Social competence is vitally important for adolescents. Social difficulties experienced during youth, such as rejection by peers, predict significant difficulties later in life, including dropping out of school, criminality, and psychological disorders (Parker and Asher 1987). Concurrently, poor social functioning has been implicated in the maintenance of a number of psychological problems, including internalizing difficulties such as depression (e.g., Rudolph et al. 2000) and externalizing symptoms such as aggressive behavior (e.g., Dodge et al. 1985; Gaffney and McFall 1981). These associations lend urgency to the development of valid theoretical and measurement models of youth social competence. This essay outlines current thinking concerning definitions of this construct and the types of factors associated with variability in social competence. Measures of social competence are placed within this theoretical framework, and implications for intervention are discussed briefly.
Definitions and Theoretical Models of Youth Social Competence
A large body of work has been devoted to operationalizing and measuring youth social competence (see Ladd 2005 for review). Although there is significant heterogeneity in definitions of social competence (Dodge 1985), there is increasing consensus that the construct reflects effectiveness in interpersonal interactions (see Rose-Krasnor 1997). Moreover, theorists have identified four sources of variability in interpersonal effectiveness: (1) individual, (2) behavior, (3) situation (i.e., the interpersonal circumstances in which behavior is embedded), and (4) judge (i.e., who is evaluating the behavior; see Dirks et al. 2007a). To date, researchers have focused primarily on the first two factors. Clearly, characteristics of individuals will contribute to their social success. Researchers have identified a variety of individual- or child-level variables that are associated with social competence, such as having a sense of humor (Masten 1986). At the extreme end of this approach are trait models of social competence, which locate interpersonal effectiveness entirely within the individual (e.g., Vaughn et al. 2000). In other words, competence is a property of youth, who each possess this trait to a lesser or greater extent. This approach is appealing, perhaps to developmental psychologists in particular, as it provides a unifying construct that can be assessed across the life span. On the other hand, trait approaches to competence have been challenged both theoretically and clinically. Theoretically, McFall (1982) noted that the logic underlying this approach is circular: people behave competently because they are competent, but they are deemed competent because they behave competently. Clinically, localizing competence entirely within youth is problematic because it does not suggest targets for intervention. Once individuals who are struggling socially have been identified, how can clinicians help them to achieve social success?
One way to solve this problem is to examine the behaviors in which youth are engaging. Social behaviors are another source of variability in youth social competence, and social-skills models of competence equate behaviors and social competence (see McFall 1982). Numerous studies have examined social behaviors associated with good and problematic outcomes in the peer group (see Ladd 2005). For example, in general, aggressive and avoidant behaviors are associated with rejection by peers, whereas sociable actions are associated with peer acceptance (see Newcomb et al. 1993). Within a social-skills approach to competence, youth who engage in “good” behaviors, would be seen as competent, whereas youth engaging in “problematic” behaviors would be seen as incompetent.
The challenge associated with locating competence exclusively in social behaviors becomes apparent almost immediately: How do investigators decide which behaviors are competent? Different researchers have suggested different criteria that might form the basis of these evaluations. As described earlier, some people have argued that interpersonal effectiveness is the benchmark for competence (see Rose-Krasnor 1997). Others have posited more specific criteria, such as meeting a goal (e.g., Erdley and Asher 1999). Although there is variability among researchers concerning how these judgments should be formed, there is a general agreement that social competence is an evaluative construct (see Dirks et al. 2007a). This idea of competence as an evaluation is reflected in McFall’s (1982) definition of competence, which states that the construct of social competence “reflects somebody’s judgment, on the basis of certain criteria, that a person’s performance on some task is adequate” (McFall 1982, p. 13).
This definition of competence implicates the four sources of variability described previously: individual, behavior, situation, and judge. Despite their acknowledged theoretical importance, far less empirical work has examined situation- and judge-level factors. Social situations can affect behavior in at least two ways. First, they will influence the type of behaviors in which a person engages. Different situations will press for different actions; in general, youth should, and do, respond differently when they are shoved by a peer than when a peer says hello to them (see Shoda et al. 1994). Even within a relatively homogeneous class of situations, youth behavior shows marked specificity. For example, youth are significantly more likely to report that they would use physical aggression in response to physical provocation by a peer, compared to relational and verbal provocation (Dirks et al. 2007b). Second, not only will situations affect the behaviors that youth enact, but the social context of a behavior will likely also influence the perceived competence of that action. For example, peers evaluate children who have hit someone who hit or pushed them first more positively than they do children who have used physical aggression unprovoked (Willis and Foster 1990). Such data hint that the same behavior, enacted in two different situations, may be perceived as more or less competent.
The other key feature that will influence judgments of competence is the identity of the person making them, or the judge. If competence is a judgment, then it is possible that the perceived competence of an action will vary depending upon who is evaluating. Very few empirical studies have examined this issue. Although inter-rater discrepancies in evaluations of youth social competence have been well-documented (see Renk and Phares 2004), the methodology of these studies does not allow for conclusions about the judge specificity of the perceived competence of specific behaviors. In general, these studies have assessed the extent to which peers, parents, and teachers agree about (1) the competence of a target child; or (2) the extent to which a target child engages in behaviors that are pre-judged to be competent (e.g., prosocial behaviors) or incompetent (e.g., aggression). In other words, they have assessed the extent to which there is agreement about whether a youth is liked or what a youth is like (see Parker and Asher 1987).
Such investigations leave unanswered the question of the extent to which important people in the social environment concur about the competence of specific behaviors. For example, do peers, parents, and teachers agree that physical aggression is an incompetent action? Work with adolescents suggests they may not. For example, one study found that in a sample of lower-income high school students, aggressive-disruptive behavior was associated positively with perceived popularity (Luthar and McMahon 1996). This finding suggests that at least some peers may view aggression as an appropriate and effective interpersonal strategy. Teachers, however, likely will not. Engaging in physical and verbal aggression are common reasons students are suspended from school (Mendez and Knoff 2003), suggesting disapproval of such behaviors among educators.
A recent study did in fact find significant differences between early adolescents’ and teachers’ judgments of the effectiveness of different responses to physical, verbal, and relational provocation by a peer (Dirks et al. 2010). In this study, youth and their teachers were presented with a number of possible responses to scenarios involving peer provocation, including physical, verbal, and relational aggression (i.e., damaging or threatening the aggressor’s social relationships), seeking an explanation for the provocation, telling the aggressor that his/her actions are unacceptable, and telling an adult. Participants rated how well each response would “work to solve the problem.” As expected, youth evaluated physically, verbally, and relationally aggressive responses to be more effective than did teachers, whereas teachers evaluated responses involving seeking an explanation to be more effective than did youth. Importantly, within the group of youth judges, some aggressive responses were viewed to be as effective as assertive strategies. For both boys and girls, ending one’s relationship with the aggressor, a strategy that could be construed as relationally aggressive (e.g., Delveaux and Daniels 2000), was deemed to be as effective as seeking an explanation or stating that the aggressor’s actions were not acceptable. Furthermore, boys also evaluated physical aggression to be as effective as these strategies.
Illuminating these inter-judge discrepancies in evaluations of behavior may provide insight into the reinforcement contingencies that exist in youth’s social environments. Ultimately, such data may aid in the development of more targeted interventions designed to improve youth’s social functioning. For example, although physical aggression may be viewed as effective by some peers, the consequences of such actions that will result from adult disapproval can be severe (e.g., suspension or expulsion from school). Furthermore, such actions are likely to cause significant harm and distress to others. This discrepancy between peer support, on the one hand, and the possible negative consequences for individuals themselves, as well as the targets of their behavior, on the other, poses a unique challenge for interventionists. In such situations, it may be helpful for clinicians to work with youth to consider who the most important judge in a given situation is. Alternatively, youth may need assistance crafting responses that are deemed to be acceptable by both peers and adults, and that do not cause harm to others around them.
Measurement of Youth Social Competence
Taken together, the empirical evidence supports increasingly the theoretical supposition that situation- and judge-level factors will play a key role in youth social competence. This more nuanced view of competence is typically not reflected in many of the instruments used to measure this construct. Researchers often assess social competence in one of two ways: sociometric strategies and behavioral approaches (i.e., nominations or rating scales). Sociometric techniques are used to determine how well-liked a child or adolescent is. A number of different approaches are used to obtain this information (see Foster et al. 1993). When working with adolescents, researchers typically use nomination procedures. Students are asked to identify the classmates that they like most and least, and these nominations form the basis of classifications such as popular (receives many liked and few disliked nominations) and rejected (receives many disliked and few liked nominations; Inderbitzen 1994). Sociometric techniques provide very valuable information concerning individuals’ popularity with their peers. The limitations of these approaches have also been widely documented. For example, sociometric analyses assess popularity with respect to a particular reference group, typically classmates at school. Adolescents often have friends in multiple contexts (e.g., at their part-time jobs, in their neighborhood); as such, sociometric procedures may not provide complete information regarding their social functioning (Inderbitzen 1994).
More generally, sociometric measurement indicates whether or not youth are liked (Parker and Asher 1987), but provides no information about what they may be doing to earn this designation (Bierman and Welsh 2000). In other words, these techniques provide data about individuals, but not about their behaviors. To address this limitation, researchers will often assess youth behavior directly. In general, this is done by having peers nominate classmates who fit specific behavioral descriptions (e.g., aggressive, avoidant; Chung and Asher 1996). Alternatively, people knowledgeable about the target individual, such as parents, teachers, or the youths themselves, may be asked to complete behavior rating scales. In general, these types of measures ask informants to rate how often youth engage in a variety of different behaviors. When working with adolescents, it is essential that rating scales assess behaviors that are relevant and important for youth of this age. The types of behaviors required to negotiate successfully the social tasks of this period, which include increased experiences with the opposite sex, as well as establishing autonomy from parents, are different than the interpersonal demands placed on younger children. Given these differences, simple adaptations of measures created for children at other developmental stages are not appropriate. This reasoning led Inderbitzen and Foster (1992) to develop the Teenage Inventory of Social Skills (TISS). This self-report measure of social skills asks youth to rate the extent to which different behavioral descriptions apply to them. Items include “I ask other [kids] to go places with me” and “I laugh at other [kids] when they make mistakes.”
The TISS, as well as other rating scales that are used with adolescents, such as the Child Behavior Checklist (CBCL; Achenbach and Rescorla 2001), provide a detailed picture of the types of behaviors in which adolescents are engaging. As such, these assessments are a valuable source of information about adolescents’ social skills. When these data are used to inform conclusions about social competence, however, two challenges emerge. First, in general, rating scales do not provide information about the social circumstances in which behaviors are embedded. Some individual items on a rating scale may include contextual information. For example, the TISS contains items such as “I tell classmates I’m sorry when I know I have hurt their feelings” and “I thank other [kids] when they have done something nice for me” (italics added). Such situational details, however, are generally lost when researchers sum up across items to form a total score (Wright et al. 2001). In doing so, researchers are treating situational variability as a source of error, rather than as potentially useful information. As a result of both the items included and the methods of scoring, then, behavioral rating scales do not account for the situation specificity of youth social behavior.
Second, this approach to measurement also does not allow for the possibility that the competence of the behaviors assessed may vary as a function of who is judging them. Rating scales assess the frequency with which youth engage in a predetermined set of behaviors. To draw conclusions about social competence from such data requires that judgments be made concerning the effectiveness of a particular action. For example, the conclusion that an adolescent who engages in aggressive behaviors frequently and assertive behaviors infrequently is not competent is predicated on the suppositions that aggressive actions are incompetent and assertive ones are effective. These blanket judgments can be problematic, as the effectiveness of these actions will vary as a function of who is evaluating the behavior. As described previously, aggressive behaviors are viewed as effective by some peers (Dirks et al. 2010). As such, deciding that youth who engage in these behaviors are not competent may be underestimating their social effectiveness with classmates.
Adding to the complexity is the reality that youth social behaviors are very nuanced, and seemingly minor differences may have a major effect on interpersonal success. For example, in a study of how early adolescents respond to provocation by a peer, Dirks et al. (2007b) found that a significant number of participants gave responses combining aggression and assertiveness. For example, many youth generated “hostilely assertive” responses, which combined verbal aggression and seeking an explanation (e.g., saying “What’s your problem?” as opposed to the less aggressive “Why did you do that?”). Previous work has treated such responses as aggressive. In two other studies examining youth responses to a variety of peer-provocation scenarios, the researchers coded responses based on the most aggressive response present (Hughes et al. 2004; Peets et al. 2007). Within this framework, a response combining verbal aggression with an assertive response would be coded only as verbal aggression. Subsequent work has demonstrated that both peers and teachers are sensitive to the difference between a verbally aggressive response and a response that combines verbal aggression and assertiveness, with both groups viewing the latter type of response as significantly more effective (Dirks et al. 2010). Thus, treating such behaviors as aggressive may underestimate youth social competence. Such findings highlight the importance of obtaining judgments of the competence of youth behavior from the relevant people in their social environment.
To summarize briefly, four sources of variability have been implicated in youth social functioning: individual, behavior, situation, and judge. For the most part, measures focus on individual- and behavior-level factors. Failing to capture situation- and judge-level characteristics, however, may result in a picture of youth social functioning that is at best incomplete, and at worst, misleading. Social competence is inherently an evaluation, and as such it is influenced by the conditions under which behaviors are enacted, as well as who is judging those behaviors. By not attending to these contextual and evaluative issues, researchers may be over- or underestimating youth social competence, as it is perceived by the people who are actually in a position to reward or punish their behavior. Furthermore, omission of situation- and judge-level factors may lead to misspecification of variability. For example, youth in lower-income environments are more likely to be targeted aggressively by peers (Dhami et al. 2005), a type of situation that will often press for aggressive responding (Dirks et al. 2007b). In the absence of contextual information, one might conclude that the problem is with the children, when in reality, the issue is that they must manage a greater number of problematic situations.
Recognizing that social competence is a multivariate evaluation influenced by characteristics of individuals, their behavior, and their social context, how best can researchers manage this complexity so that they may gain insight into the social successes and struggles of adolescents? Several investigators have suggested that social competence can be best understood with respect to key social situations or tasks (see McFall 1982; Rose and Asher 1999). Situation- or task-specific measurement provides at least two noteworthy advantages. If behaviors change as a function of situation, then the most useful and relevant information about social performance will be obtained by determining how youth respond in critical interpersonal contexts. In addition, this approach provides detailed information about when and how youth experience social difficulties. These data provide clinicians with clear targets for intervention.
If behavior is assessed with respect to key situations, it is important that we choose the right interpersonal contexts. Youth will confront an infinite number of social scenarios, but most will not yield interesting information about their social functioning. Goldfried and D’Zurilla (1969) posited that the most important situations are those that are commonly occurring, difficult to manage, and critical (i.e., performing inadequately will have negative consequences). Several research teams have set out to identify such situations in populations of adolescents. In general, all of these investigations have used the behavioral analytic approach (Goldfried and D’Zurrilla 1969). Working within this framework, investigators create an inventory of problematic situations by asking members of the population of interest to generate relevant scenarios. Freedman et al. (1978) and Gaffney and McFall (1981) developed what were perhaps the first taxonomies of problematic situations for adolescent boys (Adolescent Problems Inventory, API) and girls (Problem Inventory for Adolescent Girls, PIAG), respectively. Adolescents, as well as individuals who frequently interact with youth (e.g., parents, teachers) were asked to identify problematic situations that are relevant in the lives of teenagers. Situations not deemed by participants to be commonly occurring and difficult were not included in the final inventory. The final taxonomy covered a variety of social contexts, such as school (e.g., “A gym teacher picks on you, makes you do extra push ups”), family relationships (e.g., “Your father gets upset when you ask to borrow the car”), and academics (e.g., “You feel hopelessly lost in a geometry class”).
Employing methods similar to those utilized to create the API and PIAG, Cavell and Kelley (1992, 1994) developed the Checklist of Adolescent Problem Situations (CAPS) and the Measure of Adolescent Social Performance (MASP). On each measure, the final set of items included situations representing a number of different facets of adolescent life, including relationships with peers (e.g., “Friend ignores you,” “You were friendly to someone and now they won’t go away”), siblings (“Sibling borrows something of yours without asking,” “Sibling enjoys teasing you and making you mad,”), and parents (“Parents refuse to discuss a decision they say is final,” “Parents are too busy to take you where you want to go.”) The types of situations most relevant to adolescents change over time, as does researchers’ awareness of the kinds of problematic circumstances that arise in adolescents’ social lives. For these reasons, the CAPS and the MASP capture a number of situations not included in the earlier measures. For example, the CAPS contains several items involving relational aggression. Given the rapid changes that occur in the societal contexts in which adolescent development is embedded, it is important to update situation inventories regularly. For example, the widespread availability of personal computers and the internet has created a new set of challenging interpersonal contexts for adolescents (e.g., cyber-bullying; Ybarra and Mitchell 2004).
One domain not covered in detail by the CAPS and the MASP is relationships with opposite sex peers. Adolescence is marked by a steady transition from the almost exclusively same-sex peer groups of childhood to social networks comprised increasingly of both males and females (Grover et al. 2007). Relationships with members of the opposite sex will present adolescents with new and challenging interactions to manage, such as responding to conflict with a romantic partner and sexual harassment (Grover and Nangle 2003; Wolfe et al. 2001). Such situations were identified in the Measure of Adolescent Heterosocial Competence (MAHC; Grover et al. 2005). The researchers asked 150 adolescents to generate as many “difficult” situations with the opposite sex as they could. The final measure contained 40 situations. A number of different themes were reflected, including dating situations (e.g., asking for a date; turning a date down), initiating a friendship/relationship (e.g., calling someone that you like), and situations involving drugs and alcohol (e.g., physical contact with another person when drinking).
Although these types of situations, as well as those included in the CAPS and the MASP, are relevant for many adolescents, it is important to note that the types of problematic situations adolescents must manage will vary as a function of environmental features. A notable example of this is adolescents living in economically disadvantaged circumstances. These youth may be confronted with a number of situations – such as witnessing violence, being approached by drug dealers, or being asked to join a gang – that might not occur as frequently in more advantaged environments. When there are theoretical reasons to expect that the situations identified as commonly occurring, difficult to manage, and critical may be different for a particular group, it will be necessary to generate a new taxonomy of situations. For this reason, Farrell et al. (1998, 2006) have conducted studies aimed at identifying important situations in the lives of lower-income adolescents. These researchers conducted focus groups with lower-income, urban sixth graders to create the Interpersonal Problem Situation Inventory for Urban Adolescents (IPSIUA; Farrell et al. 1998). Participants in this study did identify situations not brought up in other investigations. For example, conflicts with teachers included having a teacher falsely accuse them or tell lies about them. These urban adolescents also described challenges associated with living with a single parent and concerns about other students bringing weapons to school. Farrell et al. (2006) conducted a similar study with economically disadvantaged seventh and eighth graders, as well as their parents and school personnel. This investigation again highlighted the unique challenges associated with living in urban poverty, and the importance of developing contextually appropriate situation taxonomies.
Situation-based inventories have been used to assess social competence in two ways. It has been suggested that simply knowing how frequently adolescents experience these situations and how difficult they find them to be will predict their social adjustment (e.g., Cavell and Kelley 1994). Adolescents who respond ineffectively to interpersonal situations are more likely to generate new social problems, and as such, will experience challenging situations at a higher rate than their more socially effective peers (see Rudolph et al. 2000). The IPSIUA assesses the frequency with which adolescents experience difficult interpersonal situations, and the CAPS measures both frequency and adolescents’ perceptions of the difficulty of social situations. Both of these measures show significant associations with other indices of social functioning, as well as psychopathology. For example, on the IPSIUA, higher frequency ratings were associated positively with anxiety, violent behavior, and drug use (Farrell et al. 1998). On the CAPS, adolescents who were unpopular (as assessed with sociometric procedures and teacher nominations) perceived situations associated with school and making friends to be more difficult and frequently occurring than did their more popular peers (Cavell and Kelley 1994).
Of course, the most detailed picture of adolescent social functioning will emerge if researchers determine not only how often youth experience situations, and how difficult they perceive these encounters to be, but also how they respond when these challenges befall them, and whether or not these responses are viewed to be effective by relevant judges. Within the behavioral-analytic framework, after situations are identified, members of the population of interest are asked to generate responses to the situations, usually by reporting what they would “say or do” if the situation happened to them. Following this, relevant judges evaluate the competence of different responses. Thus, the final measure allows researchers to assess how youth respond to specific interpersonal challenges, as well as the perceived competence of their chosen social strategies.
The developers of the API, PIAG, MASP, and MAHC all took these steps. For the most part, the researchers relied upon adult “experts” (e.g., psychologists, parents, and teachers) to evaluate the competence of responses. In addition, these measures emphasized agreement among judges when developing items. For example, Freedman et al. (1978) discarded items for which there was significant disagreement among judges regarding the competence of responses. In general, then, researchers have focused on one group of judges, and within this group, treated differences between raters as error.
Recent data suggest, however, that these inter-judge differences in evaluations of competence reflect, at least in part, valuable signal, rather than being attributable entirely to measurement-related noise. As described previously, different groups in youth’s lives may have differing opinions about the competence of a given action (Dirks et al. 2010); as such, adolescents’ perceived competence will vary systematically depending upon who is providing the ratings. Interestingly, Gaffney and McFall (1981) obtained ratings of effectiveness from both adults and teenage girls during the development of the PIAG. The results indicated that ratings of effectiveness provided by the adolescents, which reflect social competence as perceived by peers, did not discriminate between delinquent and nondelinquent girls. In other words, delinquent girls were not viewed by their age mates to be less competent than their nondelinquent peers. From an intervention perspective, knowing that adolescent girls with behavior problems struggle more from the perspective of adults than youth is critical. If peers do not perceive behaviors as problematic, or if they identify them as competent, it may be difficult to get youth to stop engaging in these actions, even if they are causing problems with adults.
For this reason, when developing situation-based measures of social competence, it will be useful to identify who the key judges for each situation are, and to maintain their unique perspectives when determining the competence of responses. In doing so, it will be important to base the selection of judges on theoretical grounds. Different judges will be relevant for different situations (e.g., Cavell and Kelley 1992); for example, for situations occurring at school, both peers and teachers are likely in a position to consequate youth behavior. It will also be important to utilize both theoretical and empirical criteria when deciding whether to combine judges’ ratings. It seems plausible, theoretically, that teachers would form a homogeneous group: they have similar professional experiences and encounter youth in similar circumstances. Empirical data point to a similar conclusion: teacher ratings of competence show very little variability (Dirks et al. 2010). The evaluations of peers are more disparate, and similar variability is likely to be evident in parent evaluations as well.
When such discrepancies are present among a class of judges, it may be necessary to break the groups down further along theoretically meaningful dimensions. For example, many researchers have noted that societal norms will influence perceptions of competence (see Chen and French 2008), suggesting the importance of considering cultural factors when identifying judges. Relatedly, previous work has also suggested that socioeconomic factors may be associated with both youth and parent perceptions of competence (e.g., Dodge et al. 1994; Luthar and McMahon 1996). Youth and parents in an urban, economically disadvantaged environment are likely to have very different perceptions of behavioral effectiveness than those living in a more affluent suburban neighborhood. In the case of parents, it may be useful to design a complementary measure that asks parents to evaluate the competence of responses given by their child. For youth, it is the judgments of their own parents (not parents, on average) that are most likely to influence their behavior. Given that it is often possible to obtain data from parents when conducting assessments with youth, it may be feasible to determine youth’s competence from the perspective of their own parents.
By developing measures that allow competence to vary as a function of the situations in which youth are acting, as well as who is evaluating their behaviors, researchers will be bringing their assessment strategies in line with their current theoretical understanding of social competence, which emphasizes that competence is an evaluative construct influenced by both situation- and judge-level factors. The recognition that evaluations of competence depend, at least in part, on characteristics of both situations and judges, is likely to pay important dividends for educators and clinicians trying to help youth experiencing social difficulties. For example, it may be important for interventionists to help youth develop strategies that will allow them to manage key social situations effectively (or, at least, in ways not perceived as grossly ineffective) from the perspective of the different groups in their social environment. When negotiating their social worlds, adolescents must consistently solve challenging multi-constraint problems: They must generate solutions to very difficult social circumstances when the key people in their lives will often not agree about the efficacy of their solutions. To the extent that measures of social competence, and ultimately, interventions targeting social competence, capture and address these complexities, researchers and clinicians will be in the best position to help adolescents succeed socially.