Deferred Imitation and Social Communication in Speaking and Nonspeaking Children With Autism

Abstract

Deferred imitation and early social communication skills were compared among speaking and nonspeaking children with autism and children developing typically. Overall, the children with autism showed a lower frequency on measures of deferred imitation and social communication compared with typically developing children. Deferred imitation was observed at a significantly lower level among the speaking and nonspeaking groups of children with autism. However, when comparing the speaking autism group with the typical group, many differences in observed social communication disappeared. These results underscore the importance of considering children’s verbal ability in autism research and clinical practice, and indicate that there are specific difficulties in deferred imitation in autism but that the social communication deficits that are observed are greatly influenced by low level of verbal ability.

Keywords

autism social communication deferred imitation joint attention

Children with autism usually have severe impairments in social interaction and communicative ability. Communicative problems are evidenced in social attention and regulation problems, and about a quarter of the children with autism never develop spoken communicative language (Hus, Pickles, Cook, Risi, & Lord, 2007). Joint attention is one of the prerequisites for the development of language, which is typically delayed in children with autism (e.g., Carpenter, Pennington, & Rogers, 2002). Another early emerging ability, which has not yet been well investigated in children with autism, is deferred imitation, although this capacity is related to both language (Heimann, Strid, et al., 2006; Rose, Feldman, & Jankowski, 2009) and cognitive competence (Strid, Tjus, Smith, Meltzoff, & Heimann, 2006) in typically developing children.

Deferred Imitation

Deferred imitation is imitation of a previously seen action that is no longer accessible by working memory but must be stored in long-term memory. This means that to be successful on a deferred imitation task, children need to form a representation of what they saw previously, store that representation in memory, and use it at a later time point. Deferred imitation of novel activities on objects has been extensively studied among typically developing infants and has been suggested to tap early recall memory in addition to being a test of imitation (e.g., Heimann & Meltzoff, 1996; Heimann & Nilheim, 2004). Deferred imitation emerges between 6 and 9 months of age (Barr, Dowden, & Hayne, 1996; Meltzoff, 1988), and researchers have shown that the ability to imitate after a delay increases with mental age (Hayne, Boniface, & Barr, 2000). Heimann, Strid, et al. (2006) reported a positive relation between deferred imitation observed at 9 months and communication measures at 14 months. In a follow-up study when these children were 4 years old (Strid et al., 2006), deferred imitation and joint attention were found to predict later cognitive competence (i.e., children with low scores on deferred imitation and joint attention in infancy also had lower cognitive levels later in childhood). Researchers also have shown a positive relation between deferred imitation at 12 months and receptive language at 36 months (Rose et al., 2009).

Only a few researchers have explicitly investigated the role of deferred imitation in autism. The findings reported show that children with autism perform at a lower level as compared with matched controls (Dawson, Meltzoff, Osterling, & Rinaldi, 1998; Whiten & Brown, 1998). However, in a study comparing immediate and deferred imitation in children with autism, Rogers, Young, Cook, Giolzetti, and Ozonoff (2008) found similar levels of difficulties on both tasks. If upheld, this finding suggests that the reduced performance on deferred imitation tasks reflects problems with imitation and not with long-term memory. Important connections between deferred imitation and growth of language ability also have been reported in children with autism (Munson, Faja, Meltzoff, Abbott, & Dawson, 2008; Toth, Munson, Meltzoff, & Dawson, 2006). These researchers found that children performing better on deferred imitation at 4 years acquired communication skills (receptive and expressive skills) at a higher rate between 4 and 6 years.

Social Communication

Social communicative skills involve the ability to coordinate the attention between a social partner and a proximal object or event (Mundy & Sigman, 2006). One aspect of social communication is the ability for triadic sharing of attention between a child, another person, and an object or event (joint attention). Joint attention requires an understanding of other people as intentional agents (Tomasello, 1995) and provides a foundation for later communicative development (Baldwin, 1995). The relationship between joint attention and language development has been well investigated (Bakeman & Adamson, 1984; Carpenter, Nagell, & Tomasello, 1998), and several researchers have shown the predictive validity of these early nonverbal skills for later language acquisition (Brooks & Meltzoff, 2005; Morales et al., 2000; Smith & Ulvund, 2003; Tomasello & Farrar, 1986).

Another social communicative skill, which requires a child and another person to engage in joint focus on an object, is requesting (Mundy, Kasari, Sigman, & Ruskin, 1995). This is a nonverbal social skill used to regulate another’s behavior, as when a child is using gestures when trying to get a toy that is out of reach. Social communication also includes social interaction, such as turn taking, which can be with and without objects (Heimann, Strid, et al., 2006). In this early developing ability, two persons are in the focus of each other’s attention, or they share attention about an object.

The communicative problems in autism are rooted in deficits within these early, basic social communication functions (Dawson et al., 2004; Mundy, Sigman, & Kasari, 1994). For example, joint attention is one of the earliest difficulties detected in autism (Charman, 2003), and it has been suggested that difficulties with joint attention reflect an inability to understand the basic foundations of communication (Sigman & Kasari, 1995). Functionally, when there is a dearth of episodes of joint attention, the child may have fewer opportunities to learn language, because such episodes help him or her link words with the correct objects (Baldwin, 1995). Impairments in nonverbal requesting and social interaction (including turn taking) have been demonstrated in children with autism, but they are less pronounced as compared with the problems related to joint attention (Carpenter et al., 2002; Mundy et al., 1994).

The ability to use functional speech in autism has been related to the development of interaction skills, including joint attention. Pry, Petersen, and Baghdadli (2009) showed that children with autism who developed from nonspeaking by 5 years to speaking at 8 years were the ones with the highest interaction competence. Different aspects of social communication, such as joint attention, requesting, and social interaction, capture the ability and motivation to share experiences with others and to understand and respond to communicative intentions. These are all aspects that are of great importance for social and communicative development in childhood.

Aims of the Study

In the present study, we focused on how deferred imitation and social communication abilities relate to language use in speaking and nonspeaking children with autism in comparison with typically developing children. We hypothesized that children with autism as a group would perform at a lower level regarding deferred imitation, joint attention, requesting, and social interaction skills than typically developing children but that speaking children with autism would perform at a similar level regarding deferred imitation, joint attention, requesting, and social interaction skills as typically developing children when matched for language age.

Method

Participants

The study was approved by the ethical committee at the University of Gothenburg. A total of 20 children (18 boys and 2 girls, M age = 66.8 months) with autistic disorder according to criteria from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association, 1994) and 22 typically developing children (12 boys and 10 girls, M age = 34.7 months) participated in the study. Chronological, vocabulary, and mental ages for the groups are listed in Table 1. In all, 19 of the children with autism were diagnosed by specialized neuropsychiatric teams (the Child Neuropsychiatry Clinic, Sahlgrenska University Hospital and the neuropsychiatric team, Kungälv Hospital). The neuropsychiatric work-up comprised clinically validated instruments, either the Diagnostic Interview for Social and Communication Disorders (DISCO; Wing, Leekam, Libby, Gould, & Larcombe, 2002) or the Autism Diagnostic Interview–Revised (ADI-R; Lord, Rutter, & Le Couteur, 1994). One child was diagnosed by a pediatrician, and no details of the clinical evaluation were available. For this child, the level of autistic behavior was measured with the Childhood Autism Rating Scale (CARS; Schopler, Reichler, & Renner Rochen, 1988) to validate the diagnosis; scores indicated mild–moderate autism. The typically developing children were recruited from day care centers situated in central Göteborg, with families predominantly of middle to high socioeconomic status. Information about the study was given to parents through their children’s day care center, and parents who were interested in participating contacted the department. The children were selected based on how well they were matched to children with autism for receptive vocabulary. For 5 children (4 in the autism group and 1 in the typical group), receptive vocabulary was not available or possible to assess; therefore, expressive vocabulary score was used instead. There was no significant difference in mental age between the groups. However, the children with autism had a significantly higher chronological age compared with the typically developing children (Table 1).

Table 1.

Participant Descriptive Data for Total Group and for Nonspeaking and Speaking Children With Autism.

	Total group		Autism group
	Autism^a	Typical^b	Nonspeaking^c	Speaking^d
Characteristic	M (SD)	M (SD)	M (SD)	M (SD)
Chronological age	66.8 (17.32)	34.7 (5.20)*	59.4 (19.70)^e*	70.8 (15.24)*^e
Vocabulary age	29.7 (15.99)	35.0 (13.15)	13.6 (1.62)^e*	38.3 (13.11)
Mental age	45.2 (19.36)^f	37.0 (8.95)	24.7 (8.31)^e,g*	54.7 (15.00)^e*
Male/female	18/2	12/10	6/1	12/1

Note. All ages are given in months. Group comparisons conducted via Mann–Whitney U test.

n = 20. ^bn = 22. ^cn = 7. ^dn = 13. ^eCompared with the typical group. ^fn = 19 due to data loss. ^gn = 6 due to data loss.

p < .01.

The children with autism were divided into two groups, a speaking and a nonspeaking group. A child was considered as nonspeaking if he or she did not use language for communication during any of the observations, including the free play with the parent, and if the parent indicated that this level of communication was typical for the child. This means that the child could use sounds other than words or simple single words, and still be in the nonspeaking group. The definition was made according to the definition of phrase speech used in ADI-R, the spontaneous use of at least two words in combination, one of which must be a verb (Joseph, Tager-Flusberg, & Lord, 2002). A total of 7 children with autism were considered “nonspeaking,” and 13 children were considered “speaking.” The categorization was made by the first author. A researcher not involved in the project categorized the speaking status of 5 (25%) randomly selected children, and 100% agreement was obtained. The means of chronological age, mental age, and vocabulary age of these two groups are presented in Table 1.

Estimating Vocabulary

Children with autism

The third edition of the Peabody Picture Vocabulary Test (PPVT-3; Dunn & Dunn, 1997) and a Swedish version of the MacArthur Communicative Development Inventories (SECDI; Berglund & Eriksson, 2000; Eriksson & Berglund, 1999) were administered to all children. Due to the large heterogeneity in the group-language level—ranging from nonverbal to well-spoken children—we used the most appropriate test to estimate language age in the children. The test was chosen according to a hierarchic procedure and used to estimate receptive vocabulary when considered reliable (n = 9). For 11 children, the PPVT was not appropriate or possible to use. The reasons were that the child refused to point at the pictures, did not understand the instructions, or did not cooperate. The second test of choice was SECDI, which is a parent report that comes in two different versions. The first version measures receptive vocabulary and is appropriate for children between 8 and 16 months. This version of the SECDI was used to estimate receptive vocabulary for 7 children. In all, 1 child had a receptive vocabulary age above 16 months, and for this child, we used the second version of the SECDI (appropriate for children between 16 and 28 months), which also included expressive vocabulary. For the last 3 children in the autism group, the SECDI was not possible to use because of ceiling effects, and instead, the Kaufmann Expressive Vocabulary subscale (Kaufman & Kaufman, 1983) was the source for estimating language age.

Typically developing children

The PPVT was used to measure receptive vocabulary for 21 of the typically developing children. In all, 1 child was assessed with the Kaufman because he did not cooperate on the PPVT and was above the ceiling on the SECDI.

Estimating Mental Age

Children with autism

If possible, mental age estimation was based on tests used as part of the diagnostic work-up. This was true for 15 of the children where clinically valid test results from the Griffiths Developmental Scale (Ahlin-Åkerman & Norberg, 1991; Griffiths, 1970) and Wechsler Pre-School and Primary Scale of Intelligence–Revised (WPPSI-R, Swedish version; Wechsler, 1999) could be used. However, for 4 children, the clinical test results were either not available (n = 2) or considered unreliable according to clinicians who had made the assessment (n = 2), and the children’s mental level was estimated by the research team using Raven’s Progressive Matrices (Raven, 1976). For 1 child with missing clinical evaluation, we were unsuccessful in administering Raven’s matrices, and this child is not included in the calculation of the mean mental age for the autism group.

Typically developing children

The typically developing children were assessed using the McCarthy Scales of Children’s Abilities (McCarthy, 1972). This test yields a General Cognitive Index based on verbal, perceptual performance, quantitative, and memory scores. The General Cognitive Index was used to calculate mental age of the participants in this group.

Procedure

All children visited the Department of Psychology, University of Gothenburg, twice, with 1 or 2 days between the visits. Each visit lasted approximately 1 to 1.5 hr, and every child was accompanied by at least one parent. The parent decided whether he or she wanted to stay in the observation room or watch the child through a one-way mirror. The child was observed in the same room and by the same experimenter on both occasions. During these 2 days of observation, several different areas were investigated. Only the abridged versions of the Early Social Communication Scales (ESCS; Mundy, Hogan, & Doehring, 1996), deferred imitation (Meltzoff, 1985, 1988), and language estimation are presented. The ESCS was the first observation on the 2nd day for all children. The deferred imitation tasks were modeled on the first visit, and the child’s response was observed on the second visit. All language tests were carried out on the first visit.

The parts of the visits for deferred imitation and the ESCS were videotaped by two cameras placed in the corners of the room for later scoring. In addition, an imitation procedure between the child and the experimenter, and a 24-min free play between the parent and child also were videotaped although the results are not presented here. However, we used this free play observation when deciding the speaking or nonspeaking status of the child (see details above). The reason that these observations were videotaped is that they needed to be scored from tapes. All families received a symbolic gift and a copy of parts of the videotaped observations.

Deferred imitation

Five actions with objects were used to measure deferred imitation. The objects, which were replicas of the ones originally used with typically developing infants by Meltzoff (1985, 1988), are described in Table 2. During Visit 1, the experimenter and the child sat at a table, and the experimenter showed an action on each object 3 times in succession. The “observation-only-design” developed by Meltzoff (1995) was used. This means that during each presentation, the child was not allowed to handle the objects, making sure that the child did not develop any motoric experience of the action. A delay was then imposed until the child came for his or her second visit approximately 2 days later (M = 50.6 hr, SD = 8.55). The experimenter then put the object in front of the child, and a 20-s response time was allowed after the child had touched the object.

Table 2.

Description of Objects, Target Actions, and Scoring Criteria for Objects Used in Deferred Imitation Procedure.

Object	Description	Target actions	Scoring criteria^a
Conventional response
Egg	Plastic egg with small beads inside	Shaking egg	Moving egg side to side 3 times
Collapsible cup	Cup that collapsed when pressing downward on it	Making cup collapse by pressing it	Collapse cup more than halfway
Beads	Cup and string of beads	Putting string in the cup	Putting string in the cup
Nonconventional response
Hinge	Wooden rectangle connected by hinge to larger rectangular base	Pushing rectangle over with elbow so it lays flat on top of base	Folding vertical flap down with greater arc than 45^o toward base by using elbow
Box	Black box with hidden black button on top that made beeping sound when pressed	Pressing button with pen	Making sound by using pen

The child had to complete the target action within 20 s from the first time the child touched the object.

For three of the objects (i.e., the conventional tasks), we used the same scoring criteria as in the original studies (Table 2). For two objects (i.e., the nonconventional tasks), the demonstration and scoring were slightly changed so that the actions depicted required nonconventional responses. First, a pen was used, instead of a finger, to press a button on a box. Second, the elbow was used instead of the hand to press the rectangle on a hinge. This modification was made to ensure that the children were imitating the actions, and not only reaching for the goal (Williams, Whiten, & Singh, 2004). For the three conventional tasks, the easiest or most obvious way to reach the goal was by imitating the previously seen action. However, for the two nonconventional tasks, the goal can be reached more easily by changing the action (i.e., using a finger to press the button on the box and using the hand to press down the rectangle).

The first and second authors coded all deferred imitation tasks (see Table 2 for scoring criteria). The isolated response period was used for the coding, and the second author had not participated in the actual data collection. Observer agreement was assessed by Cohen’s kappa (κ = .89).

Social communication

To capture nonverbal joint attention, requesting, and social interaction behaviors, the ESCS (Mundy et al., 1996) was administered during the second visit. The ESCS is a structured play situation designed to capture joint attention, behavioral requests, and social interaction behaviors. For each area, the child either initiates an action or responds to a behavior initiated by the experimenter. The child and the experimenter sit at a table where some toys are visible for the child and others are on the floor behind the experimenter. The toys are designed to encourage social interaction between the two persons. During testing, the experimenter presents three wind-up toys, three mechanical toys, opportunities to play turn-taking games with a ball and a car, and opportunities to play social games with a comb, hat, and glasses, and introduces tickle games. In addition, posters are placed on the walls, two on each side and one behind the child, and the child is encouraged to look at the posters after the experimenter has said the child’s name and pointed at him or her. A similar trial is conducted with a picture book.

Initiating joint attention (IJA) includes low-level and high-level behaviors. Low-level behaviors are scored when the child makes eye contact with the experimenter while touching a toy or is alternating eye contact between an active toy and the experimenter. High-level behaviors are scored when the child points at a toy before the experimenter points or shows a toy to the experimenter. The total frequency of high- and low-level behaviors is reported. Responding to joint attention (RJA) measures how many times the child follows the direction of the experimenter’s pointing. Each child gets up to 12 opportunities to follow pointing (pointing to pictures in a book or to posters on the wall). The exact number depends on the child’s cooperation; thus, a ratio score was calculated (i.e., the number of looks divided by the number of opportunities).

Initiating object request (IOR) includes eye contact with the experimenter when an active toy has stopped or disappeared. Other behaviors scored within this category are reaching for a toy, giving a toy to the experimenter, and pointing for request. Responding to object request (ROR) is scored when the child gives a toy to the experimenter after a request. For IOR and ROR, the total frequency of behaviors was recorded.

Initiating social interaction (ISI) includes turn taking and teasing behavior on the child’s part. Turn taking is scored when the child initiates a turn-taking sequence. The child is given two different toys (a car and a ball) and is encouraged with gestures to roll the toys to the experimenter. Responding to social interaction (RSI) is scored when the child makes eye contact with the experimenter, bangs at the table or vocalizes after being tickled, maintains turn taking (either with the car or the ball after turn taking has been initiated by the child or experimenter), or responds to an invitation: The child is invited to play with a hat, a comb, and glasses, which means that the experimenter initiates play and says, “Now it is your turn.” For ISI and RSI, the total frequency of these behaviors was scored.

A research assistant coded all ESCS videos using the observation program from Noldus (observer). The videos showed the child and the experimenter as well as the table between them. The observations were coded following the guidelines from the ESCS manual (Mundy et al., 1996). The first author coded 10% of the observations selected on a random basis, revealing an agreement of κ = .76 and an intraclass correlation of .99 (p < .001).

Statistical analysis

Nonparametric analyses (Mann–Whitney U test) were used for all group comparisons.

Results

Deferred Imitation

Overall, children with autism performed on a significantly lower level on the deferred imitation tasks as compared with the typical group (Table 3). This also was evident when comparing the nonspeaking and speaking subgroups with the typical group. In contrast, separate analyses of conventional and nonconventional tasks showed similar group differences for the conventional tasks only. No significant group differences were found for the nonconventional deferred imitation tasks.

Table 3.

Correct Response (Ms and SDs) on Deferred Imitation Tasks for T and A Groups and for NS and S Children With Autism.

	Typical^a	Autism^b	A (NS)^c	A (S)^d	T vs. A	T vs. A (NS)	T vs. A (S)
Deferred imitation tasks	M (SD)	M (SD)	M (SD)	M (SD)	p	p	p
Overall (five tasks)	3.74 (0.86)	2.80 (1.01)	2.29 (0.95)	3.08 (0.95)	<.01	<.01	<.05
Conventional (three tasks)	2.55 (0.51)	1.84 (0.69)	1.58 (0.53)	2.00 (0.74)	<.01	<.001	<.05
Nonconventional (two tasks)	1.23 (0.69)	0.95 (0.51)	0.71 (0.49)	1.08 (0.49)	ns	ns	ns

Note. T = typical; A = autism; NS = nonspeaking; S = speaking. Group comparisons conducted using Mann–Whitney U test.

n = 22. ^bn = 20. ^cn = 7. ^dn = 13.

Social Communication

As shown in Table 4, the children with autism had significantly lower means on initiating and RJA and social interactions, as well as initiating behavioral requests. There was no difference in responding to requesting (ROR) when compared with the typical group.

Table 4.

Correct Response (M and SDs) on Nonverbal Communication Measures for T and A Groups and for NS and S Children With Autism.

	Typical^a	Autism^b	A (NS)^c	A (S)^d	T vs. A	T vs. A (NS)	T vs. A (S)
Measure	M (SD)	M (SD)	M (SD)	M (SD)	p	p	p
Joint attention
Initiating (IJA)	20.23 (9.72)	9.05 (9.39)	2.00 (3.61)	12.85 (9.41)	<.001	<.001	<.05
Responding (RJA) ratio	0.96 (0.07)	0.77 (0.27)	0.43 (0.07)	0.95 (0.09)	<.05	<.001	ns
Requesting
Initiating (IOR)	17.14 (8.34)	8.80 (7.22)	3.43 (3.96)	11.69 (6.99)	<.01	<.001	<.05
Responding (ROR)	5.95 (3.59)	4.95 (4.08)	3.14 (4.26)	5.92 (3.80)	ns	<.05	ns
Social interaction
Initiating (ISI)	2.32 (1.25)	1.15 (0.99)	.29 (0.49)	1.62 (0.87)	<.01	<.001	ns
Responding (RSI)	13.68 (7.56)	8.45 (8.40)	1.86 (2.19)	12.00 (8.39)	<.05	<.001	ns

Note. T = typical; A = autism; NS = nonspeaking; S = speaking; IJA = initiating joint attention; RJA = responding to joint attention; IOR = initiating object request; ROR = responding to object request; ISI = initiating social interaction; RSI = responding to social interaction. Group comparisons conducted using Mann–Whitney U test.

n = 22. ^bn = 20. ^cn = 7. ^dn = 13.

When the group of speaking children with autism was compared with the typical group, the groups performed at the same level on all measures except IJA and IOR. Comparing the nonspeaking children with autism with the typical group, the autism group performed significantly lower on all variables.

Gender

Because the typical group and the autism group were not matched on gender, a separate analysis was conducted for boys only. The comparison between the boys with autism (n = 18) and the boys who were developing typically (n = 12) on social communication revealed a similar result as for the total group comparison, except for RJA and RSI, which no longer reached the level of significance. This means that the boys with autism and the typically developing boys differed on all initiating, but none of the responding, variables. When analyzing deferred imitation between boys only, identical results as for the total group comparison were obtained.

Discussion

We investigated deferred imitation and social communication among speaking and nonspeaking children with autism in comparison with typically developing children. Deferred imitation, IJA, and IOR were observed at a lower level in speaking children with autism as compared with a language-matched comparison group of typically developing children matched for receptive vocabulary. In addition, RJA, ROR, and responding as well as ISI were affected among the nonspeaking children with autism.

Deferred Imitation

Few researchers have investigated deferred imitation in relation to language competence in children with autism. The present results revealed that the children with autism performed lower on the deferred imitation task regardless of their ability to use spoken language or not. This is in contrast to earlier results reporting a connection between deferred imitation and language acquisition in autism (Munson et al., 2008; Toth et al., 2006). In the present study, conventional and nonconventional deferred imitation tasks were used. Although the conventional tasks differed between the groups, no group difference was found on the nonconventional tasks. This could be due to low statistical power, because only two nonconventional tasks were included in the study. Another, more speculative, interpretation is that children with autism, in contrast to typically developing children, may fail to see any difference between nonconventional and conventional tasks; that is, they treat all tasks as unique actions without relying on previous experience. In this case, the typical children would show a disadvantage in the nonconventional tasks, whereas the children with autism, by failing to recognize the odd behavior in the nonconventional tasks, would treat both tasks in a similar fashion. This means that both groups would have similar problems on the nonconventional tasks, but only the children with autism would have the same problem on the conventional tasks. This possibility needs further investigation using a more varied set of tasks and larger groups of children, but if this holds, it will give us a deeper understanding about the imitation deficit seen in autism. Imitation of nonconventional tasks shows that the child is not only imitating the goal of the action but rather the action itself. Charman and Baron-Cohen (1994) found that imitation of “novel actions” (i.e., strange actions that could be compared with our nonconventional actions) was difficult for children with autism when compared with other action imitation tasks. However, 12 out of 20 children with autism (M verbal MA = 46 months) imitated the novel action; the task was equally difficult for children with developmental delays.

It is possible that lower performance on the total deferred imitation score in the autism group to some extent depends on an “imitation deficit” and not on a general memory problem, although views differ regarding whether, or to what degree, such a deficit exists (e.g., Heimann, Ullstadius, Dahlgren, & Gillberg, 1992; Nadel, 2002; Williams et al., 2004). Researchers have suggested that imitation of actions on objects (as in the deferred imitation tasks used in this study) is less impaired in this group as compared with imitation of body movements (Stone, Ousley, & Littleford, 1997) and role reversal imitation (Carpenter, Tomasello, & Striano, 2005). Problems with imitation are more profound in younger children with autism (Charman et al., 1997; Gopnik, Capps, & Meltzoff, 2000), suggesting that imitation is not an absolute deficit in children with autism, and more dependent on developmental level. In this study, being able to use spoken language did not help the children with autism perform better on the deferred imitation tasks, even if the actions being imitated were simple actions on objects. Thus, deferred imitation might be especially demanding for children with autism. However, Rogers et al. (2008) suggested that children with autism do not have more problems with deferred imitation than with other imitation tasks, which raises the question whether the difficulties seen in deferred imitation tasks should be interpreted as imitation or memory deficits. These seemingly contradictory findings warrant future examination before we can conclude that there is a specific deficit in deferred imitation in children with autism.

Social Communication

Our findings replicate those of earlier studies (e.g., Charman, 2003; Mundy et al., 1994) demonstrating that problems with IJA are a robust finding among all children with autism, regardless of whether they are able to use spoken language. In addition, this also held for IOR. For RJA, requesting, and social interaction, as well as ISI, the performance of children with autism increased with language ability. When comparing the nonspeaking autism group and the typical group, the children with autism performed at a lower level on all social communication variables measured in this study.

Joint attention depends more on social motivation and sharing experiences as compared with other communicative gestures. Sharing experiences with others involves a sharing of positive affect in typically developing children, but this may not be the case among children with autism (Kasari, Sigman, Mundy, & Yirmiya, 1990). This probably makes joint attention more socially rewarding for typically developing children and could explain why children with autism are less motivated to engage in joint attention (Mundy & Sigman, 2006). Joint attention also involves a change of own and others’ minds, something that is especially difficult for children with autism to understand. Other social communicative gestures, such as requesting, are followed by a visible change in own or other people’s behavior (e.g., they give you a toy; Gergely, 2004; Mundy et al., 1996) and might therefore be easier to understand and produce. In this study, IJA and IOR were difficult for all children with autism, whereas RJA was without any problems for the group of speaking children. This suggests that initiating and RJA rely on different abilities and that they should be investigated separately in autism research. Joint-attention intervention for children with autism has been used with positive results. Kaale, Smith, and Sponheim (2012) randomly assigned preschool children with autism to an 8-week joint-attention intervention or to an ordinary preschool program. The children in the joint-attention intervention group significantly increased their joint-attention skills in play with their teachers and also showed more joint engagements in play with their mothers. Results showing how difficulties with joint attention differ among individual children with autism could be used to develop more tailored interventions. The fact that the children with autism have difficulties with IJA could be addressed by increasing their motivation for communication by exposing them to fun and challenging language material. An example of this is using computers with motivating multimedia material resulting in more relevant and engaging communication with their teachers (Tjus, Heimann, & Nelson, 2001). Another strategy to be explored is to use imitation as an intervention strategy, which has been especially effective for nonspeaking children with autism, the group where IJA is a core problem (Heimann, Laberg, & Nordøen, 2006).

Individual Differences

Our results highlight the importance of considering individual differences within the autism diagnosis spectrum (Mundy & Sigman, 2006). Whereas the nonspeaking children with autism evidenced reduced performance on all social communicative measures and deferred imitation, the speaking children with autism displayed a pattern that was more similar to that observed among the typical group. In this study, there were notable individual differences within the autism group on social abilities but more general difficulties in deferred imitation, which indicate difficulties with recall memory and/or imitation tasks. It is important to note that the children with autism were older compared with the typically developing children, which means that they performed on a lower level than what was expected from their chronological age. This suggests that prototypic deficits in autism may not be absolute but rather be related to the developmental level of the child (e.g., Mundy et al., 1994). The present results also have methodological implications, suggesting that the developmental level of the child and the heterogeneity of the group are important factors to consider in research on autism.

All participating children performed at similar levels on ROR. In the ESCS procedure (Mundy et al., 1996), a child may either respond to a gestural command (an open palm) or to a verbal command (“Give it to me”), but these responses were not scored separately in the present study. Thus, if the child did not respond to the nonverbal command, the experimenter would give the child a verbal command, which means that the child did not need to understand communicative gestures to respond; the child just needed to understand and follow a command. This might explain the similar performance by children with autism and the typically developing children on this task, because the task of interpreting an open hand as a command was absent in the task.

Strengths and Limitations

There are two main strengths of our study. First, we found that speaking and nonspeaking children with autism show a different pattern of difficulties within the area of social communication but more similar problems with deferred imitation. This shows the importance of considering individual differences in all autism research. Another strength is that the study included conventional and nonconventional tasks of deferred imitation. It should be expected that children with autism would be less affected by the nonconventional actions in contrast to typical children. That is, the performance on nonconventional tasks should be more similar between the groups compared with the performance on conventional tasks. This was also the result. It is important to note that the differences in performance on conventional and nonconventional tasks need to be replicated with larger groups and more items. However, by a detailed investigation of deferred imitation, the present study adds to our understanding of the nature of the imitation deficit in autism.

Research highlighting the heterogeneity among children with autism helps us to understand which areas of difficulties need to be targeted in interventions promoting language development. We interpret our data to conclude that the pattern of difficulties within the area of social communication differs among individual children with autism depending on their use of speech or not. One aspect that was difficult for all children with autism was IJA, and this needs to be highlighted when planning communication treatment (i.e., the objective must be trying to increase the children’s motivation for communication; for example, Heimann, Laberg, et al., 2006; Tjus et al., 2001). A more detailed understanding of how different communicative and imitation skills manifest in individual children is also important in understanding the unique challenges in children with autism.

Some important limitations of the present study need to be considered. First, the number of boys and girls in the autism group differed (18 boys, 2 girls), as compared with the group of typically developing children (12 boys, 10 girls). Analyses including only boys revealed a similar pattern of result as observed for the total group, except for RSI and RJA. The difference in the number of boys and girls in the two groups is important because one would expect that typically developing girls about 3 years of age have developed better language skills than boys at the same age (Bornstein, Hahn, & Haynes, 2004). However, by matching the groups on language age, we tried to minimize this discrepancy as much as possible. A second limitation concerns the use of ESCS with children at different language levels. Some of the typically developing children and some of the speaking children with autism were above the chronological age recommended by the ESCS manual. This could have affected the results because these children might reduce their gestural communication and rely more on verbal language for social communication. However, it is unlikely that this was the case for the children with autism, because the speaking children had higher scores on all social communication measures compared with the nonspeaking children.

Conclusion

For children with autism, we found that most social communicative abilities depend on the child’s ability to use spoken language and that difficulties within these areas are not present to the same extent in all children with autism. We also found that deferred imitation in autism did not improve for speaking children as compared with children unable to use spoken language for communication purposes. Children with no or very limited speech tended to show impairments that are not observed among speaking children. This suggests that deferred imitation could be used as an additional measure when evaluating or assessing social and linguistic development in autism. A further exploration of individual differences in autism will not only help us identify important subgroups but may also lead to better predictions of outcome and more effective intervention strategies.

Footnotes

Acknowledgements

The authors wish to thank Katarina Nilheim for generous assistance in completing this study as well as the parents and children who participated.

Authors’ Note

Preliminary versions of data presented in this article have been presented at the XIIth European Conference on Developmental Psychology, Tenerife, August 2005 and the BPS Developmental Section Conference, London, September 2006.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Support for this research was provided by research grants from the Swedish Council for Working Life and Social Research, Stockholm, Sweden, to Mikael Heimann (Grant 2001-1113), Erland Hjelmquist (Grant 2005-1700), and Tomas Tjus (Grant 2008-0518). The work of Mikael Heimann was also supported by the European Science Foundation Cooperation in Science and Technology Action (ESF COST Action) BM 1004 Enhancing the Scientific Study of Early Autism (ESSEA).

References

Ahlin-Åkerman

Norberg

(1991). Griffiths utvecklingsskalor [Griffiths Developmental Scales]. Göteborg, Sweden: Psykologiförlaget [in Swedish].

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Bakeman

Adamson

L. B.

(1984). Coordinating attention to people and objects in mother-infant and peer-infant interaction. Child Development, 55, 1278–1289.

Baldwin

D. A.

(1995). Understanding the link between joint attention and language. In Moore

Dunham

P. J.

(Eds.), Joint attention: Its origins and role in development (pp. 131–158). Hillsdale, NJ: Lawrence Erlbaum.

Barr

Dowden

Hayne

(1996). Developmental changes in deferred imitation by 6- to 24-month-old infants. Infant Behavior & Development, 19, 159–170.

Berglund

Eriksson

(2000). Communicative development in Swedish children 16-28 months old: The Swedish early communicative development inventory–Words and sentences. Scandinavian Journal of Psychology, 41, 133–144.

Bornstein

M. H.

Hahn

C.-S.

Haynes

O. M.

(2004). Specific and general language performance across early childhood: Stability and gender considerations. First Language, 24, 267–304.

Brooks

Meltzoff

A. N.

(2005). The development of gaze following and its relation to language. Developmental Science, 8, 535–543.

Carpenter

Nagell

Tomasello

(1998). Social cognition, joint attention, and communicative competence from 9 to 15 months of age. Monographs of the Society for Research in Child Development, 63, 1–143.

10.

Carpenter

Pennington

B. F.

Rogers

S. J.

(2002). Interrelations among social-cognitive skills in young children with autism. Journal of Autism and Developmental Disorders, 32, 91–106.

11.

Carpenter

Tomasello

Striano

(2005). Role reversal imitation and language in typically developing infants and children with autism. Infancy, 8, 253–278.

12.

Charman

(2003). Why is joint attention a pivotal skill in autism? Philosophical Transactions of the Royal Society B, 358, 315–324.

13.

Charman

Baron-Cohen

(1994). Another look at imitation in autism. Development and Psychopathology, 6, 403–413.

14.

Charman

Swettenham

Baron-Cohen

Cox

Baird

Drew

(1997). Infants with autism: An investigation of empathy, pretend play, joint attention, and imitation. Developmental Psychology, 33, 781–789.

15.

Dawson

Meltzoff

A. N.

Osterling

Rinaldi

(1998). Neuropsychological correlates of early symptoms of autism. Child Development, 69, 1276–1285.

16.

Dawson

Toth

Abbott

Osterling

Munson

Estes

Liaw

(2004). Early social attention impairments in autism: Social orienting, joint attention, and attention to distress. Developmental Psychology, 40, 271–283.

17.

Dunn

(1997). Peabody Picture Vocabulary Test (3rd ed.). Circle Pines, MN: American Guidance Service.

18.

Eriksson

Berglund

(1999). Swedish early communicative development inventory—Words and gestures. First Language, 19, 55–90.

19.

Gergely

(2004). The development of understanding self and agency. In Goswami

(Ed.), Blackwell handbook of childhood cognitive development (pp. 26–46). Oxford, UK: Blackwell.

20.

Gopnik

Capps

Meltzoff

A. N.

(2000). Early theories of mind: What the theory can tell us about autism. In Baron-Cohen

Tager-Flusberg

Cohen

D. J.

(Eds.), Understanding other minds: Perspectives from developmental cognitive neuroscience (pp. 50–72). Oxford, UK: Oxford University Press.

21.

Griffiths

(1970). The abilities of young children. A study in mental measurement. London, England: University of London Press.

22.

Hayne

Boniface

Barr

(2000). The development of declarative memory in human infants: Age-related changes in deferred imitation. Behavioral Neuroscience, 114, 77–83.

23.

Heimann

Laberg

K. E.

Nordøen

(2006). Imitative interaction increases social interest and elicited imitation in non-verbal children with autism. Infant and Child Development, 15, 297–309.

24.

Heimann

Meltzoff

A. N.

(1996). Deferred imitation in 9- and 14-month-old infants: A longitudinal study of a Swedish sample. British Journal of Developmental Psychology, 14, 55–64.

25.

Heimann

Nilheim

(2004). 6-months olds and delayed actions: An early sign of an emerging explicit memory? Cognition, Brain and Behavior, 8, 249–254.

26.

Heimann

Strid

Smith

Tjus

Ulvund

S. E.

Meltzoff

A. N.

(2006). Exploring the relation between memory, gestural communication, and the emergence of language in infancy: A longitudinal study. Infant and Child Development, 15, 233–249.

27.

Heimann

Ullstadius

Dahlgren

S.-O.

Gillberg

(1992). Imitation in autism: A preliminary research note. Behavioural Neurology, 5, 219–227.

28.

Hus

Pickles

Cook

E. H.

Jr. Risi

Lord

(2007). Using the Autism Diagnostic Interview-Revised to increase phenotypic homogeneity in genetic studies of autism. Biological Psychiatry, 61, 438–448.

29.

Joseph

R. M.

Tager-Flusberg

Lord

(2002). Cognitive profiles and social-communicative functioning in children with autism spectrum disorder. Journal of Child Psychology and Psychiatry, 43, 807–820.

30.

Kaale

Smith

Sponheim

(2012). A randomized controlled trial of preschool-based joint attention intervention for children with autism. Journal of Child Psychology and Psychiatry, 53, 97–105.

31.

Kasari

Sigman

Mundy

Yirmiya

(1990). Affective sharing in the context of joint attention interactions of normal, autistic, and mentally retarded children. Journal of Autism and Developmental Disorders, 20, 87–100.

32.

Kaufman

A. S.

Kaufman

N. L.

(1983). Kaufman assessment battery for children. Circle Pines, MN: American Guidance Service.

33.

Lord

Rutter

Le Couteur

(1994). Autism Diagnostic Interview–Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders, 24, 659–686.

34.

McCarthy

(1972). McCarthy Scales of Children’s Abilities. New York, NY: Psychological Corp.

35.

Meltzoff

A. N.

(1985). Immediate and deferred imitation in fourteen- and twenty-four-month-old infants. Child Development, 56, 62–72.

36.

Meltzoff

A. N.

(1988). Infant imitation and memory: Nine-month-olds in immediate and deferred tests. Child Development, 59, 217–225.

37.

Meltzoff

A. N.

(1995). What infant memory tells us about infantile amnesia: Long-term recall and deferred imitation. Journal of Experimental Child Psychology, 59, 497–515.

38.

Morales

Mundy

Delgado

C. E. F.

Yale

Messinger

Neal

Schwartz

(2000). Responding to joint attention across the 6- through 24-month age period and early language acquisition. Journal of Applied Developmental Psychology, 21, 283–298.

39.

Mundy

Hogan

A. E.

Doehring

(1996). A preliminary manual for the abridged Early Social Communication Scales (ESCS). Unpublished manuscript.

40.

Mundy

Kasari

Sigman

Ruskin

(1995). Nonverbal communication and early language acquisition in children with Down syndrome and in normally developing children. Journal of Speech & Hearing Research, 38, 157–167.

41.

Mundy

Sigman

(2006). Joint attention, social competence and developmental psychopathology. In Cicchetti

Cohen

(Eds.), Developmental psychopathology, theory and methods (2nd ed., Vol. 1, pp. 293–332.). Hoboken, NJ: Wiley.

42.

Mundy

Sigman

Kasari

(1994). Joint attention, developmental level, and symptom presentation in autism. Development and Psychopathology, 6, 389–401.

43.

Munson

Faja

Meltzoff

Abbott

Dawson

(2008). Neurocognitive predictors of social and communicative developmental trajectories in preschoolers with autism spectrum disorders. Journal of the International Neuropsychological Society, 14, 956–966.

44.

Nadel

(2002). Imitation and imitation recognition: Functional use in preverbal children with autism. In Prinz

Meltzoff

A. N.

(Eds.), The imitative mind: Development, evolution, and brain bases (pp. 42–62). New York, NY: Cambridge University Press.

45.

Pry

Petersen

A. F.

Baghdadli

(2009). Developmental changes of expressive language and interactive competences in children with autism. Research in Autism Spectrum Disorders, 3, 98–112.

46.

Raven

J. C.

(1976). Coloured progressive matrices. Oxford, UK: Oxford Psychologists Press.

47.

Rogers

S. J.

Young

G. S.

Cook

Giolzetti

Ozonoff

(2008). Deferred and immediate imitation in regressive and early onset autism. Journal of Child Psychology and Psychiatry, 49, 449–457.

48.

Rose

S. A.

Feldman

J. F.

Jankowski

J. J.

(2009). A cognitive approach to the development of early language. Child Development, 80, 134–150.

49.

Schopler

Reichler

R. J.

Renner Rochen

(1988). The Childhood Autism Rating Scale. Los Angeles, CA: Western Psychological Services.

50.

Sigman

Kasari

(1995). Joint attention across contexts in normal and autistic children. In Moore

Dunham

P. J.

(Eds.), Joint attention: Its origins and role in development (pp. 189–203). Hillsdale, NJ: Lawrence Erlbaum.

51.

Smith

Ulvund

S. E.

(2003). The role of joint attention in later development among preterm children: Linkages between early and middle childhood. Social Development, 12, 222–234.

52.

Stone

W. L.

Ousley

O. Y.

Littleford

C. D.

(1997). Motor imitation in young children with autism: What’s the object? Journal of Abnormal Child Psychology, 25, 475–485.

53.

Strid

Tjus

Smith

Meltzoff

A. N.

Heimann

(2006). Infant recall memory and communication predicts later cognitive development. Infant Behavior & Development, 29, 545–553.

54.

Tjus

Heimann

Nelson

K. E.

(2001). Interaction patterns between children and their teachers when using a specific multimedia and communication strategy: Observations from children with autism and mixed intellectual disabilities. Autism, 5, 175–187.

55.

Tomasello

(1995). Joint attention as social cognition. In Moore

Dunham

P. J.

(Eds.), Joint attention: Its origins and role in development (pp. 103–130). Hillsdale, NJ: Lawrence Erlbaum.

56.

Tomasello

Farrar

M. J.

(1986). Joint attention and early language. Child Development, 57, 1454–1463.

57.

Toth

Munson

Meltzoff

A. N.

Dawson

(2006). Early predictors of communication development in young children with autism spectrum disorder: Joint attention, imitation, and toy play. Journal of Autism and Developmental Disorders, 36, 993–1005.

58.

Wechsler

(1999). Wechsler Pre-School and Primary Scale of Intelligence [Swedish version] (Rev. ed.). Stockholm, Sweden: Psykologiförlaget.

59.

Whiten

Brown

J. D.

(1998). Imitation and the reading of other minds: Perspectives from the study of autism, normal children and non-human primates. In Bråten

(Ed.), Intersubjective communication and emotion in early ontogeny (pp. 260–280). Cambridge, UK: Cambridge University Press.

60.

Williams

J. H. G.

Whiten

Singh

(2004). A systematic review of action imitation in autism spectrum disorder. Journal of Autism and Developmental Disorders, 34, 285–299.

61.

Wing

Leekam

S. R.

Libby

S. J.

Gould

Larcombe

(2002). The Diagnostic Interview for Social and Communication Disorders: Background, inter-rater reliability and clinical use. Journal of Child Psychology and Psychiatry and Allied Disciplines, 43, 307–325.