Introduction

In recent years, the development of civic engagement among young people has garnered significant scholarly attention. Civic engagement refers to knowledge, values, attitudes and behaviors related to involvement in local community and broader society. Examples include knowledge of political systems, a sense of efficacy and social responsibility, prosocial behavior, political participation and civic activism. Civic engagement in youth predicts future civic behaviors such as adult volunteerism and voting behavior (Hart et al. 2007; Hart and Gullan 2010; Reinders and Youniss 2006; Schmid 2012) making engaged youth critical to maintaining and/or improving the functioning of local, national and global communities and their institutions (Ginwright 2010; Lerner 2004). Accordingly, civic engagement is appropriately seen as an essential component of positive youth development (Lerner 2004; Yates and Youniss 1996), especially for youth of color from historically disenfranchised and low-resourced backgrounds (Christens 2012; Kirshner and Ginwright 2012; Hope and Jagers 2014; Yates and Youniss 1996).

The existing research literature offers a mixed picture of civic engagement among people of color. There is a long history of struggle for civil and human rights by people of color and those from low-income backgrounds in the U.S. (Zinn 2003). However, the American Political Science Association (2004) highlighted growing disparities in adult political participation across racial, socioeconomic, and generational lines, with older, more wealthy Whites being the most politically active. On the other hand, a recent report indicates that compared to White adults, Black adults had higher levels of voter turnout in recent national elections (File 2013). Similarly, scholars have noted a civic achievement gap, where youth of color report lower levels of civic knowledge, political skills, positive civic and political attitudes, and traditional forms of political participation (e.g. voting, contacting elected officials) than White youth (Levinson 2007; National Center for Education Statistics (NCES), 2011). Other research suggests that some Black youth may be more skeptical of traditional government structures and as a result tend to engage in other forms of civic behavior such as participating in youth-led social justice movements, providing family financial assistance, participating in community service through religious organizations, and participating in politically-motivated cultural and artistic expression through poetry and hip hop (Ginwright 2010; Smetana and Metzger 2005; Watts and Flanagan 2007). Further, a recent study of civic engagement profiles (attitudes and behaviors) among a low-income, racially/ethnically diverse group of urban middle school students found between and within racial group variations that warrant further examination (Voight and Torney-Purta 2013). Gaining greater understanding of the development of civic engagement among youth of color is important for basic and applied research interested in these youths’ civic engagement as they grow into adulthood.

The current study contributes to the research literature on the civic engagement of youth of color by examining some possible mechanisms that promote civic engagement during the early adolescent period. Rather than using a race comparative frame, this study focused on a sample of youth of color to examine individual and school factors that might shed light on civic behaviors of participating students. We consider the role of school experiences in what Torney-Purta and Amadeo (2011) refer to as emergent participatory citizenship in their framing of civic engagement among middle and high school-age youth. Rather than couching civic engagement solely in terms of voting behaviors, emergent participatory citizenship among young adolescents is characterized, more generally, by any civic action and attitudes that benefit others. Thus, the concept of emergent citizenship allows for an opportunity to consider the types of civically-minded behaviors and attitudes that are developmentally appropriate and contextually accessible to middle school students. Moreover, it invites researchers to explicate how schools might foster emergent participatory citizenship among its students by providing an environment supportive of youths’ thoughts, beliefs and actions that are meant to benefit others in school and neighbors settings.

Theoretical Framework

The extant literature offers several theoretical frameworks for the examination of the development of civic engagement. For example, Watts and colleagues (Watts et al. 2003; Watts and Guessous 2006; Watts and Flanagan 2007) advance the Sociopolitical Development model (SPD) that focuses on both the individual and contextual factors that promote social-justice activism among young adults. Flanagan (2004) highlights the political socialization model of civic engagement, which suggests that parents and other significant adults transmit to children civic ideologies, understandings and practices based on current social relations. Lerner et al. (2014) suggest that the individual-environment transactions within many such frameworks (e.g., Benson et al. 2010; Damon 2008; Spencer 2006; Zaff et al. 2011) are best understood as variations of relational developmental systems metatheory that undergirds the positive youth development model.

In an effort to understand the roles of schools in fostering emergent citizenship in early adolescence, we drew from Torney-Purta and Amadeo’s (2011) developmental niche theory for emergent participatory citizenship. The framework takes a sociocultural (Vygotskian) approach to civic engagement among youth. This framing of emerging citizenship proves useful as it allows for the explicit consideration of the cultural and historical forces that exist within an adolescent’s everyday life in a variety of distal and proximal social contexts (e.g., schools) (Torney-Purta and Amadeo 2013; Torney-Purta and Amadeo 2011; Torney-Purta and Barber 2011). Specifically, through social interaction young people learn about themselves, their social partners, and what activities and tools can support action that benefits others (i.e., emergent citizenship). The developmental niche framework’s attention to context (i.e., developmental niche and sociocultural process) allows us to also consider how adolescents’ interactions with other socialization agents at the school site (e.g. teachers and peers) may influence their actions as they relate specifically to their civic engagement. It allows for the examination of the psychological and behavioral implications of the various contexts within which adolescents develop.

The concept of the developmental niche was initially proposed by Whiting and Whiting (1975) and originally focused on the influence of culture in the creation of micro- and macro-level settings in the parental socialization of infants. Super and Harkness (1986, 1999) elaborated on the developmental niche as it relates to socialization by specifying three dimensions of the niche in which children exist. The first dimension of the developmental niche is made up of the physical and social settings (e.g. social media, family, school) that comprise their day-to-day life. The second dimension of the developmental niche is the historically rooted customs and beliefs that influence the role of child in society. The third dimension consists of the individual beliefs of adults about the lives of children and their developmental processes. It is within the first dimension of the developmental niche theory that we consider the relationship of a student’s day-to-day experiences in schooling to their civic behaviors (i.e., emergent citizenship).

Civic Engagement Among Middle School Youth

In order to provide a context for the literature on civic engagement among young adolescents, it is important to note that much of this literature on civic engagement, broadly speaking, can be organized in terms of Westheimer and Kahne’s (2004) typology of citizenship, which includes personally responsible, participatory, and justice-oriented citizenship forms. In brief, the personally responsible citizen is thought to exemplify good character by displaying prosocial attitudes and behaviors, and they often promote the common good (e.g., by being helpful in their local community). The participatory citizen is actively engaged in local clubs, traditional civic organizations, social institutions and political activities. Finally, the justice-oriented citizen engages in critical analysis of sociopolitical forces and takes action to fight and remedy various forms of inequity and injustice.

Research investigating civic engagement at the middle school level (Geller et al. 2013; Voight and Torney-Purta 2013) has argued that during a period of emergent citizenship, young people can have various citizenship beliefs and attitudes, but their behaviors are limited largely to those associated with the aforementioned personally responsible and participatory citizenship forms. For example, Voight and Torney-Purta (2013) used latent class analysis to identify civic engagement groups that they termed civic moderates (low attitudes/low behaviors), social justice sympathizers (high attitudes/low behaviors) and social justice actors (high attitudes/high behaviors). Black students were more likely to be social justice actors or civic moderates, whereas Latinos were more likely to be social justice sympathizers and less likely to be social justice actors. Although Voight and Torney-Purta (2013) opted to use the label “social justice” for two of the clusters, the behavioral items more closely reflect (developmentally appropriate) personally responsible and participatory behaviors. Geller et al. (2013) examined the influence of personally responsible and participatory civic behaviors of perceptions of school climate among a diverse group of urban middle school students. They found that students’ personally responsible civic behaviors were significantly associated with more positive perceptions of the consistency and fairness of school rules, democratic climate, and student–teacher relationships. Furthermore, being a leader in a school group had positive associations with perceived student relationships, as well as with being a leader in a neighborhood club or group.

The Role of School Climate

As Fine et al. (2004) noted, schools are “…intimate places where youths construct identities, build a sense of self, read how society views them, [and] develop the capacity to sustain relations and forge the skills to initiate change” (p. 2198). The Civic Mission of Schools Report (2003) highlighted the potential importance of schools in cultivating youth civic development. It is not surprising, then, that schools have been referred to as mini-polities (Flanagan et al. 2011) and are seen as a developmental niche (Torney-Purta and Amadeo 2011) “where the younger generation can explore what it means to be a member of a political community and can practice the rights and obligations associated with membership in that community” (Flanagan et al. 2011, p. 102). This can be accomplished through classrooms or school-wide curricular content and relational processes.

Schools vary in the extent to which they uphold a democratic culture and provide opportunities for students to engage in personally responsible and participatory citizenship. One reason for this is that schools vary in their climate, which connotes the sense of safety, relationships, teaching and learning, institutional environment and school improvement processes (Thapa et al. 2013). Several studies have investigated the association of school climate to civic engagement (Torney-Purta 2002; Lenzi et al. 2014). Research indicates that multiple dimensions of school climate might be relevant for middle school students’ civic engagement. These include students’ relationships with adults (i.e., administrators, teachers and staff), student–student relationships, student participation in decision-making, perceptions of a democratic school climate and school safety (Lenzi et al. 2014; Torney-Purta 2002; Flanagan et al. 2007).

Adolescents’ positive perceptions of the inclusiveness of their school’s climate is known to relate to favorable outcomes, including reduced self-regulation problems and decreased problem behaviors (Loukas and Murphy 2007); decreased depression and behavior problems (Way et al. 2007); and increased student engagement and academic achievement (Wang and Holcombe 2010). Moreover, Wegman, Geller, and colleagues (e.g., Karakos et al., in press; Geller et al. 2013) found that student civic behavior is positively associated with more positive perceptions of teacher-student relations, student–student relations, fairness of school rules and democratic school processes. The present study examined the extent to which middle school youths’ school climate perceptions were related to their personally responsible and participatory civic behaviors and attitudes, and whether this association could be partly due to youths’ beliefs about schooling (i.e., school connectedness and academic self-efficacy).

School Climate and Emergent Participatory Citizenship: Indirect Links

Research supports the notion that a positive school climate (i.e. an environment in which a student feels safe and supported) can facilitate a sense of school connection (Eccles and Roeser 2011), which signifies students’ personal sense of “acceptance, respect and inclusion by adults and peers within the school environment” (McMahon et al. 2009; p. 269). It reflects students’ perceptions of how they fit within the school’s academic and social context. Generally speaking, issues of relatedness are particularly salient during early adolescence (Connell and Wellborn 1991), a time when school connection tends to be relatively low (Anderman 2003; Eccles and Roeser 2010). Feeling connected to school can be even more challenging for youth of color in racially/ethnically diverse settings given, for example, the possibility that adults and peers may hold negative racial and class-based stereotypes and lack of culturally responsive content and/or pedagogy (Booker 2006). However, school contexts that convey a sense of belonging may support the adoption of prosocial norms and motivations that contribute to constructive civic behaviors among early adolescents (Lenzi et al. 2012).

In addition, it is possible that school climate perceptions may inform students’ emergent citizenship through their link with their sense of academic self-efficacy, or their beliefs about their ability to successfully complete educational assignments and tasks in this setting. Academic self-efficacy derives from Bandura’s more general notion of self-efficacy, which is the belief in one’s capacity to exercise control over his or her performance and environmental demands (Bandura 2000). Self-efficacy beliefs are influenced by personal experiences and vicarious learning processes. As such, it follows that a school climate that features modeling of desired academic and social behaviors and supportive teacher and peer relationships would increase students’ sense of connection and academic self-efficacy beliefs (Schunk and Zimmerman 2007). McMahon et al. (2009) found such relationships in their study of school climate dimensions and content area-specific efficacy among ethnically diverse, urban fourth and fifth graders.

Voight and Torney-Purta (2013) found that middle school students with more civically engaged profiles (positive civic attitudes and/or behavior) had more desirable school outcomes (academic and behavioral) than students with the lowest civic attitudes and behaviors. Directionality of these relationships could not be discerned in that study. Therefore, it seems plausible that having better academic status in school contributes to student’s positive interactions with others. Indeed, in a series of experimental studies with French college students, Poortvilet and Daron (2014) reported that academic self-efficacy beliefs mediated the relationship between academic goals and helping peers. Thus, in this study we examined the relationship between students’ academic efficacy beliefs and civic engagement behaviors.

The Current Study

The present study sought to contribute to a growing literature on emergent participatory citizenship (Torney-Purta and Amadeo 2011) by examining school factors that contribute to personally responsible and participatory civic attitudes and behaviors among ethnically and racially diverse middle school students. It is particularly important to understand how and why youth of color become engaged citizens because of the legacy of sociopolitical marginalization for this segment of the U.S. population. By extension, a democracy requires prosocial interpersonal behaviors, participation in traditional political systems and system change efforts when traditional avenues do not provide adequate remedies to asymmetric social and economic relations. The present study examined personally responsible and participatory civic engagement behaviors and attitudes as a latent construct reflecting emergent citizenship. We sought to understand whether and how school climate is associated with civic engagement behaviors and attitudes, directly and indirectly, by fostering a positive sense of connection to the school and/or engendering feelings of self-efficacy in the academic domain. Accordingly, we examined the association of youths’ sense of school connection and academic self-efficacy with their personally responsible and participatory civic behaviors in school and community contexts—both concurrently and longitudinally. Based on prior theory and research, we expected school connectedness and academic self-efficacy beliefs to mediate, or help explain, the linkage between school climate and emergent participatory citizenship.

Methods

Participants

Data were drawn from Time 1 and Time 2 of a larger longitudinal study of social/emotional development among youth in ethnically diverse middle schools in the Midwest. The present study focused on data from one school at which school connectedness was assessed as part of the project. Of the full sample at this school at Time 1 (N = 492), students were excluded because they were missing all survey items (19) including auxiliary information on the constructs of interest (23). One 17-year old student was removed from the sample given this individual’s age was 3 standard deviations above the mean. Finally, only students who self-identified as African American, Multiracial or Mixed, Latino, Asian American or Pacific Islander, Other, or Native American were selected for the present study, resulting in the exclusion of another 125 students. Thus, the cross-sectional sample included 324 6th–8th grade students, of which 43 % identified as African American, 20 % identified as Multiracial or Mixed, 18 % identified as Latino, 12 % identified as Asian American or Pacific Islander, 6 % identified as Other, and 1 % identified as Native American. The age of the sample ranged from 11 to 15 (M = 12.33, SD = .97). The sample grade levels were as follows: 6th (33 %), 7th (38 %), and 8th grade (29 %). The sample was fairly balanced in terms of gender, as 48 % selected “Boy” as their gender self-identification, and 52 % selected “Girl.” 61.1 % of the cross-sectional sample reported receiving free or reduced-fee lunch at school. In addition, the longitudinal study sample included students who were in 6th or 7th grade at Time 1 (n = 232); this subsample was equally diverse with 42.2 % identified as African American, 19.8 % identified as Multiracial or Mixed, 19.4 % identified as Latino, 11.6 % identified as Asian American or Pacific Islander, 11.6 % identified as Other, and 5.2 % identified as Native American. In terms of gender, 44.8 % selected “Boy” as their gender self-identification, and 54.3 % selected “Girl.” Finally, of the longitudinal sample 62.1 % reported receiving free or reduced-fee lunch at school.

Procedure

Time 1 data collection occurred in the spring semester of 2014 and Time 2 occurred sixth months later. Student surveys were voluntary and administered by teachers during homeroom; students were assured of the confidentiality of their responses (i.e., by stating, “Your individual answers will be private and will never be shared with anyone at this school” on the survey cover sheet). All completed surveys were de-identified (all names removed and replaced with ID codes) by an external consultant who was not affiliated with the research team. Because the data were collected as part of the school’s assessment of its own practices and all participant information was de-identified, the project was determined to be Exempt by the IRB at the authors’ university.

Measures

School Climate

Students’ perceptions of school climate were indicated with six items (e.g. “Teachers go out of their way to help students here;” “Students enjoy doing things with each other in school activities;” α = .83; Brand et al. 2003). These were Likert-type items (1 = Strongly Disagree, 5 = Strongly Agree), with higher values indicating more positive perceptions of school climate.

School Connectedness

Students’ sense of connection to school was indicated with four items (e.g., “I feel like I am part of this school;” α = .86; McNeely et al. 2002). These were Likert-type items (1 = Strongly Disagree, 5 = Strongly Agree), with higher values reflecting greater school connectedness.

Academic Self-Efficacy

Academic self-efficacy was indicated with four Likert-type items (1 = Strongly Disagree, 5 = Strongly Agree) items. Students were asked to report their perceptions of their academic self-efficacy (e.g., “I am certain I can figure out how to do even the most difficult classwork;” α = .90; Patrick et al. 1997). Higher values indicated greater academic self-efficacy.

Civic Engagement

Civic engagement was indicated by seven Likert-type items (1 = Never, 5 = Always). The items were adapted from an existing civic engagement scale that assesses personally responsible and participatory attitudes and behaviors (Voight and Torney-Purta 2013) and were used to indicate emergent citizenship. A sample behavioral item is, “In the past 12 months, how often have you…[e.g., participated in community service projects (like tutoring or neighborhood cleanup)]?. “How important is to… (e.g., make sure all people are treated fairly)”?” is an example of an attitudinal item (T1 α = .80 and T2 α = .86). Higher values indicate higher rates of participation and endorsement of civically engaged activities and attitudes.

Demographic Covariates

To account for the potential confound of experience in the school setting, we included grade level at Time 1 and free/reduced-priced lunch eligibility in all analyses. Grade level was coded 1 for a given grade (e.g., sixth) and 0 for all others, such that 3 dummy codes were created to reflect the three grades (one for 6th, one for 7th, and one for 8th grade); however, two dummy codes (7th and 8th) were included in the model such that 6th graders were the omitted (reference) group in both the cross-sectional and the longitudinal analyses. Free/reduced-priced lunch eligibility was a dummy-coded variable obtained from the school where 1 = eligible and 0 = not eligible, and the not eligible group is the omitted (reference) group.

Analysis Plan

To answer our major research question, a latent variable mediation model was fitted in Mplus 7.2 to examine the hypothesized model using cross-sectional (n = 324) (see Fig. 1) and longitudinal data (see Fig. 2). For the longitudinal test of our mediation model, we drew on data from the 6th and 7th grade cohorts only (n = 232); the ethnic and racial diversity of this longitudinal subsample was virtually identical to that of the cross-sectional sample (information available upon request). For both the cross-sectional and longitudinal analyses, full information maximum likelihood was used to deal with missing data (Schafer and Graham 2002). Fit indices were inspected to determine if they met the recommended thresholds of RMSEA below .05 (Kline 2005) and CFI and TLI values of at least .90 (Hu and Bentler 1999). To determine if particular mediation pathways were significant, the bias-corrected bootstrap confidence intervals for each indirect pathway were inspected (Hayes 2013).

Fig. 1
figure 1

Concurrent model of school climate, academic beliefs, and civic engagement. Note FRPL = Free/Reduced Price Lunch status. Standardized estimates shown. Bold lines indicate a significant indirect path, and dashed line represents non-significant path. Manifest indicators of latent constructs not shown for ease of presentation

Full size image
Fig. 2
figure 2

Longitudinal model of school climate, academic beliefs, and civic engagement (6th and 7th Grade Cohorts). Note FRPL = Free/Reduced Price Lunch status. Standardized estimates shown. Bold lines indicate a significant indirect path, and dashed line represents non-significant path. Manifest indicators of latent constructs not shown for ease of presentation

Full size image

Results

Preliminary Results

Means, standard deviations, and bivariate correlations for all continuous variables are presented in Table 1. At Time 1, school climate, academic self-efficacy, school connectedness, and civic engagement behavior were significantly and positively correlated with each other (r range .26–.50, all ps < .01) concurrently. Correlations were similar and also significant between Time 1 school climate, academic self-efficacy, and school connectedness and Time 2 civic engagement; r range .23–.35, all ps < .01).

Table 1 Descriptive statistics for primary study variables at time 1
Full size table

Examination of Hypothesized Model

Cross-Sectional Analysis

Fit indices for the latent variable mediation model using the cross-sectional data suggest that the hypothesized model fit the observed data well [CFI = .94; TLI = . 93; RMSEA = .05, 90 % CI (.04, .06)]; Fig. 1 provides standardized coefficients for significant paths in this model. With regard to the primary study objectives, the model results suggest that, regardless of grade level free and reduced lunch status, more positive views of school climate are associated with higher values on student civic engagement. In the model, students’ perceptions of school climate were significantly and positively associated with greater feelings of school connectedness, which were in turn significantly and positively associated with civic engagement. In addition, school climate was significantly and positively associated with greater academic self-efficacy, but academic self-efficacy was not significantly associated with civic engagement. Tests of mediation indicated there was one significant indirect pathway between school climate and civic engagement behavior. Specifically, there was an indirect association between school climate and civic engagement via its association with sense of school connectedness, indirect association estimate = .17 (95 % CI .08, .32). The aforementioned indirect association via school connectedness accounted for 50 % (.17/. 34) of the total effect. As such, partial support was found for the prediction that students’ perceptions of a positive school climate are linked to civic engagement in part through their role in students’ sense of connection to school.

Longitudinal Associations

As with the cross-sectional analysis, we found support for the proposed mediation model using the longitudinal data from the 6th and 7th grade cohorts only (n = 232); Fig. 2 provides standardized coefficients for significant paths in this model. Fit indices for the latent variable mediation model using the longitudinal data suggest that the hypothesized model was an adequate fit to the observed data [CFI = . 93; TLI = . 92; RMSEA = .05, 90 % CI (.04, .06)]. With regard to the primary study objectives, the model results suggest that at Time 2, adjusted for grade level and free and reduced lunch status, more positive views of school climate were not directly associated with student civic engagement. Similarly to the cross sectional analysis, students’ perceptions of school climate were significantly and positively associated with greater feelings of school connectedness (b = .86, se = .24, p = .000), which were in turn significantly and positively associated with civic engagement. In addition, school climate was significantly and positively associated with greater academic self-efficacy (b = .80, se = .15, p = .000). As in the cross-sectional model, academic self-efficacy was not significantly associated with civic engagement at Time 2 (b = .10, se = .10, p = .26). Tests of mediation indicated there was one significant indirect pathway between school climate and civic engagement. Specifically, there was an indirect association between school climate and civic engagement via its association with sense of school connectedness, indirect association estimate = .23 (95 % CI .09, .48). Finally, school connectedness accounted fore 64 % of the total effect (.22/.36) of school climate on civic engagement.

Discussion

Through the use of the emergent citizenship framework (Torney-Purta and Amadeo 2011), this study examined the influence of school factors on early adolescents’ civic engagement. This study contributes to a growing literature that aims to examine the proximal contexts (i.e. schools) with regard to optimizing processes related to civic engagement among children and youth from diverse income and racial/ethnic backgrounds. In the present study, we focused on civic engagement behaviors in school and neighborhood settings and first sought to understand whether school climate was directly associated with such behaviors. We then examined possible indirect mechanisms for such an association. The present results suggest that among middle school youth, perceptions of school climate are not directly associated with civic engagement; however, results of our cross-sectional analysis suggest that students’ sense of positive school climate is indirectly associated with civic engagement through its association with such students’ feelings of connection to school. Moreover, the longitudinal analyses showed that earlier school climate perceptions indirectly predicted later civic engagement as well.

There is some evidence pointing to connections between aspects of school climate and specific civic behaviors (Geller et al. 2013). However, there was no direct effect of school climate on civic engagement in our cross-sectional or longitudinal analyses using an aggregate measure of civic behaviors and attitudes. When we consider this finding in relation to the authors’ conceptualization of the school as a developmental niche (Torney-Purta and Amadeo 2011) or mini-polities (Flanagan et al. 2011), we can better unpack why a students’ civic engagement is not directly influenced by solely an adolescent’s positive perception of a school climate. As previous research has indicated, perceptions of school climate can influence individual level processes such as school connection as well as academic efficacy beliefs (McMahon et al. 2009, Eccles and Roeser 2011). Indeed, our findings indicate that more positive school climate perceptions were positively related to both school connectedness and academic self-efficacy, but only school connectedness was significantly and directly associated with civic engagement in this sample. These results thus provide further empirical support for Torney-Purta and Flanagan and their respective colleagues’ assertions (Flanagan et al. 2011; Torney-Purta and Amadeo 2011) that students’ sense of being a part of and a valued member of the school community, in particular, likely inform their normative beliefs and dispositions toward helping others inside and outside of school (e.g., in their neighborhood).

In order to test our propositions regarding the mediating role of academic self-efficacy and school connectedness, we examined the hypothesized model in two subsamples—one cross-sectional and the other longitudinal—of a larger study of social/emotional development among middle school youth of color. In both the cross-sectional and longitudinal models, we tested whether perceptions of school climate had an indirect influence on youths’ civic engagement through the individual level processes of school connection and academic self-efficacy. Although we found support for a model in which school connectedness mediates the role of school climate in civic engagement, we did not find similar support for mediating role of academic self-efficacy. This suggests that when students feel they attend a school in which adults are supportive and in which their peers enjoy getting to know each other and working together, they are more likely to feel they themselves are part of, happy, and close to others in that setting. These feelings of connectedness promoted more engagement in behaviors such as helping others in their school or neighborhood.

Taken together, these findings indicate that merely having a positive perception of the school environment is not enough to influence civic engagement behaviors. The results also suggest that it not simply feeling efficacious or competent in the domain of school that will contribute to youths’ engagement in actions that support or help out others in and out of school. Rather, it appears that when youth actually feel a sense of connection to the school setting it may promote their engagement in behaviors oriented toward helping out in that space. On the other hand, believing that one can academically succeed in school may not necessarily mean that one feels a sense of obligation to improve school or their neighborhood. This supports the theoretical contention that through social interaction in developmental niches, young people learn about themselves, their social partners, and what practices and tools can support their emergent citizenship (Torney-Purta and Amadeo 2011). When conceptualized as a developmental niche for emergent citizenship, the middle school environment is a context where adolescents’ emergent participatory citizenship behaviors can be enacted as well as unpacked.

Limitations and Future Directions

The present study has some limitations that can help to inform future directions and lines of research. First, our longitudinal study occurred during a short time-interval. However it still provided robust, partial support for hypothesized pathway; future work should look at longer time lags. Second, while the sample was comprised of middle school youth of color and was diverse on several dimensions (i.e., race/ethnicity, grade level, gender and family income) and we gained important insights into the growth of civic engagement among these students, we did not have sufficient numbers of students at each grade level within these various categories to conduct multi-group analyses. Thus, we could not determine whether and in what ways these subgroups might differ on the study variables. Future research might include larger samples that would allow for such comparisons. The inclusion of White youth from comparable economic backgrounds could further illuminate the contributions of racial group membership and socioeconomic status to the predictors of emergent citizenship behaviors of early adolescents in the U.S. Extending our work to international samples, especially in the African diaspora and Latin American countries would provide insights into broader cultural and historical contexts.

Consistent with recent research on civic engagement at the middle school level, this study operationalized civic engagement in terms of personally responsible and participatory civic behaviors and attitudes. These represent the most likely civic behaviors and attitudes in this age group. However, future work should include justice oriented attitudes and behaviors to shed light on their onset and relationships among various forms of civic engagement during the middle school years. First, both school climate and civic engagement are multi-dimensional constructs. Consistent with Geller et al. (2013), future research should examine how discrete elements of school climate might be associated with specific school and neighborhood civic engagement behaviors. Second, it would be informative to consider the contributions of classroom experiences, to include pedagogical practices and academic content, to various civic attitudes and civic behaviors (Hope and Jagers 2014). Finally, future research should examine the ways in which setting level civic behaviors might influence individual level outcomes. Specifically, it seems likely that the degree to which other students report being helpful in school and neighborhood settings could influence perceptions of these contexts and shape individual civic attitudes and behaviors.

Implications for Practice

Finally, the present study has implications for future applied research and evaluation. Data from this study was taken from a larger, multi-year evaluation of a middle school approach to academic, social and emotional learning. Subsequent studies should build on the current findings, examining how teacher adoption of democratic practices such as morning advisory, student voice and choice in content-specific classes and student centered discipline influence student outcomes, with a focus on youth civic and school engagement. This affords us the opportunity to investigate intersections between school and civic engagement (Lawson and Lawson 2013) and the contributions of teachers and peers to these distinct, but overlapping, developmental outcomes. By extension, this research has relevance for youth organizing efforts, which include youth participatory action research in classroom and out-of-school time settings (Kirshner and Ginwright 2012). This work highlights the use of systematic research to guide and evaluate youth efforts to identify and solve local concerns that negatively affect them and their community in varying contexts. Most of the research in this area has appropriately focused on group processes and collective action. Less attention has been given to the characteristics of participating youth such as their previous or concurrent civic activities, attitudes or behaviors and how these might help inform how they experience youth organizing processes and outcomes. This might be particularly relevant in schools, which tend to operate under greater restraints than do out-of-school time settings (Ozer et al. 2013; Taft and Gordon 2013). This seems like an important consideration to be folded into to future research in this area.