, NJ: John Wiley & Sons. 73 Chawarska, K. & Volkmar, F. (2005). Autism in infancy and early childhood. In F. R. Volkmar, P. Paul, A. Klin, & D. Cohen (Eds.), Handbook of Autism and Pervasive Developmental Disorders: Vol. 1. Diagnosis, Development, Neurobiology, and Behavior. (pp. 223-246). Hoboken, NJ: John Wiley & Sons. Church, C., Alisanski, S., & Amanullah, S. (2000). The social, behavioral, and academic experiences of children with Asperger syndrome. Focus on Autism and Other Developmental Disabilities, 15, 12-20. Cohen, I. L. (2003). Criterion-related validity of the PD Behavior Inventory. Journal of Autism and Developmental Disorders, 33, 31-45. Cohen, I. L., Schmidt-Lackner, S., Romanczyk, R., & Sudhalter, V. (2003). The PD Behavior Inventory: A rating scale for asesing response to intervention in children with pervasive developmental disorder. Journal of Autism and Developmental Disorders, 33, 47-53. Cohen, I. L. & Sudhalter, V. (2005). The PD-Behavior Inventory. Lutz, FL: Psychological Asesment Resources. Coolican, J., Smith, I. M., & Bryson, S. E. (2010). Brief parent training in pivotal response treatment for preschoolers with autism. Journal of Child Psychology and Psychiatry, 51, 1321-1330. Crocket, J. L., Fleming, R. K., Doepke, K. J., & Stevens, J. S. (2007). Parent training: Acquisition and generalization of discrete trials teaching skils with parents of children with autism. Research in Developmental Disabilities, 28, 23-36. 74 Dib, N. & Sturmey, P. (2007). Reducing student stereotypy by improving teachers? implementation of discrete-trial teaching. Journal of Applied Behavior Analysis, 40, 339-343. DiSalvo, C. A. & Oswald, D. P. (2002). Per-mediated interventions to increase the social interaction of children with autism: Consideration of Per Expectancies. Focus on Autism and Other Developmental Disabilities, 17, 198-207. Feldman, M. A., Case, L., Rincover, A., Towns, F., & Betel, J. (1989). Parent education project II: Increasing afection and responsivity in developmentaly handicapped mothers: Component analysis, generalization, and efects on child language. Journal of Applied Behavior Analysis, 22, 211-222. Fernandez, M. A. & Eyeberg, S. M. (2009). Predicting treatment and follow-up atrition in parent-child interaction therapy. Journal of Abnormal Child Psychology, 431- 441. Frankel, R. M., Leary, M., & Kilman, B. (1987). Building social skils through pragmatic analysis: asesment and treatment implications for children with autism. In D. J. Cohen, A. M. Donnelan, & R. Paul (Eds.), Handbook of Autism and Pervasive Developmental Disorders (pp. 333-359). Chichester: John Wiley & Sons. Ghaziuddin, M. & Zafar, S. (2008). Psychiatric comorbidity of adults with autism spectrum disorders. Clinical Neuropsychiatry, 5, 9-12. Gilet, J. N. & LeBlanc, L. A. (2006). Parent implemented natural language paradigm to increase language and play in children with autism. Research in Autism Spectrum Disorders, 1, 247-255. 75 Ginsburg, G. S., La Greca, A. M., & Silverman, W. K. (1998). Social anxiety in children with anxiety disorders: Relation with social and emotional functioning. Journal of Abnormal Child Psychology, 26, 175-185. Gren, J., Gilchrist, A., Burton, D., & Cox, A. (2000). Social and psychiatric functioning in adolescents with Asperger syndrome compared with conduct disorder. Journal of Autism and Developmental Disorders, 30, 279-293. Hering, S., Gray, K., Tafe, J., Tonge, B., Sweney, D., & Einfeld, S. (2006). Behaviour and emotional problems in toddlers with pervasive developmental disorders and developmental delay: asociations with parental mental health and family functioning. Journal of Intelectual Disability Research, 50, 874-882. Howlin, P. (1997). Prognosis in autism: do specialist treatments afect long-term outcome? European Child and Adolescent Psychiatry, 6, 55-72. Ingersoll, B. & Gergans, S. (2007). The efect of a parent-implemented imitation intervention on spontaneous imitation skils in young children with autism. Research in Developmental Disabilities, 28, 163-175. Ingersoll, B. & Hambrick, D. Z. (2011). The relationship betwen the broader autism phenotype, child severity, and stres and depresion in parent of children with autism spectrum disorders. Research in Autism Spectrum Disorders, 5, 337-344. Jones, K. M., Wickstrom, K. F., & Friman, P. C. (1997). The efects of observational fedback on treatment integrity in school-based behavioral consultation. School Psychology Quarterly, 12, 316-326. Joshi, G., Pety, C., Wozniak, J., Henin, A., Fried, R., Galdo, M., Kotarski, M., Wals, S., & Biederman, J. (2010). The heavy burden of psychiatric comorbidity in youth 76 with autism spectrum disorders: A large comparative study of a psychiatricaly refered population. Journal of Autism and Developmental Disorders, 40, 1361- 1370. Kaminski, J. W., Vale, L. A., Filene, J. H., & Boyle, C. L. (2008). A meta-analytic review of components asociated with parent training program efectivenes. Journal of Abnormal Child Psychology, 36, 567-589. Kaufman, A. S. & Kaufman, N. L. (2004) The Kaufman Brief Inteligence Test-2. Minneapolis, MN: Pearson Asesments. Ken, D., Couzens, D., Musprat, S., & Rodger, S (2010). The efects of a parent-focused intervention for children with a recent diagnosis of autism spectrum disorder on parenting stres and competence. Research in Autism Spectrum Disorders, 4, 229- 241. Koegel, R. & Frea, W. D. (1993). Treatment of social behavior in autism through the modification of pivotal skils. Journal of Applied Behavior Analysis, 26, 369-377. Koegel, R. L., O?Del, M. C., & Koegel, L. K. (1987). A natural language teaching paradigm for nonverbal autistic children. Journal of Autism and Developmental Disorders, 17, 187-200. Koegel, R. L., Schreibman, L., Briten, K., Burke, J. C., & O?Neil, R. E. (1982). A comparison of parent training to direct child treatment. In R. L. Koegel, A. Rincover, & A. I. Egel (Eds.) Educating and Understanding Autistic Children. Houston: College Hil Pres. 77 Koegel, R. L., Symon, J. B., & Koegel, L. K. (2002). Parent education for families of children with autism living in geographicaly distant areas. Journal of Positive Behavior Interventions, 4, 88-103. Kohler, F. W., Strain, P. S., Hoyson, M., & Jamieson, B. (1997). Merging naturalistic teaching and peer-based strategies to addres the IEP objectives of preschoolers with autism: An examination of structural and child behavior outcomes. Focus on Autism and Other Developmental Disabilities, 12, 196-206. Lafasakis, M. & Sturmey, P. (2007). Training parent implementation of discrete-trial teaching: efects on generalization of parent teaching and child corect responding. Journal of Applied Behavior Analysis, 40, 685-689. La Greca, A. M. & Lopez, N. (1998). Social anxiety among adolescents: Linkages with peer relations and friendships. Journal of Abnormal Child Psychology, 26, 83-94. Landa, R. J. (2008). Asesment of social communication skils in preschoolers. Mental Retardation and Developmental Disabilities Research Reviews, 11, 247-252. Lavie, T. & Sturmey, P. (2002). Training staf to conduct a paired-stimulus preference asesment. Journal of Applied Behavior Analysis, 35, 209-211. Lavigne, J. V., LeBaily, S. A., Gouze, K. R., Binns, H. J., Keler, J., Pate, L. (2010). Predictors and correlates of completing behavioral parent training for the treatment of oppositional defiant disorder in pediatric primary care. Behavior Therapy, 41, 198-211. LeBlanc, L. A., Esch, J., Sidener, T. M., & Firth, A. M. (2006). Behavioral language interventions for children with autism: Comparing applied verbal behavior and naturalistic teaching approaches. The Analysis of Verbal Behavior, 22, 49-60. 78 LeBlanc, M., Riciardi, J. N., & Luiseli, J. K. (2005). Improving discrete trial instruction by paraprofesional staf through an abbreviated performance fedback intervention. Education and Treatment of Children, 28, 76-82. Lecavalier, L., Leone, S., & Wiltz, J. (2006). The impact of behaviour problems on caregiver stres in young people with autism spectrum disorders. Journal of Intelectual Disability Research, 50, 172-183. Leyfer, O. T., Folstein, S. E., Bacalman, S., Davis, N. O., Dinh, E., Morgan, J., Tager- Flusberg, H., & Lainhart, J. E. (2006). Comorbid psychiatric disorders in children with autism: Interview development and rates of disorders. Journal of Autism and Developmental Disorders, 36, 849-861. Litle, L. (2001). Per victimization of children with Asperger spectrum disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 995- 996. Locke, J., Ishijima, E. H., Kasari, C., & London, N. (2010). Lonelines, friendship quality and the social networks of adolescents with high-functioning autism in an inclusive school seting. Journal of Research in Special Education Neds, 10, 74- 81. Lovaas, O. I. (1987). Behavioral treatment and normal educational and intelectual functioning in young autistic children. Journal of Counseling and Clinical Psychology, 55, 3-9. Lovaas, O. I., Koegel, R., Simons, J. Q., & Long, J. S. (1973). Some generalization and follow-up measures on autistic children in behavior therapy. Journal of Applied Behavior Analysis, 6, 131-166. 79 Lovaas, O. I. & Taubman, M. T. (1981). Language training and some mechanisms of social and internal control. Analysis and Interventions in Developmental Disabilities, 1, 363-372. Lydon, H., Healy, O., & Leader, G. (2011). A comparison of video modeling and pivotal response training to teach pretend play skils to children with autism spectrum disorder. Research in Autism Spectrum Disorders, 5, 872-884. Lugnegard, T., Halerback, M. U., & Gilberg, C. (2011). Psychiatric comorbidity in young adults with a clinical diagnosis of Asperger syndrome. Research in Developmental Disabilities, 32, 1910-1917. Maestro, S., Muratori, F., Cavalaro, M. C., Pei, F., Stern, D., Golse, B., & Palacio- Espasa, F. (2002). Atentional skils during the first 6 months of age in autism spectrum disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 1239-1245. Matila, M., Hurtig, T., Hapsamo, H., Jussila, K., Kuusikko-Gauffin, S., Kielinen, M., Linna, S., Ebeling, H., Bloigu, R., Joskit, L., Pauls, D. L., & Moilanen, I. (2010). Comorbid psychiatric disorders asociated with Asperger syndrome/High- functioning autism: A community- and clinic-based study. Journal of Autism and Developmental Disorders, 40, 1080-1093. Matson, J. L., Dempsey, T., & LoVullo, S. V. (2009). Characteristics of social skils for adults with intelectual disability, autism and PD-NOS. Research in Autism Spectrum Disorders, 3, 207-213. 80 Matson, J. L., Matson, M., L., & Rivet, T. T. (2007). Social-skils treatments for children with autism spectrum disorders: An overview. Behavior Modification, 31, 682- 707. McConachie H. & Diggle, T. (2007). Parent implemented early intervention for young children with autism spectrum disorder: a systematic review. Journal of Evaluation in Clinical Practice, 13, 120-129. McEvoy, M. A. & Odom, S. L. (1987). Social interaction training for preschool children with behavioral disorders. Behavioral Disorders, 12, 242-251. Mortenson, B. P. & Wit, J. C. (1998). The use of wekly performance fedback to increase teacher implementation of a prereferal intervention. School Psychology Review, 27, 613-627. Mueler, M. M., Piaza, C. C., Moore, J. W., & Keley, M. E. (2003). Training parents to implement pediatric feding protocols. Journal of Applied Behavior Analysis, 36, 545-562. Mullen, E. M. (1995). Mullen Scales of Early Learning. Circle Pines, MN: American Guidance Service Inc. Myles, B. S., Bock, S. J., & Simpson, R. L. (2001). Asperger syndrome diagnostic scale. Austin, TX: Pro-Ed. Nigro-Bruzzi, D. & Sturmey, P. (2010). The efects of behavioral skils training on mand training by staf and unprompted vocal mands by children. Journal of Applied Behavior Analysis, 43, 757-761. Noel, G. H., Wit, J. C., Gilbertson, D. N., Rainer, D. D., & Freland, J. T. (1997). Increasing teacher intervention implementation in general education setings 81 through consultation and performance fedback. School Psychology Quarterly, 1, 77-88. Odom, S. L. & Strain, P. S. (1986). A comparison of peer-initiation and teacher- antecedent interventions for promoting reciprocal social interactions of autistic preschoolers. Journal of Applied Behavior Analysis, 19, 59-71. Osterling, J. & Dawson, G. (1994). Early recognition of children with autism: A study of first birthday home videotapes. Journal of Autism and Developmental Disabilities, 24, 247-257. Osterling, J. A., Dawson, G. & Munson, J. A. (2002). Early recognition of 1-year-old infants with autism spectrum disorder versus mental retardation. Development and Psychopathology, 14, 239-251. Palmen, A., Didden, R., & Korzilius, H. (2010). Efectivenes of behavioral skils training on staf performance in a job training seting for high-functioning adolescents with autism spectrum disorders. Research in Autism Spectrum Disorders, 4, 731-740. Parsons, S. & Mitchel, P. (2002). The potential of virtual reality in social skils training for people with autistic spectrum disorders. Journal of Intelectual Disability Research, 46, 430-443. Parsons, M. B., & Reid, D. H. (1995). Training residential supervisors to provide diagnostic fedback for maintaining staf teaching skils with people who have severe disabilities. Journal of Applied Behavior Analysis, 28, 317-322. 82 Peters-Schefer, N., Didden, R, Korzilius, H. (2012). Maternal stres predicted by characteristics of children with autism spectrum disorder and intelectual disability. Research in Autism Spectrum Disorders, 6, 696-706. Pierce, K. & Schreibman, L. (1997). Multiple peer use of pivotal response training to increase social behaviors of clasmates with autism: results from trained and untrained peers. Journal of Applied Behavior Analysis, 30, 157-160. Prizant, B., Weatherby, A., & Rydel, P. J. (2000). Communication intervention isues for young children with autism spectrum disorders. In B. Prizant & A. Weatherby (Eds.) Language Isues in Autism and Pervasive Developmental Disorder: A Transactional Developmental Perspective (pp. 193-224). Baltimore, MD: Brookes. Rao, P. A., Beidel, D. C., & Murray, M. J. (2008). Social skils interventions for children with Asperger?s syndrome of high-functioning autism: A review and recommendations. Journal of Autism and Developmental Disorders, 38, 353-361. Reagon, K. A. & Higbee, T. S. (2009). Parent-implemented script fading to promote play-based verbal initiations in children with autism. Journal of Applied Behavior Analysis, 42, 659-664. Reichow, B. & Volkmar, F. R. (2010). Social skils interventions for individuals with autism: evaluation for evidence-based practices within a best evidence synthesis framework. Journal of Autism and Developmental Disorders, 40, 149-166. Reichow, B., Volkmar, F. R., Cicheti, D. V. (2008). Development of the evaluative method for evaluative and determinig evidence-based practices in autism. Journal of Autism and Developmental Disorders, 38, 1311-1319. 83 Rocha, M. L., Schreibman, L., & Stahmer, A. C. (2007). Efectivenes of training parents to teach joint atention in children with autism. Journal of Early Intervention, 29, 154-172. Roeyers, H. (1996). The influence of nonhandicapped peers on the social interactions of children with a pervasive developmental disorder. Journal of Autism and Developmental Disorders, 26, 303-320. Rogers, S. J. (2000). Interventions that facilitate socialization in children with autism. Journal of Autism and Developmental Disorders, 30, 399-409. Rogoff, B. (1990). Apprenticeship in thinking. New York: Oxford University Pres. Ryden, E. & Bejerot, S. (2008). Autism spectrum disorders in an adult psychiatric population. A naturalistic cross-sectional controlled study. Child Neuropsychiatry, 5, 13-21. Sanders, M. R. & Woolley, M. L. (2005). The relationship betwen maternal self-eficacy and parenting practices: implications for parent training. Child: Care, Health, & Development, 31, 65-73. Sarokoff, R. A. & Sturmey, P. (2004). The efects of behavioral skils training on staf implementation of discrete-trial teaching. Journal of Applied Behavior Analysis, 37, 535-538. Sarokoff, R. A. & Sturmey, P. (2008). The efects of instructions, rehearsal, modeling, and fedback on acquisition and generalization of staf use of discrete trial teaching and student correct responses. Research in Autism Spectrum Disorders, 2, 125-136. 84 Schwartz, I. S. & Baer, D. M. (1991). Social validity asesments: Is current practice state of the art? Journal of Applied Behavior Analysis, 24, 189-204. Schepis, M. M., Reid, D. H., Ownbey, J., & Parsons, M. B. (2001). Training support staf to embed teaching within natural routines of young children with disabilities in an inclusive preschool. Journal of Applied Behavior Analysis, 34, 313-327. Scholper, E., Van Bourgondien, M. E., Welman, G. J., & Love, S. R. (2010) Childhood Autism Rating Scale-2 nd Edition (CARS2). Los Angeles, CA: Western Psychological Services. Schreiber, C. (2011). Social skils interventions for children with high-functioning autism spectrum disorders. Journal of Positive Behavior Interventions, 13, 49-62. Schreibman, L. (2000). Intensive behavioral/psychoeducational treatments for autism: Research needs and future directions. Journal of Autism and Developmental Disorders, 30, 373-378. Schreibman, L. & Koegel, R. L. (1996). Fostering self-management: Parent delivered pivotal response training for children with autistic disorder. In E. D. Hibbs & P. S. Jensen (Eds.), Psychosocial treatment for child and adolescent disorders: Empirically based strategies for clinical practice (pp. 525-552). Schultz, T. R., Schmidt, C. T., & Stichter, J. P. (2011). A review of parent education programs for parents of children with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 26, 96-104. Shanley, J. R. & Niec, L. N. (2010). Coaching parents to change: The impact on in vivo fedback on parents? acquisition of skils. Journal of Clinical Child and Adolescent Psychology, 39, 282-287. 85 Sparow, S. S., Cicheti, D. V. & Bala, D. A. (2005) The Vineland Adaptive Behavior Scales, Second Edition. Circle Pines, MN: American Guidance Service, Inc. Stahmer, A. C. (1995). Teaching symbolic play skils to children with autism using pivotal response training. Journal of Autism and Developmental Disorders, 25, 123-141. Stahmer, A. C. (1999). Using pivotal response training to facilitate appropriate play in children with autistic spectrum disorders. Child Language Teaching and Therapy, 15, 29-40. Sterling-Turner, H. E., Watson, T. S. & Moore, J. W. (2002). The efects of direct training and treatment integrity on treatment outcomes in school consultation. School Psychology Quarterly, 17, 47-77. Sterling-Turner, H. E., Watson, T. S., Wildmon, M., Watkins, C. & Litle, E. (2001). Investigating the relationship betwen training type and treatment integrity. School Psychology Quarterly, 16, 56-67. Stewart, K. K., Car, J. E., & LeBlanc, L. A. (2007). Evaluation of family-implemented behavioral skils training for teaching social skils to a child with Asperger?s disorder. Clinical Case Studies, 6, 252-262. Stokes, T. F. & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10, 349-367. Strauss, K., Vicari, S., Valeri, G., D?Elia, L., Arima, S., & Fava, L. (2012). Parent inclusion in Early Intensive Behavioral Intervention: The influence of parental stres, parent treatment fidelity and parent-mediated generalization of behavior targets on child outcomes. Research in Developmental Disabilities, 33, 688-703. 86 Symon, J. B. (2005). Expanding interventions for children with autism: Parents as trainers. Journal of Positive Behavior Interventions, 7, 159-173. Tantam, D. (2000). Psychological disorder in adolescents and adults with Asperger syndrome. Autism, 4, 47-62. Tantam, D. (2003). The chalenge of adolescents and adults with Asperger syndrome. Child and Adolescent Psychiatric Clinics, 12, 143-163. Thorp, D. M., Stahmer, A. C., & Schreibman, L. (1995). Efects of sociodramatic play training on children with autism. Journal of Autism and Developmental Disorders, 25, 265-282. Tonge, B., Bereton, A., Kiomal, M., MacKinnon, A., King, N., & Rinehart, N. (2006). Efects on parental mental health of an education and skils training program for parents of young children with autism: A randomized controlled trial. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 561-569. Trembath, D., Balandin, S., Togher, L., & Stanclife, R. J. (2009). Per-mediated teaching and augmentative and alternative communication for preschool-aged children with autism. Journal of Intelectual and Developmental Disability, 34, 173-186. Wang, P. (2008). Efects of a parent training program on the interactive skils of parents of children with autism in China. Journal of Policy and Practice in Intelectual Disabilities, 5, 96-104. Ward-Horner, J. & Sturmey, P. (2008). The efects of general-case training and behavioral skils training on the generalization of parents? use of discrete-trial 87 teaching, child correct responses, and child maladaptive behavior. Behavioral Interventions, 23, 271-284. Weis, M. J. & Haris, S. L. (2001). Teaching social skils to people with autism. Behavior Modification, 25, 785-802. Werner, E., Dawson, G., Osterling, J., & Dinno, N. (2000). Brief report: Recognition of autism spectrum disorder before one year of age: A retrospective study based on home videos. Journal of Autism and Developmental Disorders, 30, 157-162. Werba, B. E., Eyeberg, S. M., Boggs, S. R., & Algina, J. (2006). Predictind outcome in parent-child interaction therapy: Succes and Atrition. Behavior Modification, 30, 618-646. White, S. W., Koenig, K., & Scahil, L. (2007). Social skils development in children with autism spectrum disorders: A review of the intervention research. Journal of Autism and Developmental Disorders, 37, 1858-1868. Wolfberg, P. J. & Schuler, A. L. (1993). Integrated play groups: A model for promoting the social and cognitive dimensions of play in children with autism. Journal of Autism and Developmental Disorders, 23, 467-489. Wolfberg, P. J. & Schuler, A. L. (1999). Fostering peer interaction, imaginative play and spontaneous language in children with autism. Child Language Teaching and Therapy, 15, 41-52. Wood, A. L., Luiseli, J. K., & Harchik, A. E. (2007). Training instructional skils with paraprofesional service providers at a community-based habilitation seting. Behavior Modification, 31, 847-855. 88 Yang, T., Wolfberg, P. J., Wu, S., & Hwu, P. (2003). Supporting children on the autism spectrum in peer play at home and school: Piloting the integrated play groups model in Taiwan. Autism, 7, 437-453. Zercher, C., Hunt, P., Schuler, A., & Webster, J. (2001). Increasing joint atention, play and language through peer supported play. Autism, 5, 374-398. 89 Footnote 1 Reichow, Volkmar, and Cicheti (2008) presented a method for evaluating empirical evidence in order to establish evidence based practices in autism. Each study should be evaluated acording to a series of primary and secondary quality indicators. The primary quality indicators listed include, but are not limited to, the availability of participant characteristics, adequate definitions of independent and dependent variables, the presence of a comparison (group design) or baseline (single subject design) conditions, and appropriate data analysis techniques (i.e., statistical tests for group designs and visual analysis in single subject designs). Examples of secondary quality indicators include blind raters, treatment fidelity, and evaluation of generalization and maintenance. The evidence provided by the study is then rated as ?Strong,? ?Adequate,? or ?Weak? acording to the number of primary and secondary quality indicators included in the study. Please se Tables 1-3 of the manuscript (Reichow et al., 2008; pp. 1313-1314) for a more thorough description of the quality indicators and criteria for rating the strength of research.