The Social Role of Imitation in Autism Implications for the Treatment of Imitation Deﬁcits

Home , Imitation

Infants & Young Children Vol. 21, No. 2, pp. 107–119 Copyright c 2008 Wolters Kluwer Health | Lippincott Williams & Wilkins The Social Role of Imitation in Autism Implications for the Treatment of Imitation Deﬁcits

Brooke Ingersoll, PhD

Individuals with autism exhibit significant deficits in imitation skills. This article reviews the importance of imitation in typical development, focusing on the social function of imitation and its role in the development of social communication skills. Second, it reviews evidence suggesting an association between imitation deficits and social communication impairments in children with autism. Third, it discusses limitations of the current method for teaching imitation that targets only the learning function of imitation. Finally, it describes a new imitation intervention designed to teach the social use of imitation in young children with autism. Key words: autism, early intervention, imitation, social communication

HILDREN WITH AUTISM exhibit signifi- found effect on learning and development Ccant impairment in imitation skills. These in children with autism (Meltzoff & Gopnik, deficits have been reported on a variety of 1994; Rogers & Pennington, 1991), making it tasks including symbolic and nonsymbolic an important focus of intervention. body movements, symbolic and functional This article reviews the importance of imi- object use, vocalizations, and facial expres- tation in development. In particular, it focuses sions (for review, see Smith & Bryson, 1994; on the social function of imitation and its role Williams, Whiten, & Singh, 2004). In typi- in the development of social communication cal infants, imitation emerges early in devel- skills. Second, it reviews evidence suggesting opment (Meltzoff & Moore, 1977) and plays an association between imitation deficits and a crucial role in the development of cog- social communication impairments in chil- nitive and social communication behaviors, dren with autism. Third, it discusses limita- such as language, play, and joint attention tions of the current method for teaching im- (Rogers & Pennington, 1991). This associa- itation that targets only the learning function tion, as well as evidence for the specificity of of imitation. Finally, it describes an imitation the imitation deficit in autism (eg, Charman intervention designed to teach the social use et al., 1997; Rogers, Hepburn, Stackhouse, of imitation in young children with autism. & Wehner, 2003; Stone, Ousley, & Littleford, 1997), has led some researchers to propose ROLE OF IMITATION IN DEVELOPMENT imitation as a primary deficit that has a pro- In typical infants, imitation emerges early in development (Meltzoff & Moore, 1977) and serves 2 distinct functions: a learning Author Afﬁliation: Lewis & Clark College, Portland, Oregon. function, through which infants acquire new skills and knowledge, and a social function, This study was supported by grants from the Organiza- tion for Autism Research and Cure Autism Now. through which infants engage in social and emotional exchanges with others (Uzgiris, Corresponding Author: Brooke Ingersoll, PhD, 105B Psychology Bldg, Michigan State University, East 1981). It is through this social use of imita- Lansing, MI 48824 ([email protected]). tion that typically developing infants acquire 107 108 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008 the social communication skills that are found more sophisticated play skills (Morrison & to be deficient in children with autism. Kuhn, 1983). Imitation of peers serves to in- For example, in typical infants, early face- crease and refine peer interactions during to-face interactions with caregivers are often early childhood and remains a strong elicitor characterized by mutual or reciprocal imita- of social interest throughout childhood. tion in which both the caregiver and the in- In sum, the social use of imitation in infant engage in imitation of the others’ vocal- fancy and early childhood is associated with izations and facial expressions. It is through the development of more sophisticated social these reciprocal imitation games that infants communication skills. This research would communicate social interest in their partner suggest that a disruption in the early social (Nadel, Guerini, Peze, & Rivet, 1999; Uzgiris, use of imitation might have a significant im- 1981, 1999), develop a sense of shared affec- pact on the development of other social com- tive experience (Malatesta & Izard, 1984), and munication skills, a hypothesis proposed by engage in conversational turn-taking eventu- others (Meltzoff & Gopnik, 1994; Rogers & ally necessary during spoken communication Pennington, 1991). (Trevarthen, Kokkinaki, & Fiamenghi, 1999). Toward the end of the first year, play be- RELATIONSHIP BETWEEN IMITATION tween the infant and caregiver becomes more AND SOCIAL COMMUNICATION object focused and the infant begins to imitate BEHAVIOR IN AUTISM the caregiver’s actions with toys (eg, Uzgiris, 1990). In the second year, imitation games Indeed, the research on autism suggests a often involve affective gestures (Kuczynski, relationship between imitative performance Zahn-Waxler, & Redke-Yarrow, 1987). Among and other social communication skills, includ- mother-child dyads, imitation remains one of ing language, play, and joint attention. There the most common, stable patterns of interac- is considerable support for an association be- tion throughout early childhood (Halliday & tween imitation and language abilities in chil- Leslie, 1986). Reciprocal imitation serves to dren with autism. Dawson and Adams (1984) express interest and engagement between the found that children with autism categorized as child and caregiver (Waxler & Yarrow, 1975) high imitators verbalized to the experimenter and is a strategy through which the child significantly more than children categorized learns conventional actions with toys (Uzgiris, as low imitators. Sigman and Ungerer (1984) 1990) and affective gestures (Kuczynski et al., found that children with autism exhibited spe- 1987). cific deficits in vocal and gestural imitation Reciprocal imitation also plays a key role compared with typical children and children in early peer interactions. Performance of with developmental delay. These skills were the same act on the same object initiates correlated with receptive language for all 3 interactions between toddlers (Mueller & groups, and vocal imitation was correlated Lucas, 1975) and often results in maintained with expressive language in typical children or increased social interaction including coun- and children with autism. terimitation (Eckerman & Stein, 1990; Grusec A longitudinal study of imitation and lan- & Abramovitch, 1982). Sustained reciprocal guage in children with autism found an asso- imitation is the predominant mode of so- ciation between gestural imitation and the de- cial interaction and preverbal communication velopment of expressive language 6 months between same-aged toddlers (Baudonniere, later in young children with autism (Stone 1988; Eckerman, 1993). These imitative ex- et al., 1997). In another longitudinal study, changes appear to foster continued social in- Stone and Yoder (2001) found that motor im- teraction by communicating a common un- itation ability at age 2 significantly predicted derstanding of ongoing activities (Eckerman, language outcomes at age 4, suggesting the 1993) and play a role in the acquisition of strong correlation between motor imitation Imitation and Autism 109 and language development in children with relationship between the ability to imitate ac- autism. In contrast, after controlling for devel- tions with objects and the development of opmental age, Rogers et al. (2003) did not find higher level play behaviors; however, the re- a relationship between imitation skills (oral, lationship may be mediated by developmental object, or body) and concurrent language age age. in young children with autism. This finding Imitation deficits may further disrupt the may be due to the use of different imitation development of peer play, as early peer in- tasks (oral-motor vs verbal) or may suggest teractions are heavily based on reciprocal that the relationship between imitation and imitation with toys (Eckerman & Didow, language is mediated by other factors. 1996; Eckerman & Stein, 1990). Indeed, Stone The research also supports a relationship and Lemanek (1990) found that parents of between imitation and play skills in autism. A preschool-aged children with autism reported variety of studies have indicated that young that their child displayed significantly less children with autism are impaired on the imitation of another child at play compared imitation of functional and arbitrary actions with parents of children with developmental with play materials (eg, Charman et al., 1997; delays. 1998; DeMyer et al., 1972; Stone et al., Several studies have found a correlation be- 1997). Hammes and Langdell (1981) found tween imitation and joint attention in autism. that children with autism performed signifi- In nonverbal children with autism, Curcio cantly worse than children with learning dif- (1978) found that higher gesture imitation ficulties on imitation tasks involving 1 real performance predicted more sophisticated object and 1 imaginary object or pantomime communicative gestures. In another study, of an action with 2 imaginary objects. Libby, Abrahamsen and Mitchell (1990) found that Powell, Messer, and Jordan (1997) found that vocal imitation was highly correlated with the children with autism demonstrated specific number of pragmatic functions including joint difficulties with the imitation of a series of attention that children with autism used dur- pretend acts that formed scripts compared ing spontaneous communication. Carpenter, with typical children and children with Down Pennington, and Rogers (2002) found ob- syndrome. These findings are notable be- ject imitation and coordinated joint attention cause children with autism also display simi- were correlated in preschool-aged children lar deficits in their spontaneous play (Jarrold, with autism, with object imitation preceding Boucher, & Smith, 1993). the development of joint attention. Rogers In addition, Stone et al. (1997) found that et al. (2003) found that both object imitation for children with autism, imitation of actions and oral imitation were correlated with initiat- with objects at 2 years of age was highly cor- ing joint attention in the Early Social Commu- related with the development of play skills nication Scales (Mundy, Hogan, & Doehring, 1 year later, suggesting the ability to imitate 1996) in young children with autism after con- functional and symbolic actions is related to trolling for developmental age. the development of play skills. Ingersoll and In a more direct analysis of the relation- Schreibman (2006) found an increase in the ship, Ingersoll and Schreibman (2006) demon- use of spontaneous pretend play in 2 young strated that teaching object imitation skills to children with autism after teaching them to young children with autism increased coor- imitate actions with objects. However, Rogers dinated joint attention. Interestingly, Whalen, et al. (2003) found that object imitation was Schreibman, and Ingersoll (2006) found that not correlated with a concurrent measure of training joint attention initiations (coordi- play skills in children with autism after con- nated joint attention, showing, and pointing) trolling for developmental age, whereas it was in young children with autism resulted in in- for children with developmental disabilities. creases in object imitation. These findings These findings suggest that that there is a suggest that in autism, imitation and joint 110 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008 attention behaviors are related and increases multiple, successive trials. The adult selects in one positively affect the other. It may be specific subskills, such as individual nonver- that joint attention and imitation, particularly bal actions, from meaningful and arbitrary ac- with objects, both involve triadic engagement tions the child is not yet performing. The child and that either one can support the use of the is taught to imitate in response to the adult’s other. discriminative stimulus “Do this” through the In summary, children with autism exhibit use of explicit prompting, prompt fading, and significant deficits in imitation that are associ- contingent reinforcement with food or other ated with impairments in other social commu- artificial reinforcers. nication skills. It unclear whether imitation is Although this method is successful for mediating these relationships directly (Rogers teaching verbal and nonverbal imitation & Pennington, 1991), or whether they are due in a controlled setting (Baer, Peterson, & to some other developmental variable that is Sherman, 1967; Lovaas, Berberich, Perloff, also reflected in the measurement of imitation & Schaeffer, 1966; Metz, 1965), it has been skills. For example, imitation has also been criticized for several reasons. First, the adult- shown to be highly correlated with devel- directed nature of the instruction and tight opmental age and autistic symptoms (Rogers stimulus control can compromise the spon- et al., 2003), as well as social responsivity taneous use of skills (Carr, 1981). Second, (McDuffie et al., 2007; Rogers et al., 2003), all the highly structured teaching environment of which may affect social communication de- (Lovaas, 1977) and use of artificial reinforcers velopment more generally. It is also possible (Koegel, O’Dell, & Koegel, 1987) can prevent that engagement in specific social communi- generalization to the natural environment (eg, cation behaviors influences the development Spradlin & Siegel, 1982). Third, imitation is of imitation skills. Regardless of the direction taught in isolation, rather than in the context of the relationship, since imitation serves both of co-occurring social communicative behav- as learning tool and as social strategy, its dis- iors, making it unrepresentative of natural ruption is likely to have a profound effect adult-child interactions, and potentially limit- on learning and development (Rogers, 1999; ing its use by parents and other family mem- Rogers & Pennington, 1991). This possibil- bers (Schreibman, Kaneko, & Koegel, 1991). ity highlights the importance of interventions Most important, this approach targets only that teach imitation early in development. the learning function of imitation. Although imitation in this form may be useful for LIMITATIONS OF CURRENT IMITATION teaching other behaviors such as self-help INTERVENTIONS skills, it likely does not serve as a build- ing block for more advanced social com- The structured behavioral approach for munication behaviors. Teaching imitation’s teaching imitation skills, often referred to as social function may be especially important, discrete trial training, is a commonly used given recent research that suggests that chil- approach in early intervention programs. This dren with autism are particularly impaired approach targets imitation as a learning skill, in their spontaneous social use of imitation rather than a social skill, which, once estab- (Ingersoll, in press; McDuffie et al., 2007; lished, is used to teach other more complex Rogers et al., 2003). For example, Whiten and behaviors (Lovaas, Freitas, Nelson, & Whalen, Brown (1998) found that although individ- 1967). The learning environment is highly uals with autism were able to imitate after structured and adult directed, usually with being briefly taught to imitate through a se- the child and adult facing each other at a ries of prompts, they did not imitate spon- table (eg, Lovaas, 1987; Maurice, Green, & taneously, as well as young typical and de- Luce, 1996). Imitation is broken down into velopmentally delayed children. The authors discrete subskills, which are presented over suggested that individuals with autism are Imitation and Autism 111 capable of imitation, as evidenced by their im- significantly better in the structured-elicited itation performance in the elicited condition, and the spontaneous-instrumental condition but lack the social motivation to imitate spon- than in the naturalistic social condition (Stone taneously. et al., 2004). In a related study, Ingersoll (in Hobson and Lee (1999) compared chil- press) found that young children with autism dren with autism and developmental delay were less likely to imitate in a naturalistic on an object imitation task in which the ex- setting compared with a structured setting; perimenter modeled the action in either a however, typically developing children did “harsh” or “gentle” style. They found that al- not exhibit this discrepancy. This finding sug- though children with autism imitated as many gests that for children with autism, the ability goal-directed actions as developmentally de- to imitate in a structured setting does not layed children, they did not imitate the ex- translate to imitation in more naturalistic play perimenter’s style. They suggested that chil- settings. It also suggests that children with dren with autism have difficulty with self- autism are particularly impaired in their abil- other knowledge, which could lead to lack ity to imitate when the function of imitation of self-identification via imitation (Hobson & is purely social. McDuffie et al. (2007) also Lee, 1999), suggesting that imitation deficits found that imitation in structured-elicited and in autism may reflect difficulty with the social spontaneous-instrumental conditions were use of imitation. associated with attention-following skills, Ingersoll, Schreibman, and Tran (2003) whereas performance in the naturalistic so- found that children with autism were more cial condition was associated with reciprocal likely to imitate actions with objects that pro- social interaction, suggesting imitation serves duced a sensory effect in the form of flashing different functions in different contexts and lights and sounds than those that did not. Typ- may be mediated by different underlying ically developing children matched for men- skills (Rogers et al., 2003). tal age did not show this discrepancy and im- In summary, the current method for teach- itated all actions equally well. The typically ing imitation to young children with autism developing children also used more social may not adequately address the social func- behaviors during imitation than the children tion of imitation. It is likely that the social with autism. The authors suggested that typ- use of imitation is involved in the develop- ical children are motivated to imitate by the ment of other social communication skills social feedback (ie, eye contact with exper- (Rogers et al., 2003). Thus, intervention pro- imenter, exchange of positive affect) they re- grams that promote the social use of imita- ceive during the interaction, whereas the chil- tion would be most effective at promoting the dren with autism, who are not motivated by development of other social communication social feedback, prefer to imitate only when skills (McDuffie et al., 2007). provided with a nonsocial reward (ie, sensory feedback). TARGETING THE SOCIAL USE Stone, Ulman, Swanson, McMahon, and OF IMITATION Turner (2004) examined immediate imitation skills in young children with autism Reciprocal imitation training (RIT) is a in 3 conditions, a structured-elicited condi- naturalistic imitation intervention designed to tion, a naturalistic social condition, and a teach the social use of imitation to young chil- spontaneous-instrumental condition in which dren with autism during play (Table 1). It is children observed an experimenter activate based on a naturalistic behavioral model, and a mechanical device to produce lights and thus shares several features with other natu- sounds and were then given the opportunity ralistic behavioral approaches such as pivotal to imitate without instructions. Their study response training (Koegel et al., 1987, 1989), found that the children with autism imitated incidental teaching (Hart & Risley, 1968; 112 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008 imitative and spontaneous language. leading and following play partner. Enhancescomponents social of imitation. consistently varied, it does notdependence. lead to prompt lead enhances motivation and encouragesimitation. spontaneous Gradually introducing novel actionsplay builds skills. spontaneous gesture use of meaningful gestures. with motor planning, and encouragesvia future negative imitation reinforcement. reinforcement (praise) and tangible reinforcement (continued access to materials). Prepares child to imitate partnerreciprocity. and teaches Provides appropriate language models. Increases Teaches reciprocal turn-taking, allows child to take turns Verbal marker provides cue to imitate but, since it is Beginning with familiar actions and following the child’s Gestures are modeled within context to encourage Provides child with information on what to do, assists Encourages future imitation via positive social Increases eye contact and coordinated joint attention. The partner praises the child The partner imitates the child’s The partner describes what the The partner uses physical guidance to The partner models an action paired with a verbal marker that describes the action up to 3 times. object the child is engagedmore with. imitative, As the the partner child begins becomes actions. to intersperse novel functionally related to the child’sgestures play. Descriptive include conventional, affective, object, attribute, and action gestures. child is attending to and/orlanguage doing (eg, using “Dog simplified is walking”) or sound effects. once a minute on averagehas during successfully play. Once imitated the the child partner,returns the to partner imitating child and using linguistic mapping. encourage the child to imitatethe the child modeled does action not if spontaneousmodel. imitate after the third after imitating and allows thethe child toys. continued access to actions with toys, gestures/body movements,vocalizations and at the same time as the child. Linguistic mapping: Model: The partner begins modeling familiar actions with the The partner models a descriptive gesture that is Prompting: Contingent reinforcement: Contingent imitation: Goals, techniques, and rationale for reciprocal imitation training language responsivity Teach imitation The partner implements the imitation training procedure Increase imitative GoalIncrease child Teaching component Rationale Table 1. Imitation and Autism 113

McGee, Krantz, Mason, & McClannahan, forth or spinning its wheels, the adult might 1983), and milieu teaching (Alpert & Kaiser, model rolling it back and forth, while the child 1992; Kaiser, Yoder, & Keetz, 1992), in- is spinning the wheels. Initially, familiar ac- cluding following the child’s lead, explicit tions that are self-stimulatory, such as spinning prompting, reinforcing attempts, and natural the wheels of a car, may be modeled unless reinforcement. It also incorporates several they are disruptive. Starting with familiar ac- intervention techniques drawn from the tions, increases the child’s natural motivation developmental literature such as contingent to complete the action. Once the child begins imitation and linguistic mapping (eg, Warren, to imitate familiar actions, novel actions are in- Yoder, Gazdag, & Kim, 1993). troduced, such as crashing the car into a wall The goal of RIT is to teach imitation skills or placing a person in the car. Over time, self- within ongoing social interactions with an stimulatory actions are replaced with more adult. This approach uses several naturalistic appropriate actions. However, at all times, im- strategies to teach imitation that are designed itation is taught around the child’s current fo- to increase social responsiveness and intrinsic cus of interest. In addition, the adult uses so- motivation. First, the adult contingently im- cial praise and contingent imitation once the itates the child’s actions with toys, gestures, child imitates. These social reinforcers often body movements, and vocalizations during occur during natural adult-child imitative in- play. Duplicate sets of developmentally appro- teractions and are likely to continue outside priate toys are used in each session to facili- the treatment environment; thus, it is hoped tate contingent imitation. Research indicates that natural, social reinforcement will main- that contingent imitation enhances social tain imitation over time. responsiveness and coordinated joint atten- The second goal is for imitation to be- tion (Escalona, Field, Nadel, & Lundy, 2002; come spontaneous such that the child imi- Ingersoll & Schreibman, 2006; Lewy & Daw- tates when he or she sees others perform in- son, 1992; Tiegerman & Primavera, 1984), teresting actions rather than in response to which helps the child attend to the adult a specific verbal demand. Therefore, specific during modeling, increasing the likelihood commands such as “Do this” are avoided. In- of an imitative response. During contingent stead, modeling of the action is paired with a imitation, the adult describes the actions distinct, verbal marker that draws the child’s that the adult and child are performing using attention to the action. The verbal marker pro- simplified language. This form of indirect vides a sound effect or describes the modeled language stimulation has been shown to action and is consistently varied to prevent increase spontaneous and imitative language the child from becoming prompt dependent. in young children with autism (Ingersoll, For example, in placing a person in the car, Dvortcsak, Whalen, & Sikora, 2005; Ingersoll the adult might say, “Vroom, vroom”or “Boy is & Schreibman, 2006). driving.”Although the verbal marker provides Once the child begins to show an aware- a description of the action, specific verbal di- ness of the adult’s contingent imitation, the rections or commands (eg, “Push the car”)are child is taught to imitate the adult’s behavior. avoided. There are several goals in teaching the child The third goal is for imitation to be gen- to imitate the adult. The first goal is to main- eralized. In typical development, mutual tain imitation in the natural environment by engagement in the same activity conveys becoming intrinsically motivating. It is impor- meaning rather than the exact duplication of tant for the child to see imitation as an ef- an action. Therefore, the attempt to imitate fective strategy for both learning and inter- is more important than the accuracy of the acting. Therefore, actions are modeled that action. Thus, all attempts made by the child to are familiar and directly related to the child’s imitate are reinforced with praise. In addition, current play. For example, if the child typi- the adult models a variety of interesting ac- cally plays with a car by rolling it back and tions from the beginning rather than training 114 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008 a specific action to criterion. By targeting a dren in this study ranged in age from 29 to variety of actions at once, it is possible to keep 45 months (mean chronological age [CA] = the child interested and achieve generalized 36.6 months). All children exhibited signifi- responding with limited intervention. cant developmental delay (mean mental age If the child does not imitate on his or her [MA] = 19.1 months) language age and 2 of own after 3 opportunities, the adult uses phys- the children were nonverbal (mean language ical guidance to assist the child in complet- age [LA] = 16 months) at intake. The children ing the action. This guidance serves several received three 1-hour sessions per week of purposes. First, it helps the child learn that RIT for 10 weeks targeting object imitation. he or she is expected to imitate the model. Generalization probes were conducted at the Second, it provides the child with a mo- end of treatment and at 1-month follow-up to tor plan to complete new actions that may assess generalization of skills to novel environ- be especially helpful for those children who ments. During baseline, the children exhib- have difficulties with motor planning. Third, ited low rates of spontaneous imitation. After it can be seen as a form of negative rein- the onset of treatment, they showed substan- forcement in that children can learn to avoid tial gains in their use of object imitation. Four physical manipulation by completing the act of the 5 children maintained these gains after spontaneously. the removal of treatment and generalized their This intervention was originally designed ability to imitate actions with objects to novel to teach object imitation (Ingersoll & Schreib- play materials, a therapist, and a setting. One- man, 2006), but has recently been modified month follow-up data indicated that all chil- to target gesture imitation (Ingersoll, Lewis, & dren maintained higher than baseline rates of Kroman, 2007). When teaching gesture imita- object imitation. Changes in object imitation tion, conventional (eg, waving as a greeting; were also evident on a structured imitation palms up to indicate “Where did it go?”), af- assessment and during a naturalistic, struc- fective (eg, clapping to indicate “good job”; tured observation with the therapist and the hands on hips to indicate “mad”),and descrip- caregiver. tive (eg, arms out to indicate “big”;point in air In addition, the children made gains in to indicate “up”) gestures are specifically tar- other nontargeted, social communicative be- geted. Rather than modeling an action with haviors. All children increased their imitative an object, the adult models a gesture related language. Four of the children used more to the child’s play with a related verbal de- coordinated joint attention and pretend play. scriptor. For example, if the child is rolling For most of the children, increases in the a car back and forth, the adult might model nontargeted, social communicative behaviors a sweeping hand motion (“Car is fast”) or a maintained after treatment was withdrawn circular hand motion (“Wheels are spinning”). and at 1-month follow-up. Perhaps, more Because object imitation is easier (DeMyer exciting, na¨ıve observers blind to treatment et al., 1972) and more intrinsically motivating, status rated the participants as exhibiting the adult typically targets object imitation be- significantly more appropriate social commu- fore focusing on gesture imitation. nication skills and appearing more typically developing on a 7-point Likert-type scale at RESEARCH ON EFFICACY posttreatment than pretreatment, suggesting that this treatment led to socially valid There have been several studies conducted changes for the participants. on the effectiveness of this approach. The first A second study targeted the imitation of study used a multiple baseline design across meaningful gestures (Ingersoll et al., 2007). 5 young children with autism to examine the This study used a multiple baseline design effectiveness of RIT for teaching object imita- across 5 young children with autism who had tion (Ingersoll & Schreibman, 2006). The chil- difficulty with the imitation and spontaneous Imitation and Autism 115 use of gestures. The children in this study ing a variety of social communication skills in were slightly older (mean CA = 41.1 months) their child, simple and easy to use, as well as and had slightly higher mental ages (mean enjoyable for them and their child. These find- MA = 24.4 months) than the children in the ings suggest that RIT is appropriate as a parent object imitation study and all were verbal training intervention. (mean LA = 23.4 months) at intake. Children Hwang and Hughes (2000) used a similar received two 1-hour sessions of RIT target- intervention that incorporated contingent im- ing gesture imitation per week for 10 weeks. itation, expectant waiting, and communica- Generalization probes were conducted once tive temptations to target eye contact, motor a week throughout treatment and at 1-month imitation, joint attention in 3 young children follow-up to assess generalization of skills. with autism. They found that the children in- During baseline, the children exhibited lit- creased their use of eye gaze and imitation of tle to no episodes of gesture imitation or familiar actions that generalized to novel con- spontaneous gesture use. With the onset of texts. This study suggests that related inter- treatment, all children exhibited an increase ventions are also effective for increasing social in their imitation of gestures. The imitation communication skills in young children with generalized to novel environments and main- autism. tained at 1-month follow-up. In addition, 3 Given the simplicity of the treatment tech- children exhibited substantial gains, whereas niques, this intervention would likely be suit- the other 2 exhibited small but consistent able as a peer-mediated strategy. Peers or gains in their spontaneous use of gestures. siblings could be taught to use RIT strate- Na¨ıve observers who were blind to treatment gies with children with autism. In fact, given status rated the children as using more ap- the importance of reciprocal imitation dur- propriate social communication skills during ing peer play, it may be more important to treatment than baseline. These findings sug- teach young children with autism to imitate gest that RIT is effective for teaching gesture their peers than adults. A variety of research use and may have more global effects on so- has focused on teaching children with autism cial communication. and other developmental disabilities to imi- In a third study, 3 mothers were taught to tate specific peers behaviors (eg, Apolloni, implement RIT techniques with their child Cooke, & Cooke, 1977; Tryon & Keane, 1986). twice a week for 10 weeks in a clinic set- The majority of these interventions involve ting (Ingersoll & Gergans, 2007). The children the adult training the peer to model specific ranged in age from 31 to 42 months (mean CA actions, while the adult prompts the target = 36.7 months; mean MA = 17.3 months), child to imitate. A similar approach that in- and 2 of the children were nonverbal (mean corporates training peers to imitate the child LA = 11 months) at intake. The mothers of with autism may increase social interaction the 2 nonverbal children were taught to use in addition to imitation. In some preliminary RIT to teach object imitation and the mother work, we found that typically developing chil- 1 of the verbal child was taught to use RIT to dren as young as 2 2 could be taught to con- target both object and gesture imitation in a tingently imitate the behavior of a peer with multiple-baseline design. Generalization was autism and that the use of contingent imita- assessed in the families’ homes at the end tion corresponded to increases in the toddler of treatment and during 1-month follow-up. with autism’s use of coordinated joint atten- Findings indicated that the parents learned to tion (Ingersoll & Stahmer, 2002). use the intervention strategies and their chil- This approach may also be adapted for use dren improved their imitation skills. The chil- in group settings. In one study, Garfinkle and dren’s and parents’ skills both generalized to Schwartz (2002) taught 4-year-old children the home and maintained over time. Parents with autism to imitate their typically devel- rated the intervention as effective for improv- oping peers using a small group intervention. 116 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008

Each child in the group took turns being the nication behaviors, it is unclear whether the “leader” during which time he or she mod- children become motivated to engage in im- eled preferred actions with toys. The other itation for purely social purposes. It is quite children in the group were then prompted by possible that their imitation is driven by some the adult to imitate the leader’s behavior with other function (eg, obtain continued access to a duplicate set of toys and were praised for desired materials). Future research is needed successful imitation. Generalization sessions to determine whether RIT truly increases the taken during free play indicated that the in- social use of imitation in young children with tervention resulted in increases in social in- autism. teractions between the children with autism In addition, although RIT seems more likely and the typical peers. This study indicates that to produce social imitation and facilitate so- peer interventions incorporating contingent cial communication development than the tra- imitation and prompts for imitation may be ef- ditional structured approach, research com- fectively used in preschool classrooms. paring the 2 interventions has not yet been In sum, RIT is a novel method for teach- conducted. Future research is needed to de- ing the social use of imitation to young chil- termine which interventions are most likely dren with autism. The intervention is very to produce the best long-term outcomes in efficient, producing meaningful changes in imitation skills, whether they teach similar or imitation within 20 to 30 hours of interven- different imitative functions, as well as which tion. In addition, this intervention facilitates interventions have the broadest impact on the use of other social communication skills in development. Finally, additional research is young children, which makes it a particularly needed to determine the optimum intensity, effective intervention at this age. Although best treatment environment, and for which this approach produces spontaneous imita- children the intervention is most likely to be tion in conjunction with other social commu- effective.

REFERENCES

Abrahamsen, E., & Mitchell, J. (1990). Communica- Ackerman, D. Alexander, P. Firestone, M. Perkins, & tion and sensorimotor functioning in children with A. L. Egel (Eds.), The me book: Teaching manual for autism. Journal of Autism and Developmental Dis- parents and teachers of developmentally disabled orders, 20, 75–85. children (pp. 153–161). Baltimore: University Park Alpert, C. L., & Kaiser, A. P. (1992). Training parents as Press. milieu language teachers. Journal of Early Interven- Charman, T., Swettenham, J., Baron-Cohen, S., Cox, A., tion, 16, 31–52. Baird, G., & Drew, A. (1997). Infants with autism: An Apolloni, T., Cooke, S. A., & Cooke, T. P. (1977). Estab- investigation of empathy, pretend play, joint atten- lishing a normal peer as a behavioral model for devel- tion, and imitation. Developmental Psychology, 33, opmentally delayed toddlers. Perceptual and Motor 781–789. Skills, 44, 231–241. Charman, T., Swettenham, J., Baron-Cohen, S., Cox, A., Baer, D. M., Peterson, R. F., & Sherman, J. A. (1967). The Baird, G., & Drew, A. (1998). An experimental in- development of imitation by reinforcing behavioral vestigation of social-cognitive abilities in infants with similarity to a model. Journal of the Experimental autism: Clinical implications. Infant Mental Health Analysis of Behavior, 10, 405–416. Journal, 19, 260–275. Baudonniere, P. (1988). Evolution in mode of social ex- Curcio, F. (1978). Sensorimotor functioning and commu- change in 2, 3, and 4 year old peers. European Bul- nication in mute autistic children. Journal of Autism letin of Cognitive Psychology, 8, 241–263. and Childhood Schizophrenia, 8, 281–292. Carpenter, M., Pennington, B. E., & Rogers, S. J. (2002). Dawson, G., & Adams, A. (1984). Imitation and social re- Interrelations among social-cognitive skills in young sponsiveness in autistic children. Journal of Abnor- children with autism. Journal of Autism and Devel- mal Child Psychology, 12, 209–226. opmental Disorders, 32, 91–106. DeMyer, M., Alpern, G., Barton, S., DeMyer, W.,Churchill, Carr, E. G. (1981). Sign language. In O. I. Lovaas, A. D., Hingtgen, J., et al. (1972). Imitation in autistic, Imitation and Autism 117

early schizophrenic, and non-psychotic subnor- spontaneous imitation skills in young children with mal children. Journal of Autism and Childhood autism. Research in Developmental Disabilities, 28, Schizophrenia, 2, 264–287. 163–175. Eckerman, C. O. (1993). Toddlers’ achievement of coor- Ingersoll, B., Lewis, E., & Kroman, E. (2007). Teaching dinated action with conspecifics: A dynamic systems the imitation and spontaneous use of descriptive ges- perspective. In L. B. Smith & E. Thelen (Eds.), A dy- tures to young children with autism using a natural- namic systems approach to development: Applica- istic behavioral intervention. Journal of Autism and tions (pp. 333–357). Cambridge, MA: The MIT Press. Developmental Disorders, 37, 1446–1456. Eckerman, C., & Didow, S. (1996). Nonverbal imitation Ingersoll, B., & Schreibman, L. (2006). Teaching recipro- and toddlers’ mastery of verbal means of achieving cal imitation skills to young children with autism us- coordinated action. Developmental Psychology, 32, ing a naturalistic behavioral approach: Effects on lan- 141–152. guage, pretend play, and joint attention. Journal of Eckerman, C. O., Stein, M. (1990). How imitation begets Autism and Developmental Disorders, 36, 487–505. imitation and toddlers’ generation of games. Devel- Ingersoll, B., Schreibman, L., & Tran, Q. (2003). The effect opmental Psychology, 26, 370–378. of sensory feedback on immediate object imitation in Escalona, A., Field, T., Nadel, J., & Lundy, B. (2002). Brief children with autism. Journal of Autism and Devel- report: Imitation effects on children with autism. opmental Disorders, 33, 673–683. Journal of Autism and Developmental Disorders, Ingersoll, B., & Stahmer, A. (2002, May). Teaching peer 32, 141–144. interaction skills in toddlers with autism: Effects of Garfinkle, A. N., & Schwartz, I. S. (2002). Peer imita- contingent imitation training. In A. Stahmer (Chair), tion: Increasing social interactions in children with The role of typical toddlers in the early social devel- autism and other developmental disabilities in inclu- opment of children with autism. Symposium con- sive preschool classrooms. Topics in Early Child- ducted at the annual meeting of the Association for hood Special Education, 22, 26–38. Behavior Analysis, Toronto, Canada. Grusec, J., & Abramovitch, R. (1982). Imitation of peers Jarrold, C., Boucher, J., & Smith, P. (1993). Symbolic play and adults in a natural setting: A functional analysis. in autism: A review. Journal of Autism and Develop- Child Development, 53, 636–642. mental Disorders, 23, 281–307. Halliday, S., & Leslie, J. (1986). A longitudinal semi-cross- Kaiser, A. P., Yoder, P. J., & Keetz, A. (1992). Evaluat- sectional study of the development of mother-child ing mileu teaching. In S. F. Warren & J. Reichle interaction. British Journal of Developmental Psy- (Eds.), Causes and effects in communication and chology, 4, 211–222. language intervention (pp. 9–48). Baltimore: Paul Hammes, J., & Langdell, T. (1981). Precursors of sym- Brooks. bol formation and childhood autism. Journal of Koegel, R. L., O’Dell, M. C., & Koegel, L. K. (1987). A nat- Autism and Developmental Disorders, 11, 331– ural language teaching paradigm for nonverbal autis- 346. tic children. Journal of Autism and Developmental Hart, B. M., & Risley, T. R. (1968). Establishing use of de- Disorders, 17, 187–200. scriptive adjectives in the spontaneous speech of dis- Koegel, R. L., Schreibman, L., Good, A., Cerniglia, L., Mur- advantaged preschool children. Journal of Applied phy, C., & Koegel, L. (1989). How to teach pivotal Behavior Analysis, 1, 109–120. behaviors to children with autism: A training man- Hobson, R. P., & Lee, A. (1999). Imitation and identifica- ual. Santa Barbara: University of California. tion in autism. Journal of Child Psychology & Psy- Kuczynski, L., Zahn-Waxler, C., & Radke-Yarrow, M. chiatry, 40, 649–659. (1987). Development and content of imitation in the Hwang, B., & Hughes, C. (2000). Increasing early social- second and third years of life: A socialization perspec- communicative skills of preverbal preschool chil- tive. Developmental Psychology, 23, 276–282. dren with autism through social interactive training. Lewy, A. L., & Dawson, G. (1992). Social stimulation and Journal of the Association for Persons With Severe joint attention in young autistic children. Journal of Handicaps, 25, 18–28. Abnormal Child Psychology, 20, 555–566. Ingersoll, B. (in press). The effect of context on imitation Libby, S., Powell, S., Messer, D., & Jordan, R. (1997). Im- skills in children with autism. Research in Autism itation of pretend play acts by children with autism Spectrum Disorders. and Down’s syndrome. Journal of Autism and De- Ingersoll, B., Dvortcsak, A., Whalen, C., & Sikora, D. velopmental Disorders, 27, 365–383. (2005). The effect of a developmental, social prag- Lovaas, O. I. (1977). The autistic child: Language devel- matic language intervention on expressive language opment through behavior modification. New York: skills in young children with autism spectrum disor- Irvington. ders. Focus on Autism and Other Developmental Lovaas, O. I. (1987). Behavioral treatment and normal ed- Disabilities, 20, 213–222. ucational and intellectual functioning in young autis- Ingersoll, B., & Gergans, S. (2007). The effect of a parent- tic children. Journal of Consulting & Clinical Psy- implemented naturalistic imitation intervention on chology, 55, 3–9. 118 INFANTS &YOUNG CHILDREN/APRIL–JUNE 2008

Lovaas, O. I., Berberich, J. P., Perloff, B. F., & Schaeffer, E. (2003). Imitation performance in toddlers with B. (1966). Acquisition of imitative speech by autism and those with other developmental disor- schizophrenic children. Science, 151, 705–707. ders. Journal of Child Psychology and Psychiatry, Lovaas, O. I., Freitas, K., Nelson, & Whalen, C. (1967). The 44, 763–781. establishment of imitation and its use for the develop- Rogers, S., & Pennington, B. (1991). A theoretical ap- ment of complex behavior in schizophrenic children. proach to the deficits in infantile autism. Develop- Behaviour Research and Therapy, 5, 171–182. mental Psychology, 3, 137–162. Malatesta, C. Z., & Izard, C. E. (1984). The ontogenesis Schreibman, L., Kaneko, W. M., & Koegel, R. L. (1991). of human social signals: From biological imperative Positive affect of parents of autistic children: A com- to symbol utilization. In N. A. Fox & R. J. Davidson parison across two teaching techniques. Behavior (Eds.), The psychobiology of affective development Therapy, 22, 479–490. (pp. 161–206). Hillsdale, NJ: Erlbaum. Sigman, M., & Ungerer, J. (1984). Cognitive and language Maurice, C., Green, G., & Luce, S. C. (Eds.). (1996). skills in autistic, mentally retarded and normal chil- Behavioral intervention for young children with dren. Developmental Psychology, 20, 293–302. autism: A manual for parents and professionals. Smith, I., & Bryson, S. (1994). Imitation and action in Austin, TX: Pro-Ed, Inc. autism: A critical review. Psychological Bulletin, McDuffie, A., Turner, L., Stone, W., Yoder, P.,Wolery, M., & 116, 259–273. Ulman, T. (2007). Developmental correlates of differ- Spradlin, J. E., & Siegel, G. M. (1982). Language training in ent types of motor imitation in young children with natural and clinical environments. Journal of Speech autism spectrum disorders. Journal of Autism and and Hearing Disorders, 47, 2–6. Developmental Disorders, 37, 401–412. Stone, W., & Lemanek, K. (1990). Parental report of social McGee, G. G., Krantz, P.J., Mason, D., & McClannahan, L. behaviors in autistic preschoolers. Journal of Autism E. (1983). A modified incidental-teaching procedure and Developmental Disorders, 20, 513–522. for autistic youth: Acquisition and generalization of Stone, W., Ousley, O., & Littleford, C. (1997). Motor im- receptive object labels. Journal of Applied Behavior itation in young children with autism: What’s the Analysis, 16, 329–338. object? Journal of Abnormal Child Psychology, 25, Meltzoff, A. N., & Gopnik, A. (1994). The role of imita- 475–485. tion in understanding persons and developing a the- Stone, W., Ulman, T., Swanson, A., McMahon, C., & ory of mind. In S. Baron-Cohen, H. Tager-Flusberg, Turner, L. (2004, May). Structured vs. naturalistic & D. Cohen (Eds.), Understanding other minds; object imitation in young children with ASD. Poster Perspectives from autism (pp. 335–366). Oxford: presented at the annual meeting of the International Oxford University Press. Meeting for Autism Research, Sacramento, CA. Meltzoff, A. N., & Moore, M. K. (1977). Imitation of facial Stone, W. L., & Yoder, P. J. (2001). Predicting spoken lan- and manual gestures by human neonates. Science, guage level in children with autism spectrum disor- 198, 75–78. ders. Autism, 5, 341–361. Metz, J. R. (1965). Conditioning generalized imitation in Tiegerman, E., & Primavera, L. (1984). Imitating the autis- autistic children. Journal of Experimental Child Psy- tic child: Facilitating communicative gaze behavior. chology, 4, 389–399. Journal of Autism and Developmental Disorders, Morrison, H., & Kuhn, D. (1983). Cognitive aspects of 14, 27–38. preschoolers’ peer imitation in a play situation. Child Trevarthen, C., Kokkinaki, T., & Fiamenghi, G. (1999). Development, 54, 1041–1053. What infants’ imitations communicate: With moth- Mueller, E., & Lucas, T. (1975). A developmental analysis ers, with fathers and with peers. In J. Nadel of peer interaction among toddlers. In M. Lewis & L. & G. Butterworth (Eds.), Imitation in infancy A. Rosenblum (Eds.), Friendship and peer relations (pp. 127–185). Cambridge, UK: Cambridge Univer- (pp. 223–257). New York: Wiley. sity Press. Mundy, P., Hogan, A., & Doehring, P. (1996). A man- Tryon, A. S., & Keane, S. P. (1986). Promoting imita- ual for the abridged Early Social Communication tive play through generalized observational learning Scales (ESCS). Miami: University of Miami. in autistic-like children. Journal of Abnormal Child Nadel, J., Guerini, C., Peze, A., & Rivet, C. (1999). The Psychology, 14, 532–549. evolving nature of imitation as a format for commu- Uzgiris, I. (1981). Two functions of imitation in infancy. nication. In J. Nadel & G. Butterworth (Eds.), Imita- International Journal of Behavioral Development, tion in infancy (pp. 209–233). Cambridge, UK: Cam- 4, 1–12. bridge University Press. Uzgiris, I. (1990). The social context of infant imitation. In Rogers, S. (1999). An examination of the imitation deficit M. Lewis & S. Feinman (Eds.), Social influences and in autism. In J. Nadel & G. Butterworth (Eds.), Im- socialization in infancy (pp. 215–251). New York: itation in infancy (pp. 254–279). Cambridge, UK: Plenum Press. Cambridge University Press. Uzgiris, I. (1999). Imitation as activity: Its developmen- Rogers, S. J., Hepburn, S., Stackhouse, T., & Wehner, tal aspects. In J. Nadel & G. Butterworth (Eds.), Imitation and Autism 119

Imitation in infancy (pp. 186–206). Cambridge, UK: nal of Autism and Developmental Disorders, 36, Cambridge University Press. 655–664. Warren, S. F.,Yoder, P.J., Gazdag, G. E., & Kim, K. (1993). Whiten, A., & Brown, J. (1998). Imitation and the read- Facilitating prelinguistic communication skills in ing of other minds: Perspectives from the study young children with developmental delay. Journal of of autism, normal children, and non-human pri- Speech and Hearing Research, 36, 83–97. mates. In S. Braten (Ed.), Intersubjective com- Waxler, C., & Yarrow, M. (1975). An observational study munication and emotion in early ontogeny of maternal models. Developmental Psychology, 11, (pp. 260–280). New York: Cambridge University 485–494. Press. Whalen, C., Schreibman, L., & Ingersoll, B. (2006). The Williams, J. H. G., Whiten, A., & Singh, T.(2004). A system- collateral effects of joint attention training on so- atic review of action imitation in autistic spectrum cial initiations, positive affect, imitation, and sponta- disorder. Journal of Autism and Developmental Dis- neous speech for young children with autism. Jour- orders, 34, 285–299.

Current articles- http://depts.washington.edu/isei/iyc/iyc_comments.html

Back to lists of previous articles: http://depts.washington.edu/isei/iyc/iyc_previous.html