Abstract
Educating students for democratic citizenship in an increasingly pluralistic society (i.e., civic education) is an essential task of higher education (Hurtado, 2007; Parks, 2011; Reason et al., 2013; Rockenbach, 2020). Facilitating college students’ civic development necessitates attending to their moral development as well. Lagemann and Lewis (2012) highlight this connection in defining morality as “the capacity to make and explain value judgments about concepts such as fairness, social justice, freedom, and equality, conceived as both democratic ideals and lived commitments” and assert that “civic education cannot flourish if intellect is privileged over morality and action” (p. 43). The nexus of civic and moral education must take into account students’ prosocial development: the development of attitudes, dispositions and behaviors that sustain positive engagement within one's community and larger society (Brandenberger & Bowman, 2012). Yet an increasing focus on accountability, performance metrics, and career preparation has led to questions about whether colleges and universities are fulfilling their promised commitment to students’ development of moral and prosocial responsibility (Clydesdale, 2015; Lagemann & Lewis, 2012; Parks, 2011; Rockenbach, 2020).
Educating for moral and prosocial responsibility necessitates an intentional focus on the development of a set of interrelated values and capabilities thought essential to effective citizenship (Reason et al., 2013; Rockenbach, 2020). The American Association of Colleges and Universities (AAC&U's) Core Commitments initiative, for example, “emphasizes five dimensions of personal and social responsibility as indispensable capacities for all students” (O’Neill, 2012, p. 2). Of the five Core Commitments, two (Contributing to a Larger Community and Developing Competence in Ethical and Moral Reasoning and Action) focus specifically on moral and prosocial responsibility (O’Neill, 2012), adding weight to the importance of fostering college students’ development along these dimensions. Complementing formal academic initiatives, student affairs educators seek to foster students’ development along “cognitive, physical, moral, social, emotional, career, spiritual, personal, and intellectual dimensions” in the co-curricular realm [American College Personnel Association (ACPA), 2006; emphasis added].
While the intersection of higher education and moral development has been explored in general terms (King & Mayhew, 2002), limited research has examined specific educational practices that may be effective in promoting moral sensibilities and prosocial orientation across the curriculum and co-curriculum. The purpose of the present study, which was part of a larger, three-year, multi-institutional study of educational initiatives to improve moral and prosocial education, 1 was to identify college experiences that are associated with college students’ moral and prosocial development. Of particular interest were associations between participation in six types of community engagement experiences (public service, service learning, domestic service, international service, community-based research, and service club) and six moral and prosocial outcomes (including moral identity, perspective taking, two measures of purpose-related life goals, and two measures of prosocial intentions).
Conceptual Framework: College Students’ Development of Moral and Prosocial Responsibility
The following research question guided our exploration: Which college experiences are associated with college students’ development of moral and prosocial outcomes? To answer this question, we organized our conceptual framework in congruence with the principles of Astin’s (1993) I-E-O model, which is frequently used to examine the relationship between college students’ experiences and outcomes (e.g., see Bryant et al., 2012; Furco et al., 2016; Mayhew et al., 2016). The model posits that students bring with them to college a set of input (I) factors consisting of pre-college experiences and demographic and background characteristics. These input factors shape the experiences students have in the college environment (E) and the learning and development outcomes (O) that result from those experiences. In the context of this study, we included students’ demographics and attitudes and beliefs as input factors, their experiences and involvement in college as environment factors, and moral and prosocial outcomes as outcome factors (see Figure 1).
Conceptual framework for understanding college students’ development of moral and prosocial outcomes.
“O” Factors: Moral and Prosocial Outcomes
Morality refers to prescriptions of right and wrong that are not based on consensus or authority but rather stem from conceptions of fairness, justice, rights, and human welfare (Nucci, 2001). As students move through college, they encounter a variety of moral issues and situations that may prompt them to consider their own and others’ conceptions of morality. Indeed, traditional-age college students show growth in moral reasoning or judgment ability, which concerns the ability to make assessments about moral issues, during college (Brandenberger, 2005; King & Mayhew, 2002; Mayhew et al., 2016; Pascarella et al., 2011). Prosocial moral reasoning, specifically, concerns “reasoning about moral dilemmas in which one person's needs or desires conflict with those of others” absent formal authority, policies, or laws to guide reasoning (Eisenberg et al., 1995, p. 1179).
College is a time for the development not only of moral reasoning, but also of moral identity—the perception of oneself as a moral person (Aquino & Reed, 2002). Research suggests that moral identity functionally connects moral reasoning with moral and prosocial action (e.g., Aquino & Reed, 2002; Blasi, 1983; Hardy & Carlo, 2005; King & Mayhew, 2002). As Hardy and Carlo (2005) note, “… when morality is important and central to one's sense of self and identity, it heightens one's sense of obligation and responsibility to live consistent with one's moral concerns” (p. 234). We therefore chose moral identity as one of the moral and prosocial outcomes included in this study.
As moral identity represents only one aspect of moral and prosocial development, we included additional moral and prosocial outcomes in this study to account for multiple aspects of development in this area. Perspective taking, defined as “the tendency to spontaneously adopt the psychological point of view of others” (Davis, 1983, pp. 113–114), is a component of both intellectual and moral development (Gurin et al., 2004) and is essential for students’ preparation “as responsible global and local citizens” (Dey et al., 2010, p. v).
Given the connection between moral development and prosocial action, we also sought to investigate the sense of responsibility or commitment students have to engage with their communities in working toward social betterment and justice. To measure commitment to prosocial action, we selected three scales that have been validated through prior research. The first of these, Importance of Social Action Engagement, is “an outcome that indicates students’ desire to take actions in their communities and relationships in order to end social injustices” (Nelson Laird et al., 2005, p. 468). It was originally developed to assess whether diversity-related courses foster students’ achievement of democratic outcomes; in this study, we utilized it to examine whether community engagement experiences might do the same. The second, Social Justice Behavioral Intentions, is a construct intended to be “predictive of social justice-related behaviors” (Torres-Harding et al., 2012, p. 85). The third focuses on the extent to which students value engaging with their communities to work toward social progress as a life goal (Life Goals – Civic Purpose), which we drew from the Cooperative Institutional Research program (CIRP) survey designed by the Higher Education Research Institute (HERI). Together, these three scales provide different ways of measuring students’ commitment toward prosocial action.
Finally, we included a single item (also drawn from CIRP) to assess the importance students place on developing a meaningful philosophy of life (Life Goals – Meaningful Philosophy of Life). Although this outcome doesn't directly assess moral development, developing a life of purpose and meaning should be an essential outcome of a college education (Clydesdale, 2015; Lagemann & Lewis, 2012; Parks, 2011). We reasoned that if students are developing in their moral identities and commitment toward prosocial action as a result of their college experiences, they may also be developing a more meaningful philosophy of life.
“E” Factors: College Experiences and Involvement
Two of the most promising practices in fostering moral and prosocial outcomes are community service (Kim & Morgül, 2017; King & Mayhew, 2002; Youniss & Yates, 1997) and service learning (Boss, 1994; Furco et al., 2016; Lies et al., 2012; Lovat & Clement, 2016). One aim of community engagement and service activities is to develop students’ critical consciousness of the systemic issues underlying social, political, and economic inequalities and their sense of personal and collective agency and efficacy in addressing these issues (Youniss & Yates, 1997). Additionally, well-designed service learning can increase students’ intercultural competence and appreciation for diversity (Conway et al., 2009; Deardorff & Edwards, 2013; Eyler & Giles, 1999; Furco et al., 2016), empathy and perspective-taking (Astin et al., 2000; Lundy, 2007; Plante et al., 2009), and commitment to diversity and social justice (Niehaus, 2017), among other moral and prosocial outcomes. Similarly, although evidence of service learning's impact on moral reasoning per se is equivocal, participation in service learning is positively associated with prosocial outcomes such as social justice, civic responsibility, and future participation in community service (Mayhew et al., 2016).
Although our primary interest was in understanding whether community engagement experiences are associated with college students’ moral and prosocial development outcomes, we also included in our conceptual framework college experiences that may reflect students’ moral, prosocial, and religious attitudes and beliefs. These include drinking behavior (Radimer & Rowan-Kenyon, 2019; White et al., 2008), academic program of study (Brandenberger & Bowman, 2015; Myyry & Helkama, 2001), and frequency of attending religious services (Brandenberger & Bowman, 2015).
“I” Factors: Attitudes and Beliefs and Demographics
Our conceptual framework includes additional factors associated with college students’ moral and prosocial development in our models; these represent input factors per Astin’s (1993) model and include students’ attitudes and beliefs as well as demographics. The connection between religious or spiritual identity and moral and prosocial outcomes, as noted above, motivated our decision to include religious background and religiosity as input-level predictors. Students’ attitudes and beliefs concerning social inequity and responsibility toward others are also associated with students’ prosocial orientations and behaviors (Bowman, Brandenberger, Mick, et al., 2010). Therefore we included a set of prosocial attitudes and beliefs that could be measured with existing scales, which we describe in our discussion of measures. We included a measure of well-being as well, which has also been found to relate to prosocial behaviors (Bowman, Brandenberger, Lapsley, et al., 2010).
Methodology
We obtained approval by the Institutional Review Board at the University of Notre Dame prior to beginning data collection. We collected all study data via an online survey consisting of 90 questions (primarily Likert-style) addressing the predictor and outcome measures as well as 17 questions addressing respondents’ demographics, background, and college experiences.
Participants, all of whom attended a medium-sized, Catholic research university located in the Midwestern United States, were recruited between spring 2014 and spring 2016 and drawn from multiple subsamples. First, we sent an email invitation with a personalized survey link to all seniors enrolled in two interdisciplinary minor programs of study that address social and justice issues: one focused on Catholic social tradition (CST) and the other on poverty studies. Second, we utilized the university's online research coordination system, which allows students to receive extra credit in their business or psychology classes for participating in research; participants in this subsample anonymously chose to complete our survey as one option among others. Finally, we sent email invitations with personalized survey links to a random sample of 2,000 juniors and seniors provided to us by the university's institutional research office.
Altogether, 749 students completed at least part of the survey. Of these, 680 students completed at least some of the Likert-type questions used for quantitative analyses, and we dropped the responses of five students who completed the survey twice, yielding an analytic sample of 675.
Dependent Variables
Survey items asked students to reflect upon and self-assess six moral and prosocial outcomes, described previously: Moral Identity (Aquino & Reed, 2002), the perception of oneself as a moral person and identification with a community of real or abstract others who share moral traits; two Life Goals measures, drawn from HERI's CIRP survey, asking respondents to indicate the personal importance of two specific goals, Meaningful Philosophy of Life and Civic Purpose; Perspective Taking (a subscale of the Interpersonal Reactivity Index; Davis, 1983); Importance of Social Action Engagement (Nelson Laird et al., 2005), the responsibility a respondent feels to engage in prosocial activities such as promoting racial tolerance and respect, creating environmental impact awareness, and speaking up against social injustice; and Social Justice Behavioral Intentions (a subscale of the Social Justice Scale; Torres-Harding et al., 2012), a respondent's intent to work toward social justice in the future. Table 1 presents sample items, factor alphas, and descriptive statistics for these six outcome factor variables.
Measures, Sample Items, Alpha Values, and Descriptive Statistics for Outcome Factor Scale Variables.
Independent Variables
Students’ attitudes and beliefs concerning social inequity and responsibility toward others are also associated with students’ prosocial orientations and behaviors (Bowman, Brandenberger, Mick, et al., 2010). Well-being has also been found to relate to prosocial behaviors (Bowman, Brandenberger, Lapsley, et al., 2010). Therefore, our survey included six attitude and belief scales: (1) a subscale of the Belief in a Just World measure (Dalbert, 1999), assessing belief that the world is a fair and just place and that people generally get what they deserve in life; (2) Social Dominance Orientation (Pratto et al., 1994), the belief that there is a natural social hierarchy and that equality among groups is not desirable; (3–5) three triune ethical orientations: Security, Engagement, and Imagination (Narvaez, 2008, 2014), which are “motivational orientations that are rooted in evolved unconscious emotional systems shaped by experience that predispose one to react to and act on events in particular ways” (Narvaez, 2008, p. 95); and (6) mental health and well-being (PHQ-4; Kroenke et al., 2009). Table 2 presents sample items, factor alphas, and descriptive statistics for these six predictor factor scale variables.
Measures, Sample Items, Alpha Values, and Descriptive Statistics for Predictor Factor Scale Variables.
We collected indicators of demographic and background characteristics, including gender, race, parents’ education, parents’ income, political views, and academic standing, to serve as control variables. Some prior research has shown these characteristics, particularly gender and college major, to predict students’ likelihood of participating in community engagement activities, especially service learning and public service (Astin et al., 2000; Cruce & Moore, 2007; Haski-Leventhal et al., 2008). Additionally, since religious identity may be predictive of moral and prosocial outcomes (Astin et al., 2011; Brandenberger & Bowman, 2015; Clydesdale, 2015; Parks, 2011), we included religious background and religiosity (“How religious do you consider yourself to be?”) as input-level predictors, and frequency of attending religious services as a college experiences predictor.
At the level of college experiences and involvement, our primary predictors of interest were students’ community engagement experiences (specifically, participation in six types of experiences: public service, service learning, domestic service, international service, community-based research, and service club, each measured dichotomously), given the potential impacts of community engagement opportunities noted previously. We also included measures of other college experiences that may reflect students’ moral, prosocial, and religious attitudes and beliefs, such as drinking behavior (Radimer & Rowan-Kenyon, 2019; White et al., 2008) and academic program of study (Brandenberger & Bowman, 2015; Myyry & Helkama, 2001). Descriptive statistics for all independent variables are reported in Table 3.
Descriptive Statistics for Independent Variables.
Data Analyses
To facilitate analytic interpretations of regression coefficients, and in some cases due to characteristics of the sample (e.g., gender, race, and religion), we created dichotomized versions of many of the categorical demographic, background, and experience variables. For analyses, we kept academic standing and survey year as categorical variables; religiosity, parent income, and parent education (sum of both parents) were continuous variables. We standardized all continuous variables, including scores on each of the factor scales for attitudes and beliefs and outcomes, enabling interpretation of coefficients as effect sizes.
To answer our research question, we ran a series of blocked regression models, with one set of models for each of the six moral and prosocial outcomes. This technique enabled us to determine “whether the variables after the first block predict the outcome above and beyond the variables that were included in the previous block,” thereby separating the effects of the input variables from those of the college experiences variables (Mayhew et al., 2016, p. 619).
For each set of models, we entered demographic and family background characteristics in the first block, under the assumption that students could not exercise control over these variables; these included gender, race, parents’ income, and parents’ education. The second block included background characteristics over which students could conceivably exercise some control but which likely temporally preceded college experiences: religion, political identification, academic standing, and religiosity. We also included a categorical survey year variable in this block as a control for possible cohort effects. The third block in our regression models included attitude and belief variables: mental health and well-being, Belief in a Just World, the three triune ethic measures, and Social Dominance. Each factor scale variable was created by summing a participant's responses to the questions comprising that scale and then standardizing the summed score. Together, these first three variable blocks comprise the input (“I”) factors per Astin’s (1993) I-E-O model (see Figure 1). We recognize that because our participants’ attitudes and beliefs were not measured at the start of college, they may not constitute true “I” factors given that many participants (almost 74%) had been in college at least one year at the time of the survey. Our decision to input these variables in block 3 acknowledges that these variables straddle the line between “I” and “E” factors. 2
The fourth block contained variables relating to students’ general college experiences, including dichotomous variables for whether participants frequently attended religious services (at least once per week or “often”), were heavy drinkers (seven or more drinks per week), were nondrinkers, were enrolled in a STEM academic program, or were enrolled in a business academic program. In the fifth block, we entered the five moral and prosocial outcome variables not serving as the dependent variable in the model, to control for the effects of these other outcomes relative to the dependent variable of interest. The sixth and final block in each set of models included dichotomous variables related to community engagement experiences: participation in public service (at least one hour per week), enrollment in either the poverty studies or CST minors (a single variable), and a set of five variables indicating students’ participation in specific campus engagement experiences: service learning, domestic service, international service, community-based research, or a service club, as well as a dichotomous variable indicating non-participation in any of these five campus engagement activities. These final three variable blocks correspond to the environment (“E”) factors of Astin’s (1993) I-E-O model (see Figure 1).
We checked for possible multicollinearity among variables in each of the six models by assessing the variance inflation factor (VIF) and tolerance (1/VIF) by running the vif command in Stata after the final regression block. For all six models, tolerance values for all variables, including the five moral and prosocial outcome variables serving as controls in each model, were well above the recommended 0.1 minimum, indicating that these variables were not multicollinear. Post-hoc power analysis using the G*Power program (Faul et al., 2009) revealed a strong power (1 – β method) of 1.000 for each of the six final blocked regression models.
Limitations
Although we collected data over the course of three years, the data are cross-sectional, so we cannot make any claims about how students change over their time in college regarding the moral and prosocial outcomes we examined. Therefore, our results only point toward relationships between students’ experiences in college and their moral and prosocial development; we cannot make conclusions about causality. Additionally, we know that students’ attitudes and beliefs change quite a bit over their time in college, and that attitudes and beliefs at college entry influence the experiences students choose to be involved in, which then in turn further shape their attitudes and beliefs (Mayhew et al., 2016). A longitudinal design would have allowed us to measure students’ attitudes and beliefs at multiple points in time, thereby more accurately documenting the effect of attitudes and beliefs on both our independent variables of interest (participation in community engagement experiences) and the dependent variables (the six moral and prosocial outcomes).
Additionally, our data were collected from students at a single institution, and the characteristics of this institution (specifically, that it is private and faith-based) are not representative of the type of institution that most U.S. college students attend, nor might the students who responded to our survey represent the “typical” college student. Students attending a religiously affiliated institution may exhibit greater religiosity than students at non-sectarian institutions, and previous research has found a connection between religiosity and prosocial behaviors and development (Astin et al., 2000, 2011; Brandenberger & Bowman, 2015).
Yet we also believe that the characteristics of the institution at which we conducted this research offer some advantages. For example, as a Catholic university, students’ moral and prosocial development, the principles of social justice, and community engagement receive attention in the curriculum and co-curriculum in ways they may not at other institutions. Previous research has found that a campus environment in which civic engagement is an expectation and a norm are more likely to demonstrate a commitment to civic involvement (Barnhardt et al., 2015; Bryant et al., 2012). Therefore, this university arguably provides an especially information-rich context in which to explore the research question guiding our study.
Results
The results we present here are drawn from the models in which all six blocks of variables were entered. Table 4 presents standardized regression estimates for these six models (one model for each of the six moral and prosocial outcomes: Moral Identity, Perspective Taking, Importance of Social Action Engagement, and Social Justice Behavioral Intentions, and two Life Goals: Meaningful Philosophy of Life and Civic Purpose). The variance (R2) explained by each of the six models when all six blocks of variables were entered ranged from a low of 21.1% for the outcome Life Goals: Meaningful Philosophy of Life to a high of 57.4% for the outcome Social Justice Behavioral Intentions. 3 We utilize our conceptual framework (see Figure 1) to organize this section, first discussing the relationships among participants’ college experiences and involvement [environment or “E” factors, per Astin’s (1993) model] with the six moral and prosocial outcomes, then discussing relationships among participants’ attitudes and beliefs and demographics (both representing input or “I” factors) with the outcomes. In presenting our results, we use words such as “predict” not to indicate direction of causality (which cannot be established with non-experimental, cross-sectional survey data like ours), but rather to reflect the specification of our model (i.e., that independent variables predict dependent variables).
Final (Sixth) Block Standardized Regression Estimates Predicting Six Moral and Prosocial Outcomes.
Notes: All continuous independent and dependent variables are standardized. Robust standard errors in parentheses. *p < .05, **p < .01, ***p < .001
College Experiences and Involvement (“E” Factors)
Our research question focused on identifying the college experiences associated with students’ development of six moral and prosocial outcomes. These college experiences reflect the “E” (college environment) component of Astin’s (1993) I-E-O model, which guided the organization of our conceptual framework. We were primarily interested in the relationship between students’ participation in six types of community engagement experiences and our six outcomes, but we also examined relationships with other college experiences, including academic program, drinking behavior, and frequency of attending religious services, as these might reflect students’ moral and prosocial orientations. Regarding the latter set of experiences, we only found two statistically significant relationships for the models in which all variable blocks were entered: enrollment in the business program was negatively associated with Social Justice Behavioral Intentions (B = −0.266, p < .01) and positively associated with Moral Identity (B = 0.191, p < .05).
Regarding community engagement experiences, we found that participation in public service was the only engagement experience to predict any of the six outcomes. Participating in at least one hour per week of public service positively predicted Moral Identity (B = 0.239, p < .01) as well as Life Goals – Civic Purpose (B = 0.212, p < .01), but it negatively predicted Life Goals – Meaningful Philosophy of Life (B = −0.337, p < .001). None of the other types of community engagement experiences (i.e., service learning, domestic service, international service, community-based research, or service club) was related to any of our outcomes. Additionally, there was no relationship between not participating in any of the engagement experiences and the six moral and prosocial outcomes, although it is worth noting that fewer than 8% of participants reported not participating in any of the engagement activities.
Attitudes and Beliefs (“I” Factors)
Although not the focus of our research question, our models included a set of factor variables measuring students’ attitudes and beliefs as controls since these have been shown to be associated with moral and prosocial outcomes. Along with student demographic characteristics, discussed next, these variables comprise the “I” (inputs) component of Astin's I-E-O model. Social Dominance was the most consistent predictor, having negative relationships with five of the six outcomes (its relationship with Life Goals – Meaningful Philosophy of Life was nonsignificant). Effect sizes (B) ranged from −0.112 for Moral Identity (p < .05) to −0.150 for Life Goals – Civic Purpose (p < .001). Consistent with triune ethical theory (Narvaez, 2008, 2014), the Security Ethic was least predictive of the six moral and prosocial outcomes, having a relationship only with one, Perspective Taking, for which it was a negative predictor (B = −0.118, p < .01). The Ethic of Engagement positively predicted three outcomes: Moral Identity, for which it was a relatively strong predictor (B = 0.391, p < .001), as well as Perspective Taking (B = 0.188, p < .001) and Life Goals – Meaningful Philosophy of Life (B = 0.088, p < .05). The Ethic of Imagination also positively predicted three outcomes: Perspective Taking (B = 0.138, p < .05), Importance of Social Action Engagement (B = 0.139, p < .01), and Social Justice Behavioral Intentions (B = 0.090, p < .05). Belief in a Just World predicted two of the outcomes, both positively: Moral Identity (B = 0.135, p < .001) and Perspective Taking (B = 0.091, p < .05). Well-being was not associated with any of the six moral and prosocial outcomes.
Demographics (“I” Factors)
As with attitudes and behaviors, students’ demographic and background characteristics were not a focus of this study and were included in the models as control variables; these variables are also included within the “I” (inputs) component of Astin’s (1993) I-E-O model. For each of the six models in which all six blocks of variables were entered, few demographic and background characteristics were significantly related to the outcomes. Identifying as male negatively predicted two outcomes, Importance of Social Action Engagement (B = −0.224, p < .01) and Social Justice Behavioral Intentions (B = −0.163, p < .01), while identifying as a student of Color positively predicted one outcome, Perspective Taking (B = 0.233, p < .05). Parental income was associated with only one outcome, Importance of Social Action Engagement, although the coefficient magnitude (which can be interpreted as effect size since all continuous variables were standardized for analyses) was comparatively small (B = 0.091, p < .05). Religiosity positively predicted Moral Identity (B = 0.105, p < .05); however, this was the only outcome associated with religiosity. We found a negative relationship between survey year and Social Justice Behavioral Intentions (B = −0.202, p < .01), suggesting a possible cohort effect for this outcome alone. None of the demographic and background variables predicted either of the two Life Goals outcomes, Meaningful Philosophy of Life or Civic Purpose. Aside from the exceptions noted here, most relationships between the demographic and student background characteristics variables and the outcomes became nonsignificant once all variables were entered into each of the six models, suggesting that the experiences students have in college, as well as their attitudes and beliefs, are more strongly associated with moral and prosocial outcomes.
Discussion
The purpose of this study was to identify the college experiences—particularly community engagement experiences—that are associated with a set of six moral and prosocial outcomes (Moral Identity, Perspective Taking, Importance of Social Action Engagement, and Social Justice Behavioral Intentions, and two Life Goals: Meaningful Philosophy of Life and Civic Purpose). In the language of Astin’s (1993) I-E-O model, our objective was to understand the relationships between these college environment (“E”) factors with our outcomes (or “O” factors) while controlling for student-level input (“I”) factors (i.e., demographics and attitudes and beliefs).
We want to highlight several key points suggested by our results. First, with a few exceptions as discussed above, we found that students’ demographic and background characteristics in and of themselves are not strongly associated with moral and prosocial outcomes. Rather, the attitudes and beliefs students hold seem to have a much greater relationship to these outcomes. Our finding here reflects that of Rockenbach et al. (2014), who reached a similar conclusion regarding the predictive value of attitudes and values, rather than demographic factors, on community service participation (one form of prosocial behavior). From an educational perspective this is encouraging, as attitudes and beliefs are more malleable than demographic factors and can be shaped by curricular and co-curricular interventions. Yet in our blocked regression models, we found that the relationships of the six attitude and belief predictors with our six outcomes generally remained stable after the community engagement experience variables were introduced in the sixth and final block. It is worth noting that our model specification assumed that attitudes and beliefs temporally preceded community engagement experiences—that is, these blocks of variables entered into the models prior to the community engagement experiences variable block, which we entered into the model last so that we could isolate the relationships of our primary predictors of interest on the dependent variables when accounting for all other controls. In reality, however, it is likely that attitudes and beliefs both influence and are influenced by students’ participation in community engagement experiences.
Our results suggest that at least as measured by our survey, community engagement experiences overall were not directly related to the six moral and prosocial outcomes. Only public service (participating in at least one hour per week) was associated with any of the outcomes; it had a positive relationship with two (Moral Identity and Life Goals – Civic Purpose) and a negative relationship with one (Life Goals – Meaningful Philosophy of Life). These positive relationships make sense; Moral Identity in part assesses the extent to which individuals participate in activities that align with the characteristics of being helpful, generous, and compassionate, while Life Goals – Civic Purpose assesses the importance an individual places on working toward social betterment. The negative relationship between public service and Life Goals – Meaningful Philosophy of Life is more difficult to explain. This outcome consisted of a single statement asking participants the extent to which developing a meaningful philosophy of life is personally important to them. It's possible that the negative relationship reflects a developmental process; that is, perhaps participating in public service causes students to question the meaning of their life, which may ultimately foster their moral development in the longer term. Public service, as conceptualized for this study, does not involve any academic or co-curricular components (as would other forms of community engagement such as service learning or community-based research) but rather is something students engage in on their own. Thus, the lack of structured opportunities for reflection individually and with others may prevent students from processing and resolving the cognitive dissonance they experience when engaged in public service.
Regarding college experiences other than community engagement, only enrollment in the business program was related to any of the outcomes, negatively with Social Justice Behavioral Intentions and positively with Moral Identity. The negative relationship with Social Justice Behavioral Intentions is congruent with prior studies that have found business students are less likely to exhibit prosocial attitudes and behaviors than students in other academic programs, particularly the social sciences (Brandenberger & Bowman, 2015; Meier & Frey, 2004; Myyry & Helkama, 2001); such attitudes and behaviors would seem to be associated with a commitment to working toward social justice. That majoring in business was associated with a higher score on the Moral Identity outcome than majoring in arts, humanities, or social sciences (the comparison group for our analyses) is perhaps more surprising in light of prior research. Again, this result may reflect a developmental process in which students taking classes in the arts, humanities, and social sciences encounter more ideas that challenge or question their moral identity—that is, their self-conception as someone who is caring, compassionate, fair, friendly, generous, helpful, hardworking, honest, and kind—than students taking business classes.
We offer two possible explanations for our broad finding that few college experiences were associated with the six moral and prosocial outcomes we measured. First, the cross-sectional design of our survey made it difficult to identify possible selection effects; perhaps students who chose to participate in engagement experiences would have scored highly on the outcome measures prior to participation, and as a result participation had little effect (i.e., the ceiling effect). Additionally, it is possible that for some students, moral foundations and commitment to social justice preceded, rather than resulted from, their participation in engagement experiences. Our models assumed that the six moral and prosocial outcomes are “downstream” outcomes affected by participation in engagement activities. Yet the analytic challenge is that students’ scores on these six outcome measures may also predict their decision to participate or not participate in engagement activities, and because the direction of causation remains unclear, we cannot assert with confidence that college students’ development of an increased sense of commitment to their communities is the result of participation in multiple engagement activities. The cross-sectional design also prevented us from capturing the possible effects of developmental processes in progress; as noted above, some of the negative relationships we found (e.g., with Life Goals – Meaningful Philosophy of Life) may reflect the developmental regression that often precedes growth, as has been observed for other developmental outcomes among college students (e.g., Narvaez, 2008; Pizzolato & Olson, 2016). We recommend, therefore, that future research in this area utilize longitudinal designs (see Bowman, Brandenberger, Lapsley, et al., 2010; Bowman et al., 2011, for research design potentials).
A second possible explanation for the limited association between engagement activities and moral and prosocial development in this study may be due to the nature of our measures. While the survey asked students to indicate the level of hours per week in which they engage in public service, the distribution of responses necessitated creating a dichotomous variable. Intensity and quality of community engagement experiences vary widely, and previous research has established that these factors are crucial determinants of students’ prosocial development (Clayton et al., 2013; Handy et al., 2010; Rockenbach et al., 2014). Similarly, students were only asked to indicate whether they had participated in any of the five college engagement activities (in a “check all that apply”-style survey item), which provides no indication of the quality or duration of those experiences. As others have noted—especially regarding service learning—the quality and duration of the engagement experience shape the learning that results (Clayton et al., 2013). Future research sensitive to such factors is an important next step.
What are the implications of this research for educational practice? Aquino and Reed (2002) posit that moral identity is a precursor to moral behavior, and we found a relationship between moral identity and Social Justice Behavioral Intentions. Yet there was no direct relationship between Social Justice Behavioral Intentions and public service participation; in fact, the coefficient was nearly zero. This suggests that engaging in public service may lead at least some college students to experience gains in moral identity, which in turn enhances their commitment to prosocial behaviors. Social Justice Behavioral Intentions was the only moral and prosocial outcome that was positively related to all five of the other outcomes, suggesting that increasing students’ commitment in this area can enhance their general moral and prosocial development. The salience of social justice orientations within our findings is telling. Cipolle (2010) outlines a theory of critical consciousness and posits that understanding of and orientation to social justice is at the highest level of student development. Students entering with a charity model of social contribution (at a less-developed level on the Cipolle scheme) may engage differently with public service opportunities than those with more of a justice orientation, adding complexity to the input-experience-outcome research model (Astin, 1993) and with implications for pedagogy (see Moely et al., 2008).
Conclusions
The research presented here used broad empirical tools and multiple measures to examine the relationship between six types of community engagement experiences and students’ moral and prosocial outcomes during college. Our findings indicate that students’ existing attitudes and orientations to prosocial and moral concerns may play a more salient role in determining which types of community engagement experiences they are attracted to (e.g., public service or service learning) and how these experiences may influence their development. While community engagement experiences have been shown to predict moral and prosocial outcomes, our research suggests that one size does not fit all. Research designed to examine interactions across student attitudes (at entry), college experiences, and follow-up opportunities would be a natural next step. Toward that end, research on academic momentum—how prior learning, life experience, and sequenced educational activities may maximize development—is instructive (e.g., see Martin et al., 2013; Niehaus, 2017).
That moral identity and commitments to social justice were more frequent, yet not unidirectional, predictors in this study reinforces the importance of such constructs and suggests the need for further research on how they may serve as mediators of development. While the college years are a prime opportunity (both contextually and developmentally) for student ethical and prosocial growth, many additional research questions arise. What kinds of experiences are best for students oriented toward charity, social change, or justice considerations? What specific opportunities for moral action (through intentional engagement) may enhance perspective taking, moral judgement, and motivation? What cognitive and personality factors may conscribe or prompt development in the moral domain? Increased attention to prosocial and moral development during the college years and employment of longitudinal designs have the potential to address such questions and improve educational practice, ultimately helping to fulfill higher education's effectiveness in “improv[ing] social conditions by enhancing the character strengths of graduates” (Rockenbach, 2020, p. 7).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Teagle Foundation (grant number 197695).
