Anna Bortoli & P. Margaret Brown

Early Intervention Unit

Department of Learning and Educational Development

The University of Melbourne

Address for Correspondence:

Anna Bortoli

Early Intervention Unit

Department of Learning and Educational Development

The University of Melbourne

Parkville 3052

Australia

During social interaction, children need to be able to maintain extended periods of social engagement with their peers. This requires that children are able to attune themselves to the social environment and manage their behaviour independently, this is sometimes viewed as being socially competent. These two aspects of child behaviour have been the foci of studies investigating the social interactions of children with disabilities. Although a number of studies have investigated the duration and types (Antia, 1982; Guralnick, 1992; Lederberg, 1991; Rice, Sell & Hadley, 1991) of social experiences of these children and the strategies used to enter and maintain social interaction (Messenheimer-Young & Kretschmer, 1994; Roberts, Brown & Rickards, 1996) the child's attentional abilities during social engagement has been largely ignored.

Social competence has been defined as displaying appropriate social behaviour when interacting with others, and having the ability to resolve conflict in an acceptable and effective manner (Antia & Kreimeyer, 1992; Berk, 1996). Other definitions have included behaviours, which are goal directed to objects and other resources, so that one succeeds and benefits from the social situation (Berk, 1996; Guralnick, 1990). A further definition is provided by Krasnor and Rubin (1985) who suggest that social competence is being effective in achieving social goals. The literature also suggests that social competence is having the ability to utilise environmental personal resources in order to achieve an outcome (Ross & Rogers, 1990). Other definitions suggest that social competence is more than just having a number of available strategies for use, but rather having the ability to respond with an appropriate behaviour at an appropriate time that is within the social context (Ross & Rogers, 1990).

The task of participating in a social situation and responding to social cues of other peers present as a major challenge for children who have a disability. Social difficulties are not always easy to detect (Bowler, 1996) but, for children whose disability is clearly identifiable, we are able to observe the range of difficulties these children are faced with during social interaction. According to Krakow and Kopp (1982), these difficulties relate to inattention to selected tasks, distractibility, impulse control and inappropriate behaviour in social settings.

Depending on the nature and severity of the condition, children with a disability will have difficulty in processing information from their social environment. Given that successful social interaction depends on good communication skills (Craig, 1993) and to this we can include cognition, children who have deficits in one or both of these developmental domains are likely to be at risk during social interaction with a peer who has no disability. The effects of these difficulties lead to problems in maintaining social interactions with their peers (Arnold & Tremblay, 1979; Guralnick & Weinhouse, 1984; Levy-Shiff & Hoffman, 1985), they are less successful in their social bids to peers (Guralnick, Connor, Hammond, Gottman & Kinnish, 1995), they develop fewer friendships (Guralnick & Groom, 1985), and they are less willing to risk entry attempts in play activities (Remine, 1996).

Research in the inclusion movement suggest that the ability of children with a disability to establish and maintain social relationships is central to their social acceptance and social integration in the inclusive setting (Fujiki & Brinton, 1994; Guralnick, 1990, 1992; Guralnick et al., 1995; Odom, McConnell & McEvoy, 1992; Odom, McConnell, McEvoy, Peterson, Ostrosky, Chandler, Spicuzza, Creighton, Favazza, 1999). These studies have revealed that the absence of specific skills will affect interaction with peers. According to Odom et al. (1992) these deficits in social interaction may be a result of one or more developmental domains. These are communication, interpreting auditory and visual information, attending, and organising information and the knowledge gained from the environment. Cognitive models or frameworks measuring social competence continued to lead the research on the origins of social skills deficits in children (Ladd, 1999). At the core of these frameworks for social competence are underlying social cognitive competencies such as language and cognition (Leffert & Sipperstein, 1996).

These frameworks for social competence have been developed to measure the social behaviours and skills of young children when engaged in social interaction with their peers. These theoretical frameworks help educators understand how children acquire skills, process social information and how interactional difficulties develop between children. The information processing framework initially developed by Dodge (1985) and which has been reformulated into social information processing (Crick & Dodge, 1994) has set the theoretical foundations for these frameworks.

Some frameworks have adopted a social goal perspective (Rose-Krasnor, 1985), while others have embraced a social task perspective (Dodge, McClaskey & Feldman, 1985; Putallaz & Gottman, 1981). By framing social competence within social goals and social tasks that are present in the social environment, it enables us to see how children integrate skills across developmental domains, and how children organise their knowledge across sequences of events (Guralnick, 1990, 1992). The demands of child-child interactions inform educators of the likely problems that children may encounter during social exchange.

Rose-Krasnor (1985) proposed a social problem-solving framework. This researcher suggests that, in a social situation, children set certain goals. Based on the children's assessment of the social environment in which these goals are to be pursued, children will access and select specific cognitive strategies. These strategies are transformed into behaviours, and further events are assessed to determine whether the initial goal was achieved. The final result will determine whether further modifications to the original goal and strategy are warranted.

In a framework by Ladd and Crick (1989) which is similar to Rubin and Krasnor (1986), these researchers propose that, in a social situation, children pursue certain goals, which are driven by self-perceptions and emotions. In this framework the basic unit is social exchange. Social exchange is defined as an interaction between a child and peer (Erdley & Asher, 1999). The focus of the social exchange is on the action that precede behavioural enactment and the action involved in response to the behaviours. The goals are viewed as precursors to the behavioural enactment and according to Erdley and Asher (1999), these goals determine the particular behavioural strategy to be selected by the child. According to Ladd and Crick (1989) children's attributions and feelings will also determine the outcome of their social exchange. The child may persist with or revise the goal as the social exchange continues.

Other researchers have developed frameworks that adopt a social task perspective on social competence. Putallaz and Gottman (1981) adopted a social task perspective during social interaction by focusing on how children respond to the task of gaining entry into an established activity or interaction with a peer. These researchers considered that gaining successful entry into a group during play is a fundamental task for children. Peer group entry is not an easy task to master during most play activities and it appears that almost 50% of initial attempts to enter a group are rejected or ignored (Corsaro, 1981).

Included in social cognitive competencies is the area of attention (Kirchner & Knopf, 1974; Ruff, Cappozoli & Saltarelli, 1996; Tipper, Bourque, Anderson & Brehaut, 1989; Tomporowski & Tinsley, 1996). When observing the dynamic nature of social interaction, these social cognitive capacities make a significant contribution to the child's ability to attend and to participate in that interaction. These frameworks have, therefore, helped to explain the relations among children's social cognitive processes, behaviour and peer acceptance (Crick & Dodge, 1994; Ladd & Crick, 1989) which enable the child to produce social behaviour that matches a particular situation (Guralnick, 1992). By including attention as an underlying cognitive process to social competence, we are able to view social competence from both a social cognition and an attention perspective. This is more likely to help educators understand the nature and effects of disability during social interaction in the inclusive environment (Leffert & Sipperstein, 1996).

The frameworks presented in the literature have been intended for the measurement of social competence of children who have communicative abilities and who do not have cognitive delays. These frameworks have identified several social-cognitive behaviours that are necessary in order for a child to have successful social interaction. Strategy generation and implementation is a common behaviour reflected in these frameworks, and for strategies to be generated and carried out, social communicative skills need to be integrated, organised and sequenced during the course of the social task (Guralnick, 1992). Several studies suggest that these frameworks may be considered as too sophisticated for use with children who have difficulties in cognition and difficulties in communication (Guralnick, 1992; Odom et al., 1992; Rose-Krasnor, 1985). For a child with a disability it is this process of integration, organisation and sequencing that may present challenges to the child, rendering it difficult for them to create strategies. These difficulties of processing and integrating information experienced by the child with a disability are a result of not having the ability to attend to information in the environment. In order to understand the nature of the problem experienced by children with disabilities during interaction, Odom et al. (1992) suggest that a refinement of these existing frameworks and approaches may be necessary for children with a disability.

Given that attention problems have been identified in several studies of children with disabilities (Bedi, Halperin & Sharma, 1994; Feagans, 1986; Feagans, Kipp & Blood, 1994; Hindley & Kroll, 1998; Kasari, Mundy, Yirimiya & Swanson, 1990; Keogh & Margolis, 1976; Ruskin, Kasari, Mundy & Sigman, 1994; Smith, Quittner, Osberger & Miyamoto, 1998) it is likely that the ability of children with disabilities to encode appropriate social cues will be significantly affected, which would distort Dodge's (1986) information processing model and impact on strategic behaviour identified by Putallaz (1983) and Rose-Krasnor (1985). It is therefore suggested that having the ability to attend should be considered as a necessary social cognitive tool in the frameworks designed to measure social competence.

Children are born into a world full of interesting and complex stimuli, and in order to be able to adapt to environment changes, children are required to have diverse skills such as being able to orient themselves to the situation, responding to social overtures either verbally or gesturally (Warner-Rogers, Taylor, Taylor & Sandberg, 2000). To acquire these skills, children will be required to have the ability to interpret, to select and to enact upon relevant information from the environment. A further skill is necessary in which children are required to direct actions and behaviours to relevant stimuli with the presence of other irrelevant stimuli (Tipper et al., 1989). Researchers share the view that attention is considered to be an important factor for optimum cognitive development and learning, particularly with the presence of competing stimuli (Lewis & Baldini, 1979; Ruff et al., 1996; Warners-Rogers et al., 2000).

Attention is regarded as fundamental to human thinking (Berk, 1996) and it is considered to be a central process (Kopp & Vaughn, 1982; Uttal, 2000). Attention is fundamental to learning and strategic information processing (Miller & Harris, 1998) and therefore, it is significant to a child's development because it assists the child in learning and understanding the object and social world (Krakow & Kopp, 1982). Most educators and researchers generally consider attention as a precursory state to the learning of most tasks. Attention has been described, as attending to a task or goal which can be internally or externally motivated (Gibson & Rader, 1979; Ruff & Rothbart, 1996). The area of attention has therefore been useful when investigating learning difficulties and information processing (Keogh & Margolis, 1976; Kopp & Vaughn, 1982). While the attention system is operating at an optimum level, performance on tasks and interactions with other individuals are likely to be more successful.

More recent work on attention has focussed on the importance of considering the interaction between the individual and the qualities of the environment (Dykeman, 1998). For instance, Berk (1996) suggests that attention determines the source of information to be considered for most tasks and problems and that it is critical to determining the outcome of one's actions and decisions. Attention has been defined as the ways in which individuals relate to information in the environment, how this incoming information is processed (Dykeman, 1998), and then how individuals act upon this information (Kopp & Vaughn, 1982). Others have seen the attention process as an individual's ability to regulate stimuli through the initiation and inhibition of the individual's behaviour (Dykeman, 1998; Kanwisher & Downing, 1998). Although this process involves each sensory modality, most studies have focused on visual or auditory processing (Dykeman, 1998; Kanwisher & Downing, 1998; Miller & Harris, 1988; Rose, Feldman, Wallace & McCarton, 1989).

Individuals who are described as having attention problems are those who either demonstrate inappropriate and disruptive motoric activity or, have difficulty in the selection and organisation of a task and its components (Keogh & Margolis, 1976). These two factors, level of activity and planning ability, are likely to interact with each other. As a result, problems in one may lead to the development of difficulties in the other. Individuals who have problems in attention have been reported to have difficulties making decisions particularly during problem-solving situations and to be less able to maintain their attention once a task has begun (Keogh & Margolis, 1976). As a consequence, individuals with attention problems often act on impulse in learning and social situations and thus risk their chances of success or achievement in these two situations.

Children with disabilities are often reported as displaying behaviours that are characteristic of attention problems (Keogh & Margolis, 1976). These include problems with attention span, self-regulation, activity level, impulse control, distractibility, inappropriate role behaviour directed to parents and peers, and poorly controlled behaviour in social settings (Douglas, 1980; Keogh & Margolis, 1976; Kopp & Vaughn, 1982; Teeter & Semrud-Clikeman, 1997). In the main, studies have focused on the attention behaviour of children with additional needs; such as, children with Down Syndrome (Krakow & Kopp, 1982; 1983), children with a hearing impairment (Feagans, 1986; Lederberg, 1993; Wood, Wood, Griffiths, & Howarth, 1986), children with learning disabilities (Keogh & Margolis, 1976), and children described as hyperactive (Douglas, 1980). Attention has particularly been identified as a major issue for children belonging to these clinical groups and is suggested as being a critical factor for other groups of children.

Work by Krakow and Kopp (1982 and 1983), comparing the sustained attention behaviours of children with disabilities and the attention of children without disabilities, found the attention of children with Down Syndrome to be focussed exclusively on toys to the point of not engaging with others or monitoring the environment. These children glanced around the room less frequently than did their peers. In contrast, the children without a disability were able to monitor and appraise the environment whilst remaining engaged in play. There was an imbalance between their engagement with the play materials and their interaction with the environment.

The attention problems of children with a hearing impairment have been described as being particularly related to divided attention (Lederberg, 1993; Wood et al., 1986). During play, a child without a hearing impairment is more able to deal with the incoming simultaneity of visual and auditory information. A hearing impairment however, forces the child involved in a task to divide their attention sequentially between the environment and the activity in which they are involved. Wood et al., (1986) describe the need for a person with a hearing impairment to shift their visual attention from the environment to their partner in order to receive their partner's information. In the case of children and play, children with a hearing impairment must divide their attention sequentially between the environment, the toys they are playing with and the peers interacting with them. The act of communicating and problem solving during play becomes a greater challenge for the child with a hearing impairment because of the necessity to divide attention (Lederberg, 1993).

Further evidence for the importance of auditory acuity to attention has come from a study of the effects of otitis media, suggesting that even a temporary hearing loss can result in long lasting effects on attention. This study by Feagans (1986) found that children with otitis media had problems in attending which persisted into the period after the condition had been resolved. Feagans suggests that these changes in attention patterns had then become habit-forming behaviour as a result of the difficulties inherent in having an intermittent hearing loss.

Studies investigating the attention of children with autism have found that their attention span is maintained for activities that interest the child (Attwood, 1998; Wing, 1998) enabling the children to sustain attention for a considerable length of time. Attention to tasks and activities that are outside their range of interest, however, is likely to be much shorter, fleeting or non-existent (Wing, 1998). This behavioural trait displayed by children with autism has been described in the literature as selective attention (Dykeman, 1998; Sohlberg & Mateer, 1989). When challenged by a task, therefore, these children are often seen to be distracted (Wing, 1998), resulting in poor on-task performance. Recent studies of the cerebella of autistic patients (Courchesne, Akshoomoff, Townsend & Saitoh, 1995; Yamaguchi, Tsuchiya & Kobuyashi, 1998) suggest that the cerebellum is responsible for controlling attention shifts in individuals. Based on the anatomical, physiological and behavioural studies of patients with autism, the literature suggests that patients with autism have particular difficulty in executing attention shifts. The impaired attention shifts have been found to occur both between auditory and visual modalities and within the visual modality (Akshoomoff & Courschesne, 1992). Managing attention and executing attention shifts during social engagement are therefore, likely to be difficult tasks for children with disabilities.

Research investigating social attention has done so within the context of adult-infant interaction (Adamson & Bakeman, 1991; Kasari, Sigman, Mundy & Yirimiya, 1990). The studies investigating social attention of young infants suggest that they will learn to acquire social competency skills through social interaction and the developing relationship with adults (Carpenter, Nagell & Tomasello, 1998; Tomasello & Farrar, 1986). According to Carpenter et al. (1998), by twelve months of age infants are able to look where adults are looking and the infant is able to use the adult as a social reference point. These researchers also suggest that at this same age the infant will begin to direct the adult's attention to other sources using intentional communication by gesture, which will soon be replaced by attempts to vocalise.

When we view social competence from a relationship and attachment perspective, the child is seen within the context of vertical and horizontal relationships. Hartup (1989) suggests that experience in both vertical and horizontal relationships is necessary for optimum growth and development. Vertical and horizontal relationships serve different functions in the child's development (Hartup, 1989). Protection and security is provided for the child for many years during the vertical relationships and continues until the child is able to function independently. A range of basic social skills will emerge from this relationship. With horizontal relationships, children are placed in situations in which they can elaborate the basic skills and abilities with others that are more or less similar to themselves in age. Within these relationships, children are exposed to the complexities of interaction and competition, with a view to developing proficiency in these areas. It is in the horizontal relationships, according to Hartup, that the child is able to achieve intimacy in social relations. According to Hartup (1989) social competence emerges mainly from experience in close relationships and these social experiences are likely to serve as the foundation to the development of social attention.

Another view to social competence is the social co-construction theory (Valsiner, 1988). The social co-construction theory describes the abilities and difficulties of children during play within a developmental theory (Valsiner, 1988). According to theorists of co-construction (Valsiner, 1988), knowledge is gained through the contributions of the environment and the individual. Valsiner proposes that individual children's actions have no meaning in isolation, and that their actions only acquire meaning when negotiations and interpretations between partners are present. Winegar (1988) further suggests that a child's performance in social interaction is a result of knowledge and understanding of social scripts. However, the child's performance during social interaction will be further facilitated by the child's ability to organise and process the information from the environment (Winegar, 1988).

The co-constructivist theory suggests that to be successful in one's play interactions, a child must enter the play situation with a similar view of how to act and what to say in this situation with their playmate. This theory also suggests that during play, children are in a situation that requires them to adapt their language, and actions to the continuous changes in the physical and social situation as the interaction occurs between partners. This theory further reinforces the importance of attention and its contribution to successful social interaction between playmates.

Special educators have long been concerned with the attention abilities of children who have disabilities (Keogh & Margolis, 1976). Since attention is considered a cognitive process (Kopp & Vaughn, 1982) which is critical to learning, it is also considered to be sensitive to differences between children with disabilities and children who do not have a disability. For instance, Keogh and Margolis (1976) suggest that "it is the nature of the attentional disruption or the ineffectiveness of attention for particular learning or skills" (p. 20) that separates these two groups of children.

Investigations into the social interaction of children with disabilities continue to dominate research in this area. The investigations have documented the difficulties these children encounter during social engagement opportunities with their peers. Several studies share similar findings and have found that children with disabilities will have occasions where there is no interaction with a peer, and when there is an opportunity for interaction, they will have difficulty in maintaining the social exchange (Antia & Kreimeyer, 1988; Arnold & Tremblay, 1979; Craig, 1993; Dykman, Ackerman, Clements & Peters, 1971; Fujiki & Brinton, 1994; Gertner, Rice & Hadley, 1994; Guralnick, 1980, 1990; Guralnick et al., 1995; Levy-Shiff & Hoffman, 1985; Minnett, Clarke & Wilson, 1994; Ross & Rogers, 1990; Rubin, 1986).

Further studies have documented the outcomes of these difficulties during social interaction. These studies have found that, in the inclusive setting children with disabilities are likely to spend more time interacting with teachers than children without a disability (Brophy & Hancock, 1985). The children with a disability are more likely to interact with a peer of the same status that is, another child with a disability, than with peers who do not have a disability (Arnold & Tremblay, 1979; Guralnick, 1980). Antia (1994) found that children who engage in high levels of interaction with teachers may, therefore, engage in low levels of interaction with peers who do not have a disability.

Research on teacher-child interaction which focused on the teacher within the larger context of social interaction found that children with a disability received more direction, and teachers used more directed play with children who have a disability (Brophy & Hancock, 1985). In preschool settings, children with a disability received more prompts and more positive reinforcement by their teacher (Guralnick, 1981) and children with a disability spent more time alone with the teacher (Guralnick, 1991). These studies show that teachers tend to dominate the classroom interaction and, as a result, children with a disability may become dependent on adults for interaction (Antia & Kreimeyer, 1994). According to Ross and Rogers (1990), adults are better able than peers to understand and adapt to the needs and interests of a young child. For the child with a disability, interactions with adults do not create the same demands for negotiations of social problems, misunderstanding or conflicts that interactions with children do.

Studies investigating the social interactions of children with a hearing impairment have found that children who have a hearing impairment also have higher rates of interaction with teachers than their hearing peers (Antia, 1994). This researcher found that in an inclusive classroom, young children without a hearing impairment interacted with teachers between 3-6% of recorded intervals, while children with a hearing impairment interacted with teachers 5-25% of recorded intervals.

A study by Minnett et al. (1994) investigated the play behaviour and communication between hearing impaired and hearing children in an integrated preschool during outdoor play. The results of this study found that all children preferred to play and communicate with children of the same hearing status. In relation to the participants with a hearing impairment, the amount of social play differed between the two communication environments and the context of interaction was related to the behaviour and communication of children who had a hearing impairment. The results of this study suggest that outdoor play may not necessarily provide social benefits for children with a hearing impairment as children are far more likely to prefer same-hearing status in that context.

In a study by Guralnick (1980), the social interactions between preschool children at different developmental levels were investigated. The researcher observed the frequency and nature of communicative interaction for the different subject groups during morning integrated free play periods. The results of this study found that children without a disability and children with a mild disability interacted with each other more frequently than did the groups of children with a mild disability. The children with no disability and children with a mild disability interacted less frequently with children who had a moderate disability and children who had a severe disability. The participants with a moderate disability and children with a severe disability interacted with all four developmental groups. However, the play encounters of these participants were brief and poorly organised. This study suggests that developmental status is a factor that will determine social interaction between children with and without disability in a inclusive setting.

Given the results of the studies of children with a hearing impairment and children with an intellectual disability it suggested that children with disabilities are more likely to be observed in relationships that are vertical. Moving into horizontal relationships is dependent upon several factors such as, level of disability, linguistic competence and hearing status. Researchers suggest that it is only with equals that children have the opportunity to try out their skills, attempt to negotiate and exert control over their social world (Rubin, 1978). This view is further supported in another study by Antia (1985) who also found that children with hearing impairment tended to interact more frequently with peers who also had a hearing impairment in the preschool. This study also suggests that teachers should plan for social integration by teaching children with a hearing impairment social communicative skills such as, initiating and maintaining conversation and extending invitations.

Attention can be selective or it can be focused to a particular event or object. According to Ruff and Rothbart (1996), the effectiveness of an object or an event to hold a child's attention is not only dependent on the physical characteristics of the target, but also on the child's state. Internal changes occur in individuals which elicit different kinds of behaviour when presented with specific stimuli. These internal conditions are referred to as states and they can persist for different lengths of time (Ruff & Rothbart, 1996). Changes in state will involve changes in either internal or external physical behaviours. An attention state is likely to involve both increased inhibition of responses to competing stimulation (Tipper, 1992) and enhanced effort and concentration (Kahneman, 1973). Physiological studies of the brain indicate that attentional states and the shifts between them are controlled by the cerebellum in the brain (Allen, Buxton, Wong & Courchesne, 1997; Askshoomoff & Courchesne, 1992; Yamaguchi et al., 1998).

Attention is referred to as a state which will vary between children (Ruff & Rothbart, 1996). It is also dynamic in nature, as the presentation of competing stimuli will influence the shift between different states. More recent investigations of different states of attention have described these states in children as being distractible, being alert, being focused, being able to sustain, being selective and finally being able to alternate attention. Dykman et al. (1971) used the terms alertness, stimulus selection, focusing and vigilance to describe different sub-states of attention. Keogh and Margolis (1976) suggest that, while attention is made up of different partially independent components, these components also have specific functions, which may influence learning and performance differentially. These researchers suggest that individuals bring abilities to the task of maintaining attention such as being vigilant, being focused, sustaining attention and managing divided attention. This multi-component view of attention is in keeping with information processing studies that have identified a variety of functions linked to attention such as selectivity, focusing, sustaining concentration or vigilance, switching attention, distractibility, modulating the intensity of attention and attention to memory processes such as rehearsal, retrieval and coding (Parasuraman & Davies, 1984).

In relation to the management of attention states, Ruff and Rothbart (1996) suggest that attention needs to be regulated. It has been suggested that central executive functions which are in the frontal regions of the cortex become operational when individuals are required to allocate attention, control concentration and effort and plan for sequenced tasks (Stuss & Benson, 1984). The terms "executive" and "higher level" have been used in the literature to describe the thought processes. Ruff and Rothbart (1996) refer to higher level as the integration of many processes in the control of attention and behaviour. Given the discussion so far about attention, higher level gives a more accurate explanation of what is happening when an individual is using these processes.

The organisation and control of many activities which surround our lives becomes important especially when a range of tasks need to be addressed, and when individuals are required to interact with others. In this situation, individuals are required to prioritise goals for each day and to plan their actions when interacting with others. The executive or higher control is the mechanism which now determines the prioritising of goals, how an individual sequences the tasks and how the individual times the actions to meet goals (Hughes, 1998; Ruff & Rothbart, 1996). This higher level control is an active process and becomes refined through experience in dynamic interactive activities.

This paper suggests that attention is made up of different states which are managed by a higher lever control system. Individuals are able to move through different states of attention or they are able to plan, prioritise goals and enact on tasks while the higher level order system is working at an optimum level. More recent descriptions of attention have highlighted levels or states which are pivotal to memory, learning and cognition (Dykeman, 1998; Sohlberg & Mateer, 1989). For this paper the states of inattention, aware/alert, sustain, focus and divided attention will be discussed.

The ability to maintain attention will vary between individuals (Keogh & Margolis, 1976). Some clinicians interpret distractibility, or inability to sustain attention, as a sign of neurological impairment (Douglas, 1980). Distractibility or inattention is considered as a deficit in focused attention, and the distractors are task irrelevant stimuli whose presence or introduction impacts on the participant's performance (Bedi et al., 1994).

Given that attention is a particularly vulnerable aspect of cognitive processing (Berk, 1996; Krakow & Kopp, 1982), a child who is inattentive or easily distracted is therefore, more likely to miss significant information in the learning of a task or miss social overtures in an interaction with a peer. The impact of the deficits in attention on learning and social behaviour is significant and requires further investigation (Lyon, 1996), especially within the context of children with disabilities.

Alertness has been described in the literature as the development and maintenance of optimal sensitivity to the environment. The ability to prepare and sustain alertness is considered to be an important attentional function (Posner & Petersen, 1990). It has also been proposed that the ability to maintain and sustain a state of alertness is considered to be a major challenge for many individuals (Posner & Raichle, 1994).

Successful performance in learning and in social situations is dependent upon the child's ability to attend to the relevant stimulus (Kirchner & Knopf, 1974). Studies investigating the alert state of participants have conducted vigilance tasks. The literature suggests a possible relationship between vigilance and achievement in the learning situation.

The ability to remain engaged in a play or task situation is critical to learning, and behaviours in early infancy can give some insight into cognitive performance in later childhood (Douglas, 1980; Kopp & Vaughn, 1982; Krakow & Kopp, 1982, 1983). Kagan and Kogan (1970) suggest that while the infant engages in play and exploration of toys and objects, sustained attention of the infant becomes more noticeable to caregivers during the latter part of the first year of life. This suggests that the infant's experiences with objects and their environment contribute to the development of sustained attention.

Sustained attention involves being able to stay on task in a vigilant manner over a period of time (Bedi et al., 1994; Mirsky, 1996). This would require an individual to sustain their attention on task relevant inputs while ignoring or filtering out irrelevant inputs. Sustained attention is usually measured by decreases in performance over time (Halperin, 1996). For learning and performance to be successful, individuals are required to sustain their attention for an optimal period of time. This has been demonstrated by the investigations on rhesus monkeys using a computer game and joystick. In order to keep the target on track, the monkeys were required to sustain attention. Rumbaugh and Washburn, (1996) found that while rhesus monkeys were able to pursuit the target for more than sixty seconds, the monkey's attention was sustained for a longer period of time.

Focused attention is having the ability to attend to one specific stimulus while ignoring irrelevant stimulus. It is assumed by Ruff and Rothbart (1996) that, during the focus state, selectivity to a target becomes restricted to fewer elements and that the degree of effort directed at the target increases. This would suggest that the attention to be deployed in this state would vary from being highly concentrated effort to low concentration effort. Focus therefore, refers to having the capacity to concentrate on a specific task and to being able to screen out distracting stimuli (Mirsky, 1996). For example, a child may be interacting with a playmate and ignores other playmates passing by or calling out.

When individuals are engaged in two different tasks at once, they have to divide attention between the tasks and allocate mental resources to each (Iidaka, Anderson, Kapier, Cabeza & Craik, 2000; Styles, 1998). This definition of divided attention highlights the issue as to how much can be attended to at once (Wenger & Townsend, 2000), and how information is organised (Uttal, 2000). This ability to attend to more than one stimulus can therefore, be considered as a general estimate of attentional capacity (Uttal, 2000).

Divided attention has been measured using neuroimaging techniques. These techniques have enabled researchers to measure activity in the human brain while performing multiple tasks. A multiple task environment requires an individual to perform two or more tasks at the same time (Stit & Fisk, 1999). An example of a multiple task is while a child is playing with blocks in the block corner, the child is building with blocks and talking to a playmate simultaneously.

Attention is a necessary tool to organise behaviour in the presence of competing stimuli (Ruff & Lawson, 1996) and so, children are required to manage or shift their attention (Posner, Petersen, Fox & Raichle, 1988). This suggests that attention is a state that is not static, it will involve an individual moving between states of being inattentive to a state of being engaged.

The ability to shift one's focus of attention is considered by most educators to be important, especially when one is required to attend to more than one object of interest. For most children, this skill is acquired through a range of experiences in social situations and through the development of specific cognitive skills. For children with an intellectual disability, the ability to shift focus cannot be managed easily, as it places specific cognitive demands on them (Kasari, Freeman, Mundy & Sigman, 1995).

The literature has identified that children with disabilities have difficulties becoming socially engaged, and frameworks of social competence have been developed to measure children's capacities during social interaction. Research suggests that the application of these models have been intended for the measurement of social competence for children who do not have delays in communication and cognition. In order to understand the problems experienced by children with a disability, a refinement of these existing frameworks may be necessary, as they are considered to be too sophisticated for use with children who have a disability.

A framework of social attention was for this study. This social attention framework attempts to measure the child's opportunities for social interaction and their states of attention. The first level is that of the social engagement opportunity. The second level measures the type of social engagement opportunity. The third level measures the child's state of attention during social engagement opportunities. The social attention framework is shown in Figure 1.

Figure 1: Social attention framework

In this framework it is proposed that the stimuli from the social environment are likely to affect the performance of children during social interaction. The social environment in this framework is never static and is shown by the broken lines. The competing stimuli from the social environment may influence whether children will have opportunities to become socially engaged, the type of social engagement opportunities and how children attend during social engagement opportunities. According to Guralnick (1992) interaction with a peer involves higher order processes, such as monitoring a peer's responses that are likely to assist in the selection of specific social strategies. Guralnick (1992) suggests, during a social task, the higher order processes will involve the integration, organisation and sequencing of social behaviours and therefore influence how children become socially engaged and possibly determine their social attention state. For children with a cognitive or hearing impairment the higher order processes may be affected, thus impacting on how they become socially engaged and how they attend to the social task.

At Level 1 data were collected on the social engagement opportunities that were available with a peer or a group of peers. Of interest here were the number and duration of these opportunities.

For Level 2, the types of social engagement opportunities were coded. Four conditions were identified. These were disengaged, proximal, visual and interactive. Any one of these conditions is likely to lead to social interaction.

The third level was the child's attention state. Five states of attention were identified. These were inattentive, aware-alert, sustain, focus and divided. In this level, the child could move from a state of being inattentive to a higher order level of showing divided attention. To maintain the interaction during a social engagement opportunity, children are required potentially to call on and manage different states of attention to match the requirements of the interaction and the wider context in which the interaction takes place. Given that the stimuli in the social environment are constantly changing, children have not only to attend to their own role and activity, but also need to check that they are following the group script. For example children playing mums and dads in the home corner.

It would now appear critical to factor in the issue of attention problems into any assessment of a child's level of involvement and performance during social interaction. In having a problem in attention, the child with a disability readily loses important visual and auditory information emanating from the social environment which enable the child to analyse the state of the current social context and formulate an appropriate and effective strategy to become socially engaged. The research suggests that investigators need to consider the social context (Guralnick, 1992) and to consider conditions that are likely to affect the social attention of children with disabilities during social tasks. Although the child may be experiencing difficulties in other identified areas, such as hearing, vision, communication, or interpreting information gained from the environment, the added contribution of poor attention skills may exacerbate social engagement problems.