Black students experience significant social and economic inequalities that directly impact their academic engagement and motivation (Fahle et al., 2020; Reardon, 2016). However, there are important school characteristics which also contribute to declines in Black students’ academic engagement and motivation (Appleton et al., 2006; Wang et al., 2011). For example, greater student engagement is shaped by schools that bolster students’ feelings of belonging and connection to their school (Appleton et al., 2006; Wang et al., 2011). When students feel that these needs are not met, their academic engagement and motivation decline leading to poor academic performance (Eccles et al., 1993). Research investigating the role of school characteristics on Black children’s academic engagement and motivation has primarily focused on Black adolescents (Shernoff et al., 2016; Wang et al., 2015) and is cross-sectional rather than longitudinal (Cornell et al., 2016; Robayo-Tamayo et al., 2020; Shernoff et al., 2016). Very little is known about changes in engagement and motivation during the period prior to children’s entry into middle school. It is important to explore the development of academic engagement and motivation prior to this transition period to understand the protective factors that may buffer against declines in academic engagement and motivation that are often seen during middle school (Benner & Graham, 2013; Eccles & Roeser, 2009). Further, there is limited research on how school racial composition specifically influences changes in Black children’s academic engagement and motivation in late elementary school. This study will examine the relation between school racial composition and Black children’s academic engagement and motivation in late elementary school, as well as how school quality and school economic disadvantage moderate this association.

The Importance of Academic Engagement and Motivation

Academic engagement is defined as students’ actions that facilitate productive learning, school involvement, students’ feelings of school connectedness, and students’ self-regulation and learning skills (Wang et al., 2011). In contrast, academic motivation reflects children’s beliefs about their own abilities and competencies in specific tasks based on their perception of an activity’s difficulty, importance, usefulness, and their ability to improve on the activity (Eccles et al., 1993). Research has shown positive associations between students’ academic engagement, academic motivation, and academic achievement (Appleton et al., 2006). Students who display higher levels of academic engagement are likely to display higher levels of academic motivation, which leads to increases in academic achievement (Wang et al., 2011).

Fredricks et al., (2004) found that the three dimensions of academic engagement (behavioral, cognitive, and emotional) concurrently influence students’ academic achievement. For example, students’ emotional engagement with their school declines when students do not feel connected to their school; this then leads to declines in students’ behavioral and cognitive engagement, which in turn predict lower academic success as students withdraw from school-related activities and engage in negative behaviors (Benner & Graham, 2013; Wang et al., 2015; Wang et al., 2011). The two dimensions of academic motivation, self-concept and subjective task valuing, significantly predict students’ academic performance, expectations, and competency beliefs which are both key indicators of overall school performance (Eccles et al., 1993; Wang et al., 2011). Specifically, when students experience declines in academic self-concept and subjective task valuing, students also experience declines in their overall grades due to decreased competency beliefs in their academic abilities and a decreased desire to participate in academic/school-related tasks (Brown & Jones, 2004; Chavous et al., 2018).

Race can further complicate the process of academic engagement and motivation for minoritized students. Black students’ school experiences have been associated with educational and institutional disparities from elementary school through college compared to White students (Byrd & Andrews, 2016; Loque et al., 2017; Nelson et al., 2022), as they attend schools that are not prepared to meet their diverse needs leading to negative school experiences such as disproportionate discipline, inequitable school policies, stereotypes, and microaggressions (Byrd & Andrews, 2016; Cooper et al., 2022; Fahle et al., 2020). These experiences adversely impact their academic engagement and motivation (Benner & Graham, 2013; Byrd et al., 2016; Wang & Eccles, 2013). For example, Black students’ experiences of discrimination from teachers and peers has been associated with declines in their overall school engagement (Benner & Graham, 2013; Byrd et al., 2016). Negative teacher-student relationships that stem from Black students’ experiences of discrimination are associated with lower levels of behavioral, emotional, and cognitive engagement for students from seventh to eighth grade (Wang & Eccles, 2013). Conversely, when students perceived more positive, supportive relationships with their teachers, they reported greater engagement (Louque et al., 2017). Understanding what school factors influence Black students’ academic engagement and motivation is crucial to eliminating achievement gaps between Black and White students, for creating inclusive and affirming learning environments for Black students, and to further develop our understanding of processes that promote the academic development of Black students.

Research on the constructs of academic engagement and motivation has primarily focused on children in middle school, high school, and college. This is due to the belief that children’s academic attitudes and behaviors developed during early elementary school remain salient for the duration of children’s time in school (Entwisle & Alexander, 1998). However, cross-sectional and longitudinal research has identified the critical role of additional school, child, and family characteristics that significantly predict students’ academic engagement and motivation.

Correlates of Academic Engagement and Motivation

Middle school, high school, and college students report higher academic engagement and grades when they perceive greater support from their school and a more positive school climate (Cornell et al., 2016; Robayo-Tamayo et al., 2020). In addition to school characteristics, researchers have identified a relation between child characteristics and academic engagement, specifically students’ affection or disaffection with school. Student affection for school is associated with higher levels of academic engagement for college students compared to college students with negative feelings toward their school (King et al., 2015). Similar patterns have emerged for students’ academic motivation in studies of high school and college students. For example, Brown & Jones (2004) found that high school students’ perceptions of unfair treatment at school contributed to lower academic motivation, which also led to lower GPAs for these high school students. For Black college students, Chavous et al., (2018) identified students’ perceptions of their racial identity such as private regard, public regard, and race centrality as significant predictors in their academic motivation. Positive private and public regard and race centrality were associated with higher levels of academic motivation, while negative private and public regard and race centrality were associated with lower levels of academic motivation for Black college students (Chavous et al., 2018).

Changes in Academic Engagement and Motivation

Longitudinal changes in students’ academic engagement have primarily been explored for children in middle school and high school. Minoritized students’ perceptions of their school environment such as teacher relationships and peer support contribute to changes in all three domains of children’s engagement from seventh to eighth grade, such that positive teacher relationships and greater peer support are associated with increases in students’ behavioral, emotional, and cognitive engagement (Wang & Eccles, 2013). This relation is often mediated by children’s academic motivation and their satisfaction with their school. For example, middle school students’ positive emotions toward their school and increased academic motivation predicted increases in students’ academic engagement throughout their time in middle school, while dissatisfaction with their school and decreased academic motivation predicted decreases in students’ academic engagement (King et al., 2015; Wang & Eccles, 2013).

Although both cross-sectional and longitudinal research on changes in students’ academic engagement and motivation have examined this relation for students in middle school, high school, and college, we still know very little about what occurs during late elementary school, prior to the declines of academic engagement and motivation identified in middle school (Cornell et al., 2016; Wang & Eccles, 2013). Understanding this association is important, as children’s academic attitudes and beliefs are established early, specifically during elementary school (Entwisle & Alexander, 1998). By understanding this association, we can also develop a clearer understanding of the promotive or inhibiting factors during late elementary school which can contribute to declines in children’s academic engagement and motivation often seen in middle school (Eccles & Roeser, 2009).

Furthermore, there is limited longitudinal research that explores changes in academic engagement and motivation for ethnically and racially minoritized students. Longitudinal research in this area has been limited to studies of White students from middle school through high school and the relation between low academic engagement and motivation, school dropout, low GPA, and an overall decline in grades for these students (Barber & Olsen, 2004; Dotterer et al., 2009). More research is needed on changes in Black students’ academic motivation and engagement during late elementary school to better understand the factors that may support high levels of engagement and motivation as children move into middle school.

Academic Engagement, Motivation, and School Racial Composition

Black students’ academic engagement and motivation has been directly linked to their school’s racial composition. Research is mixed, as some research shows that Black students who attend ethnically matched middle and high schools tend to have higher academic motivation compared to students who attend non-ethnically matched middle and high schools (Jones, 2018; Williams et al., 2017; Yull & Wilson, 2018). However, others have reported that attending a more diverse school can have positive impacts for ethnically and racially minoritized adolescents’ overall well-being, as these schools promote more cross-cultural friendships and acceptance of diversity (Juvonen et al., 2018). The literature on the impact of school racial composition may be mixed due to two factors. First, there is a continued focus on changes in students’ academic engagement and motivation throughout middle school, high school, and college. Research in this area has not fully explored these changes for minoritized students in late elementary school, prior to declines in academic engagement and motivation during middle school (King et al., 2015; Wang & Eccles, 2013), as this research has been limited to studies of White children (Dotterer et al., 2009). Without a comprehensive understanding of changes in academic engagement and motivation prior to these declines in middle school, research will be unable to fully capture the nuances of changes in academic engagement and motivation in relation to school racial composition.

When exploring the impact of school racial composition, it is important to note that although the school aged population of United States is becoming increasingly diverse, many public schools across the country do not reflect this diversity (Graham et al., 2022). Rather, schools appear to be more racially-ethnically segregated than they were after the Supreme Court’s decision of Brown v Board of Education (Orfield et al., 2019). Studies have shown that resegregation is evident in public schools. Although White students make up 50% of the nation’s public school enrollment, 20% of schools have few to no White students enrolled while 90% of students enrolled are Black and/or Latinx (Orfield & Frankenberg, 2014; Williams & Graham, 2020). The segregation in the racial composition of schools is due to neighborhood segregation (Green 2015; Tate 2008) and availability of resources and funding within schools (Fahle et al., 2020; Reardon, 2016).

In addition to primarily exploring these changes in children from middle school to college, research has not examined how the impact of school racial composition may vary when accounting for additional school level characteristics as moderators. Additional school level characteristics can include school quality and school economic disadvantage. Higher quality schools have consistently been associated with more positive academic outcomes for children (Fahle et al., 2020). Higher quality schools are characterized as those where minoritized students experience less discrimination and where these students feel a sense of support, connection, and belonging (Byrd & Andrews, 2016; Hughes et al., 2015; Louque et al., 2017; Nelson et al., 2022). Schools with greater economic disadvantage have been associated with more negative academic outcomes for children due to issues in funding, resources, and staffing (Donnellan et al., 2013). This resource deficit in schools that primarily serve minoritized students is well documented in the literature (Fahle et al., 2020; Reardon, 2016). For example, under-resourced and underfunded schools are more likely to have teachers with fewer years of overall teaching experience and lower education, certification, and training levels compared to well-resourced and properly funded schools (Fahle et al., 2020). Additionally, these schools may offer a less rigorous curriculum that does not challenge students, which negatively impacts their academic achievement (Fahle et al., 2020; Reardon, 2016).

The Present Study

The present study examined changes in Black late elementary school-aged children’s academic engagement and motivation in relation to school racial composition, quality, and economic disadvantage. We examined three research questions. First, does the academic engagement and motivation of Black children change over the course of late elementary school, and if so, are there similar changes across all domains of engagement and motivation? Based on previous research, we predicted that there would be significant declines in all domains of academic engagement and motivation for Black children during late elementary school. To examine this association, a latent change score model was used to model change in children’s academic engagement and motivation from Wave 5 to Wave 6. Second, we asked are these changes similar or different for Black children attending schools of differing racial composition? We predicted that Black children attending predominantly Black schools would report higher levels of academic engagement and motivation overall and would also demonstrate less decline, and possibly even increases, during this time compared to Black children attending non-Black schools. A latent change score model was also used to examine changes in children’s academic engagement and motivation from Wave 5 to Wave 6 in relation to school racial composition at Wave 5. Last, we asked if school level factors such as school quality and school economic disadvantage moderated the relation between school racial composition and academic engagement and motivation. We predicted that this association would be significantly moderated by school quality and school economic disadvantage such that Black children attending high quality schools and less economically disadvantaged schools would demonstrate less declines in academic engagement and motivation compared to Black children attending low quality schools and schools with greater economic disadvantage. Using the latent change model, we assessed two interactions using a moderation approach – the interaction between school racial composition and school quality as well as the interaction between school racial composition and school economic disadvantage. Models were fit in Mplus using structural equation modeling (Muthén & Muthén, 1998–2017).

Method

Participants

The current study is a secondary analysis of data from a longitudinal study of Black and Latinx children from low-income families. Families were recruited from a large city in the southwest United States and had a target child between 29 and 31 months of age at the first time point (Wave 1, 2009–2011), at least one parent who self-identified as Black or Latinx, and a family income below 200% of the federal poverty level. A total of 407 families were enrolled in the study, of which 183 (45%) were Black.

Starting in Wave 5, the study employed a planned missing design in which study families were assigned to be interviewed in some waves but not others. In order to be eligible for participation starting in Wave 5, families had to satisfy the following criteria: completed at least two data collection points in Waves 1–4, child had not been diagnosed with a significant developmental disability, and family had not voluntarily withdrawn from the study. Of the 183 Black families enrolled in the study, one was excluded from follow-up due to a diagnosed disability, and two children died before the follow-up. In addition, six Black families voluntarily withdrew from the study, and 22 families only completed the first time point of data collection. Of the 152 Black families eligible for follow-up, 114 were scheduled to be interviewed in Wave 5 and/or Wave 6. Of these scheduled families, 94 (82%; approximately half of the initial sample) completed their assigned data collection. Before conducting analyses to address the study aims, missing data for all Black children in the sample (excluding the one child diagnosed with a developmental disability) were imputed using auxiliary variables derived from a block of study variables using principal components analysis (Howard et al., 2015). A total of 100 imputed data sets were utilized. Density plots comparing the distribution of each variable in the imputed data set versus the distribution of the observed data were examined to verify the quality of the imputation.

Characteristics of the study sample of 182 children are displayed in Table 1. Children who identified as multiracial/multiethnic were not included in the analysis of Black children in the current study. At Wave 5, children ranged in age from 9.6 to 10.9 years (Mean = 10.33 years, SD = 0.25). On average, Wave 6 data collection occurred 8–9 months after Wave 5 (Mean = 8.6 months, SD = 1.4). At Wave 6, children ranged in age from 10.4 to 12.1 years (Mean 11.3 years, SD = 0.38). There were 100 boys (54.9%) and 82 girls (45.1%). Approximately 48.4% of families had an average household income of less than 50% of the federal poverty level. Families who did not complete data collection at Wave 5 did not differ in terms of family income, child gender, or primary caregiver education compared to families who did complete data collection.

Table 1 Characteristics of study sample (N = 182)
Full size table

Measures

Academic Engagement

The child report of academic engagement scale was used to assess children’s school engagement at Waves 5 and 6 (Wang et al., 2011). It is comprised of three subscales: behavioral, emotional, and cognitive (Wang et al., 2011). In the seven item behavioral subscale, children were asked to assess how often they followed the school rules and participated in class. Children responded on a 5-point Likert scale (1 = Almost Never to 5 = Almost Always). Six of the seven items in this scale were reverse coded such that higher scores indicate higher attentiveness to schoolwork and less engagement in behavioral problems (e.g., being in a physical fight). In the 6-item emotional subscale, children were asked their feelings about the school they attended, such as their feelings of acceptance, interest, and enjoyment. Children responded on a 5-point Likert scale (1 = Strongly Disagree to 5 = Strongly Agree). Lastly, in the five-item cognitive scale, children were asked the extent of their self-regulation learning skills. Children respond to the cognitive scale items on a 5-point Likert scale (1 = Almost Never to 5 = Almost Always).

Confirmatory factor analysis (CFA) was used to examine the fit of the factor structure proposed by Wang and colleagues (2011) for the present sample. Initial model fit was poor, χ2 (150) = 303.40, p = 0.00, CFI = 0.84, RMSEA = 0.07, SRMR = 0.05. Items that did not load significantly on either scale or that cross-loaded were dropped from the model. Cross-loadings were defined as having a loading of more than 0.50 on more than one factor. On the behavioral subscale, two of the seven items were dropped, “Gets school work done on time” and “Skipped class”. None of the eight cognitive subscale items were dropped. On the emotional subscale, four of the eight items were dropped, “Schooling is not so important for kids like me”, “I learn more useful things from friends and relatives than I learn in school”, “Getting a good education is the best way to get ahead in life”, and “I learn a lot from my school-work”. To evaluate model fit, comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean residual (SRMR) were primarily used. According to Kline (2016), CFI should not go below 0.85 and RMSEA and SRMR should not go above 0.10. After drop** these items, the model fit indices improved significantly, χ2 (17) = 180.64, p = 0.00, CFI = 0.92, RMSEA = 0.05, SRMR = 0.06.

Academic Motivation

The child report of academic motivation scale was used to assess students’ perception of their achievement motivation at Wave 5 and 6 (Eccles et al., 1993). Wang and Eccles (2013) identified two subscales: academic self-concept and subjective task valuing of school learning. The academic self-concept subscale consisted of 5 items that assess students’ perceptions of their ability to learn and succeeded in certain subjects. Children were presented with a list of academic tasks and asked how good they are at the tasks listed. Children responded on a 7-point Likert scale (1 = Not at All Good to 7 = Very Good). The subjective task valuing of school learning subscale consisted of 6 items that assess how much the student values academic achievement. Children responded on a 7-point Likert scale (1 = Not an Important Reason to 7 = A Very Important Reason).

A CFA was used to examine the fit of the factor structure proposed by Wang and Eccles (2013) for the present sample. Responses were not evenly distributed across the 7 response categories. Therefore, before the model was fit, items were recoded into three-point Likert scales with 1–3 = 1, 4 = 2, and 5–7 = 3. Initial model fit was poor, χ2 (43) = 146.15, p = 0.00, CFI = 0.61, RMSEA = 0.10, SRMR = 0.07. Items that did not load significantly on either scale or that cross-loaded were dropped from the model. The academic self-concept items that were dropped included “How good at math are you?”, “How good are you in other school subjects”, and “How good are you in sports?” The subjective task valuing of school learning item that was dropped was “Why do you go to school?”. To evaluate model fit, comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean residual (SRMR) were primarily used. According to Kline (2016) CFI should not go below 0.85 and RMSEA and SRMR should not go above 0.10. After drop** these items, the model fit indices improved significantly, χ2 (17) = 21.53, p = 0.20, CFI = 0.98, RMSEA = 0.03, SRMR = 0.04.

School Racial Composition

The school racial composition variable at Wave 5 was derived from a data set for Texas schools developed by the Texas Education Agency (TEA). The current study used the percentage of students in the school that identified as Black as a measure of school racial composition. Previous research exploring school ethnic-racial composition and diversity has used cut points above 60% to determine if a school is predominantly composed of one ethnic-racial group compared to another. For example, in a study reported by Johnson (2005) exploring children’s development, schools with a population of more than 75% Black students were considered predominantly Black. Studies exploring ethnic-racial congruence in schools have developed high and low diversity ratio categories based on means and standard deviations with high diversity ratios being 1 SD above the mean (e.g., greater than 86%) and low diversity ratios 1 SD below the mean (e.g., less than 32%; Benner & Graham, 2007; French et al., 2000; Parris et al., 2018). Lastly, federal education research has labeled the school racial composition as predominantly Black when a minimum of 60% of students in a school identify as Black (Bohrnstedt et al., 2015).

Based on recommendations of previous research and the examination of frequencies for Black and Latinx youth among the minority or majority at their school, a cut point of 75% was used in the current study. If 75% or more students in the school identified as Black, the school was classified as “predominantly Black”. If less than 75% of the students in the school identify as Black, then the school will be classified as “not predominantly Black”. In the model, this variable is operationalized as a categorical variable with “not predominately Black” schools categorized as 0 and “predominately Black” schools categorized as 1.

Moderators and Covariates

Moderators and covariates included school quality, school percentage economically disadvantaged, and family income. School quality and percentage economically disadvantaged were also derived from the TEA database. TEA provided the percentage of children in the school who qualified for free or reduced lunch as an index of the proportion of economically disadvantaged children in the school. To assess school quality, we followed the work of the public policy advocacy agency Children at Risk (n.d.) to estimate a campus performance score by regressing school level pass rates for state standardized tests on the proportion of economically disadvantaged students in the school. For each school, the difference between the actual and predicted pass rate in the school (e.g., the residual) was added to the actual pass rate to create a campus performance score. Schools that had higher pass rates than what would be predicted by the economic characteristics of the school had a higher campus performance score, while those with pass rates lower than what would be predicted had a lower campus performance score. Finally, a family income-to-needs ratio was calculated at each of the first four time points by dividing family income by the federal poverty level for a family of a similar size and averaged across time points. Family income data were not collected at either Wave 5 or Wave 6. Bivariate correlations between study variables can be found in Table 2.

Table 2 Bivariate correlations
Full size table

Results

Latent Change Score Model

A latent change score model was used to model change in students’ academic engagement and motivation from age 10 (Wave 5) to age 11 (Wave 6). Two unconditional models were fit: a no change model and a constant change model. Chi-square difference tests were used to determine the best fitting model for the academic engagement and motivation subscales (Satorra, 2000). The best fitting model was used to predict the relation between school racial composition and additional study covariates with students’ academic engagement and motivation.

The unconditional constant change model was a fully saturated model and fit the data significantly better than the no change model for multiple academic engagement and motivation subscales. To answer our first research question, does the academic engagement and motivation of Black children change over the course of late elementary school, and if so, are there similar changes across all domains of engagement and motivation, chi-square difference tests across the models were used. Chi-square difference tests indicated evidence for change in two of the academic engagement subscales, emotional, \({\chi }^{2}\)Δ= 14.24, and cognitive, \({\chi }^{2}\)Δ= 16.63, as well as the one of academic motivation subscales, self-concept \({\chi }^{2}\)Δ = 55.87 (all ps <0.05). On average, the emotional and cognitive engagement subscales as well as the academic motivation self-concept subscale showed a decrease from age 10 to age 11. For both the academic engagement cognitive and emotional subscales as well as the academic motivation self-concept subscale, the test statistic exceeded the chi-square critical value (above p <0.05) indicating the constant change models fit the data significantly better than the no change models. Due to the current study’s hypotheses specific focus on change, we focus on examining the predictors of academic engagement (emotional and cognitive) and academic motivation (self-concept) that specifically evidenced change based on chi-square difference tests in the next section.

Correlates of Change in Academic Engagement and Motivation

Emotional Engagement

The results for emotional engagement are shown in Table 3. At age 10, school racial composition, school economic disadvantage, and family income did not significantly predict the intercept for students’ emotional engagement, but school quality did significantly predict the intercept for students’ emotional engagement such that higher school quality scores were associated with lower emotional engagement. There were significant main effects for school quality and school economic disadvantage on the slope of emotional engagement. Specifically, greater school quality and economic disadvantage was associated with less declines in the slope of emotional engagement.

Table 3 Parameter estimates for school predictors of latent change in academic engagement
Full size table

Additionally, in the models testing for moderation, there was a significant interaction between school quality and school racial composition on the change in emotional engagement, and once this interaction was included in the model, the main effect for economic disadvantage was no longer significant. To probe the interaction between racial composition and school quality, we plotted the interaction between school quality and school racial composition. As seen in Fig. 1, attending a predominantly Black school was associated with greater declines in emotional engagement from age 10 to age 11 for children attending a lower quality school; however, these declines were not evident for children attending higher quality schools. Children attending non-Black schools displayed declines in emotional engagement regardless of school quality, although somewhat less so for Black children attending higher quality non-Black schools. Simple slopes analyses among lower quality schools demonstrate greater decline in emotional engagement for children attending predominantly Black schools compared to children at non-Black schools (t = −2.40, p = 0.02). Further, simple slopes among higher quality schools showed no significant differences between Black children attending predominantly Black schools compared to children attending non-Black schools (t = 0.69, p = 0.49). These results conflict with our second hypotheses surrounding children attending predominantly Black schools, reporting higher levels and academic engagement and demonstrating less decline. However, they support our second hypothesis that school quality moderates the association between academic engagement and school racial composition.

Fig. 1
figure 1

Change in Emotional Engagement Across Time by School Racial Composition. Note. High school quality is represented at one standard deviation above and below the mean, and low school quality is represented at one standard deviation below the mean

Full size image

Cognitive Engagement

The results for cognitive engagement are also displayed in Table 3. School racial composition and school quality did not significantly predict students’ cognitive engagement at age 10. However, greater school economic disadvantage did significantly predict lower levels of cognitive engagement at age 10. Furthermore, school characteristics did not significantly predict changes in student cognitive engagement from age 10 to age 11. Results also indicated no significant interactions between school racial composition with school quality nor between school racial composition and school economic disadvantage. These results do not support our hypotheses surrounding children attending predominantly Black schools reporting higher levels of academic engagement and demonstrating less decline over time when moderated by additional school variables.

Academic Motivation Self-Concept

The results for academic motivation self-concept are reported in Table 4. Due to negative error variance in the academic motivation self-concept model, the residual variance of the slope was constrained to zero. Greater school economic disadvantage significantly predicted lower levels of academic motivation self-concept at age 10 while school racial composition and school quality did not. However, school racial composition and school economic disadvantage predicted significant declines in student academic motivation self-concept from age 10 to age 11. Specifically, children experienced greater declines in academic motivation self-concept when they attended predominantly Black schools (p = 0.05). In addition, counter-intuitively, higher levels of economic disadvantage predicted less steep declines in academic motivation self-concept. Lastly, the results indicated no significant interaction between school racial composition, school quality, nor school economic disadvantage. These results conflict with our second and third hypotheses surrounding children attending predominantly Black schools reporting higher levels of academic motivation and demonstrating less decline over time when moderated by additional school variables.

Table 4 Parameter estimates for school predictors of latent change in academic motivation
Full size table

Discussion

This study examined the moderating role of school racial composition on Black children’s academic engagement and motivation over the course of one year in late elementary school. To our knowledge, previous research has not examined changes in academic engagement and motivation for minoritized students in late elementary school but rather has primarily focused on children in middle school, high school, and college (Chavous et al., 2018; King et al., 2015; Wang & Eccles, 2013; Wang et al., 2011). In addition, longitudinal studies of engagement and motivation have primarily focused on White children during middle school (Dotterer et al., 2009). We developed three hypotheses based on previous research that examined the saliency of children’s academic beliefs and competencies in early elementary school and the influence of racially matched schools on Black children’s school experiences and academic trajectories (Benner & Graham, 2013; Entwisle & Alexander, 1998). First, does the academic engagement and motivation of Black children change over the course of late elementary school, and if so, are there similar changes across all domains of engagement and motivation? Second, are these changes similar or different for Black children attending schools of differing racial composition? Third, is there an interaction and moderation between school racial composition and additional school level factors of school quality and school economic disadvantage on academic engagement and motivation? Although we had hypothesized that there would be significant declines in all domains of academic engagement and motivation for Black children during late elementary school, with children attending predominantly Black schools demonstrating less decline, the research findings only partially supported this hypothesis.

First, we found that changes in emotional engagement in late elementary school differed based on the school characteristics of racial composition, quality, and economic disadvantage. At age 10 school quality significantly predicted the intercept for students’ emotional engagement. Significant main effects for school quality and economic disadvantage on the slope of emotional engagement were also identified. Consistent with the broader literature on changes in academic engagement during this developmental period (King et al., 2015; Wang et al., 2011; Wang & Eccles, 2013), most children in the sample displayed declines in emotional engagement over one year in relation to school racial composition. However, this was not the case for Black children attending higher quality predominantly Black schools, who displayed no change in emotional engagement over this period. In contrast, Black children attending lower quality predominantly Black schools displayed the steepest decline.

To understand these findings surrounding school racial composition, school quality, and emotional engagement, we must consider how school quality is operationalized in the current study. As stated previously, school quality was based on the school-level pass rates on standardized tests adjusted for the proportion of economically disadvantaged students in the school. Schools that have higher pass rates than predicted based on school economic characteristics receive higher ratings on school quality. Schools with higher ratings of school quality are challenging the narrative that posits greater economic disadvantage can lead to negative academic outcomes due to inequity in resources and less social and cultural capital (Donnellan et al., 2013). Further, research has demonstrated that students’ positive ethnic-racial identity development (Chavous et al., 2018), perceived support from teachers and peers (Kiefer et al., 2015), and sense of belonging can buffer the potential negative influence of school economic disadvantage on children (Graham, 2020), which has often been found in ethnically and racially matched schools (Graham, 2020; Graham, 2018). In the current study, emotional engagement assessed Black students’ feelings of satisfaction with their school, their sense of belonging, and the importance of excelling in school. Therefore, one possible explanation for our findings is that the students in our sample have support within their schools that protect them against the adverse influences of school economic disadvantage.

Findings regarding academic motivation self-concept approached significance and indicated that school racial composition and school economic disadvantage predicted declines in student academic motivation self-concept from age 10 to age 11 such that children attending predominantly Black schools experienced greater declines in academic motivation self-concept, while higher levels of economic disadvantage predicted increases in students’ academic motivation self-concept. Extant research has consistently emphasized the importance of ethnically and racially matched schools, as they provide students with the support necessary to bolster their social, emotional, and academic outcomes (Graham, 2018). However, there is empirical evidence of the opposite effect, that is, that school diversity can lead to improved student mental health, better school adaptation, and the development of cross ethnic and racial friendships while also decreasing students’ feelings of vulnerability and victimization in school (Graham, 2018; Juvonen et al., 2018). Our results are consistent with studies emphasizing the benefits of attending racially and ethnically diverse schools. The mixed results may be explained by additional school experiences (e.g., teacher-student relationships, peer victimization) that can moderate the role of school diversity on student outcomes. However, it is unclear why the counter-intuitive results of less steep declines in academic motivation self-concept in higher economically disadvantaged schools were found. There was a main effect of economic disadvantage on the intercept such that greater economic disadvantage was associated with lower academic motivation self-concept at age 10 which may leave less room for further declines from ages 10 to 11. There may be additional unmeasured school factors that may buffer against the negative influence of economic disadvantage (Dotter et al., 2009; Kiefer et al., 2015). Regardless, this finding requires replication.

Although our study hypotheses focused on changes in academic engagement and motivation from age 10 to 11, it is important to note that school economic disadvantage significantly predicted lower levels of cognitive engagement and academic motivation self-concept at age 10. These findings underscore the disparities in school funding and resources that occur in schools with greater economic disadvantage. The resource deficit occurring in schools that primarily serve minoritized students is well documented in the literature (Fahle et al., 2020; Reardon, 2016). Under-resourced and underfunded schools are more likely to have teachers with fewer years of overall teaching experience and lower education, certification, and training levels compared to well-resourced and properly funded schools (Fahle et al., 2020). In addition to less teacher training, these schools also may offer a less rigorous curriculum for their students and inadequate school facilities which leads to disruptions in student learning time, lower standardized test scores, and overall declines in minoritized students’ academic achievement (Fahle et al., 2020; Reardon, 2016). Taken together, these school factors contribute to the Black, Brown, and White achievement gap, as well as exacerbate educational inequalities that can lead to declines in the cognitive engagement of Black children (Fahle et al., 2020; Reardon, 2016).

Despite research indicating the interwoven nature of the dimensions of academic engagement and motivation, we did not identify any changes in students’ behavioral engagement nor in students’ academic motivation subjective task valuing from age 10 to age 11 (Fredricks et al., 2004). Additional school characteristics that are important for students’ engagement and motivation such as school climate, school equity, relationships with teachers and peers, and experiences of peer victimization and bullying that were not measured in the present study could contribute to changes in behavioral engagement and academic motivation subjective task valuing (Graham, 2020; Wang et al., 2015). These school characteristics represent school quality measures that schools are required to have under the Every Students Succeeds Act (ESSA). Without the inclusion of these measures, we are unable to explore school quality measures specific to the location of the current study and how these measures can influence student success. These null findings are also consistent with studies that identify the establishment of academic attitudes and behaviors during early elementary school which remain salient throughout students’ time in school (Entwisle & Alexander, 1998). Last, we faced the challenge of defining a specific cut point for predominantly Black schools. Future longitudinal research may benefit from further exploration of a cut at which predominantly Black schools become operational by using a larger sample as this was outside of the scope of the current study.

Conclusion

Understanding the role of school level factors in Black students’ academic engagement and motivation is important for improving academic achievement outcomes. For example, Black students who attend schools with greater economic disadvantage and lower levels of school quality are more likely to experience declines in their academic engagement and motivation (Fahle et al., 2020; Reardon, 2016). Further, students who attend racially matched schools (i.e. predominantly Black schools) with less economic disadvantage and higher levels of school quality are more likely to experience increases in their academic engagement and motivation (Jones, 2018; Williams et al., 2017; Yull & Wilson, 2018).

In this study we found that attending a predominantly Black schools was associated with declines in emotional engagement from age 10 to age 11 except when children attended higher quality schools. Our findings emphasize that predominantly Black schools that support achievement levels despite economic stressors can support positive youth engagement during this critical developmental period. Future work in this area would benefit from further exploration of school-level factors that may contribute to students’ academic engagement and motivation. By centering Black student’s experiences within the school setting, we can improve our understanding of how to develop more inclusive and affirming learning environments for Black children.