Perceived athletic competence, sociometric status, and loneliness in elementary school children.

Dunn, Janice Causgrove ; Dunn, John G.H. ; Bayduza, Angela 等

Loneliness has been characterized as an individual's "cognitive awareness of a deficiency in one's social and personal relationships ... [that is accompanied by] ensuing affective reactions of sadness, emptiness, or longing" (Asher & Paquette, 2003, p.75). Parkhurst and Hopmeyer (1999) have further defined loneliness as "a sad or aching sense of isolation ... [that corresponds with] a felt deprivation of, [and] longing for, association, contact, or closeness [with other people]" (p.58). According to Galanski and Kalantzi-Azizi (1999), when children are asked to define loneliness they typically refer to their peer group as a reference point and view loneliness as a state in which one is without friends that can result in feelings of sadness (e.g., Cassidy & Asher, 1992). Overall, loneliness for children reflects interpersonal deficits that are the result of having fewer or less satisfying personal relationships with other children than desired (Ponzetti, 1990).

Research consistently indicates that there are many psychosocial and emotional problems associated with loneliness in childhood and adolescence (Kristensen, 1995). Studies with elementary school aged children reveal that lonely children are less physically active and less fit (Page, Frey, Talbert, & Falk, 1992) and more likely to experience tension and anxiety (Page, 1991) than their non-lonely counterparts. In adolescence, loneliness has been linked to a variety of potentially health-threatening recreational behaviors including cigarette smoking and marijuana use (Page, 1990). If loneliness persists for a prolonged period of time during childhood (i.e., chronic loneliness), there is an increased risk of school drop out, depression, and alcoholism in late adolescence and early adulthood (Asher & Paquette, 2003). Clearly, prolonged loneliness has the potential to seriously undermine an individual's psychological, emotional, and physical well-being. As such, understanding the correlates of loneliness and identifying factors that can potentially decrease the likelihood of experiencing loneliness for children becomes an important research endeavor.

Research on children's loneliness has been dominated by a focus on the influence of sociometric status within the peer system (Asher & Paquette, 2003). The term "sociometric status" is often used interchangeably with other terms such as peer acceptance, social acceptance, or peer status (Weiss & Stuntz, 2004). At its most basic level, sociometric status refers to the degree to which children are liked or disliked by other children within the peer group (Gifford-Smith & Brownell, 2003). Unlike friendship, which represents a bilateral view of experiences between two specific individuals, sociometric status represents a general group-oriented unilateral view that describes how a peer group regards one of its members (Weiss & Stuntz).

Peer acceptance at school is often assessed by having children nominate peers they would most like and least like to play with, or by asking children to indicate on rating scales the extent to which they like or dislike each of their classroom peers (Cillessen & Bukowski, 2000). Irrespective of the measurement technique employed, sociometric categories are then created in an attempt to describe the child's sociometric status within the peer group. By subtracting the number of negative nominations from the number of positive nominations, a child's likeability (or level of social acceptance) by the group is calculated. Children who are rated as popular are well liked by their peers and seldom disliked (i.e., they receive many positive nominations and few negative nominations). In contrast, children who receive many least-like nominations and few most-like nominations are classified as rejected children.

Research strongly supports a link between sociometric status and loneliness in children, especially among children who are rejected by their peers. Rejected children are frequently subjected to physical and verbal harassment (Boivin & Hymel, 1997) and are often excluded from social activities by the peer group (Buhs & Ladd, 2001). Rejected children also have a tendency to disengage from the social environment as a way of avoiding further abuse (Coie & Kupersmidt, 1983). Theorists suggest that rejected children may therefore be exposed to a larger number of social isolation experiences (i.e., loneliness-provoking situations) in their daily lives (Hymel, Tarulli, Thomson, & Terrell-Deutsch, 1999) which in turn increases the degree of loneliness experienced by the child. Thus, it is not surprising that empirical research consistently finds that rejected children experience higher levels of loneliness than their more popular peers (e.g., Asher & Wheeler, 1985; Crick, Grotpeter, & Rockhill, 1999; Parker & Asher, 1993; Parkhurst & Asher, 1992).

Although many factors can potentially contribute to loneliness experiences (e.g., death of a parent or significant other, family relocation), routine rejection by peers is clearly one of the most critical factors that contribute to children's feelings of loneliness (Bullock, 1993; Crick & Grotpeter, 1996). It therefore becomes important to identify and understand factors that may increase a child's likelihood of being accepted by the peer group, which in turn decreases the likelihood of experiencing the destructive psychosocial and emotional problems that often occur in conjunction with rejection. Athletic ability or athletic competence may be one such factor (see Kunesh, Hasbrook, & Lewthwaite, 1992; Page & Scanlan, 1994; Weiss & Duncan, 1992).

Research in sport and physical activity settings has shown that children who demonstrate athletic competence and who hold positive beliefs about their own athletic abilities tend to be more popular with their peers than less skilled children (Rose, Larkin, & Berger, 1997; Weiss & Duncan, 1992). In contrast, children who are poorly skilled often become the target of criticism (Portman, 1995), ridicule and bullying (Fitzpatrick & Watkinson, 2003; Mandich, Polatajko, & Roger, 2003), and exclusion (Evans & Roberts, 1987; Mandich et al., 2003). Positions reflecting higher social standing among children within their social system (i.e., leadership positions) also are afforded to more capable athletes. For example, Evans and Roberts (1987) observed team selection among third through sixth grade boys in the playground. Interviews with the boys about this selection revealed that those who were viewed as having the highest sport ability were typically elected as captains, whereas less skilled boys were generally chosen much later in the selection process, and sometimes even excluded (i.e., "locked out") from the game entirely.

Athletic competence also has been identified as a highly valued attribute among school children. For example, Chase and Dummer (1992) asked fourth to sixth grade children to rank the importance of being good at sports, making good grades, being physically attractive, and having money in relation to their popularity with classmates. Boys ranked sports ability as most important followed by physical appearance, whereas girls ranked physical appearance as most important, followed by sport ability (also see Eccles, Wigfield, Harold, & Blumenfeld, 1993). Buchanan, Blankenbaker, and Cotten (1976) also reported that boys in elementary school viewed athletic competence as being the most important social acceptance criteria at school. Given these findings, Parkhurst and Hopmeyer (1999) speculated that deficits in athletic competence may have implications for the development of loneliness in children "if these skill deficits ... interfere with the establishment and maintenance of peer relationships" (p. 72).

Findings from the extant literature clearly suggest that children's athletic ability (whether actual or perceived) is linked to (or may influence) sociometric status (see Weiss & Stuntz, 2004, for a review). For example, Lubbers, van der Werf, Kuyper, and Offringa (2006) recently reported that higher self-perceptions of athletic competence predicted higher levels of peer acceptance among large samples of first-year junior-high boys (n = 2983) and girls (n = 3864) from the Netherlands (M age = 13 years). Boivin and Begin (1989) found that popular third- and fourth-grade children reported higher levels of self-perceived athletic competence, and received higher athletic competence ratings from teachers than rejected children. Participation in school athletic/sport teams has also been linked with students' popularity in a sample of sixth to eighth grade students (see Eder & Kinney, 1995). However, despite these findings, we could find no studies that specifically examined the relationship between perceived athletic competence and loneliness in children. Page et al. (1992) did report that elementary school children (in Grades 1 to 6) with higher levels of perceived loneliness tended to be less physically fit and less physically active than less lonely children, however, the researchers did not assess perceptions of athletic ability. Thus, the first purpose of the present study was to examine the relationship between loneliness and perceptions of athletic competence (as rated by self and peers) in elementary school children. Given the valued role that athletic competence plays for many elementary school children (Chase & Dummer, 1992; Eccles et al., 1993), and given the potential socialization opportunities that athletic competence may create (via involvement in athletic/sport teams), we hypothesized that children with higher levels of perceived athletic competence (as rated by the self and by others) would experience less loneliness than children who had lower levels of perceived athletic competence.

The second purpose of this study was to further investigate the relationship between sociometric status and perceived athletic ability in elementary school children. Although Weiss and Duncan (1992) found a strong positive correlation between peer acceptance and perceived athletic competence (as rated by the self and by teachers) among a sample of elementary-aged children, it is possible that their findings were influenced by the social context of the study. Specifically, Weiss and Duncan examined the relationship between peer acceptance and athletic competence in a sample of 8-13 year-old summer sport-camp participants. Presumably, athletic ability is a highly desirable attribute in this physical activity setting because it impacts success in almost every situation in the camp. In other words, high athletic competence likely provides multiple opportunities to gain friendships and peer-status by consistently helping one's team to be successful or competitive. In the school setting, however, athletic ability may provide these opportunities on a less frequent basis. Thus, the degree to which Weiss and Duncan's findings generalize to the regular school setting is unknown.

Weiss and Duncan (1992) also interpreted their findings along a continuum of peer acceptance, referring primarily to higher and lower levels of peer acceptance. Although children who experience higher levels of peer acceptance are typically well liked and have many friends (i.e., popular children), children who have lower levels of peer acceptance can be differentiated into those who are "neglected" and those who are "rejected." Given the importance of differentiating between neglected and rejected children (see Asher & Wheeler, 1985), the present study sought to extend Weiss and Duncan's findings by examining differences in perceived athletic competence among children who were classified as "popular" and those who were classified as "rejected" (on the basis of peer nominations).

Although Boivin and Begin (1989) did examine differences in self- and teacher-rated athletic competence levels between popular and rejected elementary school children, Boivin and Begin did not ask the children in their study to rate each others' athletic competence levels. Therefore, building upon the previous work of Weiss and Duncan (1992) and Boivin and Begin, we hypothesized that popular children would (a) give themselves higher self-rated athletic ability ratings, and (b) receive higher athletic ability ratings from peers than rejected children. In accordance with theory, we also hypothesized that popular children would report less loneliness than rejected children (Parker & Asher, 1993). Overall, this study was designed to improve our understanding of the potential role that athletic competence plays in the experience of childhood loneliness and sociometric status. We suggest that this is an important research endeavor because the results have the potential to inform teachers and/or parents about positive or healthy psycho-social experiences associated with the development of athletic competence at a young age.

Method

Participants

Children were selected from 29 classes in seven elementary schools from a western Canadian city. Information letters were distributed to 578 children in Grades 4 through 6. On average, 36% of children from each class participated in the study, with a total of 99 boys (M age = 10.08 years, SD = 1.13) and 109 girls (M age = 10.05, SD = 1.03) agreeing to participate.

Measures

Children completed self-report instruments to measure loneliness levels in school, as well as perceived athletic ability. In addition, participants rated the athletic ability of their classmates, and identified classmates who they most liked and who they least liked in order to assess the sociometric status of the in-class peer group.

Loneliness. Asher and Wheeler's (1985) modified school-specific version of Asher, Hymel, and Renshaw's (1984) Illinois Loneliness and Social Dissatisfaction Scale (ILSDS) was used to assess children's levels of loneliness in school. The scale was originally designed for use with 3rd through 6th grade children and contains 24 items (eight of which are filler items that are not scored). Using a 5-point scale (1 = Always true about me; 5 = Not at all true about me), children rate the extent to which each of the items relates to their own personal experiences in school. Ten of the 16 items are reverse scored (e.g., "I feel lonely at school", "I feel left out of things at school") and then all 16 items are summed to provide a single composite score. Higher composite scale scores reflect higher levels of school-based loneliness (Asher & Gazelle, 1999).

The ILSDS has been the most commonly used measure of loneliness in children's peer-relation studies (Asher & Gazelle, 1999). Asher and Wheeler (1985) conducted an exploratory factor analysis upon ILSDS data provided by 200 children from Grades 3 through 6 and obtained a single-factor solution. Asher and Wheeler also reported a very high level of internal consistency (Cronbach's [alpha] = .90) for the scale. Cassidy and Asher (1992) obtained a single-factor solution for ILSDS data provided by 440 kindergarten and Grade 1 children, and again reported adequate internal consistency ([alpha] = .79). A recent study by Chen et al. (2004) with Brazilian (n = 807), Canadian (n = 274), Chinese (n = 819) and Italian (n = 363) children in Grades 3 to 6 also found high levels of internal consistency for the 16-item scale:[alpha] = .84, .90, .95, and .84 respectively. In accordance with theory, construct validity evidence for the ILSDS has been established in studies where children who were classified as being "rejected" by the peer group had significantly higher loneliness scores than children who were classified as being "popular" by the peer group (e.g., Asher & Wheeler; Cassidy & Asher; Parkhurst & Asher, 1992). Overall, the ILSDS appears to possess adequate reliability and validity characteristics.

It should be noted that in addition to asking direct questions about loneliness (e.g., "I'm lonely at school"), as the instrument's name suggests, the ILSDS also contains items that focus on social dissatisfaction (e.g., "I have lots of friends in my class," "I can find a friend in my class when I need one"). However, for the sake of brevity, and because the ILSDS is treated as a unidimensional instrument from which inferences about loneliness are made (Asher & Wheeler, 1985), ILSDS results in this study will refer only to loneliness rather than to loneliness and social dissatisfaction (for a related discussion see Asher & Gazelle, 1999).

Athletic ability. Children were presented with a complete list of all students in their classroom and asked to rate each classmate's athletic ability. Specifically, participants responded to the instruction, "Please tell us how good your classmates are at playing Sports and doing physical skills during gym class" using a 5-point scale (1 = not very good, 3 =fairly good; 5 = very good). Children were verbally instructed to provide a self-rating of their own atkletic ability when they reached their own name on the class list. Although we recognized the psychometric and measurement limitations of using single item indicators to rate athletic ability, similar single-item procedures have been used successfully in previous studies that have measured perceived athletic competence in older populations (e.g., Elliot & Harackiewicz, 1996; Lubbers et al., 2006; Senko & Harackiewicz, 2002; Vallerand & Reid, 1984). Moreover, we were concerned about the potential time (and fatigue) implications associated with asking children to rate their own ability and the ability of their classmates using multiple indicators.

Given that each class had different numbers of students and different participation rates, mean scores for peer-ratings of athletic ability were computed for each child to ensure that data could be meaningfully compared across children. Thus, the peer rating of athletic ability assigned to a child was computed by taking the composite athletic ability score for the child (based upon the summation of scores provided by the children in the class who participated in the study) then dividing this composite score by the number of children who provided a rating. For example, if 11 children from a class rated Jill, and her composite score was 41, the peer-rated level of athletic ability assigned to Jill was 3.73 (i.e., 41/11).

Sociometric status. Children were presented with a complete list of the names of all students in their respective classes and given the following set of instructions:

 Imagine going on a school trip with your class. For the entire
 trip, your teacher has told you to stay in a group with three other
 people. Which three children from your class would you most like to
 have in your group? Circle the names of the three people you would
 most like to have in your group.

On a separate page, children were presented with another complete list of children's names from their class and asked to identify three children who they would least like to have in their group during the trip. This type of positive and negative peer nomination procedure has been widely used with elementary school children to obtain peer acceptance and peer rejection information (e.g., Asher & Wheeler, 1985; Cassidy & Asher, 1992; Coie, Dodge, & Coppotelli, 1982). The names within each class list were presented to children in random orders for each respondent to minimize potential presentation order effects.

The frequency of positive and negative peer nominations received by each participant was adjusted by taking into consideration the number of students in the class who provided peer nominations. For example, if Jill received five positive peer nominations from the 11 other children in her class who participated in the study, the "most-like" score assigned to Jill would have been .45 (i.e., 5/11). Similarly, if Jill received two least-like nominations, her least-like rating would have been. 18 (i.e., 2/11). An overall sociometric status score was then computed (see Cillesen & Bukowski, 2000) by subtracting the least-like rating from the most-like rating. (In Jill's case, her sociometric status score would have been .27 [i.e., .45 - .18]). Thus, the final sociometric score assigned to a child could range from -1.0 (i.e., all participants in the class gave the child a least-like nomination) to 1.0 (i.e., all participants in the class gave the child a most-like nomination). Children with negative scores receive more least-like nominations than most-like nominations, and children with positive scores receive more most-like nominations than least-like nominations. Larger negative scores are indicative of greater peer rejection, and larger positive scores are indicative of greater peer acceptance (Asher & Gazelle, 1999).

Procedures

Ethical clearance to conduct the study was obtained from the University's human ethics research board and from the school district prior to contacting individual schools. Information letters and parental consent forms were sent home with all children from every class. Only children who returned signed parental consent forms were allowed to participate in the study. Given that the research team only had access to information about the age and gender of participants, it was not possible to determine if any systematic differences existed between participants and non-participants in this study. In other words, no deviant case analysis was possible. Each child who participated in the study was removed from the classroom and individually tested in a quiet location under the supervision of a research assistant. Completion time for the questionnaire package was approximately 20 minutes. A total of 55 children (26 boys, 29 girls: Mage = 10.13 years, SD = 1.02) were re-tested to permit the assessment of test-retest reliability for the loneliness responses (i.e., ILSDS scores) and athletic ability ratings. Average time between the two test administrations was 3.91 days (SD = 1.76). The test administrator provided oral directions for the participants when requested by children to do so.

Results

Preliminary Data Analyses

Loneliness. To ensure that the ILSDS (Asher & Wheeler, 1985) was functioning in accordance with theoretical expectations, a Principal Components analysis was conducted upon the correlation matrix of the ILSDS data (N= 208). Five eigenvalues > 1.0 were obtained ([[lambda].sub.1] = 4.81, [[lambda].sub.2] = 1.33, [[lambda].sub.3] = 1.16, [[lambda].sub.4] = 1.09, [[lambda].sub.2] = 1.01), however, examination of the eigenvalue scree plot (using Cattell's [1978] scree criteria for interpretation purposes) strongly suggested the retention of a single factor. Velicer, Eaton, and Fava (2000) recommended that one of the best ways to determine the number of factors is to use parallel analysis in conjunction with Cattell's scree test. Results of a parallel analysis using procedures described by Lautenschlager (1989) indicated that a single factor be retained (i.e., only the first eigenvalue exceeded the criterion eigenvalue in the parallel analysis). Given that previous research has also suggested a single-factor solution (i.e., Asher & Wheeler, 1985), one factor was chosen to represent the latent dimensionality of the 16 items. Factor loadings ranged in magnitude from .33 to .67, indicating that all items had meaningful loadings on the factor (see Gorsuch, 1983). Internal consistency for the scale was acceptable ([alpha] = .83), and test-retest reliability (as measured by an intraclass correlation coefficient [ICC]) for the composite scale score was also acceptable ([M.sub.Time 1] = 28.43, SD = 8.41 ; [M.sub.Time 2] = 27.20, SD = 9.61 ; ICC = .77, p < .001). Overall, these results suggest that the ILSDS had adequate internal validity and reliability characteristics in this study.

Athletic ability ratings. Given that single-item indicators were employed to measure students' ratings of their own athletic ability and the athletic ability of their classroom peers, we examined the test-retest reliability of these ratings using intraclass correlation coefficients [ICC]. The ICC for self-ratings of athletic ability ([M.sub.Time 1] = 4.24, SD = .77; [M.sub.Time 2] = 4.33, SD = .75; n = 54) was .86 (p < .001), suggesting that the single-item self-ratings of athletic ability were highly reliable across time.

To compute the ICC for athletic ability ratings of peers, a classmate was selected at random from each participant's class list. This selected child's athletic ability ratings at time 1 and time 2 were then used for analytic purposes. Thus, for each of the retested children (n = 55), we used the rating for the randomly selected peer at time 1, and the rating for the same peer at time 2. The ICC for ratings of peer athletic ability ([M.sub.Time 1] = 2.47, [SD.sub.Time 1] = 1.12; [M.sub.Time 2] = 2.53, [SD.sub.Time 2] = 1.30) was .66 (p < .001). Although the ICC was statistically significant, Vincent (1995) suggests that ICCs > .70 in the behavioral sciences are desirable, so the test-retest reliability of the peer-ratings of athletic ability must be considered marginal, and subsequent results pertaining to this variable should be interpreted with some degree of caution.

Gender differences. Prior to examining the relationships among loneliness, athletic ability, and sociometric status, we screened the data for potential gender differences using a one-way MANOVA. Gender was entered as the independent variable. The four dependent variables were loneliness (i.e., composite ILSDS score), self-rating of athletic ability, peer rating of athletic ability, and sociometric status. Table I contains the means and standard deviations for the four dependent variables according to gender.

A significant multivariate test statistic was obtained: Wilks' [LAMBDA] = .699, F (4,203) = 22.06,p < .001, partial [[eta].sup.2] = .30. Follow up univariate F-tests revealed statistically significant gender differences on three of the four dependent variables. On average, boys gave themselves higher self-ratings of athletic ability (M = 4.51) than girls (M = 3.96), boys received higher ratings of athletic ability from their classmates (M = 4.11) than girls (M = 3.52), and boys received lower sociometric status ratings from classmates (M = -.02) than girls (M = .05). Given these differences, all subsequent analyses were conducted separately upon data provided by boys and girls.

Inter-Variable Relationships

Bivariate correlations (r) were computed among the four variables and are shown in Table 2. Fairly consistent patterns of correlations were observed across gender. Notably, for both boys and girls, significant negative correlations were observed between (a) loneliness and the average athletic ability rating that children received from their peers, and (b) loneliness and sociometric status. In other words, boys and girls who reported lower levels of loneliness tended to receive higher athletic ability ratings and higher sociometric ratings from their peers. In addition, girls (but not boys) also indicated that higher levels of self-rated athletic competence were associated with lower levels of loneliness (r = -.28). Collectively, these results suggest that loneliness experienced at school is related to perceptions of athletic competence (both self and peer rated) and sociometric status.

Also consistent across both sexes were the correlations between athletic ability ratings provided by peers and sociometric status. Specifically, higher levels of sociometric status were associated with higher athletic ability ratings from peers. The positive correlation between sociometric status and athletic ability (as judged by peers) was particularly strong for girls (r = .49). These results suggest that sociometric status may be linked to individuals' levels of publicly displayed athletic ability.

Predicting Loneliness

Simultaneous multiple regression analyses were conducted separately on boys' and girls' data sets to further examine the relationships between loneliness and athletic ability and sociometric status. Loneliness was regressed on self-rated athletic ability, athletic ability as rated by peers, and sociometric status. Results of these regression analyses are contained in Table 3. For boys, only sociometric status was a significant predictor of loneliness ([beta] = -.215), although the level of statistical significance for athletic ability as rated by peers was marginal (p = .06). For girls, both sociometric status ([beta] = -.243) and self-rated athletic ability ([beta] = -.224) were significant predictors of loneliness (ps < .05). These results suggest that an understanding of children's loneliness levels in school can be furthered by considering children's sociometric status and their athletic competence (whether perceived by the self in the case of girls, or judged by peers in the case of boys).

Although the bivariate correlations (Table 2) between athletic ability (as judged by peers) and sociometric status for boys and girls were quite strong (r = .33 and .49 respectively), the athletic ability rating that children received from peers was not a strong predictor of loneliness in the regression analyses. We speculated that sociometric status and athletic competence (as judged by peers) explained similar parts of the variance in loneliness scores. Consequently, we removed the sociometric status variable from the regression analyses to see if athletic ability (as rated by peers) would then predict children's loneliness scores. In these subsequent analyses, peer ratings of athletic ability became significant predictors of loneliness for both boys ([beta] = -.266, t = -2.688, p < .01) and girls ([beta] = -.263, t = -2.788, p < .01).

Differences Between Popular and Rejected Children

By definition, popular children receive a large number of most-liked peer nominations and few least-liked peer nominations from their classmates. In contrast, rejected children receive many least-liked peer nominations and few most-liked peer nominations. To examine differences between popular children and rejected children on the variables of interest in this study, an extreme-groups paradigm was adopted whereby a select number of participants were assigned to a "popular" group or a "rejected" group on the basis of their sociometric standings.

Children who had a sociometric status score > 1 SD above the mean (within their corresponding gender) were assigned to their respective gender's popular group. In contrast, children who had a sociometric status score > 1 SD below the mean (within their corresponding gender) were assigned to their respective gender's rejected group. A total of 15 boys (M = .35, SD = .11) and 11 girls (M = .34, SD = .07) met the criterion for assignment to the popular groups, and 15 boys (M = -.42, SD = .13) and 13 girls (M = -.37, SD = .12) met the criterion for assignment to the rejected groups.

Separate MANOVAs for boys and girls were conducted to examine between-group differences (i.e., popular vs rejected) on loneliness, self-ratings of athletic ability, and peer-ratings of athletic ability. Descriptive statistics for each group on the three dependent variables are contained in Table 4. Significant multivariate test statistics were obtained for both boys (Wilks' [LAMBDA] = .671, F [3, 26] = 4.25, p < .05, partial [[eta].sup.2] = .33) and girls (Wilks' [LAMBDA]= .491, F [3, 20] = 6.91 ,p < .005, partial [[eta].sup.2] = .51). Follow up univariate F-tests revealed statistically significant between-group differences on loneliness and athletic ability (as rated by peers) for both boys and girls (see Table 4). These results indicate that popular children, on average, reported less loneliness and received higher athletic ability ratings from classroom peers than rejected children.

Discussion

The first purpose of this study was to examine the relationship between loneliness and perceived athletic ability (as rated by self and peers). Results generally supported the hypothesis that children with higher levels of perceived athletic competence (as rated by the self and by peers) would experience less loneliness than children who had lower levels of perceived athletic competence. As shown in Table 2, athletic ability as rated by peers was negatively correlated with loneliness for both boys and girls. Self-rated athletic ability also was negatively correlated with loneliness for boys and girls, although the former correlation was not statistically significant. While the correlational design of this study precludes causal inferences, results of previous research conducted with children who have movement difficulties (i.e., who demonstrate low athletic ability) may shed light on these relationships.

The literature suggests that poorly skilled children are often ridiculed in physical activity situations (Mandich et al., 2003) and as a result may withdraw from participation in these settings to avoid further embarrassment and humiliation (Fitzpatrick & Watkinson, 2003). In addition, low skilled children are sometimes excluded from participation or relegated to peripheral roles in an activity (Evans & Roberts, 1987). Evans and Roberts suggested that this voluntary withdrawal or involuntary exclusion increases "the social distance between [these children] ... and their peers" (p. 27), thereby reducing opportunities to interact with peers and develop friendships. This cycle of events can lead to increased feelings of sadness and isolation (Fitzpatrick & Watkinson, 2003) that are commonly associated with loneliness (see Cassidy & Asher, 1992).

In contrast to the previous explanation, we cannot rule out the possibility that increased feelings of loneliness may actually influence an individual's self-perceptions of athletic ability. Ernst and Cacioppo (1999) suggest that loneliness provides a negative lens through which people interpret the world. In other words, lonely individuals have a tendency to engage in negative cognitive appraisals or evaluations of their circumstances (Ernst & Cacioppo). To this end, children who are higher in loneliness may be more inclined to evaluate themselves in a negative way (including evaluations of their athletic ability) in comparison to children who are less lonely. Further research is clearly required to determine the causal pathways and processes that underlie the relationships between loneliness and athletic competence observed in this study.

As noted previously, the negative correlation between girls' self-ratings of athletic ability and loneliness was statistically significant (r = -.28), but this correlation was not significant for boys (r = -.18). The means and standard deviations in Table 1 reveal a potential ceiling effect in the boys' self-ratings of athletic ability. Specifically, the mean rating for the boys was 4.51 (on a 5-point scale), with a standard deviation of .79. These descriptive statistics suggest that there is a restriction in range in the boys' self-ratings of athletic ability which in turn may have attenuated the magnitude of the correlation with loneliness. This ceiling effect may be influenced by the tendency for boys to overestimate their athletic competence. As discussed by Cole et al. (2001), boys tend to overestimate their own athletic competence because they are more self-congratulatory than girls. The significance tests in Table 1 appear to corroborate the potential validity of this position given that boys had significantly higher self-ratings of athletic ability than girls (p < .001, effect size = .68).

The second purpose of this study was to examine levels of perceived athletic ability and loneliness as a function of peer status. The hypothesis that popular children would receive higher athletic ability ratings from peers in comparison to rejected peers was supported. This is consistent with the findings of previous research showing that higher levels of peer acceptance or sociometric status correspond to higher levels of other-rated athletic competence (e.g., Weiss & Duncan, 1992). As discussed previously, athletic competence has been identified as a highly valued attribute (Chase & Dummer, 1992; Eccles et al., 1993) and an important criterion for social acceptance among school children (Boivin & Begin, 1989; Buchanan et al., 1976). Although the criteria for sociometric nominations were not assessed in this study, the current results demonstrate support for Weiss and Stuntz's (2004) conclusion that "athletic ability (actual and perceived) and popularity ... are strongly linked for children" (p. 175).

Contrary to our hypothesis (and previous research findings: Boivin & Begin, 1989; Lubbers et al., 2006; Weiss & Duncan, 1992), self-rated perceptions of athletic ability did not differ as a function of sociometric status (see Table 4). Although Boivin and Begin observed a cluster of popular children in their sample who had significantly higher self-perceptions of athletic competence than a cluster of rejected children, they also reported a cluster of rejected children who had high self-perceptions of athletic competence. The researchers suggested that this latter cluster's overly favorable self-evaluations of competence served as a self-protective mechanism to preserve and enhance self-esteem. It is possible that a similar phenomenon was operating here.

Finally, as expected, popular children reported less loneliness than their rejected peers (see Table 4). This result was similar for both boys and girls, and is consistent with previous research in the area of children's loneliness and sociometric status (e.g., Asher & Wheeler, 1985; Crick & Ladd, 1993).

The results of this study were generally consistent with expectations based on theory and previous work. Nevertheless, the study is not without limitations. First, the average classroom participation rate was quite small (i.e., 36%), introducing the possibility that certain groups of children may have been systematically under or over represented in our sample. Second, we used single item indicators to assess participants' athletic ability ratings of themselves and of their classmates. Concerns about single item indicators generally center on their lack of reliability and validity evidence. Although multiple item indicators of athletic competence would overcome this problem, the single item indicators used in this study did demonstrate adequate levels of test-retest reliability, thereby increasing our confidence in the findings. Third, we acknowledge that relying solely upon self-report data of young children may be problematic because younger children (compared to older children) are sometimes less reliable in their use of self-report protocols (see Mellor, 2004). Consequently, future research may be strengthened if sociometric ratings and athletic competence ratings were also provided by teachers who observe the children on a daily basis.

Given the negative impact of loneliness on psychological and emotional well being, identifying the relative contributions of factors that influence children's sociometric status (and loneliness) is an important research endeavor. Factors such as physical appearance, athletic competence, and social skills are believed to affect children's peer acceptance (Page & Scanlan, 1994). Although some researchers have even suggested that "athletic skills may be more central to popularity than social competence" (Boivin & Begin, 1989, p. 594), the mechanisms and processes by which these factors influence peer acceptance require more research.

Author's Note

This research was supported in part by funding from the Sport, Recreation, and Lotteries Branch of Saskatchewan Municipal Government and by a standard research grant from the Social Sciences and Humanities Research Council of Canada (SSHRC).

References

Asher, S. R., & Gazelle, H. (1999). Loneliness, peer relations, and language disorder in childhood. Topics in Language Disorders, 19 (2), 16-33.

Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55, 1457-1464.

Asher, S. R., & Paquette, J. A. (2003). Loneliness and peer relations in childhood. Current Directions in Psychological Science, 12(3), 75-78.

Asher, S. R., & Wheeler, V. A. (1985). Children's loneliness: A comparison of rejected and neglected peers. Journal of Consulting and Clinical Psychology, 53, 500-505.

Boivin, M., & Begin, G. (1989). Peer status and self-perception among early elementary school children: The case of the rejected children. Child Development, 60, 591-596.

Boivin, M., & Hymel, S. (1997). Peer experiences and social self-perceptions: A two-stage mediational model. Developmental Psychology, 33, 135-145.

Buchanan, H. T., Blankenbaker, J., & Cotten, D. (1976). Academic and athletic ability as popularity factors in elementary school children. Research Quarterly, 47, 320-325.

Buhs, E. S., & Ladd, G. W. (2001). Peer rejection as an antecedent of young children's school adjustment: An examination of mediating processes. Developmental Psychology, 3 7, 550-560.

Bullock, J. R. (1993). Lonely children. Young Children, 48, 53-57.

Cassidy, J., & Asher, S. R. (1992). Loneliness and peer relations in young children. Child Development, 63,350-365.

Cattell, R. B. (1978). The scientific use of factor analysis in behavioral and life sciences. New York: Plenum Press.

Chase, M. A., & Dummer, G. M. (1992). The role of sports as a social status determinant for children. Research Quarterly for Exercise and Sport, 63, 418-424.

Chen, X., He, Y., De Olivera, A. M., Coco, A. L., Zappulla, C., Kaspar, V., et al. (2004). Loneliness and social adaptation in Brazilian, Canadian, Chinese, and Italian children: a multi-national comparative study. Journal of Child Psychology and Psychiatry, 45, 1373-1384.

Cillessen, A. H. N., & Bukowski, W. M. (2000). Recent advances in the measurement of acceptance and rejection in the peer system. San Francisco, CA: Jossey-Bass.

Cohen, J. (1977). Statistical power analysis for the behavioral sciences (revised ed.). New York: Academic Press.

Coie, J. D., Dodge, K. A., & Coppotelli, H. (1982). Dimensions and types of social status: A cross-age perspective. Developmental Psychology, 18, 557-570.

Coie, J. D., & Kupersmidt, J. B. (1983). Abehavioral analysis of emerging social status in boys' groups. Child Development, 54, 1400-1416.

Cole, D. A., Cho, S., Martin, J. M., Seroczynski, A. D., Tram, J., & Hoffman, K. (2001). Effects of validity and bias on gender differences in the appraisal of children's competence: Results of MTMM analyses in a longitudinal investigation. Structural Equation Modeling, 8, 84-107.

Crick, N. R., & Grotpeter, J. K. (1996). Children's treatment by peers: Victims of relational and overt aggression. Development and Psychopathology, 8, 367-380.

Crick, N. R., Grotpeter, J. K., & Rockhill, C. M. (1999). A social-informantion-processing approach to children's loneliness. In K. J. Rotenberg & S. Hymel (Eds.), Loneliness in childhood and adolescence (pp. 153-175). New York: Cambridge University Press.

Crick, N. R., & Ladd, G. W. (1993). Children's perceptions of their peer experiences: Attributions, loneliness, social anxiety, and social avoidance. Developmental Psychology, 29, 1-11.

Eccles, J. S., Wigfield, A., Harold, R. D., & Blumenfeld, P. (1993). Age and gender differences in children's self- and task perceptions during elementary school. Child Development, 64, 830-847.

Eder, D., & Kinney, D. A. (1995). The effect of middle school extracurricular activities on adolescents' popularity and peer status. Youth & Society, 26, 298-324.

Elliot, A. J., & Harackiewicz, J. M. (1996). Approach and avoidance achievement goals and intrinsic motivation: A mediation analysis. Journal of Personality and Social Psychology, 70, 461-475.

Ernst, J. M., & Cacioppo, J. T. (1999). Lonely hearts: Psychological perspectives on loneliness. Applied & Preventive Psychology, 8, 1-22.

Evans, J., & Roberts, G. C. (1987). Physical competence and the development of children's peer relations. Quest, 39, 23-35.

Fizpatrick, D. A., & Watkinson, E. J. (2003). The lived experience of physical awkwardness: Adults' retrospective views. Adapted Physical Activity Quarterly, 20, 279-297.

Galanski, E. P., & Kalantzi-Azizi, A. (1999). Loneliness and social dissatisfaction: Its relation with children's self-efficacy for peer interaction. Child Study Journal, 29(1), 1-22.

Gifford-Smith, M. E., & Brownell, C. A. (2003). Childhood peer relationships: Social acceptance, friendships, and peer networks. Journal of School Psychology, 41, 235-284.

Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, N J: Erlbaum.

Hymel, S., Tarulli, D., Thomson, L. H., & Terrell-Deutsch, B. (1999). Loneliness through the eyes of children. In K. J. Rotenberg & S. Hymel (Eds.), Loneliness in childhood and adolescence (pp. 80-106). New York: Cambridge University Press.

Kristensen, K. (1995). The lived experience of childhood loneliness: A phenomenological study. Issues in Comprehensive Nursing, 18, 125-137.

Kunesh, M., Hasbrook, C. A., & Lewthwaite, R. (1992). Physical activity socialization: Peer interactions and affective responses among a sample of sixth grade girls. Sociology of Sport Journal, 9, 385-396.

Lautenschlager, G. J. (1989). A comparison of alternatives to conducting Monte Carlo analyses for determining parallel analysis criteria. Multivariate Behavioral Research, 24, 365-395.

Lubbers, M. J., van der Werf, M. P. C., Kuyper, H., & Offringa, G. J. (2006). Predicting peer acceptance in Dutch youth: A multilevel analysis. Journal of Early Adolescence, 26, 4-35.

Mandich, A.D., Polatajko, H.J., & Rodger, S. (2003). Rites of passage: Understanding participation of children with developmental coordination disorder. Human Movement Science, 22,583-595.

Mellor, D. (2004). Furthering the use of the Strengths and Difficulties Questionnaire: Reliability with younger child respondents. Psychological Assessment, 16, 396-401.

Page, R. M. (1990). Loneliness and adolescent health behavior. Health Education, 21 (5), 14-17.

Page, R. M. (1991). Loneliness as a risk factor in adolescent hopelessness. Journal of Research in Personality, 25, 189-195.

Page, R. M., Frey, J., Talbert, R., & Falk, C. (1992). Children's feelings of loneliness and social dissatisfaction: Relationship measures of physical fitness and activity. Journal of Teaching in Physical Education, 11, 211-219.

Page, R. M., & Scanlan, A. (1994). Childhood loneliness and isolation: Implications and strategies for childhood educators. Child Study Journal, 24, 107-118.

Parker, J. Ct, & Asher, S. R. (1993). Friendship and friendship quality in middle childhood: Links with peer group acceptance, and feelings of loneliness and social dissatisfaction. Developmental Psychology, 29, 611-621.

Parkhurst, J. T., & Asher, S. R. (1992). Peer rejection in middle school: Subgroup differences in behavior, loneliness, and interpersonal concerns. Developmental Psychology, 28, 231-241.

Parkhurst, J. T., & Hopmeyer, A. (1999). Developmental change in the sources of loneliness in childhood: Constructing a theoretical model. In K. J. Rotenberg & S. Hymel (Eds.), Loneliness in childhood and adolescence (pp. 56-79). New York: Cambridge University Press.

Ponzetti, J. J. (1990). Loneliness among college students. Family Relations, July, 336-340.

Portman, P. A. (1995). Who is having fun in physical education classes? Experiences of sixth-grade students in elementary and middle schools. Journal of Teaching in Physical Education, 14, 445-453.

Rose, B., Larkin, D., & Berger, B. G. (1997). Coordination and gender influences on the perceived competence of children. Adapted Physical Activity Quarterly, 14, 210-221.

Senko, C., & Harackiewicz, J. M. (2002). Performance goals: The moderating roles of context and achievement orientation. Journal of Experimental Social Psychology, 38, 603-610.

Weiss, M. R., & Duncan, S. C. (1992). The relation between physical competence and peer acceptance in the context of children's sport participation. Journal of Sport & Exercise Psychology, 14, 177-191.

Weiss, M. R., & Stuntz, C. P. (2004). A little friendly competition: Peer relationships and psychosocial development in youth sport and physical activity contexts. In M. R. Weiss (Ed.), Developmental sport and exercise psychology: A lifespan perspective (pp. 165-196). Morgantown, WV: Fitness Information Technology.

Vallerand, R. J., & Reid, G. (1984). On the causal effects of perceived competence on intrinsic motivation: A test of cognitive evaluation theory. Journal of Sport Psychology, 6, 94-102.

Velicer, W. F., Eaton, C.A., & Fava, J. L. (2000). Construct explication through factor or component analysis: A review and evaluation of alternative procedures for determining the number of factors or components. In R.D. Goffin & E. Helmes (Eds.), Problems and solutions in human assessment: Honoring Douglas N. Jackson at seventy (pp. 41-71). Boston, MA: Kluwer.

Vincent, W. J. (1995). Statistics in kinesiology. Champaign, IL: Human Kinetics.

Janice Causgrove Dunn, John G. H. Dunn and Angela Bayduza

University of Alberta

Address Correspondence To: Dr. Janice Causgrove Dunn, Faculty of Physical Education & Recreation, E-488 Van Vliet Centre, University of Alberta, Edmonton, Alberta, Canada T6G 2H9. Phone: (780)-492-0580. FAX: (780)-492-2364. e-mail: janice.causgrovedunn@ualberta.ca

Table 1. Means, Standard Deviations, and Univariate Test Statistics
for Between-Gender Comparisons

 Gender

 Boys Girls

Variables M (SD) M (SD)

Loneliness 27.64 (8.02) 27.49 (8.19)

Self-rating of
athletic ability 4.51 (0.79) 3.96 (0.83)

Athletic ability as
rated by peers 4.11 (0.59) 3.52 (0.66)

Sociometric status -0.02 (0.24) 0.05 (0.20)

 Univariate test statistics

Variables F (dj) p ES (a)

Loneliness .018 (1,206) ns .02

Self-rating of
athletic ability 24.078 (1,206) <.001 .68

Athletic ability as
rated by peers 46.578 (1,206) <.001 .94

Sociometric status 5.160 (1,206) <.05 -.32

(a) Effect size (ES) was computed using Cohen's (1977) effect
size index (d) for independent samples.

Table 2. Correlations Between Loneliness, Athletic Ability
Ratings, and Sociometric Status for Boys and Girls

 Loneliness Self-rating Athletic Socio-
 of athletic ability rated metric
 ability by peers status

Loneliness -- -.18 -.29 *** -.29 ***
Self-rating of
athletic ability -.28 *** -- .21 * .07
Athletic ability
as rated by peers -.32 *** .30 *** -- .33 ***
Sociometric
status -.32 *** .06 .49 **** -

Note. Correlations for boys (n = 99) are contained in the upper
triangular matrix, and correlations for girls (n = 109) are contained
in the lower triangular matrix.

* p < .05. ** p < .0l. *** p < .005. **** p < .001.

Table 3. Simultaneous Multiple Regression Analyses of Athletic Ability
Variables and Sociometric Status on Loneliness for Boys and Girls

Gender Variables [beta] t

Boys F(3,95)=5.185, p<.005, [R.sup.2]=.14
 Self-rating of athletic ability -.122 -1.253
 Athletic ability as rated by peers -.195 -1.897
 Sociometric status -.215 -2.132
Girls F(3,105)=7.956, p<.001, [R.sup.2]=.19
 Self-rating of athletic ability -.224 -2.416
 Athletic ability as rated by peers -.136 -1.275
 Sociometric status -.243 -2.391

 Semipartial
Gender Variables p Correlation

Boys F(3,95)=5.185, p<.005, [R.sup.2]=.14
 Self-rating of athletic ability ns -.119
 Athletic ability as rated by peers .061 -.180
 Sociometric status <.05 -.203
Girls F(3,105)=7.956, p<.001, [R.sup.2]=.19
 Self-rating of athletic ability <.05 -.213
 Athletic ability as rated by peers ns -.112
 Sociometric status <.05 -.211

Table 4. Means, Standard Deviations, and Univariate Test Statistics
for Between-Group Comparisons of Popular and Rejected Children
According to Gender

 Sociometric Status

 Popular Rejected

Gender Variables M (SD) M (SD)

Boys
 Loneliness 25.40 (5.07) 33.20 (12.04)
 Self-Rating of 4.40 (1.12) 4.44 (0.67)
 Athletic Ability
 Athletic Ability as 4.48 (0.50) 3.84 (0.60)
 Rated by Peers
Girls
 Loneliness 23.64 (5.51) 32.59 (8.92)
 Self-Rating of 4.18 (0.75) 3.85 (0.67)
 Athletic Ability
 Athletic Ability as 3.87 (0.43) 2.85 (0.78)
 Rated by Peers

 Univariate test statistics

Gender Variables F (df) p ES (a)

Boys
 Loneliness 5.346 (1, 28) <.05 0.84
 Self-Rating of 0.012 (1, 28) ns 0.04
 Athletic Ability
 Athletic Ability as 10.282 (1, 28) <.005 1.09
 Rated by Peers
Girls
 Loneliness 8.348 (1, 22) <.01 1.18
 Self-Rating of 1.304 (1, 22) ns 0.46
 Athletic Ability
 Athletic Ability as 14.773 (1, 22) <.001 1.58
 Rated by Peers

Note. Group sizes: popular boys (n = 15), rejected boys (n = 15),
popular girls (n = 11), rejected girls (n = 13).

(a) Effect size (ES) was computed using Cohen's (1977) effect size
index (d) for independent samples.