Printer Friendly

Learning disabilities and social competence: a social ecological perspective.

* The justification for mainstreaming children with mild disabilities, particularly those with learning disabilities (LD), into regular classrooms has always been based less on possible academic gains for children with disabilities and more on the potential social benefits for all children that would result from such integration (Dunn, 1968). It was expected that creating a single social group would enhance the social competence of children with mild disabilities by providing them more sophisticated social models while providing other children opportunities to interact with youngsters with disabilities, thus reducing the mystique and stigma associated with disability (Coleman, 1985). Unfortunately, empirical evidence to support mainstreaming as a method of enhancing the social competence of children with mild disabilities is scarce. Instead, research has suggested that the social interaction skills and social acceptance of children with LD remain deficient in comparison to other children (Fox, 1989). This seems true regardless of whether the judgment of the child's social competence is based on teachers' perceptions (Bursuck, 1989; McKinney, McClure, & Feagan, 1982), parents' perceptions (Gresham & Reschly, 1986; Sater & French, 1989), peer perceptions (Bryan, 1974; Garrett & Crump, 1980; Kistner & Gatlin, 1989; Vaughn, Hogan, Kouzekanani, & Shapiro, 1990), or the actual behavior of children observed in social interaction (Bryan, 1974; Bryan & Bryan, 1978). In fact, the data have so consistently linked social skills deficits and peer rejection to mild disabilities that it has been suggested that such difficulties be considered criteria for defining (LD) (Interagency Committee on Learning Disabilities, 1987).

Despite the evidence, we believe that linking social competence to a definition of LD is premature for several reasons. First, it overlooks the fact that sociometric studies comparing the social status of children with LD to their peers suggest that many of these children are accepted by their peers. Dudley-Marling and Edmiason (1985), in their review of research on the social status of children with LD, concluded that most such children enjoy relatively neutral social status. Perlmutter, Crocker, Corday, and Garstecki (1983) reported that a substantial number of children with disabilities were judged popular by their peers without disabilities. Finally, Sater and French (1989) provided some evidence that differences in social competence between accepted and rejected children with LD may be comparable to those found between liked and disliked children without disabilities. It is clear that social deficits and peer rejection are not common denominators for LD and that many children experience academic difficulties independent of social acceptance by their peers.

A second objection to including social competence within a definition of LD is the lack of evidence to suggest how they might be linked causally. Children with LD in the educational mainstream represent only a small proportion of the children who are actively rejected by their peers. Most rejected children have not been assigned disability labels or received special services. However, they do share with children labeled as having LD the characteristic of low achievement, which is predictive of lower social status (Hartup, 1983). It has been argued (Bruck, 1986) that low social status, not a disability, is linked to school failure.

Only a few studies have attempted to compare the social competence of students with and without disabilities, having first matched the groups on achievement. Bursuck (1989) contrasted low-achieving students to those with LD on three dimensions of social competence. Though finding differences between the two groups in terms of peer acceptance, Bursuck found the groups comparable on both teacher and self-ratings of various facets of social competence. Sater and French (1989) also compared small groups of low-achieving children and those with LD and reported no between-group differences with regard to sociometric status or the incidence of peer rejection. They argued that there was little evidence to conclude that children with LD experience unique social behavioral deficits that differentiate them from other children rejected by their peers. In fact, Coleman, McHam, and Minnett (1992) provide evidence that children with LD may even be more skilled than achievement-matched peers in some areas of social competence.

The present study further explored the relationship between social competence and disabilities, while addressing a major methodological weakness evident in previous research. The interpretation of past research has been clouded by the fact that socially rejected children are overrepresented in the population of children with LD, as compared to their representation in populations without disabilities. As such, when differences in the social competence of the two groups were reported, it was difficult to decide whether the effects resulted from characteristics unique to children with LD or from overrepresentation of rejected children in these samples.

To separate these issues, subjects in the present investigation included children both with and without LD; these children were matched on social status, grade, sex, and racial/ethnic variables. The proportion of children in each of three social-status categories studied (popular, rejected, and neglected) reflected the makeup of the sample with LD. This design allowed us to observe the effects of LD on social competence, independent of the child's social status, as well as to view the competencies of children from different social statuses without regard to disability. Finally, the interaction between these factors allowed us to determine if the relationship between social competence and social status was different for the two groups of children.

To study this issue in as broad a scope as possible, we took indicators of children's social competence from a wide range of sources, including the perceptions of the subjects, their peers, and their teachers. In addition, we assessed other facets of social competence through direct observation of the subjects involved in social interactions with their peers and teachers in a school setting. Finally, we collected academic grades from school records to study the possible relationships between academic and social competence. The diversity of measures reflected an attempt to assess as many facets of the child's social ecology as possible and to avoid allowing the interpretation of results to depend too heavily on a single measure and method of assessment, a common criticism of much of the previous research in this area (Coleman, Pullis, & Minnett, 1987).



Subjects for this study were taken from a larger, longitudinal study of social development in learning disabled (LD) and nondisabled (ND) children conducted in collaboration with the Dallas, Texas, Independent School District, one of the 10 largest school districts in the United States. Participants were selected from 60 classrooms in eight elementary schools. Two schools each contained black, white, or Hispanic majorities; and two schools were balanced with regard to racial/ethnic composition. All testing was conducted under the blanket approval given by parents to the school district's group-testing program with the restrictions that (a) all tests were to be group administered, (b) subjects were not required to reveal information about their families, and (c) that all subject's responses would be identified only by their six-digit student code. In addition, teachers and school administrators were not present during classroom sessions in which questionnaires were completed, and only aggregated data were available to school district personnel.

Seventy-three LD children were selected from a larger sample of over 300 LD students in regular elementary school classrooms. The only restriction was that the classroom contained more than one LD child. These children were evenly distributed across Grades 3 through 6; were 78% male; and contained 41 % whites, 3 8% blacks, and 20% Hispanics. All LD subjects had been certified as "learning disabled" by the school district, based on a discrepancy between potential and performance, and were receiving 1 or 2 hr of daily instruction in resource classrooms. The social status of each LD child was determined from data available through peer-nomination studies conducted in their regular classrooms (see "Procedures" section). A comparison sample of ND subjects was then selected from a pool of over 1,100 children using a stratified, random-sampling technique within the classroom to match these children to LD students regarding social status, sex, race, grade, and ethnicity. In a few cases, within-classroom matches were impossible; thus, a child from the same school and grade but a different classroom was selected. The final sample contained equal numbers of LD and ND students; 112 boys and 34 girls; and 46 popular, 70 rejected, and 30 neglected children.


Social Status. Social status for each child was determined from regular-class peer nominations, in which students in each regular classroom were given rosters containing the names of all boys and girls in that class. Boys were then asked to circle the names of three boys they like to play with; girls were asked to do the same with regard to girls. They were then instructed to place an X beside the names of three same-sex children they did not like to play with. Though some researchers have suggested this procedure might sensitize children to disliked classmates, recent studies have shown that the effect is minimal (Bell-Dolan, Foster, & Sikora; 1989; Hayvren & Hymel, 1984). Social-status categories for all subjects were then derived using procedures described initially by Coie, Dodge, and Coppotelli (1982) and redefined by Coie and Dodge (1983) as follows: "Like" and "dislike" nominations were standardized within classroom and sex by converting them to Z scores. Social preference (standardized liked -- standardized dislike) and social impact (standardized liked + standardized disliked) scores were then computed and restandardized within sex and classroom. Social status for each child was then determined using the following criteria:

* Popular--like > 0, dislike < 0, and social

preference > 1.

* Rejected--like < 0, dislike > 0, and social

preference < -1.

* Neglected--like < 0, dislike < 0, and social

impact < - 1.

* Controversial--like > 0, dislike > 0, and social

impact > 1.

* Average-subjects whose scores did not meet

criteria for popular, rejected, neglected, or

controversial group inclusion.

Controversial children, those who are both highly liked and highly disliked, were excluded from the study because this group is always very small in the sociometric literature, is less stable than other categories (Beck & Collins, 1985; Bukowski & Newcomb, 1984), and has seldom been studied. Children of average social status were also excluded because our research questions targeted children whose social standing among peers possessed either a positive or negative valence, which, by definition, precluded average-status subjects. The selected sample of LD and matched ND peers was 47% rejected by their regular-class peers, whereas 30% fell into the popular category and 23% were neglected.

Self-concept. All subjects completed two self-concept questionnaires. The Harter Perceived Competence Scale (Harter, 1982, 1985) contains 28 stimulus items that children endorse on a 4-point scale. The measure has been factor analyzed into four domains assessing self-competence in the following areas: cognitive, physical, and social self-concept, as well as general self-esteem. Each stimulus item is dichotomous, in that the child is provided two alternative descriptions of types of children. For example, "Some kids forget what they learn BUT other kids can remember things easily." Once the child has determined which type of child describes him or her, then the child must further decide if it is "Really true for me" or "Sort of true for me." Reports of subscale reliability in several samples have ranged from .75 to .83 (Cognitive), .75 to .84 (Social), .77 to .86 (Physical), and .73 to .82 (General) (Harter, 1982; Wylie, 1989).

The second measure of self-concept was the Self-Description Questionnaire (SDQ) (Marsh & Parker, 1984; Marsh, Relich, & Smith, 1983), a multidimensional instrument designed to measure seven facets of self-concept hypothesized by Shavelson and Bolus (1982). The measure contains 66 stimulus items divided into the following areas: (a) physical abilities, (b) physical appearance, (c) relations with peers, (d) relations with family, (e) reading abilities, (f) mathematics abilities, and (g) all school subjects abilities. Children completed the instrument by providing ratings on a 6-point Likert scale that ranged from completely false to completely true. Marsh and Parker (1984) reported internal consistency reliabilities for the seven scales in the .80s and .90s, which have been replicated by other investigators (Wylie, 1989). The two self-concept measures represent the best instruments currently available in terms of their psychometric properties and their grounding in theoretical models of children's self-perception. Numerous studies have confirmed the integrity and stability of the structure of these instruments through factor analysis, and independent evidence shows that children's performance on the measures is predictable from theoretical models. Wylie saw these criteria as the most appropriate for determining the utility of self-concept instruments.

Social Relations. Children's social relations were judged using several methods. First, each subject completed the Loneliness Questionnaire (Asher, Hymel, & Renshaw, 1984). The instrument consists of 24 items, 16 of which focus on children's feelings of loneliness, social adequacy or inadequacy, and estimation of peer status. Children respond on a 5-point scale indicating how true each statement is about them, and a single total loneliness score is yielded which can take values ranging from 16 to 80. An internal consistency reliability of .90 was obtained, based on Cronbach's Alpha.

Second, all children completed a 5-point sociometric rating scale on which they indicated the extent they liked to play with each of their classmates. Descriptors included: "I like to play with this person a lot, I kind of like to play with this person, I neither like nor dislike playing with this person, I kind of don't like to play with this person, and I don't like to play with this person at all." This procedure differed from the peer nomination technique in that each child rated all other children in the classroom, including both same-sex and opposite-sex peers.

Finally, the Affiliation Network Questionnaire, developed by the authors for the project and designed to assess peer relations outside of school, was administered. Issues addressed included how many friends the child had in the neighborhood, their ages, how often they played together, and the extent to which they argued or fought with their peers. One additional question, asking how many schools the child had attended, was included to judge the child's longevity in the neighborhood.

Teacher Ratings. Teachers were asked to complete a 23-item questionnaire for all students in their class. The instrument, a shortened version of Thomas and Chess' (1977) Teacher Temperament Questionnaire, yields three supraordinate categories: (a) task orientation, (b) adaptability, and (c) reactivity (Cadwell & Pullis, 1983). Teacher responses are on a 6-point scale ranging from "hardly ever" to "almost always," which indicate how often certain kinds of behavior occur within the classroom. Keogh, Pullis, and Cadwell (1982) reported 5-week temporal stability reliabilities for the instrument across several samples that averaged .81. They have also reported the following internal consistency reliabilities based on coefficient alpha: task orientation, .94; adaptability, .88; and reactivity, .69. These temperament characteristics have been found to be significant factors that influence teachers' perceptions and classroom decisions for both normal children (Pullis & Cadwell, 1982) and those with mild disabilities (Pullis, 1985). Teachers were also asked to use a 6-point Likert scale, ranging from significantly below average to significantly above average, to rate students' social skills, current academic performance, classroom motivation, and general intelligence.

Academic Data. Children's fall semester grades were obtained in four areas: reading, math, social studies, and science. Social studies and science grades were always assigned by regular-class teachers. Math grades for LD subjects were assigned by both regular-class and special education teachers; and reading grades for LD subjects were assigned solely by special education teachers.

Behavioral Observations. Positive and negative social behaviors given and received by each child in interaction with peers and teachers were assessed by direct observation. Frequency counts of each class of behavior were segmented into 1-min intervals during 10-min observations in each of two settings (regular classroom and physical education class). Positive social behavior included verbal praise, affiliative touch, laughing, smiling at another, and helping. Negative behavior included verbal and physical abuse, screaming, taunting, teasing, gestures, rejecting another, and disrupting others' activities.

Five observers were trained on the observation system using videotapes of classroom situations and free-play episodes. At three time intervals during the actual observations, 20% of the sample was observed concurrently by various pairs of the five observers to determine reliability. The reliability of the data was judged in terms of occurrence or nonoccurrence of the various classes of target behavior in each of the twenty 1-min frames that constituted the observation. Reliability was computed as the number of frames in which observers agreed a target behavior occurred, expressed as a proportion of the total number of observation frames (agreements and disagreements). Frames in which observers agreed on the nonoccurrence of behavior were not included. Reliability varied by behavioral category as follows: positive behavior to peers, 76%; positive behavior from peers, 76%; negative behavior to peers, 78%; negative behavior from peers, 88%; positive behavior to teacher, 80%; positive behavior from teacher, 100%; negative behavior to teacher, 100%; and negative behavior from teacher, 88%.


Self-report, peer nominations, and peer ratings were collected during the spring semester of the school year to ensure that all children were familiar with members of the class. Teachers were not present during the sessions, and all children in every classroom participated. Each questionnaire item was read aloud to the entire class by members of the research staff as the children completed the measures individually. Teachers completed their ratings for each member of the class during the spring semester and were told only that the research project was studying the social development of LD children who are mainstreamed into regular classrooms. Semester grades were taken from the fall semester before the other assessments.

Behavioral observations were made during the last 2 months of the school year, following the collection of descriptive information. Subjects were observed for 10 min in their regular classroom and 10 min in physical education classes (PE) by five doctoral students who did not know either the child's academic placement or social status. Teachers were unaware of which children were the targets for observation, and students were told that the observer was simply a visitor who was studying to be a teacher.

Collecting such an extensive set of information on a large sample necessarily required a fairly long period of time. Inevitably, data were lost for some subjects on some variables as a result of the student's moving or being absent on the day a particular measure was administered. For this reason, the sample size varies somewhat across the various analyses presented in the next section. However, this attrition does not appear to have seriously affected the balance of the samples across the characteristics on which they were matched.


A series of six multivariate analyses of variance (MANOVAS) were conducted, using group membership (LD or ND) and social status (popular, rejected, and neglected) as independent variables. The grouping of dependent variables for each analysis were as follows: 1. Academic Ability, which included numeric

semester grades in reading, math, social studies,

and science. 2. Teacher Academic, which included regular-class

teachers' perceptions of the child's

classroom motivation, academic performance,

task orientation, and general ability. 3. Teacher Social, which included the regular-class

teachers' ratings of the child's adaptability

and social skills, as well as direct

observational data concerning the frequency

of positive and negative behavior in interactions

between teacher and child. 4. Social Contact, which included the number of

neighborhood friends and how often they

played together, the incidence of arguing and

fighting with friends, the number of schools

attended (to reflect mobility), the child's ratings

of loneliness, and average classmates'

ratings of the child's social desirability. 5. Social Self-concept, which consisted of the

social factors from the Harter and the SDQ. 6. Peer Interaction, which included both positive

and negative behavior initiated by the child

and received by the child from peers.

Behavioral observations were taken in two contexts (classroom and PE class). In a preliminary analysis, differences in behavioral frequencies across the two settings were compared. The results suggested that peer interactions occurred more often in PE classes, whereas teacher interactions were more frequent in classroom settings. However, setting did not interact with either group membership or social status to produce differential outcomes. For this reason, the behavioral data were collapsed across settings.

When an analysis yielded a significant multivariate effect, the various univariate variables comprising the effect were examined. When appropriate, significant univariate effects were further decomposed through post hoc testing using the Scheffe procedure. Alpha was set at .05 for all analyses.

Academic Ability

The 2 x 3 MANOVA on academic variables yielded only a significant multivariate effect for group membership, F(4,124) = 2.86, p = .025. Subsequent analysis of individual variables revealed that ND children scored higher than LD children in two academic areas, social studies and science (see Table 1). Although math and reading scores failed to differentiate the two groups, these scores must be considered in light of the fact that most LD children received their grades in these academic areas from special education teachers. It is possible that these grades reflect different grading policies on the part of resource teachers or differences in the difficulty of curriculum within each of the two academic settings. [TABULAR DATA 1 OMITTED]

Though not surprising, these results clearly indicate the academic difficulties encountered by LD children independent of their social status.

Teacher Academic

The academic distinctions between LD and ND subjects were further emphasized by the significant multivariate effect for group membership in this analysis, F(4,114) = 12.45, p = .0001. Significant group effects were evident for all dependent variables included (see Table 1). Teachers rated ND children higher than LD children in motivation, task orientation, general intelligence, and academic performance but made no distinctions between children based on social status; nor did the two factors interact.

Teacher Social

The only significant multivariate effect in this analysis was again for group membership, F(6,114) = 3.32, p = .005. Teachers made distinctions between the two groups in terms of their social skills, reporting ND children as more skilled (see Table 1). In addition, regular-class teachers displayed more negative behavior to LD than to ND children. Though only a marginal effect, F(12,228) = 1.60,p =.09, the interaction between independent variables details the relationship between teacher interactions, group membership, and social status (see Figure 1). Univariate interactions indicated trends for the amount of positive behavior directed toward teachers, F(2,119) = 2.90, p < .06, and amount of negative behavior received from teachers, F(2,119) = 2.66, p < .07. Across all groups, LD popular children were more likely to initiate positive behavior toward teachers and receive negative behavior from teachers, whereas ND popular children initiated the fewest positive behaviors and received the fewest positive behaviors from regular-class teachers. It appears that teacher interactions with popular students vary substantially as a function of the child's group membership.

Peer Contact

The MANOVA conducted on items dealing with the peer network produced a significant overall effect for social status, F(14,238) = 3.54, p = .0001, but not for group membership or the interaction between the two factors (see Table 2). Popular children reported that they played with friends in the neighborhood more often than did rejected or neglected children. This may have been influenced by the fact that popular children are more stable within the neighborhood, moving less often than rejected children. Popular children, both LD and ND, also returned lower scores on the loneliness measure than did rejected children. Classmates of both sexes also indicated that they preferred playing with both popular and neglected children more than with rejected children. [TABULAR DATA 2 OMITTED]

Social Self-Concept

Significant overall main effects were found for both academic placement, F(2,116) = 4.64, p = .01, and social status, F(4,232) = 4.25 1, p =.002, in the 2 x 3 MANOVA. Table 1 contains significant univariate analyses for the main effect for placement on social self-concept. Scores from the SDQ peer factor were higher for LD than for ND children. A similar pattern was evident on the Harter, but the group differences were much smaller. Significant univariate analyses for the main effect of social status are shown in Table 2, which shows that both the SDQ peer relations and Harter social self-concept scores varied between the status groups. Scheffe comparisons revealed that popular children reported more positive relations with peers (SDQ) than did rejected children, and that both popular and neglected children had higher social self-concepts (Harter) than did rejected children. The overall interaction effect was nonsignificant. These differences in self-reported social competence parallel those reflected in children's reports of loneliness and contact with peers outside of school, as well as classmates' ratings of the child's likability.

Observed Behavior

Results of the 2 x 3 MANOVA conducted on observed behaviors with peers revealed a significant overall main effect for social status, F(8,274) = 2.51, p < .01, and a trend for the interaction of academic placement and social status, F(8,274) = 1.72, p < .09. Table 3 lists the significant univariates for the effect of academic placement, which included giving positive behavior to peers and receiving positive behavior from peers. Post-hoc comparisons found that popular children were more likely to direct positive behavior to peers than were rejected children, and that popular children were more likely to receive positive behavior from peers than were rejected children. [TABULAR DATA 3 OMITTED]

The two-way interaction for displaying positive behavior toward peers approached significance, F(2,140) = 2.84, p = .06, while the same effect for receiving positive behavior from peers was significant, F(2,140) = 3.42,p =.03 (see Figure 1). Exchange of positive behavior was far more variable for LD than for ND children. The interaction effects seem primarily due to the fact that popular LD children were far more likely to both give and receive positive social behavior from peers. This held true regardless of academic placement. This finding closely parallels the finding that popular LD children were also more likely than other groups to initiate positive behavior with teachers.


This investigation was designed to study differences in the social competencies of children both with and without LD, while equating the samples on peer social status. Matching the two groups on social status and other demographic characteristics was useful in separating the effects of social status from those of disability, but such matching necessarily created a sample of children without disabilities that inadequately represented the larger population of children without disabilities from which they were selected.

These subjects were taken from a larger pool of more than 1,100 elementary school children in which 15% of those without disabilities and 28% of those with LD were rejected by classmates in mainstreamed classrooms. Therefore, in matching the sample of children without disabilities to the social status of the sample with LD, the proportion of rejected children without disabilities was substantially raised. At the same time, the proportion of popular children without disabilities was lowered. Given this sampling strategy, it seems likely that the scores of the sample without disabilities are lower than would be the case if the sample was taken randomly from the larger population without disabilities. For this reason, the reader is cautioned not to generalize from this sample without disabilities to the larger population of children without disabilities.

It is clear that academic differences between children with disabilities and those without exist independent of the child's social status. Grades given by regular-class teachers in science and social studies favored students without disabilities. Though grades in math and reading did not discriminate the two groups, this finding must be tempered by the fact that some math grades and all reading grades were given by resource-class teachers. It is possible that the comparable grades given to the two groups in these academic areas may reflect a positive bias in the grading of special education teachers rather than indicating similar levels of academic success. Equally plausible is the possibility that the curriculum in the resource classroom is tailored to the child's current competencies and that better grades should be expected. Academic distinctions between the two groups of children were also evident in the ratings provided by regular-class teachers. In addition to making better grades and being considered smarter, children without disabilities were also seen as being more task oriented and having higher motivation than children with LD.

These teachers also considered children with LD less skilled socially, and the teachers engaged in more negative interactions with these students than with children without disabilities. However, both these outcomes must be viewed cautiously. Analysis of teacher ratings of social skills yielded a larger effect for social status than for group membership, although this outcome was not reported earlier because the multivariate effect for group membership was not significant. Univariate analysis of negative teacher behavior toward the child produced a significant interaction between academic placement and social status. As Figure 1 demonstrates, popular children with LD received more negative behavior from teachers than any other group, whereas there were no instances of such teacher behavior for popular children without disabilities. These behavioral data are also somewhat suspect because of the low base rate of the target behavior and the possibility that such a short observation period might not adequately capture the true rate of the behavior targeted.

Regarding ratings provided by peers or the subjects themselves, the only multivariate effect for academic placement was for social self-concept. Children with LD scored higher as a group than did children without disabilities on the Peer Relations factor of the SDQ. To demonstrate the magnitude of these differences, effect sizes (Glass, 1977) were computed for each social status group, using the standard deviation of the appropriate sample of children without disabilities as the denominator. Rejected children with LD scored .8 of a standard deviation higher than did children without disabilities from the same social status, whereas this difference fell to .45 for popular children with LD and .27 for neglected children without disabilities. Higher scores for children with LD were also evident on the Harter social factor, although the differences were much smaller.

A number of investigations have suggested that the self-concepts of children with LD may be comparable to those of children without disabilities (Chapman, 1988; Coleman, 1983; Strang, Smith, & Rogers, 1978). The basis for these findings seems couched in social comparison theory, which suggests that children use others in their immediate social comparison group (classmates) as the basis for making judgments about their own competence. In addition, the tendency to use specific peers as the basis for referential evaluation is a function of perceived similarities; that is, children seem more likely to choose peers of comparable ability as the basis for comparison (Smith, Zingale, & Coleman, 1978).

If this is true, then the resource classroom provides many children with LD with a reference group in which they may perceive their own capabilities within a more favorable light. This would seem particularly true for rejected children with LD, who represent the largest social group within such settings. Though rejected children seem less competent socially than others, using other rejected children as the basis for making decisions about their own social competence might yield more favorable results than if their social comparisons were limited solely to regular-class peers. This contention seems buttressed by the extremely low self-concept scores for rejected children without disabilities. Like rejected children with LD, those without disabilities have limited social competence but, unlike their peers with disabilities, have no second reference group within which to make self-concept-relevant social comparisons. To some extent, the provision of a second social comparison group may prove beneficial to the self-perceptions of all children with disabilities, regardless of their social status.

Self-concept differences as a function of social status were even larger than those that resulted from group membership. On the Harter, popular and neglected children were systematically higher than rejected children while on the SDQ popular and rejected children were significantly different, with neglected children falling in an intermediate level. This pattern was consistent, regardless of academic placement, and parallels the outcomes of other studies with children without disabilities. These results suggest a positive relationship between social status and perceived social self-competence (Boivin & Begin, 1989), with popular children viewing themselves as more competent than do children from other status groups (Minnett & Coleman, 1992).

Social status differences were also evident regarding children's social networks. Peers rated popular children as more socially desirable than neglected or rejected children. These ratings were from both same-sex and opposite-sex peers, whereas the social nominations used to construct the status groups were based only on same-sex nominations. Popular children also spent more time playing with peers in their neighborhoods than did rejected children, who, in turn, reported having fewer contacts with peers and being lonelier than popular children. One partial explanation for the difficulties encountered by rejected children in their peer network is the fact that their families appear to be more transient than those of children from other social statuses. Rejected children have attended a greater number of schools than popular children. It seems likely that difficulties in social competence would necessarily be exacerbated by frequent moving and having to establish new social networks.

The effects of social status were clearly evident in the behavioral observations of children engaged in social interactions, but there was no main effect for academic placement. Popular children both gave and received more positive social behavior than rejected children, with neglected children falling between the other two groups. Moreover, the interaction between academic placement and social status approached significance, suggesting that for children with LD to be considered popular by peers without disabilities requires an even higher level of pro-social behavior than for popular children without disabilities.

Once social status differences between the two groups of children are equated, it seems clear that the distinctions between the groups reside more in the academic than in the social domain. In fact, the major findings of this investigation indicate that differences in children's social competence are related to the child's social status, not disabilities. Children considered popular by same-sex peers are viewed as more competent than rejected children. This pattern is evident in teacher ratings, the child's self-perceptions, ratings provided by peers, the child's social contacts outside of school, and at the level of overt behavior. It is consistent for all children, both with and without LD.

Two points must be made. One, though children with LD are often rejected by their peers, this outcome is not inevitable. Some of these children are considered popular. As such, they share many of the social characteristics of popular children without disabilities. We must learn more about these children who, despite their academic difficulties, appear capable of succeeding in the social mainstream of public elementary education. It seems unlikely that we will be able to completely eliminate the academic difficulties encountered by children with LD; but studying children who are socially successful despite limited academic success seems a profitable avenue for identifying social skills that may be useful to other children with LD.

Even more important, we must realize that children with LD who are viewed unfavorably by their regular-class peers represent a small subset of a larger group of socially rejected children. These children, though not considered as having disabilities, encounter many of the same problems faced by the unpopular child with LD. The knowledge that special education has developed regarding social-skills training must be shared with regular education. It would seem desirable to supplement social-skills training in special education with comparable training in the educational mainstream for all children who have social difficulties--without distinction between the presence or absence of LD.

We are led to the conclusion that including social skills deficits as a defining characteristic of LD is unwarranted. We must guard against allowing the disproportionately high rate of social rejection among children with LD to lure us into the generalization that LD and social deficits are linked causally. Such a generalization diverts our attention from studying those children with LD who prosper socially in the educational mainstream, and it unduly singles out rejected children with LD from the larger group of rejected children in which they reside.


Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55, 1456-1464. Beck, S., & Collins, L. (1985, May). Children's sociometric groups: Developmental and stability issues. Paper presented at the Midwestern Psychological Association, Chicago. Bell-Dolan, D. J., Foster, S. L,. & Sikora, D. M. (1989). Effects of sociometric testing on children's behavior and loneliness in school. Developmental Psychology, 25(2), 306-311. Boivin, M., & Begin, G. (1989). Peer status and self-perception among elementary school children: The case of the rejected children. Child Development, 60, 591-596. Bukowski, W. M., & Newcomb, A. F. (1984). The stability of determinants of sociometric status and friendship choice: A longitudinal perspective. Developmental Psychology, 21(6), 1032-1038. Bursuck, W. (1989). A comparison of students with low achieving and higher achieving students on three dimensions of social competence. Journal of Learning Disabilities, 22(3), 188-194. Bruck, M. (1986). Social and emotional adjustments of learning disabled children: A review of the issues. In S. Ceci (Ed.), Handbook of cognitive, social and neuropsychological aspects of learning disabilities (pp. 200-232). Hillsdale, NJ: Lawrence Erlbaum. Bryan, T. H. (1974). Peer popularity of learning disabled children. Journal of Learning Disabilities, 7, 621-625. Bryan, T. H., & Bryan, J. H. (1978). Social interactions of learning disabled children. Learning Disability Quarterly, 1, 33-38. Cadwell, J., & Pullis, M. E. ( 1983). Assessing changes in the meaning of children's behavior: Factorial invariance of teachers' temperament ratings. Journal of Educational Psychology, 75, 553-560. Chapman, J. W. (1988). Learning disabled children's self-concepts. Review of Educational Research, 58, 347-371. Coie, J. D., & Dodge, K. A. (1983). Continuities and changes in children's social status: A five-year longitudinal study. Merrill-Palmer Quarterly, 29(3), 261-282. Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982). Dimensions and types of social status: A cross-age perspective. Developmental Psychology, 18, 557-571. Coleman, J. M. (1983). Self-concept and mildly handicapped: The role of social comparisons. Journal of Special Education, 17(1), 37-45. Coleman, J. M. (1985). Achievement level, social class and the self-concepts of mildly handicapped children. Journal of Learning Disabilities, 18, 26-30. Coleman, J. M., McHam, L., & Minnett, A. M. (1992). Similarities in the social competencies of learning disabled and low-achieving elementary school children. Journal of Learning Disabilities 25(10), 671-677. Coleman, J. M., Pullis, M. E., & Minnett, A. M. (1987). Studying mildly handicapped children's adjustment to mainstreaming: A systemic approach. Remedial and Special Education, 8(6), 19-30. Dudley-Marling, D. D., & Edmiason, R. (1985). Social status of learning disabled children and adolescents: A review. Learning Disability Quarterly, 8,189-204. Dunn, L. M. (1968). Special education for the mildly retarded: Is much of it justifiable? Exceptional Children, 34, 5-22. Fox, C. L. (1989). Peer acceptance of learning disabled children in the regular classroom. Exceptional Children, 56,50-59. Garrett, M. K., & Crump, W. D. (1980). Peer acceptance, teacher preference, and self-appraisal of social status among learning disabled students. Learning Disability Quarterly, 3, 42-47. Glass, G. V. (1977). Integrating findings: The meta-analysis of research. Review of Educational Research, 5, 351-379. Gresham, F. M., & Reschly, D. J. (Winter, 1986). Social skill deficits and low peer acceptance of mainstreamed learning disabled children. Learning Disability Quarterly, 9, 23-32. Harter, S. (1982). The Perceived Self-Competence Scale for Children. Child Development, 53, 87-97. Harter, S. (1985). Self-Perception Profile for Children. (Revision of the Perceived Competence Scale for Children.) Denver, CO: University of Denver. Hartup, W. W. (1983). Peer relations. In E. M. Hetherington (Ed.) & P. H. Mussen (Series Ed.), Handbook of child psychology: Vol. 4. Socialization, personality; and social development (pp. 103-196). New York: Wiley. Hayvren, M., & Hymel, S. (1984). Ethical issues in sociometric testing: The impact of sociometric measures on interaction behavior. Developmental Psychology, 20, 844-949. Interagency Committee on Learning Disabilities. (1987). Learning disabilities: A report to Congress. Washington, DC: Author. Keogh, B., Pullis, M. E., & Cadwell, J. (1982). A short form of the Teacher Temperament Questionnaire. Journal of Educational Measurement, 19(4), 323-329. Kistner, J. A., & Gatlin, D. F. (1989). Sociometric differences between learning-disabled and nonhandicapped students: Effects of sex and race. Journal of Educational Psychology, 81(1), 118-120. Marsh, H. W., & Parker, J. W. (1984). Determinants of student self-concept: Is it better to be a relatively large fish in a small pond even if you don't learn to swim as well? Journal of Personality and Social Psychology, 47(1), 213-231. Marsh, H. W., Relich, J. D., & Smith, I. D. (1983). Self-concept: The construct validity of interpretations based upon the SDQ. Journal of Personality and Social Psychology, 45(1), 171-187. McKinney, J., McClure, S., & Feagan, L. (1982). Classroom behavior of the learning disabled child. Learning Disability Quarterly, 16, 45-52. Minnett, A. M., & Coleman, J. M. (1992). Social status among same- and opposite-sex peers: Is self-concept related to opposite-sex peers' perceptions? Manuscript submitted for publication. Perlmutter, B. F., Crocker, J., Corday, D., & Garstecki, D. (1983). Sociometric status and related personality characteristics of mainstreamed and learning disabled adolescents. Learning Disability Quarterly, 6, 20-30. Pullis, M. E. (1985). Temperament characteristics of LD students and their impact on decisions made by resource and mainstream teachers. Learning Disability Quarterly, 8, 104-123. Pullis, M. E., & Cadwell, J. (1982). The influence of children's temperament characteristics on teachers' decision strategies. American Educational Research Journal, 19, 165-181. Sater, G. M., & French, D. C. (1989). A comparison of the social competencies of learning disabled and low achieving elementary-aged children. The Journal of Special Education, 3(1), 29-42. Shavelson, R. J., & Bolus, R. (1982). Self-concept: The interplay of theory and methods. Journal of Educational Psychology, 74(1),

3-17. Smith, M. D., Zingale, S. A., & Coleman, J. M. (1978). The influence of adult expectancy/child performance discrepancies upon children's self-concepts. American Educational Research Journal, 15, 259-265. Strang, L., Smith, M. D., & Rogers, C. M. (1978). Social comparison, multiple reference groups, and the self-concepts of academically handicapped children before and after mainstreaming. Journal of Educational Psychology, 70, 487-497. Thomas, A., & Chess, S. (1977). Temperament and development. New York: Bruner/Mazel. Vaughn, S., Hogan, A., Kouzekanani, K., & Shapiro, S. (1990). Peer acceptance, self-perceptions, and social skills of learning disabled students prior to identification. Journal of Educational Psychology, 82(1), 101-106. Wylie, R. C. (1989). Measures of self-concept. Lincoln: University of Nebraska Press.


J. MICHAEL COLEMAN (CEC TX Federation) is an Associate Professor of Human Development and Communication Sciences and ANN M. MINNETT (CEC TX Federation) is a Research Scientist in the School of Human Development at the University of Texas at Dallas.

This research was supported by Grant #H023CBD140 from the U.S. Department of Education, Office of Special Education and Rehabilitative Services. We wish to acknowledge the administrators, principals, teachers, and students of the Dallas, Texas, Independent School District who participated in this project. In addition, Ann Huffman, Eunsook Kang, Cynthia Kaye, Keri McGiboney, Laura McHam, and Ann Tubbs were instrumental in the collection and coding of the data reported here.
COPYRIGHT 1992 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Coleman, J. Michael; Minnett, Ann M.
Publication:Exceptional Children
Date:Dec 1, 1992
Previous Article:Adult adjustment among high school graduates with mild disabilities.
Next Article:Relationships among states' fiscal and demographic data and the implementation of P.L. 94-142.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters