School Psychology Quarterly Copyright 2008 by the American Psychological Association 2008, Vol. 23, No. 4, 474–495 1045-3830/08/$12.00 DOI: 10.1037/1045-3830.23.4.474 Social Competence, Social Support, and Academic Achievement in Minority, Low-Income, Urban Elementary School Children

Maurice J. Elias Norris M. Haynes Rutgers University Southern Connecticut State University

Despite living in disadvantaged urban communities experiencing social and economic hardships, many children emerge with positive outcomes. Social-emotional compe- tence and social support were hypothesized to have strong influences on academic trajectories during the critical period of academic skill acquisition. Participants were 282 third-grade students from six elementary schools in a Northwestern urban com- munity. Beyond the importance of prior levels of academic competence, considerable variance in end-of-year academic outcomes was predicted by initial levels of academic social-emotional competence and improvements in social-emotional competence and perceived teacher support over the course of the year. Noteworthy is that findings were strongest for African-American students, but methodological caveats regarding re- search with underachieving minority youth were discussed. The findings suggest that school psychologists and others designing interventions to improve achievement of disadvantaged students should address social-emotional competencies and classroom climate, especially teacher support of students.

Keywords: social-emotional competence; prevention; resilience; academic underachievement; urban minority youth

Educators are becoming increasingly aware 2004). Research has shown that early social of the potential relationships that exist between interactions and the quality of these interactions educational achievement, social-emotional provide the basis for future developmental mile- competence, and social support in elementary stones (Vygotsky, Reiber, & Carton, 1987). As school children (e.g., Elliott, Malecki, & Dema- children mature and enter the school system, ray., 2001; Welsh, Parke, Widaman, & O’Neil, teachers become increasingly important in fa- 2001; Zins, Weissberg, Wang, & Walberg, cilitating or hindering the adjustment-to-school

Maurice J. Elias, PhD, is Professor and Undergraduate Research. He is also an Associate Clinical Professor at the Internship Coordinator, Psychology Department, Rutgers Yale University Child Study Center, the Yale Department of University, Academic Director of Rutgers’ Civic Engage- Psychology and the Bush Center. Dr. Haynes is a licensed ment and Service Education Partnerships program, Presi- Psychologist. He is a Fellow of the American Psychological dent of the Society for Community Research and Action Association and holds Diplomat Status in the International (SCRA)/Division of Community Psychology (27) of APA, Academy for Behavioral Medicine, Counseling and Psycho- Director of the Rutgers Social-Emotional Learning Lab and therapy. He is a member of several professional associations the Developing Safe and Civil Schools (DSACS) prevention and serves on several editorial boards and expert panels. Dr. initiative (www.teachSECD.com), and Founding Member Haynes is the author of over 100 peer-reviewed articles, of the Leadership Team for the Collaborative for Academic, book chapters, and monographs. He has authored or coau- Social, and Emotional Learning (www.CASEL.org). He is thored several books. Fellow of APA Divisions 16, 27, 53, 9, and 12 and a past winner of the Lela Rowland Prevention Award and the The authors wish to acknowledge with appreciation that SCRA Distinguished Contribution to Practice and Minority this article is based on dissertation research conducted by Mentoring awards. Dr. Elias lectures nationally and inter- Keisha A. Mitchell while working at the Rutgers Social- nationally, has been featured on numerous TV and radio Emotional Learning Lab. Erich Labouvie contributed exten- programs, and devotes his research and writing to the area sively to all data analyses. of emotional intelligence in children, schools, and families. Correspondence concerning this article should be ad- Norris M. Haynes, Ph.D. is Chairperson and Professor in dressed to Maurice J. Elias, PhD, Rutgers University, the Counseling and School Psychology Department at Psychology Department, 53 Avenue E, Tillett Hall, Pis- Southern Connecticut State University and Founder and cataway, NJ 08854-8040, Fax 732-445-0036. E-mail: Director of the Center for Community and School Action [email protected]

474 SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 475 process (Baker, 1999; Esposito, 1999; Munsch Children who attend school in these settings & Wampler, 1993; Schaps & Solomon, 2003). are plagued by unimaginative curricula, over- Teachers provide encouragement for attempting crowded classrooms, inadequate school facili- and persevering in challenging school activities. ties, and too few teachers who have confidence By the time a child enters grade four, his or her in them and who generally expect them to learn academic trajectory has been established (e.g., (Kozol, 2005; Wilson, 1996). According to At- Eccles, Roeser, Wigfeld, & Freedman-Doan, tar et al. (1994), these children grow up expe- 1999; Freedman-Doan, Wigfeld, Eccles, Blu- riencing neighborhood disadvantage, as evi- menfeld, Arbreton, & Harold, 2000; Welsh et denced by poverty, unemployment and under- al., 2001). If one accepts that education is a employment, limited resources, substandard socially situated process, any attempt to explain housing, and high crime rates. These environ- this phenomenon should consider elements at mental risk factors have a high probability of the individual as well as the environment levels subsequently resulting in stressor-filled psycho- (see Albee, 1982; Baker, 1999; Elias, 1987). logical environments and the development of Recent research has begun to shed light on the problem behaviors in children (Gonzales et al., role of emotion recognition and regulation and 1996; Halpern, 1990; Tolan et al., 1997; Wein- related social-emotional skills in effective so- stein, 2002). In the context of establishing social cial interaction (Saarni, 2007). These skills thus relationships, additional risk factors include can be important targets for interventions de- dysfunctional families, peer modeling of antiso- signed not only to promote positive interactions cial behavior, and constricted social networks but also support one of their major distal effects, (Haggerty, Sherrod, Garmezy & Rutter, 1994; academic achievement (Elias & Arnold, 2006). Masten, 1994; Reynolds, 1998; Wilson, 1996). While there is strong theoretical reason to have In general, children who attend school in confidence in these suppositions, empirical sup- low-income areas consistently show the lowest port across diverse populations is still emerging. academic achievement and the poorest develop- Specifically, relatively little is known about the ment of social skills (Haggerty et al., 1994; Hoff implications of these relationships for elemen- & Mitchell, 2006; NCES, 2002; Pogrow, 2006; tary school children living in low-income urban Reynolds, 1999). An examination of nationwide communities (Baker, 1999; Esposito, 1999; reading scores in 2002 revealed that fourth Roeser, Eccles, & Sameroff, 2000). This study graders in central city schools performed lower examined how the social and emotional compe- than their peers who attended urban fringe/ tencies of minority, low-income, urban school large town and rural/ small town schools. In the children in the third grade are related to their same year, students who were eligible for free/ end-of-year school outcomes. reduced-price lunch programs performed lower From an ecological perspective, children’s than students who were ineligible for such pro- school outcomes are affected most strongly by grams, with only 14% of the former group per- the neighborhoods in which they live, their fam- forming at the proficient reading level, com- ily life, the schools they attend, and the re- pared to 41% of the latter. These results, when sources that are available to them personally examined by ethnic group, also revealed that and through the school (Bronfenbrenner, 1979; white and Asian/ Pacific Islander students out- Dalton, Elias, & Wandersman, 2007). The harsh performed their black, Hispanic, and American realities of living in the inner city present chal- Indian peers. Additionally, female students lenges for many children born and raised there scored consistently higher than their male peers, (Attar, Guerra, & Tolan, 1994; Gonzales, increasing a gap that has been widening since Cauce, Friedman, & Mason, 1996; Spencer, 1998 (NCES, 2002). These problems are further 2005; Tolan, Guerra, & Montaini-Klovdahl., compounded by the reality that schools in low 1997; Wilson, 1996). Consequently, discussions income urban districts also have the lowest rat- about the education process or evaluations of ings of school climate, which have been shown programs that address educational needs must to be concomitant with problems in student consider these realities when studying children achievement and socialization (Bernstein, 1992; from high-poverty urban settings (Clark, 1991; Esposito, 1999; Haynes, Emmons, & Ben- Estell, Cairns, Farms, & Cairns, 2002; Luthar, Avie., 1997; Schaps & Solomon, 2003). Recent 1995; Ogbu, 1991; Reynolds, 1998, 1999). data suggest no meaningful changes in these 476 ELIAS AND HAYNES patterns have resulted from further implemen- p. 275). According to the Collaborative for Ac- tation of NCLB reforms (Rothstein & Jacobsen, ademic, Social, and Emotional Learning (CA- 2006). SEL), social-emotionally competent students Despite the exposure to risk, for some in- exhibit key emotional (e.g., understanding and dividuals the outcomes are not as devastating managing emotions), cognitive (e.g., problem as those experienced by others (e.g., Attar et solving and goal-setting) and behavioral (e.g., al., 1994; Baker, 1999; Haynes et al., 1997; understanding and displaying socially appropri- Haynes, Troutman, & Nwachuku, 1998; Ko- ate behavior) skills across different domains of bus & Reyes, 2000; Levitt & Levitt, 1994; home, school, and the wider community (Elias Maton et al., 1996; Reyes, Gillock, Kobus & et al., 1997). In the school domain, students at Sanchez, 2000). Wright and Masten (2005) the third grade level should possess skills that define such occurrences as examples of resil- facilitate social interaction with their peers. ience. They go on to point out that recent They should be able to listen and respond ef- thinking in the field, what they refer to as a fectively to their classmates; they should de- “second wave” of research (p. 25), takes the velop sensitivity to issues related to being in- ecological, developmental, and transactional cluded or excluded from social groups; and they approach used in this study. Resilience is seen should have the capability of regularly acting in as the outcome of an ongoing set of processes ways that are “assertive, self-calming, and co- involving the individual, family, and commu- operative” (Elias et al., 1997, p. 135). nity relational networks. Among these pro- For minority, low-income students placed at- cesses, those most relevant for research and risk, these social-emotional skills are particu- practice and most likely to account for differ- larly important to achieving school success ential effects are referred to as protective (Baker, 1999; Banks et al., 2001; Luthar, 1995; processes. Protective processes are “strengths Reyes et al., 2000). The mechanisms by which or resources associated with positive individ- these skills foster resiliency are linked to inter- ual outcomes” that operate in ongoing ways to personal processes in classrooms. Children’s help people function well in society (Dalton, ability to regulate their emotions when frus- et al., 2007, p. 245). In schools, protective trated, puzzled, or dejected, or beset with per- processes can be discussed in terms of the vasive feelings of hopelessness or anger clearly school climate, that is “the quality and con- will affect the energy they can devote to learn- sistency of interpersonal interactions within ing, even when presented with rigorous and the school community that influence chil- empirically supported academic curricula. dren’s cognitive, social-emotional, and psy- Thus, possessing social-emotional skills with chological development” (Haynes et al., fluency will allow students to better focus on 1997, p. 322). academic tasks despite bringing into school the Thus, within the larger framework of resil- many interpersonal difficulties they may be ex- ience theory and research (Goldstein & Brooks, periencing outside (as well as inside) the build- 2005), the present study focuses on two protec- ing. tive processes: (1) the social-emotional skills Social-emotional Competence and School that a student possesses to foster successful Outcomes. There is both conceptual and em- adaptation despite the prevailing conditions pirical support for positing a relationship be- (e.g., Wang & Gordon, 1994), collectively re- tween social-emotional competence and school- ferred to as social-emotional competence; and related outcomes (e.g., Dalton et al., 2007; Hag- (2) the individual’s perception that resources are gerty et al., 1994; Wentzel, 1991; Zins et al., available for help in the environment, should 2004). For example, social-emotional compe- one need them, that is, perceived social support tence skills have been shown to be influential in (e.g., Baker, 1999). the developmental trajectory of children’s lives in a study of elementary school children in Italy Social-emotional Competence (Caprara, Barbaranelli, Pastorelli, Bandura, & Zimbardo, 2000). Once children have grasped Social-emotional competence “can be viewed these skills, they are more prepared to manage in terms of life skills for adaptation to diverse their emotional responses and to control aggres- ecologies and settings” (Haggerty et al., 1994, sion. A repertoire of social-emotional skills also SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 477 serves a moral function by guiding responsible tence-promotion interventions for youth in ele- decision-making processes, and applying sanc- mentary, middle, and high schools and found tions for harmful conduct (Caprara et al., 2000). that they had a significant impact on social- In addition, social-emotional competence works emotional-cognitive skills, positive self- to promote better intellectual functioning and efficacy, school bonding, and adherence to so- significantly enhances positive outcomes for cial norms; effect sizes ranged from .21 to .41. even the most disadvantaged groups (Masten, These programs also reduced negative behavior, 1994). school violence, detention/suspension, and peer In general, positive relationships have been rejection (effect sizes .21-.28) and increased found to exist between levels of academic suc- positive behavior at school, academic achieve- cess, acceptance by peers, ratings by teachers of ment test scores, and grades (effect sizes .28- responsible classroom behaviors, and educa- .47). While all studies did not look at all vari- tional outcomes (Caprara et al., 2000; Green, ables and the effect sizes are modest at best, the Forehand, Beck, & Vosk, 1980; Luthar, 1995; clear pattern of results support existing theoret- Wentzel, 1991). Elias and Clabby (1992) fol- ical views of the mechanisms by which social- lowed elementary school children over a 2-year emotional competence might influence aca- period that included their transition into middle demic performance. That is, performing well on school. Their results showed that students who tests requires social-emotional skills such as participated in a program designed to enhance self-control, cooperative interaction, and appro- social-emotional competencies showed im- priate assertiveness and problem solving on the provements on teacher ratings of behavior, so- day of the event but also in the period of prep- ciometric indices, and self-reports about their aration/ studying for test-taking and when en- social adjustment and their ability to cope with gaged in the larger set of tasks associated with everyday life situations. academic learning and homework. Even if a Gresham and Elliott’s (1990) research sup- child possesses the requisite skills, motivation ports the claim that higher or lower levels of to use them will be related to perceptions of social-emotional competence are correspond- social support for school-related activities on ingly associated with high or poor achievement. the part of peers and teachers, as well as parents. Welsh et al. (2001) showed that academic com- petence prospectively and positively influenced Classroom-Related Perceived Social social-emotional competence from first grade to Support second grade, and from second grade to third grade. Results also indicated that social- Weissberg’s study joins others suggesting the emotional competence was reciprocally related importance of social support as a protective to academic achievement from second grade to process. Perceived social support has been ar- third grade. These results suggest that once stu- gued to be an important ingredient for healthy dents become integrated into the school system, development in childhood (e.g., Cauce, Reid, the school-related competencies become preem- Landesman, & Gonzales, 1990; Elliott et al., inent in predicting later social-emotional com- 2001; Munsch & Wampler, 1993; Rosenfeld, petence. Wentzel (1991) concluded from her Richman, & Bowen, 2000). Attachment theory own research that “socially responsible behav- argues that social connectedness is required in ior appears to mediate relations between order for children to internalize social standards achievement and both interpersonal and self- and to develop respect for social institutions regulatory aspects of social competence” (p. (Baker, 1999; Deci & Ryan, 1985). In this 1076). study, we focus on perceived social support Intervention studies at the elementary level from teachers and peers. To provide ecological have found positive effects on academic perfor- focus, we review relevant studies that focus on mance, even at 6-year follow-up (Elias, Gara, urban minority children, including some on Schuyler, Branden-Muller, & Sayette, 1991; older youth samples from which inferences Hawkins, Catalano, Kosterman, Abbott, & Hill, about construct relationships might be drawn. 1999). In the most comprehensive study to date, Perceived Teacher Support and School Out- Weissberg (2005) reported on a meta-analysis comes. Dubow and Tisak (1989) focused on of 379 studies of universal preventive/compe- third through fifth grade urban and suburban 478 ELIAS AND HAYNES students and found that student ratings of per- The next section discusses the role played by ceived social support from teachers were posi- peers. tively correlated with teacher-rated competen- Perceived Peer Support and School Out- cies and grade point average, and negatively comes. Inconsistent findings have character- correlated with teacher- and parent-rated prob- ized the research linking peer support and stu- lems. In addition, they found a significant main dent outcomes. The notion that peer support effect for teacher support, indicating that stu- may have deleterious effects on school related dents with higher levels of perceived support outcomes can be traced back to one of the from their teachers had lower levels of teacher- earliest studies of social support and adjustment rated problem behaviors. among urban, low income, minority youth. Studying children in a Northeast urban center Cauce, Felner, and Primavera (1982) found that who had participated in a Head Start program high school students with high levels of per- for their preschool year, Esposito (1999) found ceived peer support had lower levels of aca- that the teacher/student relationship for urban, demic performance as measured by grade point minority (sample: 80% African American, 18% average and school absences. However, using a Hispanic, 2% Caucasian), low-income children combination of social network variables and in kindergarten, first grade, and second grade, measures of perceived peer support, these re- was significant in predicting positive school searchers also found that among low-income, outcomes (i.e., school performance). In first and predominantly African American, urban middle second grades, the teacher/student relationship school students, reciprocated best friends, per- contributed uniquely to school adjustment, even ceived peer support, and school achievement after controlling for family resources, maternal orientation were all significant predictors of education, and kindergarten school adjustment. school competence measures. Even after con- A study of 61 third, fourth, and fifth graders trolling for individuals’ cognitive skills, social network variables significantly contributed to from a public elementary school in a large met- students’ GPA. ropolitan, southeast school district (Baker, In a study of low income, urban, African 1999) provided corroborating information. All American seventh and eighth grade students, the students and teachers in the school were Gonzales et al. (1996) found that higher levels African American and 98% of the student pop- of peer support were associated with higher ulation participated in the free or reduced-price ratings of perceived academic competence. But lunch program. Low-income African American Demaray and Malecki (2002) did not find a students most at-risk for school failure experi- similar relationship in their study of urban, His- enced alienation from school, as well as poor panic middle school students, although peer academic achievement. Students who were support was a significant, positive predictor of more satisfied with school tended to have more behavioral adjustment and had a positive impact caring teacher-student relationships (thus, they on students’ self esteem. A study of predomi- perceived more teacher support), when com- nantly minority ninth graders in a Connecticut pared to others who were not satisfied with inner-city public school revealed that “adoles- school. Students who experienced less satisfac- cents’ early academic success, and early peer tion with school were three times less likely to perceptions of their leadership qualities, were receive help when they asked for it, compared each found to be linked with improvements in to students who were satisfied with school. Ad- the children’s classroom behaviors as rated by ditionally, these students had lower levels of teachers” (Luthar, 1995, p. 424). The results competence and adjustment to school when also showed that over the course of the school compared to their more satisfied counterparts. year, students who were initially rated by their In fact, the less satisfied students received twice peers as friendly showed the greatest declines in as many behavioral reprimands than the satis- academic achievement, as well as “peer-rated fied students. qualities of leadership and dependability” (p. The studies presented show the importance of 425). In addition, those who were initially rated the teacher-student relationship in determining as responsible leaders were later rated among school outcomes. However, teachers are not the the lowest for being gregarious and sociable. It only source of supportive influence in schools. appears that among adolescents in the context of SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 479

Luthar’s study, peer reputations and school suc- The Present Research cess and socially competent behavior directly oppose each other. This is a position supported Based on the literature reviewed, it was hy- in other research with at-risk populations (e.g., pothesized that direct relationships exist be- Cauce et al., 1982; Gonzales et al., 1996; Maton tween school outcomes and social-emotional et al., 1996; Ogbu, 1991). competence, perceived teacher support, and In a study of 92 first graders (53 boys, 39 perceived peer support among urban minority girls, 99% African American) from two elemen- third graders. Further, as students become more tary schools in an inner-city area of a major social-emotionally competent, their ability to Southeastern city, aggressive children low on seek (or create) supportive relationships also social-emotional competence tended to associ- should increase. This can be explored by exam- ate with other aggressive children who comple- ining ways in which changes in competence ment their problem behavior. This finding im- affect changes in the perception of classroom- plies that students who are perceived as social- related perceived support, and thereby influence emotionally incompetent may still have the school performance. Figure 1 contains a de- ability to establish strong peer support in the tailed depiction of the specific pathways that are classroom, though they may not have teacher hypothesized, each noted by a number. support. Further, clusters of academically suc- cessful boys and girls were almost as high or Method even higher in aggression than groups of stu- dents considered average in competence, behav- Participants and Setting iorally at risk, or academically lagging. Boys and girls who were aggressive and academically Data for this study were collected as part of successful were also able to establish peer rela- the assessment of a larger longitudinal action- tions and attain positive school outcomes, find- research project based on an integration of the ings that generally stand in contradiction to evidence-based Social Decision Making/Social results from nonminority populations (Estell et Problem Solving curriculum (Elias & Bruene- al., 2002). Butler, 2005) with “Talking with TJ,” (Hall- The variation in results suggests that what- mark Corporation, 1994), a video-based series ever relationships might exist between per- focused on teamwork skills. The resulting cur- ceived peer support and academic success must riculum (Rutgers Social & Emotional Labora- be viewed in terms of both development and tory, 1998) was tailored toward preventing vi- environmental context. The dynamic of individ- olence and promoting social-emotional compe- ual classrooms and schools make the likelihood tence among urban elementary school children. of a strong effect of perceived peer support Data for 282 third grade students (46% boys, unlikely. However, especially at the early ele- 54% girls) from six elementary schools in a mentary level, when it is less likely that there Northeastern urban community were analyzed will be strong organized peer pressure against in this study. Of these, 172 were Black/African school academic norms, perceived peer support American, 27 were Hispanic/Latino, 1 Native can be expected to serve as a positive protective American, 2 Caucasian, and 3 Other. Seventy- feature of urban classroom settings (Levitt & seven students were ethnically unidentified. Our Levitt, 1994; Luthar, 1995; Roeser et al., 2000). experience in the district and reports of demo- Additionally, the research literature supports graphic studies in this and related communities viewing social-emotional competence and sup- suggest that it is most likely that these students port as interrelated: “Personal characteristics were of mixed ethnicity whose status was not such as social competence can lead an individ- recorded due to lack of clarity on the part of ual to develop and access effective social sup- parents/guardians or conflicting information port across relationships” (Dubow & Ullman, contained in school records. Note that while the 1989, p. 62). It is therefore reasonable to expect community demographic includes about 20% that support in the classroom and students’ so- Caucasian population, the study sample is rep- cial-emotional competence would impact posi- resentative of the full school population. tively on school outcomes for young, minority, Socioeconomic information about the sample low-income, urban children. revealed that approximately 60% received free 480 ELIAS AND HAYNES

Time 1 (T1) Time 2 (T2)

Competence Academic School Outcomes

Change in Competence

Teacher Support Change in Teacher Support

Non Academic School Outcomes Peer Support Change in Peer Support Gender, Ethnicity

Academic School Outcomes

Non-Academic School Outcomes

Figure 1. Hypothesized Model of Social Competence, Perceived Social Support, and School Outcomes or subsidized lunch, although more who were scale and peer nominations of social preference, eligible did not apply. In the community, the and moderate correlations with self-esteem. crime index was twice the national average, and Due to constraints in the time allowed for indices of environmental health were in the in-school assessment, the current study used the lower 20% nationally. Over half of all families shortened form of the SOCSS, which was de- were single, headed by females. More gener- veloped using the 3 items with the highest factor ally, unemployment rates in the surrounding loadings for Teacher and Peer Support factors community were 9.2%. and contains high correlations with the full- items subscale scores (Dubow, Edwards, & Ip- Measures polito, 1997). Checking the internal consistency for the current sample suggested that the reli- Social Support. The Survey of Children’s ability of the teacher and peer support scales Social Support (SOCSS), developed by Dubow could be improved by deleting an item from and Ullman (1989), is one of the few multidi- each subscale. After making these statistically mensional measures of perceived social support recommended changes, the internal consistency that has been used with urban elementary school estimates of reliability were .75 and .67 teacher children from third grade to fifth grade. The and peer support subscales, respectively. original version contains 31 items correspond- Items on the modified teacher support sub- ing to perceived family, peer and teacher sup- scale include: (a) “Do your teachers make you port factors. Dubow and Ullman (1989) re- feel important?” and (b) “Do you think your ported Cronbach’s alpha for subscales ranging teachers care about you?”. The modified peer from .78-.83 and test–retest reliability (3 to 4 support subscale items, which more accurately weeks) ranging from .66-.73. Evidence for va- reflect perceived protection from peer threat, lidity includes moderate to high correlations are: (a) “Do you get picked on and teased by with the corresponding subscales of Harter’s your friends?” and (b) “Do you feel left out by (1985) social support scale for children, signif- your friends?”. Negatively scored items were icant correlations between the peer support sub- recoded to reflect a positive support dimension. SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 481

All subscales used Likert-type responses rang- Teachers were also asked to complete the ing from never (1) to always (5), with a mid- SSRS for each student in their class. A demo- point of sometimes (3). The final perceived so- graphic information sheet was attached to cial support scores were calculated by adding each social-emotional competence measure, the items for each subscale and obtaining the and teachers recorded data about the child’s mean score. age and gender. Teachers were paid at the Social-emotional Competence. The Social prevailing hourly rate to compensate them for Skills Rating Scale is an instrument with well their time. researched reliability and validity (SSRS; Gre- In May of the following year (9 months sham & Elliot, 1990) and it was used to mea- later), research assistants returned to the schools sure 3 components of competence: cooperation, and administered the same instruments as had assertion, and self-control. Competence scores been given previously. In June, at the end of were calculated by summing the responses on the elementary school year, students’ aca- each subscale and obtaining the mean score demic and nonacademic information was ob- across subscales. In total, 28 items were used to tained from their report cards. Teachers were obtain the competence score. If participants had not aware that the latter information would be more than 3 items missing, they were excluded collected. from the analyses. Cronbach’s alpha was .85. School Outcomes. School outcomes were measured using report card grades for reading Results and mathematics at the end of the first and the fourth marking periods, that is, the beginning Summary Statistics and Data Reduction and the end of the school year. Mean scores for reading and mathematics were averaged to ob- The means and standard deviations for the tain an overall indicator of academic perfor- study variables are presented in Table 1. mance. The scale used was the equivalent Table 2 provides comparisons of means by of 4.0 ϭ A, 3.0 ϭ B, 2.0 ϭ C, 1.0 ϭ D, and 0 ϭ gender and ethnic group status respectively, F. Students with any academic scores missing along with Cohen’s d as a measure of effect were excluded from the analyses. Attendance size. Results indicate significant gender differ- across corresponding periods was also used as a ences for T1 academic performance, T1 social- nonacademic indicator of positive school out- emotional competence, and T1 peer support. come. Higher scores on this variable indicate a Girls generally scored higher on academic out- greater number of absences. comes, social-emotional competence, and peer Demographic Information. Demographic support. Boys reported numerically higher lev- information was collected about students’ gen- els of perceived teacher support than girls, but der, ethnicity, and grade level. Gender and eth- there was no statistical significance found ( p ϭ nicity data are presented in Table 2. .16). Significant group differences were found by Procedure ethnic group status (African American or non- African American; the 77 ethnically unidenti- As stated above, the data for this study were fied students were excluded from these ethnicity collected as part of a longitudinal project aimed analyses, although subsequent analyses includ- at preventing youth violence. At the beginning ing them as non-African Americans yielded of the school year, trained undergraduate re- similar results). African Americans were found search assistants were sent into the schools to to have significantly lower levels of social- administer preassessments. As part of a larger emotional competence and fewer absences than assessment packet, Grade 3 elementary school their non-African American counterparts. While students were given instruments to measure no d level exceeded Cohen’s criterion for a their levels of perceived support. Research as- medium effect size, the pattern of findings sistants read each item aloud to the students as nonetheless suggested that gender and ethnic they completed the survey, and paused to allow group status be included in subsequent analyses time for appropriate responses. (Valentine & Cooper, 2003). 482 ELIAS AND HAYNES

Table 1 Descriptive Statistics for Main Study Variables N Mean SD Minimum Maximum Time 1 variables: Competence 222 1.37 .42 .28 2.00 Academic 272 2.55 .97 .00 4.00 Non-academic 278 3.48 4.00 .00 38.00 Teacher support 255 2.50 1.34 1.00 5.00 Peer support 255 1.75 1.14 1.00 5.00 Time 2 variables: Competence 227 1.35 .46 .38 2.00 Academic 282 2.50 1.07 .00 4.25 Non-academic 282 3.86 3.75 .00 21.00 Teacher support 227 1.87 1.33 1.00 5.00 Peer support 227 2.37 1.42 1.00 5.00 Change variables: Competence 198 Ϫ.02 .32 Ϫ1.17 .90 Teacher support 207 .08 1.34 Ϫ4.00 4.00 Peer support 207 Ϫ.20 1.34 Ϫ4.00 4.00

Structural Equation Modeling tion context, in the form of cultural and gender- related influences. The Mplus program (Muthe´n “Structural equation modeling (SEM) is a & Muthe´n, 2005) was used to test the latent comprehensive statistical approach to testing conceptual model (in this case, also the mea- hypotheses about relations among observed and surement model) from the program-generated latent variables” (Hoyle, 1995, p. 1). Due to its covariance matrix. It should be noted that the ability to simultaneously evaluate the pathways imputation procedure uses the SEM algorithm specified in the experimental model presented and assumes that data are missing at random in Figure 1, and to impute missing data (a (Dempster, Laird, & Rubin, 1977; Schafer & common feature of large community-based as- Graham, 2002). Further, the use of change sessments), SEM procedures were used to test scores in SEM with their baseline scores is for hypothesized relationships. In the present acceptable, provided that there is no multicol- study, latent factors were social competence, linearity and that the change score represents a academic performance, students’ perceptions of conceptually distinct variable from the status of social support, and the influences of socializa- the individual being assessed at Time 1 or 2; the

Table 2 Comparisons of Means of Predictive Variables by Gender and Ethnicity African Non-African Males Females Effect American American Effect Size (33 ؍ n) (172 ؍ Variable (Range) Mean (SD) Mean (SD) Size (n Competence (pre) (0.28–2) 1.24 (.44) 1.44 (.40) .48** 1.32 (.43) 1.50 (.40) .43* Academic (pre) (0 – 4) 2.40 (.99) 2.70 (.96) .31* 2.52 (.96) 2.50 (1.13) .02 Non-academic (pre) (0–38) 3.27 (3.37) 3.44 (4.72) .11 3.10 (2.92) 5.39 (8.05) .44** Peer Support (pre) (1–5) 2.24 (1.26) 2.69 (1.35) .35** 2.52 (1.32) 2.17 (1.33) .26 Teacher Support (pre) (1–5) 1.91 (1.24) 1.69 (1.08) .19 1.79 (1.11) 1.44 (.86) .36 * p Ͻ .05 ** p Ͻ .01 SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 483 present circumstances meet both of these as- found to be significantly associated with gender sumptions (Hoyle, 1995). (r ϭϪ.169, p Ͻ .05), end of the year (T2) academic performance (r ϭϪ.130, p Ͻ .05), Correlations and T1 perceived teacher support (r ϭ .143, p Ͻ .05). The results also indicated significant asso- As an initial step, the program-generated cor- ciations between perceived teacher support and relation matrix was analyzed. Pearson’s corre- T1 (beginning of the school year) social- lation coefficients are presented in Table 3. Hy- emotional competence (r ϭϪ.187, p Ͻ .01). pothesized relationships were generally sup- ported by the intercorrelation results. Social- Structural Equation Models emotional competence at T1 was significantly related to T1 academic performance (r ϭ .589, Three models were tested. The first model p Ͻ .01) and T2 academic performance (r ϭ tested was the hypothesized model presented in .550, p Ͻ .01), T1 nonacademic school outcome Figure 1 above. The results of the initial SEM (r ϭϪ.171, p Ͻ .05), gender (r ϭϪ.234, p Ͻ procedure are presented in Table 4. The initial .01), and ethnic group status (r ϭϪ.158, p Ͻ model solution includes the results of nonsig- .05). Social-emotional competence was not sig- nificant as well as significant hypothesized nificantly related to non academic school out- pathways. As hypothesized, competence, come at the end of the school year. change in competence, and change in teacher In terms of classroom-related perceived sup- support, are important in predicting academic port, T1 perceived peer-related support was outcomes, despite contributions made by earlier

Table 3 Intercorrelation of Main Study Variables 123456789101112 1. Gender (M ϭ 1; F ϭ 0) — 2. Ethnicity (AA ϭ 1; NonAA ϭ 0) Ϫ.116* — 3. Academic (pre) Ϫ.142** Ϫ.021 — 4. Academic (post) Ϫ.135* Ϫ.056 .758** — 5. Non-academic (pre) Ϫ.023 Ϫ.126* Ϫ.184** Ϫ.175** — 6. Non-academic (post) .045 .093 Ϫ.113* Ϫ.124* .419** — 7. Competence (pre) Ϫ.234** Ϫ.126* .608** .564** Ϫ.200** Ϫ.170** — 8. Teacher support (pre) .086 .119* Ϫ.008 Ϫ.100* Ϫ.059 .122* Ϫ.156** — 9. Peer support (pre) Ϫ.171** Ϫ.001 Ϫ.005 Ϫ.132* Ϫ.114* .104* .025 .074 — 10. Change in competence Ϫ.001 .042 Ϫ.005 .137* .039 .081 Ϫ.272** .064 .030 — 11. Change in teacher support Ϫ.119* .054 .117* .157** .047 Ϫ.015 .224** Ϫ.486** .007 Ϫ.224** — 12. Change in peer support .034 Ϫ.069 Ϫ.141** Ϫ.064 Ϫ.033 .000 .053 .003 Ϫ.433** Ϫ.068 Ϫ.122* — * p Ͻ .05 ** p Ͻ .01 N ϭ 282 Notes. AA ϭ African American, NonAA ϭ Others Non-academic ϭ number of absences 484 ELIAS AND HAYNES

Table 4 Initial Structural Equation Model Results[1] b SE t ␤ R2 Change in competence on .117 Gender (M ϭ 1; F ϭ 0) Ϫ.045 .047 Ϫ.969 Ϫ.072 Ethnicity (AA ϭ 1; NonAA ϭ 0) Ϫ.021 .063 Ϫ.330 Ϫ.025 Competence (pre) Ϫ.330 .067 Ϫ4.940* Ϫ.445 Teacher support (pre) Ϫ.002 .019 Ϫ.096 Ϫ.007 Peer support (pre) Ϫ.002 .016 Ϫ.112 Ϫ.008 Academic (pre) .082 .027 3.036* .254 Non-academic (pre) .000 .005 Ϫ.062 Ϫ.004 Change in teacher support on .302 Change in competence Ϫ.815 .287 Ϫ2.843* Ϫ.191 Gender (M ϭ 1; F ϭ 0) Ϫ.086 .179 Ϫ.477 Ϫ.032 Ethnicity (AA ϭ 1; NonAA ϭ 0) .450 .256 1.759 .124 Competence (pre) .118 .289 .409 .037 Teacher support (pre) Ϫ.568 .073 Ϫ7.784* Ϫ.480 Peer support (pre) .067 .061 1.095 .067 Academic (pre) .094 .108 .872 .068 Non-academic (pre) .015 .021 .724 .045 Change in peer support on .246 Change in competence Ϫ.044 .297 Ϫ.149 Ϫ.010 Gender (M ϭ 1; F ϭ 0) Ϫ.147 .185 Ϫ.795 Ϫ.055 Ethnicity (AA ϭ 1; NonAA ϭ 0) Ϫ.098 .266 Ϫ.371 Ϫ.027 Competence (pre) .526 .299 1.760 .167 Teacher support (pre) .111 .075 1.477 .095 Peer support (pre) Ϫ.451 .063 Ϫ7.174* Ϫ.453 Academic (pre) Ϫ.380 .111 Ϫ3.417* Ϫ.278 Non-academic (pre) Ϫ.006 .022 Ϫ.265 Ϫ.017 Academic (post) on .647 Change in competence .784 .153 5.120* .231 Change in teacher support .092 .040 2.299* .116 Change in peer support .000 .039 .000 .000 Gender (M ϭ 1; F ϭ 0) Ϫ.011 .091 Ϫ.116 Ϫ.005 Ethnicity (AA ϭ 1; NonAA ϭ 0) Ϫ.142 .131 Ϫ1.091 Ϫ.049 Competence (pre) .575 .147 3.908* .228 Teacher support (pre) .042 .044 .940 .044 Peer support (pre) .058 .036 Ϫ1.594 Ϫ.073 Academic (pre) .654 .056 11.655* .598 Non-academic (pre) Ϫ.010 .011 Ϫ.969 Ϫ.038 Non-academic (post) on .241 Change in competence .744 .796 .935 .062 Change in teacher support .195 .204 .959 .070 Change in peer support .212 .197 1.078 .076 Gender (M ϭ 1; F ϭ 0) .570 .459 1.240 .076 Ethnicity (AA ϭ 1; NonAA ϭ 0) 1.505 .662 2.274* .148 Competence (pre) Ϫ.148 .755 Ϫ.197 Ϫ.017 Teacher support (pre) .507 .224 2.263* .154 Peer support (pre) .215 .183 1.174 .077 Academic (pre) .027 .286 .096 .007 Non-academic (pre) .425 .053 8.019* .453 N ϭ 282 Notes. AA ϭ African American, NonAA ϭ Others Non-academic ϭ number of absences Results presented in this table can be interpreted like those obtained in regression analyses. For each outcome measure, R2 refers to the percent of variance accounted for by the set of independent variables listed below it; The path coefficient is denoted by (b), the standard error of the estimate (S.E.), the test of significance (t 2.00, p Ͻ.05, denoted by asterisk), and the standardized path coefficient (␤) shows the relative predictive power of the variables. SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 485 academic performance. Time 2 nonacademic ceptability of the model to the data (Hoyle, outcomes, that is, absences, are predicted by 1995). These included a chi-square/degrees-of- perceived teacher support, previous absences, freedom ratio (␹2/ df) of .751, a comparative fit and ethnic group status. The parameter esti- index (CFI) of 1.000, and a TLI (Tucker-Lewis) mates (b) are presented for each significant of 1.017. The root mean square error of approx- pathway. Overall, the model explains almost imation (RMSEA) was 0.000 (good model fit is 65% of the total variance in academic perfor- indicated by a RMSEA that is less than .04), and mance and 24% of the total variance in the the lower and upper limits of the 90% confi- number of absences. Once the structural equa- dence interval were .000 and .033, respectively, tion modeling program evaluates specified path- p ϭ .994. The ␹2(23, N ϭ 282) for the final ways, results are given that suggest ways of model was 17.269 ( p ϭ .796). The significant improving the model to best describe the data. model parameters are presented in Figure 2. Modifications are usually related to “freeing Like the initial model, the trimmed model ex- parameters that formerly were fixed or fixing plains almost 64% of the variance in academic parameters that were formerly free” (Hoyle, performance, and almost 23% of the variance in 1995, pp. 8). Based on the modification indices, absences, but in a more parsimonious manner. the model was trimmed to better fit the data by Interpretation of the model was guided by eliminating nonsignificant pathways. Figure 2 considerations outlined by Garson (2007). In presents the final model. (All students are in- general, the best fitting model (see Figure 2) cluded; for this analysis, the 77 ethnically un- was slightly different from the proposed model. identified students were classified as non- These results suggest that not all the hypothe- African American; findings did not change sized pathways are necessary, and that the ex- when they were excluded; also, there was no planation of academic and nonacademic out- difference in the pattern of construct relation- comes through social-emotional competence ships by gender.) and perceived classroom-related support Goodness-of-fit indices denote the overall ac- sources is less complex than suggested by the

Time 1 (T1) Time2(T2)

b = .619 Competence Academic School Outcomes b = -.309 b = .777 Change in Competence

b = .070 Ethnicity b = .440 b = -.823

Teacher b = -.585 Change in b = .659 Support Teacher Support

b = .079 Non Academic School Outcomes b = .559 Peer Support Change in Peer Support b = -.434

b = -.380

Academic School b = 1.660 Outcomes

b = .467

Non-Academic School Outcomes b = .434

Figure 2. Trimmed Model of Social Support and Social Competence on School Outcomes 486 ELIAS AND HAYNES original model. At the same time, the residual TLI (Tucker-Lewis) of 1.016. The root mean variance unexplained by the model suggests square error of approximation (RMSEA) that additional factors not included here may was 0.000, and the lower and upper limits of the assist in further explaining school-related out- 90% confidence interval were .000 and .056, comes. respectively, p ϭ .922. The ␹2(20, N ϭ 172) for One significant pathway required further in- the final model was 17.319 ( p ϭ .632). The vestigation. In the trimmed model (see Figure final results of the SEM procedure for the Af- 2), ethnic group status was found to be impor- rican American subsample are presented in Fig- tant in determining the level of change in per- ure 3. Not surprisingly, the model is similar to ceived teacher support, and also important in the model for the entire sample because African the students’ absence from school. Therefore, to Americans are the dominant ethnic group rep- ascertain the accuracy of the model as it relates resented in the sample. Collectively, previous to African American students, another SEM social-emotional competence, change in compe- procedure was conducted. Summary statistics tence, and previous academic performance ac- for the African American students are presented count for 64% of the variance in academic per- in Table 5. formance. Only previous absences are signifi- The goodness-of-fit indices yielded a chi- cant, determining almost 21% of the variance in square/degrees-of-freedom ratio (␹2/ df) of .866, the number of absences at the end of the school a comparative fit index (CFI) of 1.000, and a year.

Table 5 Descriptive Statistics: Correlations, Means, SDs of Main Study Variables for African American Subsample 1 2345678 91011 1. Gender (M ϭ 1; F ϭ 0) — 2. Academic (pre) Ϫ.144** — 3. Academic (post) Ϫ.157** .781** — 4. Non- academic (pre) .018 Ϫ.179** Ϫ.137* — 5. Non- academic (post) .025 Ϫ.016 Ϫ.034 .455** — 6. Competence (pre) Ϫ.312** .579** .549** Ϫ.179** Ϫ.101* — 7. Teacher support (pre) .122* Ϫ.146** Ϫ.020 Ϫ.105* .028 Ϫ.146** — 8. Peer support (pre) Ϫ.181** Ϫ.051 Ϫ.111* .099 .120* Ϫ.051 .159** — 9. Change in competence .026 .025 .116* .051 .088 Ϫ.258** .070 Ϫ.046 — 10. Change in teacher support Ϫ.202** .140* .155** .022 .036 .291** Ϫ.465** Ϫ.028 Ϫ.260** — 11. Change in peer support Ϫ.046 Ϫ.136* Ϫ.100* Ϫ.103* Ϫ.122* .055 Ϫ.056 Ϫ.390** Ϫ.054 Ϫ.056 — Means .449 2.523 2.441 3.099 3.686 1.327 1.792 2.518 Ϫ.027 .033 Ϫ.150 SD .498 .956 1.072 2.909 3.392 .427 1.109 1.318 .311 1.339 1.261 * p Ͻ .05 ** p Ͻ .01 N ϭ 172 SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 487

Time1(T1) Time2(T2)

b = .502 Academic School Competence Outcomes b = .510 b = -.296 Change in competence

b = -1.068

Teacher b = .736 Support b = -.553 Change in Teacher Support

b = .083 Non Academic School Outcomes Peer Support Change in b = -.396 Peer Support

b = -.201

Academic School Outcomes

Non-Academic b = .529 School Outcomes

Figure 3. Trimmed Model of Social Support and Social Competence on School Outcomes for African American Subsample

Discussion disadvantage communities (e.g., Baker, 1999; Caprara et al., 2000; Zins et al., 2004). Beyond The primary goal of this study was to exam- the importance of prior levels of academic com- ine the relationships that exist among social- petence, considerable variance in end-of-year emotional competence, perceived support at academic outcomes is predicted by initial levels school (particularly in the classroom), and of social-emotional competence and improve- school outcomes for minority children who live ments in social-emotional competence and per- in low-income, urban communities. This was ceived teacher support over the course of the operationalized by exploring the validity of the year. In the context of the current findings, the mediational model represented in Figure 1. The relationship of academic pretest scores to aca- discussion reviews the relationships among the study variables, examines their stability demic posttests shows that prior educational throughout the school year, and analyzes gender history is highly predictive but not determinis- and ethnic group differences. tically so. Such a finding has particularly im- portant implications in third grade, which is Empirical Relationships Among the Study toward the end of the critical period of acquisi- Variables tion of literacy skills. Many educators view this as a nodal developmental transition point after The hypothesized relationship between pre- which risk for low academic achievement in- vious social-emotional competence and school creases considerably if children are still far be- outcomes was mostly supported. In agreement hind in reading or do not move into fourth grade with the current literature, previous social- with strong, significant momentum. From the emotional competence affects academic perfor- point of view of resilience theory, the findings mance of children in at-risk, high neighborhood suggest that social-emotional skills and support 488 ELIAS AND HAYNES combine to serve as protective processes, but important in students’ end of year overall school not in ways that are uniform across cultural and performance. This finding is in opposition to the ethnic groups or sources of support. findings of previous research, although these The relationship between previous social- generally involved older children. Timing may emotional competence and nonacademic indi- be an important factor; at the beginning of the ces, that is, the number of absences, was not school year, the picture of school safety may not supported. This lack of a relationship is perhaps be as coherent or consistent as it emerges later a function of the nonacademic indicator chosen, in the school year. Further, peer support in or a function of the developmental implications high-risk, predominantly minority, and urban for these third graders. It is possible that another populations may begin to be less consistently nonacademic indicator, for example, tardiness, aligned with meeting adult goals for school with an implicit requirement for more basic achievement (Ogbu, 1991). Younger students social-emotional skills, would have been a more may feel that their peers are not capable of appropriate indicator. However, third graders in providing assistance in school success. It also this population may or may not have much may be the case that they have no reason to control regarding their attendance and their late- believe that their peers can affect their learning ness in arriving at school. As suggested by status, nor affect their attendance at school. This Gonzales et al. (1996), problems at home (in- reasoning may correspond with Feiring and cluding a lack of money and frequent mobility), Lewis’ (1991) findings that support networks in or in the surrounding community could be more the earlier stages of development may be ben- influential than social-emotional competence in eficial in developing social-emotional skills, but determining a student’s attendance at school. not as essential in school performance. Perceived increases in teacher support were found to be modestly associated with academic Stability of Relationships: Social- performance as measured at the end of the emotional Competence and Perceived school year. Perceived teacher support in the Social Support first month of the school year did not generally influence students’ end of the year academic SEM analyses showed that prior social- performance but was influential in the third emotional competence and change in compe- graders’ later degree of absence from school. tence were found to significantly affect later One reason for this disparity could be that for academic performance. It should be noted that academic advancement, students need more the level of change in competence was affected time to adjust to or to develop a stable relation- by previous academic performance and previ- ship with the new teacher; hence relationship ous competence scores. This suggests that the quality is difficult to assess at the initial stages ability of third graders to adapt in this setting is of such a relationship. However, it appears that affected by their previous experiences of school not as much time is needed for the perceived success, as well as their past experiences in quality of the teacher-student relationship to adapting to changes in the social and ecological impact on the number of absences that a student climate of the school. It supports Reyes et al.’s will have at the end of the school year. Perhaps, (2000) position that transitional support is in- as is suggested by Reyes et al. (2000), students’ fluential in determining later school perfor- perception of support at the start of their school mance for at-risk students. year is influential on school outcomes as the Although social-emotional competence year unfolds. If students enter third grade with scores were relatively stable overall from the an already low level of perceived teacher sup- beginning to the end of the school year, the port, and perceive the same (or less) support relatively large range of negative and positive than they had in second grade, it is possible that change (Ϫ1.17 and .90, respectively) indicates the immediate evaluation of support quickly that some students experienced a significant re- impacts on current and later school attendance. duction in their capacity to adapt to their social It was surprising to discover that initial per- environment, while others increased in their ceptions of peer threat—which is the more ac- ability to handle changing social requirements. curate characterization of the peer support vari- Clearly, there are important mediating factors able as actually assessed in this study—were not that can either help steer urban youth toward SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 489 positive outcomes or foster movement along ence was also important in the change in the negative trajectories. perception of teacher support (see Table 3), as Students’ mean perception of a supportive seen in Figure 2. In the urban context of the teacher-student relationship declined from 2.50 present study, non-African American students (SD ϭ 1.34) to 1.87 (SD ϭ 1.33). This suggests are the minority. These students tended to ex- that at the beginning of the school year, students hibit higher levels of competence than their generally perceived the teacher as more sup- African American counterparts, yet also had a portive than at the end of the school year. How- higher rate of absenteeism. These results are ever, in agreement with Baker (1999), less com- consistent with the view that when one is part of petent students perceive more support from a minority group in an environment where eth- their teachers when compared to their more nic/cultural tensions exist, it is not unusual to competent counterparts. This suggests that experience higher levels of environmental stress teachers do what they have been trained to do, and also perceive lower levels of support (e.g., that is, they identify their weakest students and Barker, 1999; Kobus & Reyes, 2000; Maton et provide extra attention in order to strengthen al., 1996; Munsch & Wampler, 1993). Under them. At the high school level, it seems that such circumstances, higher absentee rates students are aware of this scenario and are not would not be surprising. With regard to higher upset by it (Reyes et al., 2000). However, with competence ratings, Kozol (2005) has pointed younger children, these results suggest that de- out that African Americans have been stereo- spite the benefits to the weaker students in the typed as being very energetic, gregarious, and class, there are costs associated with the other vociferous (i.e., distinctly expressive). These students perceiving less support available from characteristics may be seen as appealing in their most valuable resource in the classroom, many contexts but may not be viewed favorably that is, their teacher. Interestingly, students’ ini- in an academically pressured educational con- tial perception of teacher support did not impact text. If the non-African American students as a on end-of-year academic performance, but group were distinctly less active and less ex- change in the perception of teacher support did. pressive (and, by extension, less disruptive of the classroom routines), then it follows that Group Differences in Social-emotional their behaviors will more likely match those Competence and Perceived Social required for social-emotionally competent chil- Support, and Their Effects on School dren in elementary school (i.e., the skills mea- Outcomes sured as being competent by the SSRS). Such possibilities call upon researchers and educators In general, the gender differences in this to be more cognizant of the ecological behav- study (see Table 2) did not seem important in ioral norms in school settings and the impact determining overall school performance when these may have on both children’s behavior and other factors were considered (see Table 3). The on adults’ ratings of their behavior. disappearance of differences based on gender in overall school outcomes should not be inter- Limitations and Future Research preted to mean that gender is an unimportant factor when considering students in this age The present findings illuminate a number of group in similar at-risk environments. Rather, complexities in the relationship between com- the results suggest that when other demographic petence and perceived support among young variables are introduced in an environment children in an urban minority population. How- where they are significant, the gender effect is ever, caution should be exercised in interpreting negligible and secondary. In addition, these results from single studies. A few limitations children may still be relatively early in the pro- may affect the analysis and interpretation of the cess of being socialized into gender-based pat- data presented in this study. The first limitation terns of school behavior. is that any changes that occurred during the On the other hand, ethnic group differences school year cannot be solely attributed to devel- were found to be important when considering opmental outcomes. The students that were the number of absences and previous levels of studied were part of a larger longitudinal project competence (see Tables 2 and 3). This differ- designed to teach social and emotional skills to 490 ELIAS AND HAYNES children. Not only were they exposed to a cur- Elliot, 1990) is widely used and has strong riculum designed to teach respect for diversity, empirical support in the literature. teamwork, and emotion-identification, but they The third and last methodological limitation were also exposed to various school-based pro- concerns an inherent problem with structural grams designed to enhance their social- equation modeling (SEM) procedures. These pro- emotional and academic competencies. Never- cedures use an iterative process to determine sig- theless, this was a constant across all groups of nificant pathways based on the variables that are children. More to the point, social-emotional included in the model. However, should other and character-related interventions have now variables be included/excluded, or the direction of become a common component of school curric- prospective relationships be hypothesized and ula and are required as part of the instructional tested in a different way, the results of significant standards in a number of states, including Illi- pathways may also change. For example, in this nois, Iowa, Massachusetts, New York, Ohio, study it was found that competence is important in Rhode Island, and the state in which this study determining later academic achievements. How- took place, New Jersey (Cohen, 2006; Kress, ever, it is also possible that the teachers’ prior Norris, Schoenholz, Elias, & Seigle, 2004). knowledge of the children’s academic perfor- As with other research with predominantly mance may influence their ratings for competence, minority, at-risk students, there were method- and their prior knowledge of children’s compe- ological concerns (Cauce, Hannan, & Sargeant, tence might influence their view of children’s ac- 1992; Reyes et al., 2000). The first methodolog- ademic potential. The best solution for disentan- ical limitation relates to the operationalization gling this potential method variance would be to of perceived peer and teacher support. The have multiple sources of data for both sets of modified measure was created based on select- variables. A longer measurement time frame ing items with the highest factor loading from would also be desirable. Regardless, even longi- the original Dubow and Ullman (1989) study. tudinal studies are, in a sense, snap shots of more After making reliability scale adjustments, the extensive ecological and developmental processes final analyses were based on scales having two and contexts within which they are embedded and items each. This may compromise the replica- which need to be captured with appropriate com- bility of the present findings, although the in- plexity. ternal consistency of the scales actually im- proved from Time 1 to Time 2. Further, the peer Future Research and Action support items focused only on a narrow aspect of the construct, peer threat; thus, findings from In addition to the methodological issues men- the present study may be inconsistent with those tioned above, future research should address of studies using the full SOCCS scale and more other competencies and support-related dimen- generally may not match those using broader sions that are important in assisting third grad- peer support scales. ers in achieving school success. The results The second methodological problem has to have shown that the model is partially validated do with teachers being the only source of judg- and that social-emotional competence and per- ments of students’ competence, and that the ceived support play important roles in school measure was based on a 3-point scale. The success, but there are still undiscovered factors literature supports that teachers are generally that affect a student’s performance at school, reliable sources of data; however, it would be especially for African American children (see best to have a comparable source of competence Figure 3). Among the variables suggested by data. This issue combined with the fact that the recent research and intervention studies are measure was based on a 3-point scale may have emotion recognition, decision-making skills, affected the magnitude of change in competence feelings of being engaged in school, and, im- scores. Teachers may miss small changes, or if portantly, parent/guardian support (Elias & Ar- they saw them, did not find appropriate grada- nold, 2006). tions available for indicating the change. A In taking an ecological perspective, future small improvement is still an improvement; research should pay greater attention to the in- however, it is not an unequivocal positive re- fluence of cultural factors on students’ social sponse. Nonetheless, the SSRS (Gresham & and learning behaviors and academic achieve- SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 491 ment. The growing presence of immigrant pop- a child with below-average age-appropriate ulations in many communities suggests assess- cognitive skills may not perform well at lan- ments of acculturation and the values being guage-based subjects, but may excel at art, or communicated to children around school behav- may earn good citizenship or merit awards on ior and achievement by parents/guardians and a regular basis. Understanding how a wider also extended family members should be stan- range of school-based accomplishments affect dard in research on culturally diverse children. educators, parents, and the child’s own judg- These are data that would be best collected ments about school engagement and support directly from family members. This research and personal capabilities, and the role of so- design element will bring needed nuance to cial-emotional competence in predicting these investigations of the relationship between fam- outcomes, may lead to the strengthening or ily support, social-emotional competence, class- deeper contextualizing of current findings. room-related support, and overall school perfor- Action implications suggest that interven- mance. The possibility that compensatory rela- tions designed to improve the academic tionships may exist among sources of support and mediate their role on academic success achievement of urban youth, particularly during (e.g., that strong parental support might reduce the elementary school years so crucial for read- the impact of low peer or teacher support or that ing acquisition, should explicitly address both weak family support for academics may be a students’ social-emotional competencies and primary mediator of academic achievement in the relationships between teachers and students young children despite teacher and peer sup- in the classroom. For school psychologists in port) is especially salient for prevention re- particular, the emerging research on the link of search and action in urban minority contexts. school performance and social-emotional fac- Time and timing are important factors in un- tors provide an important way to foster integra- derstanding the development of supportive re- tion of school psychology concerns with the lationships in classrooms. After a few months of broader mission of schools. Current approaches being together with classmates and the teacher, emphasizing rigid, prescriptive academic drills students will typically be more able to reliably and practice oriented toward the format and respond to the measures of perceived support. content of standardized tests have not meaning- An additional assessment period could be in- fully changed the academic trajectories of most cluded about halfway through the school year. underachieving youth and are not likely to do so This assessment would capture changes that in the future (Kozol, 2005). Indeed, even if test may have occurred during the few months spent scores improve, there is reason to believe that together, but because of any number of life genuine learning and the skills needed for suc- events, may have disappeared by the end of the cess in life are being compromised (Elias, school year. The time frame between assess- 2001). Similarly, those planning and imple- ments for younger children may have to be menting skill-oriented preventive interventions shortened in order to understand nuances in the in schools can begin to have more confidence in progression of their development across areas emphasizing the potentially important role that (Gonzales et al., 1996). teacher support plays in mediating students’ Lastly, future research should consider ad- ditional indicators of school performance. valuing and use of the skills they learn. There While certain measures are invested with are corresponding implications for the selection, large social value, such as report card grades training, and ongoing supervision of teachers and standardized test scores in reading and and other educators, particularly with the ways mathematics, we would benefit from knowing in which teachers distribute their attention and how school performance varies as a function support across students of different needs and of indicators selected. For example, a child abilities in their classrooms. Of course, these who is not born in the United States and has implications cannot be drawn from this study memories of another country may be better in alone, but rather from the growing related liter- world history, social studies, or geography ature on the connection of academics to social- than in reading and mathematics, especially if emotional learning (Elias & Arnold, 2006; Zins a language barrier exists. In another scenario, et al., 2004). 492 ELIAS AND HAYNES

Conclusion and adjustment in urban elementary-school chil- dren. Journal of Clinical Child Psychology, 23, The results of this study provide insight into 391–400. a relatively infrequently studied population and Baker, J. A. (1999). Teacher-Student interaction in how the social and emotional development of urban at-risk classrooms: Differential behavior, re- students in third grade in an inner-city elemen- lationship quality, and student satisfaction with tary school district affects school outcomes. school. The Elementary School Journal, 100, 57– Findings supported the proposed model that so- 70. Banks, J. A., Cookson, P., Gay, G., Hawley, W. D., cial-emotional competence and classroom- Jordan Irvine, J., Nieto, S., et al. (2001). Diversity related perceived social support are important in within unity: Essential principles for teaching and determining school success in this population. learning in a multicultural society. Phi Delta Kap- Developing social-emotional competence in pan, 83, 196–203. school is important because it gives children Bernstein, L. (1992). Where is reform taking place? who do not necessarily have these skills from An analysis of policy changes and school climate. home an opportunity to develop them in school Educational Evaluation and Policy Analysis, 14, (Baker, 1999), and to achieve greater academic 297–302. success. Perceived teacher support is important Bronfenbrenner, U. (1979). The ecology of human for academic success and school attendance. development: Experiments by nature and design. This is true for all groups, but especially true for Cambridge, MA: Harvard University Press. those students who attend school in highly dis- Caprara, G. V., Barbaranelli, C., Pastorelli, C., Ban- advantaged communities, perhaps because of dura, A., & Zimbardo, P. (2000). Psychosocial the fact that African Americans comprise the foundations of children’s academic achievement. Psychological Science, 11, 302–306. significant majority in these districts (Attar et Cauce, A. M., Felner, R. D., & Primavera, J. (1982). al., 1994). Unlike the results of studies with Social support in high-risk adolescents: Structural older age groups in this context, perceived peer components and adaptive impact. American Jour- support was not found to be influential in over- nal of Community Psychology, 10, 417–428. all school performance. Cauce, A. M., Hannan, K., & Sargeant, M. (1992). These results have implications for programs Life stress, social support, and locus of control that would seek to alter the current educational during early adolescence: Interactive effects. context for these students, such as classroom- American Journal of Community Psychology, 20, based interventions, introduction of new curric- 787–798. ula, or the introduction of additional assess- Cauce, A. M., Reid, M., Landesman, S., & Gonzales, ments. Policies and action designed to affect N. (1990). Social support in young children: Mea- learning environments for students placed at- surement, structure, and behavioral Impact. In risk must consider the role and function of the B. R. Sarason, I. G. Sarason & G. R. Pierce (Eds.), Social support: An interactional view (pp. 64–94). teachers, their social and emotional capacity to New York: Wiley. negotiate the social and academic tasks, as well Clark, M. L. (1991). Social identity, peer relations, as the quality of their interaction with peers in and academic competence of African American the classroom. Interventions designed to create adolescents. Education and Urban Society, 24, 41– environments conducive to learning and perfor- 52. mance for third graders in low-income urban Cohen, J. (2006). Social, emotional, ethical, and ac- districts appear more likely to succeed if they ademic education: Creating a climate for learning, systematically incorporate ways to develop participation in democracy, and well-being. Har- children’s social-emotional competence and vard Education Review, 76, 201–237. teachers’ supportiveness. Dalton, J. H., Elias, M. J., & Wandersman, A. (2007). Community psychology: Linking individuals and communities (2nd ed.). Belmont, CA: Wadsworth. References Deci, E. L., & Ryan, R. M. (1985). Intrinsic motiva- tion and self-determination in human behavior. Albee, G. W. (1982). Preventing psychopathology New York: Plenum Press. and promoting human potential. American Psy- Demaray, M. K., & Malecki, C. K. (2002). The chologist, 37, 1043–1050. relationship between perceived social support and Attar, B. K., Guerra, N. G., & Tolan, P. H. (1994). maladjustment for students at risk. Psychology in Neighborhood disadvantage, stressful life events, the Schools, 39, 305–316. SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 493

Dempster, A. P., Laird, N. M., & Rubin, D. B. Esposito, C. (1999). Learning in urban blight: School (1977). Maximum likelihood estimation From in- climate and its effect on the school performance of complete data via the EM algorithm (with discus- urban, minority, low-income children. School Psy- sion). Journal of the Royal Statistical Society, Se- chology Review, 28, 365–377. ries B, 39, 1–38. Estell, D. B., Cairns, R. B., Farmer, T. W., & Cairns, Dubow, E. F., Edwards, S., & Ippolito, M. F. (1997). B. D. (2002). Aggression in inner-city early ele- Life stressors, neighborhood disadvantage, and re- mentary classrooms: Individual and peer-group sources: A focus on inner-city children’s adjust- configurations. Merrill-Palmer Quarterly, 48, 52– ment. Journal of Clinical Child Psychology, 26, 76. 130–144. Feiring, C., & Lewis, M. (1991). The development of Dubow, E. F., & Tisak, J. (1989). The relation be- social networks from early to middle childhood: tween stressful life events and adjustment in ele- Gender differences and the relation to school com- mentary school children: The role of social support petence. Sex Roles, 25, 237–253. and social problem solving skills. Child Develop- Freedman-Doan, C., Wigfeld, A., Eccles, J. S., Blu- ment, 60, 1412–1423. menfeld, P., Arbreton, A., & Harold, R. D. (2000). Dubow, E. F., & Ullman, D. G. (1989). Assessing What am I best at? Grade and gender differences in social support in elementary school children: The children’s beliefs about ability improvement. Jour- Survey of Children’s Social Support. Journal of nal of Applied Developmental Psychology, 21, Clinical Child Psychology, 18, 52–64. 379–402. Eccles, J. S., Roeser, R., Wigfeld, A., & Freedman- Garson, G. D. (2007). Structural equation modeling. Doan, C. (1999). Academic and motivational path- Retrieved April 12, 2007, from http:// ways through middle childhood. In L. Balter & www2.chass.ncsu.edu/garson/pa765/structur.htm. C. S. Tamis-LeMonda. (Eds.), Child psychology: Goldstein, S., & Brooks, R. B. (Eds.) (2005). Hand- A handbook of contemporary issues (pp. 287–317). book of resilience in children. New York: Kluwer. Gonzales, N. A., Cauce, A. M., Friedman, R. J., & Philadelphia: Psychology Press. Mason, C. A. (1996). Family, peer, and neighbor- Elias, M. J. (1987). Establishing enduring prevention hood influences on academic achievement among programs: Advancing the legacy of Swampscott. African-American adolescents: One-year prospec- American Journal of Community Psychology, 15, tive effects. American Journal of Community Psy- 539–553. chology, 24, 365–387. Elias, M. J. (2001). Prepare children for the tests of Green, K. D., Forehand, R., Beck, S., & Vosk, B. Education Week, 21 life, not a life of tests. , 40. (1980). An assessment of the relationship among Elias, M. J., & Arnold, H. A. (Eds.) (2006). The measures of children’s social competence and chil- educator’s guide to emotional intelligence and ac- dren’s academic achievement. Child Develop- ademic achievement: Social-emotional learning in ment, 51, 1149–1156. the classroom. Thousand Oaks, CA: Corwin Press. Gresham, F. M., & Elliott, S. N. (1990). Social Skills Elias, M. J., & Bruene-Butler, L. (2005). Social de- Rating System manual. Circle Pines, MN: Ameri- cision making/Social problem solving: A curricu- can Guidance Service. lum for academic, social, and emotional Learning, Haggerty, R. J., Sherrod, L. R., Garmezy, N., & grades 2–3 and 4–5. Champaign, IL: Research Rutter, M. (1994). Stress, risk, and resilience in Press. children and adolescents: Processes, mechanisms, Elias, M. J., & Clabby, J. (1992). Building social and interventions. New York: Cambridge Univer- problem-solving skills: Guidelines from a school- sity Press. based program. San Francisco: Jossey-Bass. Hallmark Corporation (1994). Talking with TJ. Kan- Elias, M. J., Gara, M., Schuyler, T., Brandon-Muller, sas City, MO: Hallmark Foundation. L. R., & Sayette, M. A. (1991). The promotion of Halpern, R. (1990). Poverty and early childhood par- social competence: Longitudinal study of a pre- enting: Toward a framework for intervention. ventative school-based program. American Jour- American Journal of Orthopsychiatry, 60, 6–17. nal of Orthopsychiatry, 61, 409–417. Harter, S. (1985). Manual for the self-perception Elias, M. J., Zins, J. E., Weissberg, R. P., Frey, K. S., profile for children. Denver, CO: University of Greenberg, M. T., Haynes, N. M., et al. (1997). Denver Press. Promoting social and emotional learning: Guide- Hawkins, J. D., Catalano, R. F., Kosterman, R., Ab- lines for educators. Alexandria, VA: Association bott, R., & Hill, K. G. (1999). Preventing adoles- for Supervision and Curriculum Development. cent health-risk behaviors by strengthening protec- Elliott, S. N., Malecki, C. K., & Demaray, M. K. tion during childhood. Archives of Pediatric and (2001). New directions in social skills assessment Adolescent Medicine, 153, 226–234. and intervention for elementary and middle school Haynes, N. M., Emmons, C., & Ben-Avie, M. (1997). students. Exceptionality, 9, 19–32. School climate as a factor in student adjustment 494 ELIAS AND HAYNES

and achievement. Journal of Educational and Psy- Pogrow, S. (2006). Restructuring high–poverty ele- chological Consultation, 8, 321–329. mentary schools for success: A description of the Haynes, N. M., Troutman, M. R., & Nwachuku, U. Hi–Perform school design. Phi Delta Kappan, 88, (1998). School factors in substance abuse preven- 223–229. tion among young male African Americans. Jour- Reyes, O., Gillock, K. L., Kobus, K., & Sanchez, B. nal of Educational and Psychological Consulta- (2000). A longitudinal examination of the transi- tion, 9, 143–154. tion into senior high school for adolescents from Hoff, D., & Mitchell, S. (2006). Pay–to–play: Fair or urban, low-income status, and predominantly mi- foul? Phi Delta Kappan, 88, 223–229. nority backgrounds. American Journal of Commu- Hoyle, R. H. (Ed.). (1995). Structural equation mod- nity Psychology, 28, 519–544. eling: Concepts, issues, and applications. Thou- Reynolds, A. J. (1998). Resilience among Black ur- sand Oaks, CA: SAGE Publications. ban youth: Prevalence, intervention effects, and Kobus, K., & Reyes, O. (2000). A descriptive study mechanisms of influence. American Journal of Or- of urban Mexican American adolescents’ per- thopsychiatry, 68, 84–100. ceived stress and coping. Hispanic Journal of Be- Reynolds, A. J. (1999). Educational success in high- havioral Sciences, 22(2), 163–178. risk settings: Contributions of the Chicago Longi- Kozol, J. (2005). Apartheid in America? Phi Delta tudinal Study. Journal of School Psychology, 37, Kappan, 87, 264–275. 345–354. Kress, J. S., Norris, J. A., Schoenholz, D. A., Elias, Roeser, R. W., Eccles, J. S., & Sameroff, A. J. M. J., & Seigle, P. (2004). Bringing together edu- (2000). School as a context of early adolescents’ cational standards and social and emotional learn- academic and social-emotional development: A ing: Making the case for educators. American summary of research findings. Elementary School Journal of Education, 111, 68–89. Journal, 100, 443–471. Levitt, M. J., & Levitt, J. L. (1994). Social support Rosenfeld, L. B., Richman, J. M., & Bowen, G. L. and achievement in childhood and early adoles- (2000). Social support networks and school out- cence: A multicultural study. Journal of Applied comes: The centrality of the teacher. Child and Developmental Psychology, 15, 207–222. Adolescent Social Work Journal, 17(3), 205–226. Luthar, S. S. (1995). Social competence in the school Rothstein, R., & Jacobsen, R. (2006). What is basic? setting: Prospective cross-domain associations Principal Leadership, 7, 14–19. among inner-city teens. Child Development, 66, Rutgers Social-Emotional Learning Laboratory 416–429. (1998). Talking with TJ: A Social Decision Mak- Masten, A. S. (1994). Resilience in individual devel- ing/Social Problem Solving Curriculum for Urban opment: Successful adaptation despite risk and ad- versity. In M. C. Wang & E. W. Gordon (Eds.), Elementary Youth. Piscataway, NJ: Author, Rut- Educational Resilience in Inner-City America: gers University, Psychology Department. Challenges and Prospects (pp. 3–26). Hillsdale, Saarni, C. (2007). The development of emotional com- NJ: Erlbaum. petence: Pathways for helping children become emo- Maton, K. I., Teti, D. M., Corns, K. M., Vieira-Baker, tionally intelligent. In R. BarOn, J. Maree, and M. J. C. C., Lavine, J. R., Gouze, K. R., et al. (1996). Elias (Eds.), Educating people to be emotionally in- Cultural specificity of support sources, correlates telligent (pp. 15–36). Westport, CT: Praeger. and contexts: Three studies of African-American Schafer, J. L., & Graham, J. W. (2002). Missing data: and Caucasian youth. American Journal of Com- Our view of the state of the art. Psychological munity Psychology, 24, 551–587. Methods, 7, 147–177. Munsch, J., & Wampler, R. S. (1993). Ethnic differ- Schaps, E., & Solomon, D. (2003). The role of the ences in early adolescents’ coping with school school’s social environment in preventing student stress. American Journal of Orthopsychiatry, 63, drug use. Journal of Primary Prevention, 23(3), 633–646. 299–328. Muthen, L., & Muthen, B. (2005). Mplus: Statistical Spencer, M. B. (2005). Crafting identities and assess- analysis with latent variables: User’s guide. Los ing opportunities post-Brown. American Psychol- Angeles: Muthen & Muthen. ogist, 60, 821–830. NCES. (2002). The nation’s report card. Washing- Tolan, P. H., Guerra, N. G., & Montaini-Klovdahl, ton, DC: U.S. Government Printing Office. L. R. (1997). Staying out of harm’s way: Coping Ogbu, J. U. (1991). Low school performance as an and the development of inner-city children. In adaptation: The case of Blacks in Stockton, CA. In S. A. Wolchik & I. N. Sandler (Eds.), Handbook of M. A. Gibson & J. U. Ogbu (Eds.), Minority status children’s coping: Linking theory and intervention and schooling: A comparative study of immigrant (pp. 453–479). New York: Plenum Press. and involuntary minorities (pp. 249–285). New Valentine, J. C., & Cooper, H. (2003). Effect size York: Garland. substantive interpretation guidelines: Issues in the SOCIAL COMPETENCE AND ACADEMIC ACHIEVEMENT 495

interpretation of effect sizes. Washington, DC: the American Psychological Association, Wash- What Works Clearinghouse. ington, DC. Retrievable from http://www.casel Vygotsky, L. S., Rieber, R., & Carton, A. S. (Eds.) .org/downloads/apa08.20.05.ppt (1987). The collected works of L. S. Vygotsky, Vol. Welsh, M., Parke, R. D., Widaman, K., & O’Neil, R. 1: Problems of general psychology. New York: (2001). Linkages between children’s social and Plenum Press. academic competence: A longitudinal analysis. Wang, M. C., & Gordon, E. W. (Eds.) (1994). Journal of School Psychology, 39, 463–482. Educational Resilience in Inner–City America: Wentzel, K. R. (1991). Relations between social compe- Challenges and Prospects. Hilldale, NJ: Erl- tence and academic Achievement in early adolescence. baum. Child Development, 62, 1066–1078. Wang, M. C., Haertel, G. D., & Walberg, H. J. Wilson, W. J. (1996). When work disappears: The (1997). Fostering resilience: What do we know,“ world of the new urban poor. New York: Knopf. Principal, 77, 18–20. Wright, M., & Master, A. S. (2005). Resilience pro- Weinstein, R. (2002). Overcoming inequality in cesses and development. In S. Goldstein & R. schooling: A call to action for community psy- Brooks, (Eds.), Handbook of resilience in children chology. American Journal of Community Psy- (pp. 17–38). New York: Kluwer. chology, 30, 21–42. Zins, J. E., Weissberg, R. P., Wang, M. C., & Walberg, Weissberg, R. P. (2005). Social and emotional learn- H. J. (Eds.). (2004). Building academic success on ing for school and life success. Distinguished con- social and emotional learning: What does the re- tribution award address at the annual meeting of search say? New York: Teachers College Press.

Call for Nominations: Psychology of Violence

The Publications and Communications (P&C) Board of the American Psychological Association has opened nominations for the editorship of Psychology of Violence, for the years 2011–2016. The editor search committee is chaired by William Howell, PhD. Psychology of Violence, to begin publishing in 2011, is a multidisciplinary research journal devoted to violence and extreme aggression, including iden- tifying the causes and consequences of violence from a psychological frame- work, finding ways to prevent or reduce violence, and developing practical interventions and treatments. As a multidisciplinary forum, Psychology of Violence recognizes that all forms of violence and aggression are interconnected and require cross-cutting work that incorporates research from psychology, public health, neuroscience, sociology, medicine, and other related behavioral and social sciences. Re- search areas of interest include murder, sexual violence, youth violence, inpatient aggression against staff, suicide, child maltreatment, bullying, inti- mate partner violence, international violence, and prevention efforts. Editorial candidates should be members of APA and should be available to start receiving manuscripts in early 2010 to prepare for issues published in 2011. Please note that the P&C Board encourages participation by members of underrepresented groups in the publication process and would particularly welcome such nominees. Self-nominations are also encouraged. Candidates should be nominated by accessing APA’s EditorQuest site on the Web. Using your Web browser, go to http://editorquest.apa.org. On the Home menu on the left, find “Guests.” Next, click on the link “Submit a Nomination,” enter your nominee’s information, and click “Submit.” Prepared statements of one page or less in support of a nominee can also be submitted by e-mail to Emnet Tesfaye, P&C Board Search Liaison, at [email protected]. Deadline for accepting nominations is January 31, 2009, when reviews will begin.