Adolescents who have a purpose are more academically engaged than their peers (Hill et al., 2018; Klein et al., 2019). Purposeful adolescents compared to their counterparts get better grades (Adelabu, 2008), perform better on intelligence tests (Minehan et al., 2000), hold more positive academic self-identities (Dukes and Lorch, 1989), and are more academically motivated than are their peers (Nurmi, 1991). When an adolescent has a meaningful purpose, not only does it cultivate a future orientation, it guides and motivates present behavior toward achieving future-oriented goal(s) (Damon et al., 2003). For example, adolescents who were motivated by future goals exhibited better studying behaviors (Horstmanshof and Zimitat, 2007) and greater intrinsic interest in academics (Brown and Jones, 2004).

The BTS aspect of purpose highlights an intention to pursue aims that transcend the self and impact other people, the greater society, or the world. Adolescents who are committed to a purpose(s) that would impact others, possibly their own families, communities, and/or social groups, may be more likely to persist despite obstacles and adversity (Bronk and Finch, 2010; Sumner et al., 2018). Adolescents who experience marginalization may commit to a purpose that includes a prosocial (i.e., BTS) goal of improving the issues faced by their social group through civic or political action (Malin et al., 2015; Godfrey and Cherng, 2016). Additionally, a prosocial, self-transcendent purpose was associated with academic self-regulation and increased deeper learning behaviors (Yeager et al., 2014).

Two centrally notable variables that seem to work alongside purpose to encourage academic engagement and performance are goal orientation (Bronk et al., 2009) and self-efficacy (DeWitz et al., 2009). For example, research suggests that purposeful adolescents are more likely to be motivated toward intrinsic goals rather than extrinsic goals (Damon et al., 2003; Damon, 2009). Moreover, adolescents who are motivated by intrinsic goals tend to have a higher sense of self-esteem and self-concept (Middleton and Midgley, 1997; Turner and Patrick, 2004).

Researchers have drawn connections between the development of purpose and intrinsic vs. extrinsic motivation (Liang et al., 2016). Adolescents are bombarded by societal values that prioritize extrinsic goals (e.g., pushing ahead of others and self-focused success for the sake of power, prestige, and money). Yet, some adolescents demonstrate more intrinsic motivations that are associated with purpose. Purpose, by definition, is a life aspiration that is both personally meaningful and beneficial to society (Damon et al., 2003).

Understanding how adolescents develop these intrinsic vs. extrinsic orientations toward their long-term goals and aspirations are a key to understanding how to foster the BTS aspect of purpose (Liang et al., 2016). Dweck (1986) developed a goal orientation framework that explains students’ approaches to goals as falling into two main categories: (1) performance and (2) learning or mastery. A performance goal orientation is bound by social comparison and is characterized by a desire to gain positive judgments and/or avoid negative judgments (Dweck, 1986). This motivation differs from a genuine, intrinsic interest in the goal, and a desire to become more competent and knowledgeable (Pintrich, 2000).

Performance goal orientation includes two subcategories: (1) performance-approach and (2) performance-avoidance. Individuals with a performance-approach orientation aim to outperform their peers in order to gain recognition (i.e., play to win), while those with a performance-avoidance goal orientation try to avoid losing in order to avoid looking foolish or incompetent (i.e., play not to lose) (Elliot and Harackiewicz, 1996). Performance-avoidance goal orientation has been positively correlated with test anxiety and a reluctance to seek help when needed (Middleton and Midgley, 1997), and it is negatively correlated with learning and academic performance (Elliot and Church, 1997; Payne et al., 2007).

Research on the adaptiveness of a performance-approach orientation is mixed. Some studies suggest that when paired with a mastery goal orientation, performance-approach orientation can be adaptive, but only under certain conditions (e.g., lower fear of failure and uncertainty) (Elliot and Church, 1997; Darnon et al., 2007). More recent research suggests that adolescents, especially girls from high socioeconomic status communities experience overwhelming academic and psychosocial stress due in part to the pressure to out-achieve their peers and to compete for limited privileges (e.g., entry at prestigious universities) (Luthar et al., 2013; Spencer et al., 2018).

In contrast to this emphasis on social comparison among those with a performance goal orientation, those with a mastery goal orientation were motivated by a desire to increase their competence, knowledge, and understanding (Pintrich, 2000). Mastery goal orientation is based in high competency expectancies and intrinsic motivation (Elliot and Church, 1997), and it has been positively associated with effective learning strategies, such as self-regulatory behaviors (Payne et al., 2007) and high levels of self-efficacy and interest (Middleton and Midgley, 1997).

Research has demonstrated that adolescents who disproportionately valued and attained extrinsic goals rather than intrinsic goals (e.g., personal growth, close relationships, community involvement) experienced poorer mental health and a reduced sense of self-efficacy (Liang et al., 2016). These adolescents pay a heavy personal price when their definitions of success tie their self-worth to performance over purpose. In contrast, those who pursue intrinsic goals, such as cultivating a sense of purpose and contributing to the good of others, may be less subject to the pressures of competition and social comparison, and thus enjoy greater health and self-esteem (Lyman and Luthar, 2014; Spencer et al., 2018). Adolescents who pursued purpose over performance described being less driven by fears of failure, and more driven by passions to fulfill a calling. Thus, despite the surrounding achievement pressures, they remained centered and self-assured.

Thus, goal orientation and self-efficacy appear to play a combined role in academic engagement and outcomes (Lent et al., 1996; Ryan and Deci, 2000). More specifically, self-efficacy refers to individuals’ judgments of their ability to organize and enact behaviors required to perform well (Bandura, 1977). In other words, it is an individuals’ sense of whether they are able to perform in the way they hope. It is a dynamic set of self-beliefs that are linked to different areas of functioning in different domains, as one person cannot feel competent at all tasks (Bandura, 1997; Lent, 2012). Thus, an individual may have a strong sense of self-efficacy in one area of performance, such as an academic subject area, but a low sense of self-efficacy in another domain, such as social skills. Adolescents who are highly efficacious may have a wide array of experiences that contribute to deeper, more accurate understandings of their capabilities that inform which behaviors to further pursue. Ultimately, self-efficacy beliefs are thought to lead to corresponding behaviors. For example, a strong sense of self-efficacy regarding one’s academic capabilities is necessary to motivate individuals to engage in behaviors that lead to academic achievement (Jinks and Morgan, 1999).

One explanation for why self-efficacy is tied to behavior is that when adolescents are highly efficacious, they are more likely to persevere in pursuing their ultimate aims or long-term goals. Thus, it is expected that adolescents who are highly efficacious compared to their counterparts may be more likely to pursue long-term goals or ultimate aims that are tied to academic success. For instance, adolescents who are highly efficacious are more likely to complete their education and engage in behaviors that prepare them for a range of career options (Zimmerman, 1990; Bandura et al., 2001). More specifically, self-efficacy beliefs are associated with reduced delinquency behaviors and increased academic grades (Carroll et al., 2009). Ultimately, these beliefs have positive long-term effects on adolescents’ academic success (Hwang et al., 2016), especially in school contexts characterized by a growth mindset (Høigaard et al., 2015). This latter finding is particularly relevant in the current study given that the MPOWER program is designed to create a classroom environment that fosters a sense of BTS purpose rather than just success for the sake of social comparison. MPOWER aims to increase purpose and decrease performance-avoidance goal orientations. Moreover, the program aims to increase self-efficacy beliefs that ultimately lead to increases in academic achievement (Carroll et al., 2009; Høigaard et al., 2015; Hwang et al., 2016).

This school-based program (for more details on MPOWER, see Klein et al., 2019) was designed to guide students in exploring purpose through a weekly, 50-min classroom-based, yearlong curriculum and through supportive relationships such as additional one-to-one mentoring by a guidance counselor and a classroom community of co-participants.

Moreover, MPOWER incorporates the 4 P’s of purpose framework by using the curriculum and these formative relationships to help adolescents identify and cultivate their passion by identifying core values and considering whether these core values align with daily activities or pursuits. As adolescents identify their passion(s), they begin reflecting on their propensity by identifying their strengths and skills relevant to pursuing their purpose; if adolescents believe they do not possess the propensity, then they identify ways to develop the strengths and skills needed or consider alternative aspirations that may be a better fit for them. Finally, adolescents are also supported to explore potential prosocial benefits of their aspirations, and asked to reflect on the impact of their purpose on people’s lives.

With respect to fostering the BTS aspect of purpose, MPOWER uses a number of exercises that helped participants understand how fulfilling their aspirations may impact the world. For example, participants played an adapted version of “The Dictator Game” which has been shown to promote altruistic intentions and prosocial behaviors in youth (Benenson et al., 2007). More specifically, participants completed multiple rounds of the aforementioned game where they can give or receive money with their peers, which is tied to their ranking. The majority of participants exhibited prosocial behaviors by helping their peers at the expense of their own ranking. Upon completion of the activity (and other activities similar to it), participants explore their own desired prosocial impact or contributions they would like to make. The BTS aspect of purpose is emphasized in MPOWER because it was created to improve academic engagement for adolescents with marginalized identities. Familial struggles (e.g., precarious immigration status, low SES) may motivate adolescents to pursue a purpose that helps their family and/or themselves escape adversity associated with their marginalized backgrounds (e.g., poverty) (Liang et al., 2017; Gutowski et al., 2018). Finally, a brief psychological intervention that promoted a self-transcendent purpose (i.e., BTS purpose) improved academic performance over several months (Yeager et al., 2014).

In addition to fostering BTS intentions, participants were asked to engage in personal reflection through journaling and writing exercises which primed them to think about their core values and character strengths (Cohen et al., 2009). Experiential group activities were used as a way of increasing a sense of community and academic performance among the participants (Nagaoka et al., 2013). These activities served as a starting point for participants’ one-to-one mentoring sessions with the MPOWER instructor through which they were given an opportunity to reflect more deeply and share personal experiences that they may have been reluctant to share in the larger context of the classroom. These individual sessions focused on enabling participants to discuss how their lived experiences, families, cultures, and communities informed their long-term aspirations/purpose. Participants were taught the language and framework of purpose, and engaged in introspective reflection in order to help them create a narrative in which their past and present experiences could be connected with their future aspirations (Wilson, 2011).

The current study examined the effectiveness of the MPOWER program for increasing participants’ BTS aspect of purpose, mastery goal orientation, self-efficacy, and academic achievement. We hypothesized that students participating in the MPOWER program would show: (a) increased BTS purpose and academic achievement (i.e., grade-point average); (b) increased mastery goal orientation and self-efficacy; and (c) decreased performance-approach and performance-avoidance goal orientation.

This study was pre-approved by the Institutional Review Board at Boston College (IRB Protocol: 16.145.02-3) which required pre-approval by the participating school’s principal and district superintendent and pre-registered the measures that were used to assess pre-test and post-test scores. This study focused on a sample of 12th grade students (N = 94, control N = 42, 61.7% female) from a U.S. public high school in the northeast. As it is the only high school within the district, its students are representative of the district’s diverse working- to middle-class community. Additionally, this sample represented 30% of the schools’ student body in the 12th grade. The MPOWER program was delivered to only seniors. Figure 1 displays how participants were randomly assigned to the program (i.e., MPOWER) and control group (i.e., academic enrichment/study hall class). The sample was predominantly female due to female students exhibiting greater academic performance than male students at this high school. In order to qualify for the study hall period that MPOWER and the academic enrichment class were offered during, students must be on track to graduate and have a free period that can be filled by the elective.

The participants in the control group were given an alternative curriculum that focused on increasing academic engagement and improving academic performance. In addition to giving participants allotted time in school to complete their academic work, they were also provided a curriculum that taught them a variety of non-cognitive skills that are correlated with improved student outcomes. These non-cognitive skills included intrapersonal skills, such as task, organizational, and time management as well as social emotional skills, including stress regulation, impulse control, and capacity for optimism. Moreover, participants in the control were offered group and individual tutoring as needed. The control condition is viewed as “treatment as usual” because this is typically what students receive during this elective period; the MPOWER intervention was allowed to be piloted during the elective period.

The sample included participants who self-identified as White/Euro-American (N = 42, 64.2% female), African-American/Black (N = 11, 45.4% female), Haitian (N = 7, 71.4% female), Caribbean (N = 1, 100% female), Latino or Hispanic (N = 7, 57.1% female), Brazilian or Portuguese (N = 12, 58.3% female), Asian (N = 8, 50.0% female), Asian-Americans (N = 3, 66.7% female), Pacific Islander (N = 1, 100% female), and Native American (N = 2, 100% female). Based on parental education level, the sample represented a range of socioeconomic backgrounds (8.5% of parents had a master’s or doctoral degree, 37.2% of parents had an associate’s or bachelor’s degree, 37.2% of parents graduated from high school, 17.0% of parents did not graduate from high school). Participants were given pre-test surveys prior to their respective interventions that assessed their self-efficacy, goal orientation, and purpose; they also completed post-test surveys after their respective interventions (see section “Measures” for more detail).

In anticipation of data collection, student participants’ parents received a co-signed letter from the principal and research team, as well as informed passive consent forms. To deny consent for their child, parents were asked to return the signed letter, email a school administrator, or email a member of the research team.

The survey was administered twice during a single class period, first as a pre-test in September 2017 and then as a post-test in May 2018. The MPOWER instructor and research assistants arrived at each class with internet enabled laptops. The study was described to students, and they provided informed assent. All students whose parents had given passive consent were invited to participate in the study. The MPOWER instructor and a research assistant were present to assist students and answer questions that arose.

Students with both a pre-test and post-test were included in the final sample. Those who missed either the pre-test (N = 10) or post-test (N = 13) data collection were excluded from this study. Any other missing data were replaced using an expectation maximization (EM) algorithm to create a single new dataset where the missing values are imputed with maximum likelihood values (Acock, 2005). EM is expected to produce unbiased parameter estimates when the data are missing cases at random (MCAR) (Acock, 1997 as cited in Fox-Wasylyshyn and El-Masri, 2005). Little’s MCAR test (Little, 1988) was conducted to see whether the EM algorithm converged; our analysis suggested that our data was MCAR and converged.

Separate one-way analyses of covariance (ANCOVAs) were conducted to test the effects of the MPOWER program on the outcomes of interest. Four outcome measures (purpose, goal orientation, self-efficacy, GPA) were used with pre-test scores for each outcome included in the analyses as covariates. Covariates were included in these analyses to partial out the effects of pre-test scores. The students in this study were randomly assigned to the intervention and control group; thus, group differences were not included as covariates. ANCOVA tables include the estimated marginal means for all analyses. An alpha of 0.05 was established as the criterion for significant results.

Analyses were conducted using SPSS 28.0, which produces an index of effect sizes using the partial eta-square; however, we calculated omega-square which is considered a less biased alternative with smaller samples (Okada, 2013). Omega-square can be used like Cohen’s d, (Cohen, 1988) to assess the effect of the intervention. The metric for assessing omega square for small, medium, and large effect sizes are 0.01, 0.06, and 0.14, respectively (Fields, 2013).

A one-way ANCOVA was conducted to compare the effectiveness of the MPOWER intervention while controlling for pre-test scores, as shown in Table 1. There was not a significant difference in composite purpose post-test scores [F_{(1, 91)} = 3.32, p > 0.05, ω² = 0.012] between the conditions. Post hoc tests showed there was a trend toward a difference in the composite purpose score between the MPOWER and control group, but it was not significant (p = 0.071).

Analyses with the purpose measure subscales demonstrated that there was not a significant difference across program and control groups post-test scores for purpose meaningfulness [F_{(1, 91)} = 1.34, p > 0.05, ω² = 0.002] and purpose engagement [F_{(1, 91)} = 0.313, p > 0.05, ω² = −0.005]. However, there was a significant difference in post-test BTS purpose [F_{(1, 91)} = 8.44, p < 0.01, ω² = 0.042]. The estimated marginal means showed that MPOWER students reported significantly higher BTS purpose post-test scores, M = 4.14, 95% CI [3.94, 4.34] compared to control students M = 3.71, 95% CI [3.49, 3.93].

A one-way ANCOVA controlling for pre-test scores demonstrated a significant difference in self-efficacy post-test scores [F_{(1, 91)} = 6.69, p < 0.05, ω² = 0.052]; MPOWER students reported significantly higher self-efficacy post-test scores, M = 32.57, 95% CI [31.01, 34.12] than did control students, M = 29.48, 95% CI [27.75, 31.22].

A one-way ANCOVA controlling for pre-test GPA demonstrated a significant difference in GPA at post-test [F_{(1, 91)} = 10.42, p < 0.005, ω² = 0.001]. MPOWER students reported significantly higher GPA post-test scores, M = 3.02 than did control students, M = 2.96.