Gateways: International Journal of Community Research and Engagement
Vol. 19, No. 2
October 2026
RESEARCH ARTICLE (PEER REVIEWED)
Beyond ‘Definitional Anarchy’: A Latent Class Analysis of Institutional Definitions of Community Engagement
Andrew Pearl1, Tanvi Padalkar2, Emily Locke3, Diana Quito4, Henry R. Cunningham5
1 Kansas State University, Kansas, USA
2 University of Alabama at Birmingham, Alabama, USA
3 University of Alabama, Alabama, USA
4 Rollins College, Florida, USA
5 University of Louisville, Kentucky, USA
Corresponding author: Andrew Pearl, apearl@ksu.edu
DOI: https://doi.org/10.5130/gz44ae45
Article History: Received 01/08/2025; Revised 01/07/2026; Accepted 17/09/2026; Published 10/2026
Abstract
The Carnegie Foundation’s Elective Classification for Community Engagement was first awarded in 2006. As of the 2024 cycle, 368 institutions held the classification. At the time of this research, the American Council on Education (ACE) hosted the Carnegie Elective Classifications. We used latent class analysis to create a typology of four definitions with varying levels of emphasis placed on values like collaboration, mutual benefit and connections to scholarship. The resultant typology offers a framework for understanding how institutions communicate their values related to community engagement. Our findings provide opportunities for practical guidance and reflective opportunities for leaders seeking to institutionalise community engagement.
Keywords
Community Engagement; Latent Class Analysis; Institutional Definitions; Institutional Theory
A Latent Class Analysis of Institutional Definitions of Community Engagement
In the community engagement field, a single agreed-upon definition of community engagement has been difficult to formalise, and even the need for a singular definition has been a recurring feature of critical debate. As an inherently transdisciplinary field that can trace its origins to many different traditions, including experiential learning (Dewey 1938) or community organising (Avila 2018), many terms have been used to describe the work of community engagement. Giles (2008) illustrates this point by emphasising that the research agenda on engaged scholarship relies on overarching terms that may include similar activities, yet vary in scope, priority and motivation. Even with this attempt to establish clarity by organizing the field’s overlapping terminology, a conceptual fragmentation or ‘definitional anarchy’ (Sandmann 2008, p. 91) dominated.
Attempts to define community engagement persist today. Many factors have been shown to contribute to the institutionalisation of community engagement in higher education (Furco et al. 2009; Holland 1997), including establishing a formal, campus-wide definition of community engagement (Furco et al. 2009; Janke, Jenkins, et al. 2023). Because of this, attempts to define community engagement persist today, necessitating clarity, especially as an increasing number of colleges and universities continue to embark on the complex and adaptive challenge of embedding engagement work as central to their missions (Farner 2019; Fitzgerald et al. 2012; Reed & Swanson 2022; Welch 2016). Given the complexity of the higher education sector, it could be expected that campus-wide definitions of community engagement may vary by institutional type, control, location, size, among other factors.
Our research began in 2023 as a part of the Carnegie Elective Classifications Research Lab (Pearl, Jones-Friedrichs, et al. 2025). Our group, the Community-Higher Education Engagement Research (CHEER) Lab, is interested in how colleges and universities define, operationalise and institutionalise community engagement. For this study, we sought to provide some clarity around the persistent ‘definitional anarchy’. To do so, we examined institutional definitions of community engagement, exploring what is shared, what variations exist and what is missing. Using the definition of community engagement provided by the Carnegie Foundation as a baseline, we analysed data using LCA, a statistical method that identifies subgroups within a population, to provide insight into the heterogeneity in institutional definitions of community engagement.
Carnegie’s Elective Community Engagement Classification
Since it was first awarded in 2006, the Carnegie Foundation’s Elective Classification for Community Engagement (CE Classification henceforth) has served as a way for colleges and universities to affirm their commitments to community engagement (Driscoll 2009). The goal of the elective classification was not to be fully prescriptive in what constituted high-quality community engagement considering the variation in institutional contexts and approaches to community engagement (Driscoll 2008). Nonetheless, earning the CE Classification can serve as an indicator of an overall commitment to the public purposes of higher education (Saltmarsh & Johnson 2018).
Defining Community Engagement through Carnegie
While there is currently not a single consensus definition of community engagement in the field of higher education, the CE Classification provides a definition of community engagement in higher education that includes these characteristics: ‘collaboration between institutions of higher education and their larger communities (local, regional/state, national, global) for the mutually beneficial exchange of knowledge and resources in a context of partnership and reciprocity’ (American Council on Education, n.d.-b). The purpose of community engagement is further articulated as a way to enrich scholarship, research and creative activity; enhance educational practices and learning experiences; cultivate engaged citizens; promote democratic values and civic responsibility; respond to critical societal issues; and contribute to the public good (American Council on Education, n.d.-b).
The 2026 cycle’s application framework included the following sections (Pearl, Jones-Friedrichs, et al. 2025):
—Campus, Community and Community Engagement Context
—Institutional Identity and Culture
—Quality of Community Engagement Relationships and Academic Partnerships
—Faculty and Staff
—Curricular Engagement
—Co-Curricular Engagement
—Civic Learning and Life
—Community Engagement and Other Institutional Initiatives
In the Campus, Community and Community Engagement Context section, the application framework for the CE Classification specifically asks institutions to provide their campus-wide definition of community engagement. As a result, colleges and universities that have applied for and received the CE Classification establish contextually rich and diverse definitions of community engagement to communicate institutional commitment to a variety of internal and external constituencies. Despite this diversity, several core elements and indicators of high-quality community engagement in the application framework have become normative expectations for institutions that have applied for and received the CE Classification. These include collaborative approaches, mutually beneficial outcomes and a connection to the institution’s scholarly mission.
Theoretical Grounding
Much of the theoretical grounding for this study is guided by the analysis of Morphew and Hartley (2006) on higher education institutional mission statements. They assert, ‘they [mission statements] exist because they are expected to exist’ (p. 458), which sends a message of legitimacy to interested parties both within and outside of the institution. The authors point to the use of institutional theory to explain organisational behaviours (Meyer & Rowan 1977), including the normative function that mission statements play. There are some studies that suggest that this normativity may not be worthwhile (Newsom & Hayes 1991) or is simply symbolic (Chait 1979), with an isomorphic tendency to include common elements (DiMaggio & Powell 1983). Despite their ubiquity, however, Morphew and Hartley (2006) note that institutions strategically use their mission statements to communicate to multiple and diverse audiences. So, while institutions have mission statements because they are expected to have mission statements, they are also used intentionally to communicate to audiences that have expectations about the purpose of the institution. Similarly, institutional definitions of community engagement can be used to guide an institution’s commitment and strategic work related to community engagement. Parallel to a mission statement, formalised community engagement definitions can be used to articulate institutional values and priorities.
This study examines institutional definitions of community engagement over the last ten years to better understand what, if any, shared values, norms, emphases may be emerging. To achieve this, we have created a definitional typology of community engagement by conducting a latent class analysis using data from successful institutional applications for the CE Classification. By creating this typology, we can learn more about the variations in and commonalities among how colleges and universities define community engagement, synthesising field-based developments for defining community engagement, which can provide clarity for institutional leaders and community engagement professionals seeking to further institutionalise community engagement at their institutions.
Methods
Data
In this study, we examined a single item from the application framework for the CE Classification. There is slight variation in how the question was phrased in different cycles, but each version essentially asks for evidence of a campus definition of community engagement and relevant contextual information. As an example, the first-time framework during the 2024 cycle asked:
If your campus has an institutionally sanctioned definition of community engagement and related terms, provide them here. Describe the context for the creation of the definition, how it was approved, how it is used, and any evidence of its effectiveness in guiding community engagement on campus. If your campus does not have an institutional definition of community engagement but you are on a campus that has multiple definitions reflecting how different units and disciplines interpret community engagement, provide some description and examples here. If your institution does not have such a definition, please describe any work under way to adopt one. (American Council on Education, n.d.-a)
Our target sample included the institutions who successfully received the CE Classification during the 2015 (re-classification), 2020 (re-classification) and 2024 (first-time and re-classification) cycles. The application framework has continued to evolve and change over each cycle of the CE Classification, and there was no item related to a campus-wide definition of community engagement for first-time applicants during the 2015 and 2020 cycles. We examined the most recent data available for institutions that had submitted applications in both 2015 and 2024. When submitting an application for the CE Classification, institutions had the option of not being included in the research dataset and were therefore not included in our analysis. Our final sample included 221 institutions.
Data Collection Tool
The first step in our analysis was to create a data collection tool. Because our goal was to better understand institutional definitions of community engagement, we decided to use Carnegie’s formal definition as a baseline (see above). As a research team, we identified the following components of this definition to guide the development of our data collection tool:
—Characterising Community
—Scholarly Connections
—Collaboration
—Mutual Benefit
We created an initial data collection tool from these components and pilot-tested the tool with randomly selected definitions from our sample. Four members of the research team used the tool to independently code the same institutions. We then discussed any discrepancies in the coding and identified opportunities to improve or refine the data collection tool. We repeated this procedure until we reached agreement on the final version of the tool. In addition to the four components above, our iterative process of refining the tool also revealed that several definitions in our sample intentionally referenced critical approaches to community engagement, highlighting systemic inequalities or power hierarchies. Because of this, our research team agreed that including a related question in our data collection tool was valuable. Our final data collection tool included the following questions:
1. Does the institution specifically identify a community (geographic and/or populations) in their community engagement definition? [YES or NO]
2. Does the definition include a connection between community engagement and the institution’s scholarly mission? [YES or NO] If the answer to this question was yes, we also asked if there were specific mentions of:
2a. Teaching and learning [YES or NO]
2b. Research and creative work [YES or NO]
2c. Service [YES or NO]
3. Is a collaborative process mentioned in the definition? [YES or NO]
4. Are mutually beneficial outcomes mentioned in the definition? [YES or NO]
5. Does the institutional definition of community engagement mention or reference a critical approach to community engagement that addresses systemic inequalities, power hierarchies, etc.? [YES or NO]
Using this final version of the tool, one researcher coded the full dataset. As a validity check, four other members of the research team were each randomly assigned institutions to code and compare. Any discrepancies were discussed and resolved. These codes were then used to create a typology using latent class analysis.
Latent Class Analysis
We then used the institutional assessments to conduct a latent class analysis (LCA). Latent class analysis is a statistical method that identifies subgroups within a population of empirical data (Collins & Lanza 2010). The latent variables are not measured directly. Instead, each latent variable is indirectly measured through at least two observed variables, also referred to as the ‘manifest’ variables (Linzer & Lewis 2011). LCA models categorical responses as latent classes, with the research team assigning names to the latent classes (Collins & Lanza 2010). The interpretation of the resultant classes is based on the probability that an individual responds in a particular way to a certain item. Based on these item-response probabilities, latent classes emerge among which there are qualitative differences. The researcher then determines names for the latent classes based on the item-response probabilities and informed by theory and the purpose of the research.
Our analysis is not based on a prior hypothesis guiding the labels for the latent classes. While several models were generated, we relied on goodness of fit measures to help determine model fit (Sinha, Calfee & Delucchi 2021), explored in more detail below. Latent class analysis was performed in R Statistical Software (R Core Team 2024) using the poLCA package (Linzer & Lewis 2011). To ensure the identification of the global maximum likelihood, the models were estimated using 100 random sets of starting values (nrep = 100) using the poLCA package.
Findings
Based on our research team’s coding using the tool that we developed, the descriptive statistics for our sample are shown in Table 1.
*Note. The institutional definitions that included a scholarly connection could also indicate a connection to more than one specific approach to scholarship (teaching and learning, research and creative work and/or service). Therefore, the total number of “yes” responses are more than 179.
Determining the Model Fit
A number of goodness of fit statistics can be used to determine the appropriate number of classes from an LCA. Linzer and Lewis (2011) recommend that the Akaike information criterion (AIC) and/or the Bayesian information criterion (BIC) be used as guidance for selecting the appropriate number of classes; specifically, the point at which those criteria are minimised. As shown in Table 2, the AIC was minimised in the five-class model, and the BIC was minimised in the three-class model.
| Number of classes | AIC | BIC |
|---|---|---|
| 2 | 1780.893 | 1872.643 |
| 3 | 1699.337 | 1838.662 |
| 4 | 1679.827 | 1866.726 |
| 5 | 1669.55 | 1904.023 |
| 6 | 1680.522 | 1962.569 |
| 7 | 1691.423 | 2021.045 |
Because of this discrepancy in the number of classes, we relied on the advice of Sinha, Calfee and Delucchi (2021) that the “best fit does not always mean good fit” (p. 8). In other words, the goodness of fit measures are valuable, but the number of classes also needs conceptual alignment and logical interpretability. Our research team closely examined the output for the three, four and five class models and agreed that the four-class model had strong conceptual alignment. Further, the four-class model’s entropy measure was 0.962, suggesting a very high separation between the four classes. A table of the complete item-response probabilities is available in the Appendix. Based on those response probabilities, the research team discussed and assigned the following descriptive class names: Scholarly Partnerships, Pedagogical Partnerships, Engagement as Pedagogy and Operational Engagement.
Scholarly Partnerships
The Scholarly Partnerships class of institutional community engagement definitions indicate general commitments to both collaborative partnerships and mutually beneficial outcomes, as well as a holistic connection to scholarship, inclusive of teaching and learning, research and creative work and service. These institutions were not likely to identify a specific community in their definition, and do not indicate a critical approach to their community engagement work. The full item response probabilities are shown in Table 3.
Pedagogical Partnerships
Similar to the Scholarly Partnerships class, the Pedagogical Partnerships class of definitions refers to collaborative approaches, mutually beneficial outcomes and connections to scholarship; however, those scholarly connections are limited to teaching and learning, not research and creative work nor service. In addition, these institutions were not likely to identify a specific community in their definition, and do not indicate a critical approach to their community engagement work. The full item response probabilities are shown in Table 4.
Engagement as Pedagogy
The Engagement as Pedagogy class of definitions is similar to the Pedagogical Partnership class of definitions in their connections to the teaching and learning aspect of scholarship, but not a connection to research and creative work or service. Further, the Engagement as Pedagogy class does not mention collaborative processes or mutually beneficial outcomes or critical approaches. These institutions were also not likely to identify a specific community in their definition. The full item response probabilities are shown in Table 5.
Operational Engagement
Unlike the other three classes, the Operational Engagement class does not make a connection between community engagement and scholarship. This class is slightly more likely than not to mention collaborative approaches (61.9%) and mutually beneficial outcomes (57.1%). Like the other classes, these institutions were also not likely to identify a specific community in their definition, and do not indicate a critical approach to their community engagement work. The full item response probabilities are shown in Table 6.
In Table 7, we show the breakdown of the predicted class memberships in our sample by percentage. The largest class in our sample was Pedagogical Partnerships (36.20%), and the smallest class in our sample was Engagement as Pedagogy (11.76%). Table 7 also includes the predicted class memberships for the public institutions, private institutions, land-grant institutions and 2-year institutions in our sample.
Discussion
Through our analysis, we identified four distinct types of institutional definitions of community engagement, showing that institutions include varying areas of emphasis in their campus-wide definitions of community engagement. In the following sections, we discuss several elements from the community engagement definition and purpose of community engagement that were particularly salient in our findings.
Defining Community
In the formal community engagement definition, Carnegie advocates for a broad conceptualisation that considers local, regional/state, national and global communities. This expresses that community is primarily defined in a geographic context; however, other definitions of community also account for relational dimensions, including shared interests or identities. For example, Doberneck, Glass and Schweitzer (2010) discuss that communities often exist beyond physical boundaries in the context of community engagement, such as communities characterised by identity, profession, affiliation or area of interest, faith, or kin. In our analysis, none of the classes were more likely than not to specifically define the communities with which they partner. Many of the definitions in our sample talked about broad strategies or approaches to engagement, for example, using terms like ‘external audiences’, using engagement to ‘service society’, or generally referring to ‘the community’.
There are multiple interpretations of this approach to defining community engagement. One perspective is that specifically identifying communities may have been somewhat limiting. For instance, many colleges and universities engage with partners that span a large number of communities, and to specifically identify one, or even a small group, might inadvertently send a message of prioritising some partners over others. Alternatively, this approach could give institutions flexibility in determining the communities with which they collaborate, perhaps allowing them to be less accountable to any one community in particular. Elsewhere in the CE Classification application, institutions are expected to provide detailed examples of impactful community partnerships. Because all of the institutions in our sample were successfully classified, it is reasonable to assume that their applications identify specific community partnerships that demonstrate a commitment to those communities. Guidance from Trebil-Smith (2019) advises that community-university partners are built on relationships between people. Perhaps in a loosely-coupled system (Weick 1976), the role of the institution is to provide values-based guidance on how community engagement work is done (for example, collaboration, mutual benefit and a scholarly grounding) rather than specifically with whom that work is done. Additional research and data collection can provide further insight into the reasoning behind these decisions.
Defining Scholarly Connections to Community Engagement
According to Carnegie, the purpose of community engagement includes explicit connections to the institution’s scholarly mission through community-engaged approaches to teaching, research and service. This connection has its origins in Boyer’s (1990) model of scholarship that later culminated in his discussion of the scholarship of engagement (Boyer 1996) as an imperative for higher education. A connection between community engagement and the institution’s scholarly mission was consistently present for three of the four institutional types in our analysis: Scholarly Partnerships, Pedagogical Partnerships and Engagement as Pedagogy.
Numerous scholars have emphasised the necessity of integrating community engagement into the academic mission areas of the university to build towards institutionalisation (Cunningham & Smith 2020; Fitzgerald et al. 2012; Holland 1997; Ward et al. 2023; Wendling 2023), and not viewing it as an alternative or additional focus area. Without this explicit connection in the definition, new and emerging generations of scholars, who are motivated by the potential public impact of their work (Post et al. 2016) may fear they will not find the associated necessary motivations and career supports (Janke, Quan, et al. 2023; O’Meara 2011). The connections between community engagement and approaches to scholarly work are varied based on institutional context and priorities (Pearl, Ward, et al. 2025), but generally represent an approach to doing scholarly work that prioritises community-engaged scholarship (Sandmann 2019).
We believe it is also valuable to consider the Operational Engagement class of institutions, the only one of the four that does not connect community engagement to the scholarly mission of the institution. Based on the Carnegie definition, the purpose of community engagement is to connect to and enhance the teaching, research and service missions of the institution, and Sandmann (2008) specifically identifies this connection as a way to distinguish community engagement from more traditional public service or outreach work in higher education. The Operational Engagement class is also more likely than not to include references to mutual benefit and collaboration; however, it is to a much lesser degree than the Pedagogical Partnerships and Scholarly Partnerships classes. Although more research is needed, one possibility is that the Operational Engagement definitions are more perfunctory in nature and do not demonstrate the best practices of community engagement. It could very well be that the evidence throughout the rest of these institutions’ applications is clearly tied to scholarly endeavours, but that connection is missing from the definition of community engagement provided.
Defining Collaboration and Mutual Benefit
According to Bringle, Clayton and Price (2009), one of the defining characteristics of civic and community engagement is mutually beneficial collaboration. Collaborative processes and mutually beneficial outcomes in our context are connected; therefore, it is unsurprising that collaboration and mutual benefit were tied together in our results, with both consistently mentioned in the definitions of the Pedagogical Partnerships and Scholarly Partnerships definitional types. Albeit to a lesser degree, as discussed above, the Operational Engagement definitions were also more likely than not to include these elements.
Collaboration in community–university engagement occurs when all partners are able to contribute knowledge, skills, and experience to identify issues and priorities, pose research questions, discuss possible approaches to resolving issues and defining and assessing success ( Bringle, Clayton & Price 2009; Davis, Kliewer & Nicolaides 2017; Stanton 2012; Trebil-Smith 2019). Collaboration implies a fundamental shift from traditional models of academic expertise conducting outreach and service to co-creative and co-productive development of knowledge and solutions (Anderson et al. 2025; Davis, Kliewer & Nicolaides 2017). A collaborative approach to scholarship has been recognised by community-engaged scholars as a normative expectation for community engagement, with the recognition that the epistemic perspective is participatory by its nature (Janke, Jenkins, et al. 2023). Collaborative partnerships are not stagnant and exist on a continuum, depending on the depth and nature of the engagement (Clayton et al. 2022; Enos & Morton 2003; Key et al. 2019; Kniffin et al. 2020).
There is significant research showing why collaboration and mutual benefit warrant explicit focus in definitions. Collaboration has been shown to address power imbalances and enable co-creation. For example, some of the elements that play a part in successful and effective community-university collaboration include intentionally sharing power and addressing imbalances (Anderson et al. 2025; Key et al. 2019; Wallerstein et al. 2019), focusing on co-creation (Trebil-Smith 2019), emphasising communication (Davis, Kliewer & Nicolaides 2017; Trebil-Smith 2019) and prioritising co-defined impact (Anderson et al. 2025; Trebil-Smith 2019). Further, the concept of full participation (Sturm et al. 2011) encourages us to consider how to create spaces ‘that enable people, whatever their identity, background, or institutional position, to thrive, realize their capabilities, engage meaningfully in institutional life and contribute to the flourishing of others’ (p. 3).
Similarly, for the concept of mutual benefit, research clearly shows its importance for diverse partners. Gelmon, Holland and Spring (2018) provide a valuable resource as we intentionally consider the assessment of benefits for different partners. For example, student benefits may include academic learning, career development, commitment to service or communication skills. Faculty outcomes include professional satisfaction, career advancement or new avenues for teaching and research. These types of outcomes are especially useful to consider in a context where the next generation of scholars are increasingly motivated by the impact they can have through community-engaged work (Post et al. 2016). Gelmon, Holland and Spring (2018) also discuss potential community outcomes, including the capacity to fulfil and expand services, a range of economic and social benefits or satisfaction with the partnership.
In the context of collaboration and mutual benefit, campus leaders of community engagement need to consider whether or not a definition of community engagement is complete if it does not intentionally include references to these particular elements. As evidenced by their successful classifications, the Engagement as Pedagogy institutions may have demonstrated evidence of collaborative processes and mutually beneficial outcomes in their full applications; however, this is not represented in their campus-wide definitions of community engagement. This finding further illustrates the complex nature of colleges and universities, and how they seek to institutionalise community engagement. Some values may surface in practice but are not expressly stated in a definition that is meant to characterise a complex system made up of multiple (and perhaps, contradictory) units, histories and priorities. If a definition of community engagement can serve to communicate a set of values to both internal and external audiences, what is the message being sent if fundamental values like collaboration and mutual benefit are not present in some definitional types? Our intent is not to suggest that simply inserting ‘collaboration’ and ‘mutual benefit’ into a definition represents a fundamental shift in an institution’s approach to community engagement. Rather, we encourage continued purposeful reflection on the ways that those values show up in the work being done in partnership with communities.
Variations in Institutional Type
Some of the variation in definitional focus might be attributed to institutional type. As we show in Table 7, the proportion of institutions who fall into the different classes varies depending on their focus. Some of these variations are not surprising. For example, when we look at 2-year institutions, which tend to have more of a focus on teaching, 63.64 per cent of the institutions have definitions that align with the classes that emphasise the connection between community engagement and teaching (Pedagogical Partnerships and Engagement as Pedagogy), which is far greater than those in the full sample (47.96%) that emphasise this connection. Similarly, it is unsurprising that more than half (54.17%) of the land-grant institutions in our sample fall into the Scholarly Partnerships class of definitions. Land-grant institutions have a long history of connecting higher education to the public good that dates back to their founding, so it would follow that the definitions of community engagement at these institutions would cut across teaching, research and service, and would emphasise the importance of mutually beneficial outcomes and collaborative partnerships.
On the other hand, some of the variations in definitions across institutional types are less apparent. For example, it is notable that nearly 42 per cent of the public institutions in our sample are in the Scholarly Partnerships class, while only 20 per cent of the private institutions fall into that class. In addition, the private institutions in our sample also had the greatest percentage of institutions (24.44%) in the Operational Engagement class. Without immediately apparent explanations for these types of variations, additional research is necessary.
Communicating to Different Audiences
Based on our analyses, the Pedagogical Partnerships and Engagement as Pedagogy definitional types included a specific focus on student learning through community engagement. Many students, particularly those from historically marginalised backgrounds, are interested in and motivated to engage in service-learning experiences (Pearl & Christensen 2017a, 2017b). Emphasising this component of community engagement can serve as a valuable signal to students who are motivated to actively engage with their communities, suggesting that the institution values and emphasises these types of experiences.
Like the Pedagogical Partnerships and Engagement as Pedagogy types, the Scholarly Partnerships type also included a focus on connections between community engagement and the teaching and learning mission of the institution. However, this focus on scholarly connections to community engagement was broader and also included an explicit connection between community engagement and the institution’s mission areas of research and service. These different definitional types can be important for faculty members and prospective faculty members, who are seeking to find alignment between their personal values and the institutions at which they work (Ward et al. 2024). These signals are also valuable messages for community engagement professionals (Dostilio 2017).
Based on our analyses, the Scholarly Partnerships, Pedagogical Partnerships and Operational Engagement definitional types are likely to include a specific focus on both collaborative processes and mutually beneficial outcomes, some of them with specific approaches identified. These definitions reinforce the notion of colleges and universities doing work in partnership with communities, rather than doing work for communities, which would be more aligned with traditional approaches to institutional service and outreach. This sends a signal about what the institution values in its approach to external community partners (individuals and organisations), donors, policymakers and legislators. For example, Weerts’ (2014) research on public universities noted that an engaged approach may lead to greater than expected state financial allocations, and initiatives like the Broader Impacts Toolkit, developed by the Center for Advancing Research Impacts in Society (ARIS), highlight the external funding potential of purposefully integrating a community-engaged ethos into academic work done in concert with the public (Hotaling et al. 2024).
Future Research
As the application encompasses a vast array of details, it is crucial to explore the definitions from a multitude of angles. For instance, finding alignment between the definitions of community engagement and the evidence provided in other areas of the application can be helpful in understanding examples the institutions choose to highlight. Similarly, future studies can further delve into how ongoing community-engaged programs and projects at the institution reflect the community engagement values signaled in institutional definitions of community engagement. Our findings show both variation and homogeneity in CE definitions, which presents an avenue for future research that maps how institutions integrate CE best practices into their applications, as well as their practices on campus.
Finally, it is important to note that the question related to critical approaches to community engagement did not show up consistently in any of the groups of definitions and was only present in 13.6 per cent of definitions. Therefore, while our discussion in this paper does not focus on this question, this element is an important thread to follow, and our research team has conducted a concurrent qualitative thematic analysis that is explicitly focused on the ways that critical approaches to community engagement show up in institutional definitions (Quito et al. 2026).
Limitations
We recognise that the data was based on cross-sectional data from previous years of CE classification, which constrains the ability to fully understand the typology of definitions with varying levels of emphasis placed on core principles associated with CE classification. Further, our sample size was relatively small for a latent class analysis, which may have affected the results. We also recognise that while our decisions regarding the number of classes were guided by different goodness of fit measures, our subjectivities may have played a role in that decision-making process. Additionally, with the data being from different years, we acknowledge that this may have been influenced by real-world events shaping the application. Finally, the classes observed are based on analysis rather than giving an in-depth insight into the signalling the institutions may be projecting or the process of generating their CE classification definitions and applications.
Conclusion
While there is only one ‘official’ definition of community engagement from the Carnegie Foundation, institutional variation exists. This study provides insights about the heterogeneity in institutional definitions of community engagement. Particularly when speaking about community engagement, which is in many ways antithetical to homogeneity, it is valuable to reflect on whether it is a worthwhile goal to determine a singular definition of community engagement for an institution. Conversely, if CE is defined too broadly, do its core values and principles then become diluted? Carnegie’s universal classification system was initially intended to be a system of categorisation, yet ended up becoming a de facto rankings system (Altbach 2002). In the pursuit of prestigious designations like the Elective Classification for Community Engagement, there may be tendencies towards institutions looking similar (DiMaggio & Powell 1983) by tailoring institutional policies and responses to the application framework. We encourage community engagement professionals and administrators to resist this temptation and reflect on approaches to describing and defining community engagement in a way that is authentic and appropriate to the specific institutional context, while maintaining its essential values. The four latent subgroups in our analysis place varying levels of focus on emphasis placed on collaboration, mutual benefit and connections to scholarship, which suggests varying potential avenues for institutions to signal values to specific groups. In delving into the complexities of institutions’ definitions of community engagement, we present varying profiles of definitions that institutions can tailor to their preexisting and future engagement with communities and how they want to define that blueprint.
Acknowledgement
The research in the current study was conducted through the Community-Higher Education Engagement Research (CHEER) Co-Lab, which was established as part of the Carnegie Elective Classifications Research Lab (CECRL). Initial seed funding for the CECRL was provided by the American Council on Education.
References
Altbach, PG 2002, ‘Differentiation requires definition: The need for classification in complex academic systems’, International Higher Education, vol. 26, pp. 2–3. https://doi.org/10.6017/ihe.2002.26.6956
American Council on Education n.d.-a, Access for researchers, Carnegie Classification of Institutions of Higher Education, viewed 22 April 2025, https://carnegieclassifications.acenet.edu/elective-classifications/community-engagement/researcher-access/
American Council on Education n.d.-b, The elective classification for community engagement, Carnegie Classification of Institutions of Higher Education, viewed 10 June 2024, https://carnegieclassifications.acenet.edu/elective-classifications/community-engagement/
Anderson, N, Cousins, G, Farrow, T, Fox, C, Hall, R, Hammelman, C, Joy, C, Kilmer, R, Nilsson, I, Racine, E, Terrell, D & White, B 2025, ‘Building impactful community-university partnerships through a shared power approach’, Journal of Community Engagement and Scholarship, vol. 18, no. 1. https://doi.org/10.54656/jces.v18i1.573
Avila, M (with Gecan, M & Peters, SJ) 2018, Transformative civic engagement through community organizing, Stylus.
Boyer, EL 1990, Scholarship reconsidered: Priorities of the professoriate, Carnegie Foundation for the Advancement of Teaching.
Boyer, EL 1996, ‘The scholarship of engagement’, Journal of Higher Education Outreach and Engagement, vol. 1, no. 1, pp. 11–20.
Bringle, RC, Clayton, P & Price, M 2009, ‘Partnerships in service learning and civic engagement’, Partnerships: A Journal of Service-Learning and Civic Engagement, vol. 1, no. 1, article 1.
Chait, R 1979, ‘College mission statements’, Science, vol. 205, no. 4410, p. 957. https://doi.org/10.1126/science.205.4410.957
Clayton, PH, Camo-Biogradlija, J, Kniffin, LE, Price, MF, Bringle, RG & Pier, AA 2022, The Transformational Relationship Evaluation Scale II (TRES II) reflection framework: Version 2, IUPUI ScholarWorks. https://doi.org/10.7912/zy6w-xj66
Collins, LM & Lanza, ST 2010, Latent class and latent transition analysis, Wiley & Sons. https://doi.org/10.1002/9780470567333
Cunningham, HR & Smith, P 2020, ‘Community engagement plans: A tool for institutionalizing community engagement’, Journal of Higher Education Outreach and Engagement, vol. 24, no. 2, pp. 53–68.
Davis, KL, Kliewer, BW & Nicolaides, A 2017, ‘Power and reciprocity in partnerships: Deliberative civic engagement and transformative learning in community-engaged scholarship’, Journal of Higher Education Outreach and Engagement, vol. 21, no. 1, article 1.
Dewey, J 1938, Experience & education, Simon and Schuster.
DiMaggio, PJ & Powell, WW 1983, ‘The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields’, American Sociological Review, vol. 48, no. 2, pp. 147–160. https://doi.org/10.2307/2095101
Doberneck, DM, Glass, CR & Schweitzer, J 2010, ‘From rhetoric to reality: A typology of publically engaged scholarship’, Journal of Higher Education Outreach and Engagement, vol. 14, no. 4, pp. 5–35. https://doi.org/10.54656/RPOV2967
Dostilio, LD (ed.) 2017, The community engagement professional in higher education: A competency model for an emerging field, Stylus Publishing, LLC. https://campuscompact.presswarehouse.com/browse/book/9781945459030/The-Community-Engagement-Professional-in-Higher-Education
Driscoll, A 2008, ‘Carnegie’s community-engagement classification: Intentions and insights’, Change: The Magazine of Higher Learning, vol. 40, no. 1, pp. 38–41. https://doi.org/10.3200/CHNG.40.1.38-41
Driscoll, A 2009, ‘Carnegie’s new community engagement classification: Affirming higher education’s role in community’, New Directions for Higher Education, vol. 2009, no. 147, pp. 5–12. https://doi.org/10.1002/he.353
Enos, S & Morton, K 2003, ‘Developing a theory and practice of campus community partnerships’, in B Jacoby (ed.), Building partnerships for service-learning, 1st edn, Jossey-Bass, pp. 20–41.
Farner, K 2019, ‘Institutionalizing community engagement in higher education: A case study of processes toward engagement’, Journal of Higher Education Outreach and Engagement, vol. 23, no. 2, article 2.
Fitzgerald, HE, Bruns, K, Sonka, ST, Furco, A & Swanson, L 2012, ‘The centrality of engagement in higher education’, Journal of Higher Education Outreach and Engagement, vol. 16, no. 3, article 3.
Furco, A, Weerts, D, Burton, L & Kent, K 2009, Assessment rubric for institutionalizing community engagement in higher education. https://hdl.handle.net/11299/213717
Gelmon, SB, Holland, B & Spring, A 2018, Assessing service-learning and civic engagement: Principles and techniques, Campus Compact.
Giles, DE 2008, ‘Understanding an emerging field of scholarship: Toward a research agenda for engaged, public scholarship’, Journal of Higher Education Outreach and Engagement, vol. 12, no. 2, article 1.
Holland, B 1997, ‘Analyzing institutional commitment to service: A model of key organizational factors’, Michigan Journal of Community Service Learning, vol. 4, pp. 30–41.
Hotaling, L, Lichtenwalner, CS, McDonnell, JD, O’Connell, K & Ferraro, C 2024, ‘The ARIS Broader Impacts Toolkit: An online guide to support Broader Impact project development and evaluation for researchers and BI professionals’, Journal of Community Engagement and Scholarship, vol. 17, no. 2. https://doi.org/10.54656/jces.v17i2.644
Janke, E, Jenkins, I, Quan, M & Saltmarsh, J 2023, ‘We’re talking about process: The primacy of relationship and epistemology in defining community-engaged scholarship in promotion and tenure policy’, Michigan Journal of Community Service Learning, vol. 29, no. 1. https://doi.org/10.3998/mjcsl.2734
Janke, E, Quan, M, Jenkins, I & Saltmarsh, J 2023, ‘Persistence and proliferation: Integrating community-engaged scholarship language into 59 departments, seven units, and one university’s academic promotion and tenure policies’, Michigan Journal of Community Service Learning, vol. 29, no. 1. https://doi.org/10.3998/mjcsl.2537
Key, KD, Furr-Holden, D, Lewis, EY, Cunningham, R, Zimmerman, MA, Johnson-Lawrence, V & Selig, S 2019, ‘The continuum of community engagement in research: A roadmap for understanding and assessing progress’, Progress in Community Health Partnerships: Research, Education, and Action, vol. 13, no. 4, pp. 427–434. https://doi.org/10.1353/cpr.2019.0064
Kniffin, L, Camo-Biogradlija, J, Price, MF, Kohl, E, Dickovick, ADC, Williams, J, Goodwin, J, Johnson, KV, Clayton, PH & Bringle, RG 2020, ‘Relationships and partnerships in community–campus engagement: Evolving inquiry and practice’, International Journal of Research on Service-Learning and Community Engagement, vol. 8, no. 1. https://doi.org/10.37333/001c.18586
Linzer, DA & Lewis, JB 2011, ‘poLCA: An R package for polytomous variable latent class analysis’, Journal of Statistical Software, vol. 42, no. 10, pp. 1–29. https://doi.org/10.18637/jss.v042.i10
Meyer, JW & Rowan, B 1977, ‘Institutionalized organizations: Formal structure as myth and ceremony’, American Journal of Sociology, vol. 83, no. 2, pp. 340–363. https://doi.org/10.1086/226550
Morphew, CC & Hartley, M 2006, ‘Mission statements: A thematic analysis of rhetoric across institutional type’, The Journal of Higher Education, vol. 77, no. 3, pp. 456–471. https://doi.org/10.1080/00221546.2006.11778934
Newsom, W & Hayes, CR 1991, ‘Are mission statements worthwhile?’, Planning for Higher Education, vol. 19, no. 2, pp. 28–30.
O’Meara, K 2011, ‘Inside the panopticon: Studying academic reward systems’, in JC Smart & MB Paulsen (eds), Higher Education: Handbook of Theory and Research, vol. 26, Springer Netherlands, pp. 161–220. https://doi.org/10.1007/978-94-007-0702-3_5
Pearl, A & Christensen, RK 2017a, ‘First-year student motivations for service-learning: An application of the volunteer functions inventory’, Michigan Journal of Community Service Learning, vol. 23, no. 2, pp. 66–82. https://doi.org/10.3998/mjcsloa.3239521.0023.205
Pearl, A & Christensen, RK 2017b, ‘First-year student motivations for service-learning: An exploratory investigation of minority student perceptions’, Journal of Higher Education Outreach and Engagement, vol. 21, no. 4, pp. 117–137.
Pearl, A, Jones-Friedrichs, C, Ward, E, Cunningham, HR, Azordegan, JM, Cruz, T, Deas, F, Locke, E, Mitchell, TD, Padalkar, T, Quito, D & White, CC 2025, ‘The Carnegie Elective Classification: A methodological guide for community engagement research’, International Journal of Research on Service-Learning and Community Engagement, vol. 13, no. 1. https://doi.org/10.37333/001c.151686
Pearl, A, Ward, E, Janke, E, Rios, M & Eatman, TK 2025, ‘Rewarding faculty publicly-engaged scholarship in promotion and tenure policies at Carnegie classified community engagement institutions: A descriptive analysis’, Journal of Participatory Research Methods, vol. 6, no. 5. https://doi.org/10.35844/001c.153977
Post, MA, Ward, E, Longo, NV & Saltmarsh, JA (eds) 2016, Publicly engaged scholars: Next generation engagement and the future of higher education, 1st edn, Stylus Publishing LLC. https://doi.org/10.4324/9781003446583-1
Quito, D, Locke, E, Cunningham, HR & Pearl, A 2026, ‘Signaling critical community engagement: Exploring definitions from the 2020 and 2024 Carnegie Classification data’, Michigan Journal of Community Service Learning, vol. 32, no. 1. https://doi.org/10.3998/mjcsl.8219
R Core Team 2024, R: A language and environment for statistical computing [Computer software], R Foundation for Statistical Computing. https://www.R-project.org/
Reed, AS & Swanson, LE 2022, ‘Transdisciplinary engagement: Advancing the community engagement mission for all’, Journal of Community Engagement and Scholarship, vol. 14, no. 3. https://doi.org/10.54656/jces.v14i3.45
Saltmarsh, J & Johnson, MB 2018, The Elective Carnegie Community Engagement Classification: Constructing a successful application for first-time and re-classification applicants, Campus Compact.
Sandmann, LR 2008, ‘Conceptualization of the scholarship of engagement in higher education: A strategic review, 1996–2006’, Journal of Higher Education Outreach and Engagement, vol. 12, no. 1, article 1.
Sandmann, LR 2019, Placing scholarly engagement ‘on the desk’, Campus Compact.
Sinha, P, Calfee, CS & Delucchi, KL 2021, ‘Practitioner’s guide to latent class analysis: Methodological considerations and common pitfalls’, Critical Care Medicine, vol. 49, no. 1, pp. e63–e79. https://doi.org/10.1097/CCM.0000000000004710
Stanton, TK 2012, ‘New times demand new scholarship II: Research universities and civic engagement: Opportunities and challenges’, Journal of Higher Education Outreach and Engagement, vol. 16, no. 4, article 4.
Sturm, S, Eatman, TK, Saltmarsh, J & Bush, A 2011, Full participation: Building the architecture for diversity and community engagement in higher education, Imagining America. https://surface.syr.edu/cgi/viewcontent.cgi?article=1001&context=ia
Trebil-Smith, K 2019, Perceptions of partnership: A study on nonprofit and higher education collaboration, Iowa Campus Compact. https://compact.org/sites/default/files/2024-01/2019_IACC_PerceptionsOfPartnershipStudy.pdf
Wallerstein, N, Muhammad, M, Sanchez-Youngman, S, Rodriguez Espinosa, P, Avila, M, Baker, EA, Barnett, S, Belone, L, Golub, M, Lucero, J, Mahdi, I, Noyes, E, Nguyen, T, Roubideaux, Y, Sigo, R & Duran, B 2019, ‘Power dynamics in community-based participatory research: A multiple-case study analysis of partnering contexts, histories, and practices’, Health Education & Behavior, vol. 46, no. 1_suppl, pp. 19S–32S. https://doi.org/10.1177/1090198119852998
Ward, E, Janke, E, Wrencher, E, Goodner, H, Mitchell, T & Hempill, M 2024, ‘Faculty community-engaged scholars choosing to come and stay: Alignment between individual and institutional identities and values’, International Journal of Research on Service-Learning and Community Engagement, vol. 12, no. 1. https://doi.org/10.37333/001c.127561
Ward, E, Shonkoff, E, Carlson, C & Stuetzle, C 2023, ‘Advancing community-engaged research via the food justice research and action cluster: A transdisciplinary ecosystem model advancing community’, Metropolitan Universities, vol. 34, no. 5, article 5. https://doi.org/10.18060/27154
Weerts, DJ 2014, ‘State funding and the engaged university: Understanding community engagement and state appropriations for higher education’, The Review of Higher Education, vol. 38, no. 1, pp. 133–169. https://doi.org/10.1353/rhe.2014.0038
Weick, KE 1976, ‘Educational organizations as loosely coupled systems’, Administrative Science Quarterly, vol. 21, no. 1, p. 1. https://doi.org/10.2307/2391875
Welch, M 2016, Engaging higher education: Purpose, platforms, and programs for community engagement, Routledge. https://doi.org/10.4324/9781003444473
Wendling, L 2023, ‘Evaluating community-engaged research in promotion and tenure’, Metropolitan Universities, vol. 34, no. 5, article 5. https://doi.org/10.18060/26658
Appendix
Response Probabilities by Class
| Defined Community | Yes | No |
|---|---|---|
| Pedagogical Partnerships | 0.2623 | 0.7377 |
| Scholarly Partnerships | 0.2311 | 0.7689 |
| Engagement as Pedagogy | 0.3499 | 0.6501 |
| Operational Engagement | 0.381 | 0.619 |
| Scholarly Mention | Yes | No |
| Pedagogical Partnerships | 1 | 0 |
| Scholarly Partnerships | 1 | 0 |
| Engagement as Pedagogy | 1 | 0 |
| Operational Engagement | 0 | 1 |
| Teaching and Learning | Yes | No |
| Pedagogical Partnerships | 0.831 | 0.169 |
| Scholarly Partnerships | 0.9742 | 0.0258 |
| Engagement as Pedagogy | 0.925 | 0.075 |
| Operational Engagement | 0 | 1 |
| Research and Creative Work | Yes | No |
| Pedagogical Partnerships | 0.3071 | 0.6929 |
| Scholarly Partnerships | 0.9437 | 0.0563 |
| Engagement as Pedagogy | 0.3106 | 0.6894 |
| Operational Engagement | 0 | 1 |
| Service | Yes | No |
| Pedagogical Partnerships | 0.083 | 0.917 |
| Scholarly Partnerships | 1 | 0 |
| Engagement as Pedagogy | 0.1958 | 0.8042 |
| Operational Engagement | 0 | 1 |
| Collaboration | Yes | No |
| Pedagogical Partnerships | 0.94 | 0.06 |
| Scholarly Partnerships | 0.9603 | 0.0397 |
| Engagement as Pedagogy | 0 | 1 |
| Operational Engagement | 0.619 | 0.381 |
| Mutual Benefit | Yes | No |
| Pedagogical Partnerships | 0.837 | 0.163 |
| Scholarly Partnerships | 1 | 0 |
| Engagement as Pedagogy | 0 | 1 |
| Operational Engagement | 0.5714 | 0.4286 |
| Critical Approach | Yes | No |
| Pedagogical Partnerships | 0.1531 | 0.8469 |
| Scholarly Partnerships | 0.1045 | 0.8955 |
| Engagement as Pedagogy | 0.1514 | 0.8486 |
| Operational Engagement | 0.1429 | 0.8571 |