Journal of Early Intervention, 1999 Vol . 22, No . I, 70–86 Copyright 1999 by the Division of Early Childhood . Council for Exceptional Children The Nature and Meaning of Social Integration for Young Children With Mild Developmental Delays in Inclusive Settings

MICHAEL J. GURALNICK University of Washington

This paper examines fundamental conceptual, methodological, and outcome issues with respect to the social integration of preschool-age children with mild developmental delays in inclusive early childhood settings. Cast within a developmental-ecological framework, social integration is evaluated in terms of three constructs: (1) the connectedness of peer interactions; (2) the quality of interpersonal relationships ; and (3) the nature of adjustments that occur during social exchanges. A general model of factors that influence social integration and their interrelationships is presented as a means of organizing future intervention activities to promote social integration.

One expectation of inclusive practices is that goal of social integration has been achieved in meaningful social relationships will form be- inclusive preschool settings is examined in this tween children with and without special needs paper. To do so, a conceptual and methodolog- as they become familiar with one another in ear- ical framework is established first to character- ly childhood settings. This expectation that so- ize the nature and meaning of social integration cial integration will occur with respect to chil- in inclusive settings. This is followed by a sec- dren's peer relationships and friendships seems tion in which the available data on social inte- reasonable because inclusive practices empha- gration are organized and evaluated within this size principles and values that seek to maximize framework. Finally, two general approaches de- respect for individual differences in develop- signed to promote social integration are consid- It, ensure equal access, and foster a sense of ered, and a general model relating factors influ- .,belonging to a common community (Guralnick, encing social integration is presented. 1978, 1990). Ideally, we would hope that inclu- It is important to note that this discussion sive early childhood programs would be char- will be limited to preschool-age children with acterized by children's willingness to under- mild developmental (cognitive) delays .' This stand, go beyond, accept, and even overlook de- velopmental differences, unusual behavior pat- ' The population of children with mild developmental (cognitive) terns, or certain physical characteristics of their delays described in this paper was carefully defined using well- accepted criteria based on intelligence test scores and measures of peers, and establish productive social relation- adaptive behavior. Specific exclusionary criteria also were estab- ships. To support the development of these pos- lished with respect to children's behavior problems, sensory def- icits, motor impairments, and communication disorders . Etiology itive relationships between children with and of the delay (when known) was not considered, as the categorical without special needs, the programmatic design definition was applied uniformly. Of note, within this relatively homogenous and well-defined population, only weak associations of quality early childhood settings should ex- are obtained between peer-related social competence and intelli- emplify inclusive principles and values by pro- gence or language measures, although behavior problems are more strongly correlated (Guralnick, Connor. Hammond, Gottman, & moting full participation of all children in social Kinnish . I996a; Guralnick & Groom 1985) . Individual differences and nonsocial activities, and by adapting and in peer interactions and social integration appear to he related to more process-type factors (see text) . Other well-defined popula- accommodating to children's special needs. tions, such as children with communication disorders, are likely In view of the importance of what is certainly to exhibit patterns different from those found for children with mild developmental delays (e .g ., Guralnick, Connor, Hammond, a core issue in our field, the degree to which the Gottman, & Kinnish, 1996h).

70 JEI, 1999, 22:1

is a relatively high incidence group even at development, however, that are more difficult the preschool level ; one that generally is in- to conceptualize or to assess than the domain cluded in typical preschool or daycare set- of peer-related social development (Howes, tings. Moreover, by focusing on children with 1988). The complexity of this issue becomes mild developmental delays, a framework for apparent when we realize what must be con- examining critical developmental and context sidered: (a) the various manifestations of so- issues can be established ; something not cial integration (e.g., active acceptance, pas- readily accomplished for heterogeneous sive integration, exclusion, rejection), (b) the groups of children . In fact, it will be argued varying strengths of a relationship (e .g., ac- that both developmental and ecological per- quaintanceship, intimate friendship), (c) the spectives are essential for a complete under- different types of data (e .g., observational, standing of social integration in inclusive set- phenomenological), (d) the specific character- tings for this group of children . Such a per- istics of playmates (e .g., chronological age, spective may encourage similar approaches gender), and (e) the context in which peer in- for other well-defined groups of children. teractions take place (e .g., free-play or struc- tured activities ; dramatic play or motor-ori- THE MEANING OF SOCIAL ented activities) . Thus, the first challenge for INTEGRATION an outcomes approach is to define (with ex- pectations) and measure the dimensions of so- Unfortunately, no generally accepted criteria cial integration as represented in inclusive set- exist that can inform us about the extent to tings. Then, we must examine the available which social integration has been achieved. data in relation to this set of expectations and Consequently, a framework that carefully dimensions. specifies expectations for specific patterns of social integration outcomes is needed . These EVALUATING SOCIAL expectations can then be evaluated to deter- INTEGRATION OUTCOMES mine if specified outcomes have been realized. Of importance, establishing expectations is To establish an outcomes framework, three so- somewhat arbitrary and will vary with the cial integration constructs and corresponding purpose or value framework developed to measures are defined and examined: (a) the guide the analysis. A reasonable approach connectedness (or extent) of peer interactions; (though perhaps idealized) is to expect full in- (b) the quality of interpersonal relationships; tegration to be found across all specified so- and (c) the nature of adjustments that occur cial interaction dimensions . Other approaches during social exchanges . Data relevant to each may establish lower expectations or may an- of these three social integration constructs in ticipate differences in social integration for relation to interactions occurring between different dimensions of social interaction or children with and without developmental de- differences depending upon whether assess- lays are presented. ments are obtained from the perspective of It is important to point out that the social children with or without developmental de- integration of children with mild developmen- lays. The important point is that expectations tal delays will be evaluated in relation to typ- are made explicit. ically developing chronological age mates. Conceptually, an outcomes-based frame- Despite the fact that children with mild de- work assumes that identified patterns of social velopmental delays are less developmentally integration meaningfully represent variations advanced than typically developing children in children's social experiences . Correspond- of the same age, parents view typically de- ingly, it assumes that a methodology is avail- veloping age mates as the appropriate refer- able or can be constructed in which measures ence group for their child (Guralnick, Connor, can be derived to index identified social in- & Hammond, 1995). In addition, social inte- tegration patterns . There are few domains in gration as evaluated within the framework of

Guralnick 71

the three social integration constructs will pri- of participants found in the studies of social marily take place from the perspective of the integration has not allowed analyses of spe- typically developing children. Moreover, in cific subgroups of children, such as children this analysis, the expectation (hypothesis) is with mild developmental delays who are the that complete social integration is achieved focus of this paper. The value of "specificity," when typically developing children are con- i.e., selecting well-defined and more homo- nected to and maintain the same quality of geneous subgroups of children, has been em- interpersonal relationships with children with phasized for the domain of social integration mild developmental delays as they do with (Buysse & Bailey, 1993 ; Guralnick, 1981b) children without delays. This analysis further and for the general field of early intervention assumes that for complete social integration to (Guralnick, 1997b, 1998). In addition, existing occur, typically developing children must groups of children with and without disabili- make appropriate adjustments to the unique ties usually have been observed in these stud- developmental characteristics of children with ies, yet few efforts have been made to control delays. This ideal, but by no means unreason- for relevant child and family characteristics able, expectation is examined below. such as chronological age, gender, socioeco- Most studies of social integration have in- nomic status, or familiarity . In the absence of cluded heterogeneous groups of children with these experimental controls, it is difficult to disabilities, often consisting of children exhib- interpret findings on social integration within iting a range of motor, cognitive, behavioral, a developmental framework. and communicative disabilities. In my earlier The only series of studies in which rela- review of the social integration literature tively homogenous groups of children with (Guralnick, 1981 a), the interactions of typi- mild developmental delays can be separately cally developing children with other typically analyzed is that of Guralnick and his col- developing children formed the reference leagues (Guralnick et al ., 1998; Guralnick, point from which to evaluate the extent of in- Connor, Hammond, Gottman, & Kinnish, tegration for diverse groups of children with 1996a; Guralnick, Gottman, & Hammond, disabilities . Despite the heterogeneity of the 1996; Guralnick & Groom, 1987, 1988a ; Gur- participants and measures, one particularly alnick & Paul-Brown, 1984, 1986, 1989). consistent pattern emerged from these impor- Consequently, only results from this research tant early studies (Cavallaro & Porter, 1980; program will be presented . It is important to Guralnick, 1980; Ispa & Matz, 1978 ; Peterson note that children in this series of studies par- & Haralick, 1977; Porter, Ramsey, Tremblay, ticipated in short-term (2–4 weeks) play- Iaccobo, & Crawley, 1978 ; White, 1980). Spe- groups created specifically for research pur- cifically, the degree of social separation in- poses. Although the playgroup methodology creased as a function of the severity of the offers many advantages (e .g., precise match- child's disability . Of importance, despite con- ing of subjects, children free of reputation siderable within group variability, social sep- bias), its conclusions may be limited to spe- aration was observed even for children with cific features of the playgroup parameters. mild disabilities . Moreover, more recent stud- Thus, to assess the generalizability of the find- ies continue to document similar patterns. ings, comparisons also will be made to related Whether indexed by social exchanges, proso- research. cial behaviors, or friendships, or assessed via observational or peer sociometric measures, Connectedness diverse groups of preschool-age children with Most would agree that the extent to which disabilities are less preferred playmates by children with and without developmental de- typically developing children than are other lays are socially connected with one another typically developing children (e .g., Blackmon constitutes an important aspect of what is & Dembo, 1984; Nabors, 1997 ; Strain, 1984). meant by social integration . The critical ques- Unfortunately, the generally small number tion based on expectations established for this

72 JEI, 1999, 22:1 analysis, however, is whether all children are differences for these interactive measures, equally connected socially, irrespective of data from a recent study revealed no overlap their developmental status. From a quantita- for the 95% confidence intervals of the means tive perspective, it is possible to determine for children with and without delays . Only whether or not children with and without de- one delayed child received social interactions velopmental delays interact with one another from typically developing children above the to the extent represented simply by their avail- mean for typically developing children inter- ability in the preschool setting. As such, the acting with other typically developing chil- extent to which children show a preference for dren, and approximately 80% of typically de- playmates based on developmental level veloping children prefer other typically devel- serves as a key index of social integration. oping children to children with developmental Connectedness, however, is likely to vary in delays (Guralnick et al., 1996a). terms of the social demands placed on the The high level of social integration found play partners. Passive type play (e.g., parallel, for more passive measures, however, suggests onlooker) places the least demands ; interac- that children with mild developmental delays tive social exchanges, such as those occurring do have numerous opportunities for observa- during associative or group play, consistently tional learning in relation to typically devel- place more demands on a relationship ; and oping children. Because the children often maintaining a friendship is clearly the most play in close proximity to one another, passive demanding form of connectedness. play may well evolve into more active forms As might be expected, when available data of play. Bakeman and Brownlee (1980) ob- are examined in relation to each of the three served that parallel play often serves as a step- major dimensions of connectedness (i .e., pas- ping-stone to group play for typically devel- sive play, interactive play, friendship), differ- oping children, and this also may be the case ent patterns of social integration are found. for children with mild developmental delays For passive play measures such as parallel (Guralnick & Hammond, in press) . Findings play or onlooker behavior, social separation is for the passive measures also address a fre- found to exist between children with and with- quently stated expectation of inclusion; that of out developmental delays, but only to a rela- equal access . Based on this passive measure tively minor extent in comparison to assess- of social integration, children with and with- ments of more interactive play . In contrast, so- out delays seem to move about play areas cial separation is clearly evident for more so- freely, engaging in activities of interest with cially interactive forms of play . This is out regard to the developmental status of their particularly the case when interactive play peers. consists of extended and active social ex- The most demanding and rewarding aspects changes such as positive social interactions or of social relationships emerge when young group play (Guralnick et al., 1996a; Guralnick children establish dyadic friendships (see & Groom, 1987) . Generally, for interactive Howes, 1988; Newcomb & Bagwell, 1995) . A measures of social integration, typically de- unilateral friendship is said to occur when one veloping children interact with children with child prefers a peer (usually indicated by developmental delays about half as often as spending more time with or more frequently expected, based on the number of children interacting positively with a specific peer), but available in the two groups (typically devel- the peer does not show a similar preference oping, mildly delayed). for that child . Nevertheless, a clear preference Within group variation is certainly appar- is exhibited by one child . These unilateral ent, but perhaps the most salient observation friendships are observed among a substantial is the substantial separation that exists be- proportion of preschool children, including tween children with and without mild devel- children with mild developmental delays. opment delays for more extended and active When analyzed in terms of the developmental social exchanges. In fact, despite individual characteristics of the friend selected, typically

Guralnick 73

developing children choose peers with mild children, and they share information less often delays to a much lesser extent than would be (Guralnick & Paul-Brown, 1984, 1989) . Al- expected based on their availability (Gural- though exchanges are not generally hostile or nick et al., 1996a; Guralnick & Groom, unusually negative, disagreements occur more 1988a). When reciprocal friendships are eval- frequently when typically developing children uated (mutual preferences for one another), interact with children with mild delays than typically developing children strongly prefer when they interact with other typically devel- other typically developing children. Of note, oping children (Guralnick & Paul-Brown, children with mild delays have considerable 1989) . This suggests the existence of a stress- difficulties forming reciprocal friendships in ful relationship between children with and general (Guralnick et al . ; Guralnick & without developmental delays . In fact, inter- Groom). personal stress is particularly apparent during conflict situations . Under those circumstances, Interpersonal Relationships typically developing children are more nega- Another perspective on social integration can tive and less positive to children with mild be obtained by examining the content of direct developmental delays than to other similar age interpersonal exchanges. What is the quality typically developing children (Guralnick et of those interactions? Are children with and al., 1998). without developmental delays treated similar- Findings consistent with both the connect- ly by typically developing children? Is there edness and interpersonal relationship mea- evidence that typically developing children sures of social integration are obtained when actively isolate or reject children with devel- typically developing children are asked to rate opmental delays and does this contribute to their playmates in terms of the extent to which the patterns of social separation identified in they like to play with a particular child (Asher, the previous section on connectedness? Singleton, Tinsley, & Hymel, 1979). When It is important to note that based on a va- these peer sociometric ratings are carried out, riety of measures, the vast majority of inter- typically developing children rate children actions occurring between children with and with delays as much lower (less overall ac- without developmental delays are positive in ceptance) and children with delays generally both content and style . In fact, a composite receive fewer positive and more negative rat- measure based on social exchanges judged to ings (Guralnick et al ., 1996a; Guralnick & be negative does not indicate that children Groom, 1987). with mild delays are singled out for poor treat- ment by typically developing children (Gur- Accommodations alnick et al ., 1996a) . There are, however, some For social interactions to be productive, par- fairly subtle indicators suggesting interperson- ticipants must thoughtfully adapt not only to al difficulties. An intensive analysis based on the context but to the unique characteristics, an utterance-by-utterance evaluation of social- styles, and abilities of their partner. When in- communicative exchanges occurring between teracting with children of different ages, for children with and without delays revealed a example, adjustments must be made relative number of concerns. Specifically, when typi- to the cognitive and linguistic characteristics cally developing children request children of the playmate; an adjustment that is ob- with delays to do something, they direct a served to occur even for preschool children smaller proportion of joint requests such as (e.g., Shatz & Gelman, 1973) . Children with "let's" or "we," and they justify their re- mild developmental delays, however, pose un- quests less frequently . In addition, typically usually difficult challenges for their typically developing children direct a greater proportion developing chronological age mates . That is, of strong, particularly unmitigated directives discrepancies existing between a peer's chro- (e.g., "Do this!" "Give me that!") to children nological age and his or her developmental with delays than to other typically developing characteristics must somehow be resolved.

74 JEI, 1999, 22:1

Moreover, appropriate adjustments by typical- mental level rather than developmental status. ly developing children to their playmate's cog- Although one might argue that even social nitive, linguistic, and socioemotional charac- isolation of younger children is inappropriate, teristics reflect not only a sensitivity to com- such chronological age separation may well plex individual differences, but also a willing- be normative (e.g ., Strayer, 1980) . Conse- ness to work hard to initiate and maintain quently, this circumstance provides a devel- connectedness to diverse types of children. opmental explanation for the observed pat- Interestingly, despite evidence of the inter- terns of social separation involving similar personal stress noted above, when typically age children with mild developmental delays. developing children interact with develop- If this explanation constitutes a valid alterna- mentally delayed children they make a num- tive to separation based on the existence of a ber of important adjustments that appear to be child's disability (i .e., delay), then establishing responsive to differing developmental char- expectations for social integration outcomes in acteristics of their companions, particularly inclusive settings should consider these find- their cognitive and linguistic levels . Specifi- ings. That is, the expectations established for cally, to better ensure appropriate communi- social integration might be more appropriately cation, typically developing children use more linked to patterns based on children's devel- directives, clarify messages more often, and opmental level. Again, the choice of expec- rely more upon multiple modes of communi- tations for social integration will depend on cation, particularly nonverbal strategies when the purposes and value framework articulated interacting with children with developmental by those conducting the analysis. delays (Guralnick & Paul-Brown, 1984, To rule out this developmental explanation, 1989). These findings are consistent with typ- a matched group of younger typically devel- ically developing children taking more re- oping children is required . Studies of children sponsibility to organize play ; an area of spe- with delays that fail to include such a com- cial concern for children with developmental parison group (matched at minimum on the delays (Guralnick & Groom, 1985, 1987). basis of developmental level) leave open the alternative developmental explanation for any Developmental Framework social integration differences observed be- The analyses presented above clearly suggest tween chronological age mates with and with- that, from the perspective of typically devel- out developmental delays. For studies in oping chronological age mates, children with which matched groups of younger typically developmental delays are, for the most part, developing children have been included, not thoroughly integrated socially. The extent available evidence indicates that children with to which children with developmental delays mild developmental delays experience social are isolated from same age typically devel- separation for most measures even beyond oping children varies with the measure se- that which occurs for the matched group of lected, but the pattern is nevertheless appar- younger typically developing children (Gur- ent. It is possible, however, that the behavior alnick et al., 1998; Guralnick & Groom, 1987; of typically developing children toward chil- Guralnick & Paul-Brown, 1989) . Although a dren with developmental delays constitutes a developmental level explanation can account reasonable (from their perspective) pattern of for some of the differences discussed above social interactions, one consistent with the de- (e.g., typically developing children rate de- velopmental level of the children with delays. layed age mates as negatively as they rate de- Put simply, typically developing children may velopmentally matched younger typically de- be interacting with children with mild devel- veloping children and they use similar strate- opmental delays as they would with younger gies with both groups during conflicts), the typically developing children; i.e., social in- central patterns of social separation identified teraction and preference patterns therefore are earlier remain. Accordingly, it is something to occurring on the basis of a peer's develop- do with children's developmental status rather

Guralnick 75

than their developmental level that is respon- with mild delays have preferences for other sible for much of the social separation that children with delays or for typically devel- emerges when children with developmental oping children? Does the developmental status delays interact with typically developing chro- of the play partner affect responsiveness to the nological age mates in inclusive settings. social bids of children with mild delays? Do Of considerable importance is that analyses children with mild delays experience interper- of the interaction patterns of younger typically sonal difficulties to a different extent when in- developing children with the same develop- teracting with children with or without devel- mental level as the children with delays pres- opmental delays? This is dependent in part on ent a similar and occasionally even more se- the ability of children with delays to connect vere pattern of social separation from children with peers, some of whom may be reluctant with developmental delays as obtained for or even difficult play partners . The ability of typically developing chronological age mates children with developmental delays to make (Guralnick et al ., 1998 ; Guralnick & Groom, the connections, however, constitutes an im- 1987, 1988a ; Guralnick & Paul-Brown, 1989). portant index of social integration from their In many respects, younger typically develop- perspective . Many proponents of inclusion ing children may have more difficulty (per- may consider this form of social integration haps being less skillful) accommodating to to be of most significance and establish ex- children with delays than older typically de- pectations accordingly. veloping children, especially during conflicts Interestingly, available evidence indicates (Guralnick et al .). that when children with mild delays engage in Some contradictory findings, however, are more interactive social play such as group found when "older" (5 to 6 year old) children play, conversation, or rough and tumble play, with developmental delays are observed inter- they either show no preference or prefer to acting with younger typically developing chil- interact with typically developing age mates dren. Specifically, patterns of social separation (Guralnick et al., 1996a; Guralnick & Groom, are not as apparent as patterns of separation 1987) . Global assessments of acceptance that occur with chronological age mates (Gur- based on peer sociometric measures, however, alnick & Paul-Brown, 1986; Ispa & Matz, reveal that children with mild delays do prefer 1978). Although matching on developmental typically developing children. Similarly, for level was not used in these studies of older unilateral friendships, when children with children with delays, children with and with- mild delays reveal a preference, it is for typ- out delays were certainly more similar to one ically developing children (Guralnick et al ., another developmentally because there was a 1996; Guralnick & Groom, 1988a). Children chronological age discrepancy. Additional with mild delays also offer more joint direc- work using appropriate matching is needed to tives (let's, we) to typically developing chil- determine if kindergarten-age children with dren than to other children with mild delays mild delays exhibit different patterns of social (Guralnick & Paul-Brown, 1989) . For more integration with younger typically developing passive measures such as parallel play, how- children than do preschool-age children with ever, a preference for one group or another is mild delays. generally not evident (Guralnick et al., 1996a; Guralnick & Groom, 1987) . Similarly, only Perspective of Children with minor differences as a consequence of peers' Developmental Delays developmental status are apparent with respect The preceding discussion has focused exclu- to responsiveness to the social bids of children sively on social integration from the perspec- with delays, although there is some tendency tive of typically developing children . Similar for the responsiveness of typically developing analyses, however, can be carried out from the children to children with delays to decrease perspective of children with mild develop- over time (Guralnick et al., 1996a ; Guralnick mental delays. When playing, do children & Groom, 1987).

76 JEI, 1999, 22:1

From an interpersonal perspective, similar self-esteem of children with mild develop- to the experiences of typically developing mental delays? If so, do these adverse effects children, children with mild delays have fewer hold only for certain dimensions (e .g., friend- difficulties with typically developing children ships) of social separation? On the other hand, than with other children with mild delays. is it possible that the subjective experiences They use fewer strong directives, share infor- of social separation as defined objectively re- mation more, have fewer disagreements, and sult in only transient levels of discomfort? are less negative and more positive during Even if we had the tools, it would be dif- conflicts with typically developing children ficult to isolate the psychological impact of than with children with mild delays (Gural- social separation on children with mild devel- nick et al ., 1998; Guralnick & Paul-Brown, opmental delays because social separation oc- 1989). curs in the context of many positive and per- haps beneficial social experiences involving Psychological Meaning of Social typically developing children . We know that, Separation when given a choice, children with mild de- Taken together, evidence suggests that an ab- velopmental delays often prefer and seek out sence of complete social integration exists typically developing chronological age mates, from the perspective of typically developing and their experiences with typically develop- children, especially for measures that require ing children are more productive than those more intensive or intimate social relation- that occur with other mildly delayed children ships. Instances in which typically developing (e.g., more positive, less negative, share in- children overtly reject children with develop- formation more) (see Guralnick et al ., 1998). mental delays are found, but this is not a dom- Moreover, it does not appear that the presence inant pattern. In contrast, it appears that social of a large proportion of typically developing separation occurs primarily through a process children in inclusive settings suppresses the of exclusion. As such, typically developing social relationships or the social interactions children simply prefer other typically devel- of children with mild developmental delays. oping children and may ignore or avoid chil- Both unilateral and reciprocal friendships oc- dren with delays . For the most part, despite cur at the same rate in inclusive or segregated more frequent disagreements, special difficul- settings (Guralnick et al ., 1996a). In addition, ties occurring during conflicts, and the use of although the proportion of negative interac- more demanding forms of speech, social ex- tions is increased slightly in inclusive settings, changes between children with and without the social interaction levels of children with developmental delays are generally cordial, delays have consistently been found to be and there is little evidence suggesting an un- higher in inclusive as opposed to segregated usually negative style of interacting . In addi- settings (Buysse & Bailey, 1993 ; Guralnick et tion, typically developing children appear to al., 1996a; Guralnick & Groom, 1988b). recognize differences in their playmates' abil- ities and adjust interactions accordingly . Typ- Generality of Patterns of Social ically developing children, however, do rate Separation children with mild delays less positively and To what extent do these patterns of social sep- more negatively when judging children with aration generalize to different contexts? Every whom they like to play. study, or even series of studies, is constrained As important as these objective assessments by its own unique research strategies and lim- of social integration are, we have few insights ited set of parameters. The generality of the into the personal meaning for those children playgroup strategy is of particular concern. being socially separated along the various di- Despite offering important controls over fa- mensions identified earlier. Does objectively miliarity among children and subject selec- defined social separation correspond to sub- tion, legitimate questions can be raised about jective experiences of rejection and impair the the representativeness of the social interaction

Guralnick 77

patterns and the stability of those patterns over changes in these patterns occur over time. For time. These critical issues are discussed next. example, the proportion of social interactions With regard to whether free-play social in- of children with disabilities directed toward teractions occurring in the playgroups are rep- typically developing children did not change resentative of free-play interactions occurring over a 2 month period, despite continued in- in more typical preschool programs, available tensive efforts to promote social interaction evidence suggests that the playgroups are in- and social integration (Jenkins, Odom, & deed ecologically valid settings . In particular, Speltz, 1989) . Similarly, time did not have any social play patterns among children develop impact on the preference patterns of older in the playgroups as would be expected based children with mild delays and typically de- on the developmental literature (Guralnick & veloping children (Guralnick, 1980). In a Groom, 1987) . Moreover, children with de- more recent study, Diamond, LeFurgy, and velopmental delays exhibiting more socially Blass (1993) described an inclusive preschool competent behavior during free-play in the program that provided specific teaching about playgroups as evaluated through observational disability and diversity issues . Despite these measures are judged to be more socially com- efforts, peer sociometric measures revealed petent in community settings as evaluated by that children with disabilities were rated lower their parents (Guralnick & Hammond, in in acceptance than same-age typically devel- press) . The general absence of constraints in oping children, and that status remained free play and the use of toys and materials throughout the school year. Interestingly, ev- similar to those in other studies further mini- idence for more complete social integration mizes the possibility that the playgroups pro- has been obtained from studies in which chil- duce an unusual pattern of social interactions. dren with mild disabilities were older (es- The short-term nature of the playgroups pecially 1 year or more) than their typically (2–4 weeks), however, leaves open the possi- developing classmates and demonstrated so- bility that social integration patterns may cial play skills similar to those of typically change over time through processes related to developing children (Ispa & Matz, 1978). children becoming more familiar with one an- Other studies show that, over time, some pos- other. Moreover, most quality inclusive pre- itive social integration changes occur for chil- schools make concerted efforts to both foster dren with disabilities but not for typically de- social integration and to encourage acceptance veloping children, although it is unclear what of individual differences. Over time, children level of social integration is actually achieved may respond favorably to these efforts, there- (Dunlop, Stoneman, & Cantrell, 1980) . How- by producing social integration patterns dif- ever, even under seemingly ideal conditions ferent from those obtained in short-term play- (e.g., high levels of social skills exhibited by groups. children with disabilities and a supportive, As noted earlier, none of the published non- quality program), social separation is apparent playgroup studies separate out the social in- well into the school year (Ispa, 1981) . Relat- tegration patterns of children with mild de- edly, preference patterns for typically devel- velopmental delays. Virtually all studies in- oping children, especially negative relation- cluded children with heterogeneous types and ships, either remain stable or become increas- severity of disability . Moreover, the programs ingly selective and negative over the course included children with disabilities who often of the preschool year (Ramsey, 1995). were older (usually by 1 year) than the typi- Accordingly, in the series of studies utiliz- cally developing children . Nevertheless, as ing the playgroup methodology, patterns of discussed earlier, even for studies that primar- social integration for children with mild de- ily included children with mild disabilities, velopmental delays appear to be robust . Sim- evidence supports the overall social integra- ilar findings are obtained across a wide range tion patterns found in the playgroups . Related of program types, subject samples, times of studies further suggest that only minor the year, measures, and efforts to promote so-

78 JEI, 1999, 22:1

cial integration and social skills. This is not tup, 1983 ; Howes, 1988 ; Rubin & Lollis, to say that quality inclusive programs are not 1988). able to promote social integration over time. There also exists the possibility that in- There is every reason to believe that programs creased social integration in classroom set- with well-developed inclusive philosophies tings may have beneficial effects on social in- and practices, integrated thoughtfully within tegration in community settings . There is cer- general preschool activities, will yield more tainly room for improvement as research has positive social integration outcomes . Yet, free- shown that the community-based peer social play settings in which the nature of social ex- networks of young children with developmen- changes and choice of play partner are not tal delays are more limited than those of typ- constrained by teachers' activities or limited ically developing children (Byrne, Cunning- through play structures, yield considerable so- ham, & Sloper, 1988 ; Guralnick, 1997a ; Lew- cial separation even in these quality programs. is, Feiring, & Brooks-Gunn, 1987 ; Stoneman, Why this might be the case and which ap- Brody, Davis, & Crapps, 1988) . In addition, proaches have the potential to reduce existing creating linkages between relationships levels of social separation are discussed next. formed with peers at preschool or daycare set- tings and those in their neighborhood are es- pecially problematic for young children with PROMOTING SOCIAL developmental delays (Guralnick, 1997a). INTEGRATION Similarly, even in inclusive settings, an all too common phenomenon is that social separation Despite the fact that the psychological and re- among parents tends to occur in accordance lated implications of social separation in in- with the disability characteristics of their chil- clusive settings are not fully understood, ef- dren (Bailey & Winton, 1989; Stoneman, forts to maximize social integration remain 1993). important for ideological and developmental reasons. From an ideological perspective, by Ecological Approaches encouraging the development of meaningful What factors contribute to the patterns of so- social relationships between children with and cial integration that have been described? without disabilities during the early years, a Classroom-based state-of-the-art inclusive foundation for constructing an inclusive com- practices can certainly play an important role. munity throughout the life span is established. Quality programs, teachers well trained in in- Consistent with this expectation, research has clusive practices, and innovative curricula can indicated that parents perceive that inclusive foster social integration (see Lamorey & early childhood settings do, in fact, promote Bricker, 1993) . Clearly, classroom environ- acceptance of children with special needs, en- ments, established by state-of-the-art practic- hance sensitivity to individual differences, and es, can remove barriers to full social integra- encourage the development of peer relations tion and help set expectations about behavior and friendships among all children (Bailey & and attitudes for all children . In turn, positive Winton, 1987 ; Guralnick, 1994 ; Guralnick et changes in relevant policy and corresponding al., 1995 ; Turnbull, Winton, Blacher, & Sal- administrative or organizational environments kind, 1982) . From a developmental perspec- governing inclusive preschools can be bene- tive, increased social integration implies in- ficial (Peck, 1993), as these more distal eco- creased positive social experiences with peers. logical factors are among those that can po- Should this be realized, it is likely that chil- tentially affect the quality of inclusive pre- dren with developmental delays will reap nu- school practices . It remains to be seen, how- merous developmental benefits in relation to ever, whether these approaches will be cognitive, communicative, and general pro- sufficient to dramatically alter the patterns of social skills as well as foster an emerging social separation described (see also File, sense of self (Bates, 1975 ; Garvey, 1986 ; Har- 1994).

Guralnick 79

It is perhaps the larger ecology of home, of children with disabilities continue to ex- community, societal beliefs, attitudes, and ac- press concerns related to possible rejection of tions toward individuals with disabilities that their child (Bailey & Winton, 1987 ; Gural- can exert a more substantial influence on chil- nick, 1994; Guralnick et al., 1995). Of partic- dren's social integration patterns (see Stone- ular concern is that their child's own behavior man, 1993). Models analyzing how multiple may contribute to this rejection in the pre- ecological factors affect inclusive processes school setting (Guralnick et al ., 1995). In the have been presented previously (Guralnick, following section, this and related issues are 1982; Peck, 1993) . For typically developing considered in the context of a more child-ori- children in particular, the transmission of at- ented focus to promote social integration. titudes from children's families and commu- nities with respect to an appreciation of di- Child-Oriented Approaches versity, for example, is a complex process. A primary reason why altering friendship pat- American society has made extraordinary ad- terns or the extent to which diverse groups of vances in the past generation in fostering the children engage in more extended play epi- inclusion of people with disabilities . However, sodes is likely to be most difficult is due to inclusion for people with disabilities is certain the fact that children's selection of play part- to be affected by larger questions of how so- ners is generally based on common interests, ciety is able to address the nature of diversity abilities, experiences, and interaction styles in general ; an issue that remains contentious. (e.g., Rubin, Lynch, Coplan, Rose-Krasnor, & The concern, of course, is that far too many Booth, 1994) . Unfortunately, it has been well typically developing children enter preschool documented that even children with mild de- settings with limited experiences with children velopmental delays manifest peer-related so- with disabilities and a set of attitudes and ex- cial interaction difficulties that extend beyond pectations about children differing from them- those expected on the basis of their develop- selves that are inconsistent with fostering so- mental level (e .g., Guralnick & Groom, 1987; cial integration. The extent to which these ini- Guralnick & Weinhouse, 1984) . These social tial expectations are altered by state-of-the-art competence problems are apparent even in in- inclusive practices and by contemporary ex- clusive settings. Accordingly, it is possible periences with children with disabilities in that these unusual peer interaction difficulties preschool settings and in the community to serve to limit the degree to which social in- establish a new level of social integration re- tegration occurs with typically developing mains a critical question for the future. chronological age mates. Despite the association likely to exist be- The reasons for the peer-related social com- tween these ecological factors and social in- petence problems characteristic of children tegration outcomes, only limited experimen- with mild developmental delays are not fully tally-based information is available on this known. Their more limited social experiences matter. Moreover, how these factors might af- with peers, especially the relative absence of fect the various dimensions of social integra- more in-depth relationships with typically de- tion identified in this paper has not been ad- veloping children in home and community dressed. It is quite possible that instances of settings both prior to and during the preschool overt rejection and even certain aspects of ex- years, may be partially responsible (Gural- clusionary patterns directed toward children nick, 1997a; Stoneman, 1993). In addition, with developmental delays can be altered greater difficulties in regulating emotions dur- through changes in the larger ecology . Wheth- ing play with peers (Guralnick et al., 1998; er friendship patterns between children with Guralnick & Paul-Brown, 1989), or with re- and without delays can be altered by those gard to specific social-cognitive processes same factors is clearly a far more difficult is- (Guralnick, 1992) are possible explanations. sue. Indeed, despite recognizing many poten- Children with developmental delays struggle tial advantages of inclusive settings, parents with the important social tasks of peer group

80 JEI, 1999, 22 :1

entry, resolving conflicts, and maintaining interrelationships as they might influence the play. More limited language abilities is anoth- three social integration constructs identified in er alternative (even when controlling for de- this paper as assessed in free-play situations. velopmental level), although language level at As illustrated in the model, the ultimate de- this age within subgroups of mildly delayed gree to which children with developmental de- children is not a strong correlate of peer-re- lays, or perhaps children with disabilities in lated social competence (e .g., Guralnick et al., general, experience social integration in pre- 1996a). school settings is dependent upon four main Consequently, child-oriented intervention factors. First are expectations with respect to programs that can successfully foster the peer- establishing potential social relationships with related social competence of children with children with disabilities by typically devel- mild developmental delays may serve to re- oping children that occur prior to participation duce the degree of social separation observed in inclusive settings . These expectations are in inclusive settings. Unfortunately, it has formed through yet little understood processes been remarkably difficult to achieve sustained which transmit attitudes by parents and other and generalizable changes in the peer-related family members toward people with disabili- social competence of children with disabilities ties, through a more general family atmo- (e.g., Odom, McConnell, & McEvoy, 1992). sphere regarding acceptance and even celebra- It is likely that a comprehensive child-focused tion of diversity, and specific experiences with approach involving school, family, and com- children (or people in general) with disabili- munity components will be needed to achieve ties (see Prior Expectations in Figure 1). meaningful outcomes (Guralnick & Neville, These specific experiences with individuals 1997). with disabilities are often arranged, or fail to be arranged as the case may be, by parents. A GENERAL MODEL OF As a consequence, there is likely to be a high INFLUENCES ON SOCIAL correlation among the three components of INTEGRATION "prior expectations." Not indicated in the model in Figure 1 are the overarching societal All children bring to a social setting, such as attitudes and beliefs and associated practices an inclusive preschool, a set of social inter- that can affect the three components of prior action skills, specific interests in certain social expectations. Clearly, as these larger ecologi- and nonsocial activities, expectations that oth- cal factors serve to encourage more positive ers share their rule systems, a certain level to attitudes and increase the likelihood of contact which they are willing to accept or overlook with children with disabilities, prior expecta- unexpected or even inappropriate behavior in tions will be influenced. their peers, and a respect for varying degrees Second, these prior expectations are subject of individual differences. In turn, these pat- to modification through more contemporary terns reflect prior or ongoing experiences with experiences between children with and with- peers and family members, especially parental out disabilities in the home or community. Ex- beliefs, attitudes, and behaviors and, of periences occurring in the context of recrea- course, their own unique social abilities . In tional, religious, or family organized social inclusive preschool settings, from the perspec- events that involve children with disabilities tive of typically developing children, the de- have the potential to modify these prior ex- gree to which children with mild developmen- pectations. Most often, however, these prior tal delays (at any level of peer-related social expectations continue to exert their influence competence) are socially integrated will de- over time . They not only affect contemporary pend upon these factors as well as specific ex- experiences in home and community, but may periences in the environment of the inclusive well affect whether parents select an inclusive setting itself. preschool setting for their child in the first Figure 1 attempts to depict these complex place. Similarly, these prior experiences of

Guralnick 81

Figure 1. A model depicting factors that influence social integration during free-play situations in inclusive preschool settings.

typically developing children may influence tors that can alter social integration outcomes the actual social experiences that occur with have been identified previously (Guralnick, children with disabilities in an inclusive pre- 1981b), and can serve as a framework for school setting, including those that occur dur- evaluating social integration (Buysse & Bai- ing free-play situations . These influences on ley, 1993). typically developing children by prior expec- The model further suggests (see double- tations and by contemporary experiences are headed arrow) the possibility of reciprocal in- likely to be felt most strongly during the early fluences between contemporary experiences in phases of the inclusive preschool experience, home and community settings and those oc- but also may be subject to modification over curring in the inclusive preschool setting. time. Friendships do indeed appear to be formed be- Accordingly, the third main factor influenc- tween children with disabilities and typically ing social integration consists of children's ex- developing chronological-age mates in inclu- periences in the inclusive preschool setting. sive preschools which extend to the commu- The experiences of typically developing chil- nity (Guralnick et al ., 1995). However, those dren in particular with children with disabili- linkages and perhaps the strength of the rela- ties may serve to modify negative prior ex- tionship as well are not as well developed for pectations and will likely succeed in trans- children with developmental delays (and like- forming children with more benign sets of ex- ly children with other disabilities) in compar- pectations, but outcomes will depend on the ison to typically developing children (Gural- quality of inclusive practices. Except for the nick, 1997a) . Moreover, parents report that most extreme prior expectations by typically only a small percentage of their typically de- developing children, it is here that teacher at- veloping children have friends with a disabil- titudes and behaviors, innovative curricula, ity (Guralnick, 1997a) . An interesting and po- and other practices can exert some degree of tentially important path of influence, one not influence on social integration evaluated in the depicted in the model, is whether contempo- free-play setting . Important programmatic fac- rary experiences in the inclusive preschool,

82 JEI, 1999, 22 :1

home, and community alter parental attitudes centrated efforts on the three current factors toward diversity in general and to children related to home and community experiences with disabilities in particular. with people with disabilities, the inclusive The final factor that can influence social in- preschool setting, and interventions directed tegration focuses exclusively on children with toward children with disabilities to improve disabilities. As noted earlier, it has been dif- their peer-related social competence . It is this ficult to develop successful intervention pro- activist agenda, carried out within a develop- grams to enhance children's peer-related so- mental ecological model and analyzed across cial competence, and insufficient resources are the three social integration constructs identi- devoted to social development issues in pre- fied in this paper that may provide a useful school programs (Michnowicz, McConnell, framework for future research on the nature Peterson, & Odom, 1995) . New approaches and meaning of social integration for children that consider the larger developmental-ecolog- with disabilities. ical contexts are emerging, however, and are attempting to address school, family, and REFERENCES community aspects of children's peer-related Asher, S. R., Singleton, L. C., Tinsley, B . R., & social development and social integration Hymel, S . (1979) . A reliable sociometric mea- (Guralnick & Neville, 1997) . As reflected in sure for preschool children . Developmental the figure, these interventions may enhance Psychology, 15, 443-444. the general social experiences and social com- Bailey, D. B ., Jr., & Winton, P. J. (1987) . Stability petence of a child with a disability through and change in parents' expectations about mainstreaming . Topics in Early Childhood Spe- highly specialized child-oriented programs de- cial Education, 7, 73-88. signed for both the early childhood inclusive Bailey, D. B., Jr., & Winton, P. J. (1989) . Friendship setting and the child's family and community. and acquaintance among families in a main- Should these interventions prove to be effec- streamed day care center. Education and Train- tive, there is every reason to expect to realize ing in Mental Retardation, 24, 107-113. substantial improvements in social integration Bakeman, R., & Brownlee, J . R. (1980) . The stra- in preschool settings as evaluated in free-play tegic use of parallel play : A sequential analysis. Child Development, 51, 873-878. situations. Taken together, these four factors combine Bates, E . (1975) . Peer relations and the acquisition of language. In M. Lewis & L. A . Rosenblum to influence the three constructs and related (Eds.), The origins of behavior: Vol. 4. Friend- dimensions of social integration that have ship and peer relations (pp . 259-292) . New been described in this paper. As discussed, a York: John Wiley & Sons. coherent and robust body of evidence is avail- Blackmon, A. A., & Dembo, M . H. (1984) . Pro- able that adequately characterizes and pro- social behaviors in a mainstreamed preschool. vides a realistic portrayal of the social inte- Child Study Journal, 14, 205-214. gration patterns of children with mild devel- Buysse, V., & Bailey, D . B., Jr. (1993) . Behavioral opmental delays with typically developing and developmental outcomes in young children with disabilities in integrated and segregated chronological age mates in inclusive settings. settings : A review of comparative studies . The The extent to which these social integration Journal of Special Education, 26, 434-461. patterns can be altered by any of the four fac- Byrne, E . A., Cunningham, C. C., & Sloper, P. tors in the model remains a vital question for (1988) . Families and their children with the future. To be sure, there exist overarching Down's syndrome : One feature in common. ecological forces related to societal values, London : Routledge. economic conditions, legal and legislative Cavallaro, S . A., & Porter, R . H. (1980) . Peer pref- erences of at-risk and normally developing changes, or organizational issues that can af- children in preschool mainstream classrooms. fect each of these factors, particularly prior American Journal of Mental Deficiency, 84, expectations . However, whatever the nature of 357-366. these prevailing forces may be, social integra- Diamond, K., LeFurgy, W., & Blass, S . (1993) . At- tion may benefit more from direct and con- titudes of preschool children toward their peers

Guralnick 83

with disabilities : A year-long investigation in Guralnick, M. J. (1997b) . Second generation re- integrated classrooms . The Journal of Genetic search in the field of early intervention. In M. Psychology, 154, 215-221. J. Guralnick (Ed .), The effectiveness of early Dunlop, K. H ., Stoneman, Z ., & Cantrell, M . L. intervention (pp. 3-22) . Baltimore : Brookes. (1980). Social interaction of exceptional and Guralnick, M. J. (1998). The effectiveness of early other children in a mainstreamed preschool intervention for vulnerable children : A devel- classroom. Exceptional Children, 47, 132-141. opmental perspective . American Journal on File, N. (1994). Children's play, teacher-child inter- Mental Retardation, 102, 319-345. actions, and teacher beliefs in integrated early Guralnick, M. J ., Connor, R ., & Hammond, M. childhood programs . Early Childhood Re- (1995). Parent perspectives of peer relations search Quarterly, 9, 223-240. and friendships in integrated and specialized Garvey, C. (1986). Peer relations and the growth of programs . American Journal on Mental Retar- communication . In E. C. Mueller & C . R. Coo- dation, 99, 457-476. per (Eds .), Process and outcome in peer rela- Guralnick, M. J ., Connor, R., Hammond, M., Gott- tionships (pp . 329-345) . Orlando, Florida : Ac- man, J . M., & Kinnish, K . (1996a). Immediate ademic Press. effects of mainstreamed settings on the social Guralnick, M. J. (Ed.). (1978) . Early intervention interactions and social integration of preschool and the integration of handicapped and non- children . American Journal on Mental Retar- handicapped children. Baltimore : University dation, 100, 359-377. Park Press. Guralnick, M. J ., Connor, R., Hammond, M., Gott- Guralnick, M. J. (1980) . Social interactions among man, J. M., & Kinnish, K . (1996b) . The peer preschool children . Exceptional Children, 46, relations of preschool children with communi- 248-253. cation disorders . Child Development, 67, 471- 489. Guralnick, M. J. (1981a). The efficacy of integrat- ing handicapped children in early education Guralnick, M. J., Gottman, J. M., & Hammond, M. settings: Research implications. Topics in Early A. (1996). Effects of social setting on the Childhood Special Education, 1(I), 57-71. friendship formation of young children differ- ing in developmental status . Journal of Guralnick, M. J. (1981b) . Programmatic factors af- Applied Developmental Psychology, 17, fecting child-child social interactions in main- 625-651. streamed preschool programs . Exceptional Ed- Guralnick, M. J., & Groom, J . M. (1985) . Corre- ucation Quarterly, 1(4), 71-91. lates of peer-related social competence of de- Guralnick, M. J. (1982) . Mainstreaming young velopmentally delayed preschool children. handicapped children : A public policy and eco- American Journal of Mental Deficiency, 90, logical systems analysis. In B. Spodek (Ed .), 140-150. Handbook of research on early childhood ed- Guralnick, M. J ., & Groom, J . M. (1987). The peer ucation (pp. 456-500) . New York : The Free relations of mildly delayed and nonhandi- Press/MacMillan. capped preschool children in mainstreamed Guralnick, M. J. (1990) . Major accomplishments playgroups. Child Development, 58, 1556- and future directions in early childhood main- 1572. streaming . Topics in Early Childhood Special Guralnick, M. J., & Groom, J . M. (1988a). Friend- Education, 10(2), 1-17. ships of preschool children in mainstreamed Guralnick, M. J. (1992) . A hierarchical model for playgroups . Developmental Psychology, 24, understanding children's peer-related social 595-604. competence. In S. L. Odom, S . R. McConnell, Guralnick, M. J., & Groom, J . M . (1988b) . Peer & M . A. McEvoy (Eds.), Social competence of interactions in mainstreamed and specialized young children with disabilities : Issues and classrooms : A comparative analysis . Exception- strategies for intervention (pp. 37-64) . Balti- al Children, 54, 415-425. more: Brookes. Guralnick, M. J ., & Hammond, M . A . (in press). Guralnick, M. J. (1994) . Mothers' perceptions of Sequential analysis of the social play of young the benefits and drawbacks of early childhood children with mild developmental delays . Jour- mainstreaming . Journal of Early Intervention, nal of Early Intervention. 18, 168-183. Guralnick, M. J., & Neville, B . (1997). Designing Guralnick, M. J. (1997a). The peer social networks early intervention programs to promote chil- of young boys with developmental delays. dren's social competence . In M. J. Guralnick American Journal on Mental Retardation, 101, (Ed.), The effectiveness of early intervention 595-612 . (pp. 579-610) . Baltimore : Brookes.

84 JEI, 1999, 22:1

Guralnick, M. J., & Paul-Brown, D . (1984). Com- Michnowicz, L . L., McConnell, S. R., Peterson, C. municative adjustments during behavior-re- A., & Odom, S . L. (1995). Social goals and quest episodes among children at different de- objectives of preschool IEPs : A content anal- velopmental levels . Child Development, 55, ysis. Journal of Early Intervention, 19, 273- 911-919. 282. Guralnick, M. J., & Paul-Brown, D . (1986). Com- Nabors, L. (1997) . Playmate preferences of children municative interactions of mildly delayed and who are typically developing for their class- normally developing preschool children : Ef- mates with special needs . Mental Retardation, fects of listener's developmental level . Journal 35, 107-113. of Speech and Hearing Research, 29, 2-10. Newcomb, A. E, & Bagwell, C. L. (1995) . Chil- Guralnick, M. J., & Paul-Brown, D . (1989). Peer- dren's friendship relations : A meta-analytic re- related communicative competence of pre- view. Psychological Bulletin, 117, 306-347. school children: Developmental and adaptive Odom, S. L., McConnell, S . R., & McEvoy . (Eds.). characteristics . Journal of Speech and Hearing (1992). Social competence of young children Research, 32, 930-943. with disabilities : Issues and strategies for in- Guralnick, M. J., Paul-Brown, D ., Groom, J. M ., tervention . Baltimore : Brookes. Booth, C . L., Hammond, M . A., Tupper, D . B ., & Gelenter, A . (1998). Conflict resolution pat- Peck, C. A. (1993) . Ecological perspectives on the implementation of integrated early childhood terns of preschool children with and without programs. In C. A. Peck, S. L. Odom, & D. D. developmental delays in heterogeneous play- .), groups. Early Education and Development, 9, Bricker (Eds Integrating young children with 49-77. disabilities into community programs (pp. 3- 15). Baltimore : Brookes. Hartup, W. W. (1983) . Peer relations . In E. M. Heth- erington (Ed .), P. H. Mussen (Series Ed .), Peterson, N . L., & Haralick, J. G. (1977). Integra- Handbook of child psychology : Vol. 4. Social- tion of handicapped and nonhandicapped pre- ization, personality, and social development schoolers : An analysis of play behavior and so- (pp. 103-196). New York : Wiley. cial interaction . Education and Training of the Mentally Retarded, 12, 235-245. Howes, C. (1988). Peer interaction of young chil- dren. Monographs of the Society for Research Porter, R . H ., Ramsey, B ., Tremblay, A., Iaccobo, in Child Development, 53(1, Serial No . 217). M., & Crawley, S . (1978). Social interactions in heterogeneous groups of retarded and nor- Ispa, J. (1981). Social interactions among teachers, mally developing children : An observational handicapped children, and nonhandicapped study. In G. P. Sackett (Ed .), Observing behav- children in a preschool . Journal of Applied De- ior: Vol. 1. Theory and applications in mental velopmental Psychology, 1, 231-250. retardation (pp. 311-328). Baltimore : Univer- Ispa, J ., & Matz, R . D. (1978). Integrating handi- sity Park Press. capped preschool children within a cognitively Ramsey, P. G. (1995) . Changing social dynamics in oriented program . In M. J . Guralnick (Ed.), early childhood classrooms . Child Develop- Early intervention and the integration of hand- ment, 66, 764-773. icapped and nonhandicapped children (pp. 167-190) . Baltimore : University Park Press. Rubin, K. H., & Lollis, S . P. (1988) . Origins and consequences of social withdrawal . In J . Belsky Jenkins, J . R., Odom, S . L., & Speltz, M . L. (1989). Effects of social integration on preschool chil- & T. Nezworski (Eds .), Clinical implications of attachment (pp. 219-252) . Hillsdale, N .J.: dren with handicaps. Exceptional Children, 55, 420-428. Lawrence Erlbaum. Lamorey, S., & Bricker, D . D. (1993) . Integrated Rubin, K. H., Lynch, D., Coplan, R ., Rose-Krasnor, programs : Effects on young children and their L., & Booth C . L. (1994). "Birds of a feath- parents. In C . A. Peck, S . L. Odom, & D . D. er. . .": Behavioral concordances and preferen- Bricker (Eds.), Integrating young children with tial personal attractions in children. Child De- velopment, disabilities into community programs (pp . 249- 65, 1778-1785. 270). Baltimore: Brookes. Shatz, M., & Gelman, R . (1973) . The development Lewis, M., Feiring, C ., & Brooks-Gunn, J . (1987). of communication skills : Modifications in the The social networks of children with and with- speech of young children as a function of lis- out handicaps : A developmental perspective . In tener. Monographs of the Society for Research S . Landesman & P. Vietze (Eds.), Living envi- in Child Development, 38(5, Serial No . 152). ronments and mental retardation (pp. 377- Stoneman, Z . (1993) . The effects of attitude on pre- 400). Washington, DC : American Association school integration . In C. A. Peck, S . L. Odom, on Mental Retardation. & D. D. Bricker (Eds .), Integrating young chil-

Guralnick 85

dren with disabilities into community programs Turnbull, A . P., Winton, P J ., Blacher, J ., & Salkind, (pp. 223-248). Baltimore : Brookes. N . (1982) . Mainstreaming in the kindergarten Stoneman, Z., Brody, G . H., Davis, C. H., & classroom: Perspectives of parents of handi- Crapps, J . M. (1988). Childcare responsibili- capped and nonhandicapped children . Journal ties, peer relations, and sibling conflict : Older of the Division of Early Childhood, 6, 14-20. siblings of mentally retarded children . Ameri- White, B . N. (1980) . Mainstreaming in grade school can Journal on Mental Retardation, 93, 174- and preschool: How the child with special 183. needs interacts with peers. In T. M. Field, S. Strain, P. S . (1984) . Social behavior patterns of non- Goldberg, D . Stern, & A . M . Sostek (Eds .), handicapped and nonhandicapped-developmen- High-risk infants and children: Adult and peer tally disabled friend pairs in mainstream pre- interactions (pp . 347-371) . New York : Aca- schools . Analysis and Intervention in Devel- demic Press. opmental Disabilities, 4, 15-28. Strayer, F. F. (1980) . Social ecology of the preschool peer group. In W. A. Collins (Ed.), The Min- Address correspondence to Michael J. Guralnick, nesota symposia on child psychology : Vol . 13. Ph.D., Center on Human Development and Dis- Development of cognition, affect, and social re- ability, Box 357920, University of Washington, Se- lations (pp . 165-196) . Hillsdale, NJ : Lawrence attle, WA 98195-7920, Erlbaum . Email: mjgural@u .washington.edu.

86 JEI, 1999, 22:1