Community Without Place the Impact of Distance Learning on College Student Perceptions of Community During the COVID 19 Pandemic

Community Without Place: The Impact of Distance Learning on College Student Perceptions of Community During the COVID-19 Pandemic A thesis submitted in partial fulfillment of the requirements for graduation with Honors in Sociology. 2022 ii Table of Contents Acknowledgements ............................................................................................................ iv Abstract ............................................................................................................................... v List of Tables ..................................................................................................................... vi Introduction ......................................................................................................................... 1 Setting ................................................................................................................................. 2 Defining Community .......................................................................................................... 3 Types of Community and Community Spaces ................................................................... 5 Place Attachment .............................................................................................................. 14 Importance of Community for College Students .............................................................. 16 Online Student Community .............................................................................................. 17 Methods............................................................................................................................. 21 Limitations .................................................................................................................... 23 Ethics............................................................................................................................. 24 Sample........................................................................................................................... 25 Data Analytic Technique .............................................................................................. 26 Survey Measures ........................................................................................................... 27 Independent Variable: Class Format. ............................................................................... 27 Independent Variable: Housing ......................................................................................... 27 Independent Variable: Place Attachment ......................................................................... 27 Independent Variable: Class Preferences ......................................................................... 28 Independent Variable: Extracurriculars ............................................................................ 29 Dependent Variables: Gemeinschaft and Gesellschaft .................................................. 29 Dependent Variables: Identificational and Purposive Community .............................. 30 Dependent Variable: Recent GG ....................................................................................... 31 Findings............................................................................................................................. 31 Univariate Analysis ....................................................................................................... 31 Bivariate Correlations ................................................................................................... 36 Effects of Class Format and Housing on Dependent Variables .................................... 37 Multiple Regression Analyses ...................................................................................... 39 Effects of Online Classes Beyond the First Semester ................................................... 41 Discussion ......................................................................................................................... 42 Practical Implications: Supporting Student Communities ............................................ 46 Limitations .................................................................................................................... 48 Directions for Further Research .................................................................................... 48 Conclusion ........................................................................................................................ 49 Appendix. Full Survey ...................................................................................................... 51 References ......................................................................................................................... 63 iii iv Abstract During the COVID-19 pandemic in 2020, the transition to online learning for many colleges and universities caused changes in the ways that students interacted with each other, and in the level of connection they felt to their school. Using survey data from 69 Whitman College students, this study investigates the question: for students during their first semester of college, how did starting school online, compared to in-person, affect their experience of community and connection with other students and professors? Ferdinand Tönnies concepts of gemeinschaft and gesellschaft provide a theoretical framework to explain the results. The findings reveal that students who took online classes in their first semester of college tended to feel lower levels of identification and belonging in a student community and had lower amounts of social interaction with other students. In online classes, relationships students had with each other also tended to be formal and based around academics, rather than friendly and informal. Suggestions are made for how instructors and institutions can encourage a stronger and more supportive student community in an online context. v List of Tables Table 1: Respondents class formats during their first semester at Whitman ................. 32 Table 2: Rankings of Aspects of Online and In-Person Classes .................................... 33 Table 3: Perceptions and Experience of Gemeinschaft and Gesellschaft ...................... 34 Table 4: Perception and Experience of Community ...................................................... 35 Table 5: Perceptions of Community, Gemeinschaft and Gesellschaft .......................... 36 Table 6: One-way ANOVA for Dependent Variables Across Class Format Groups ... 38 Table 7: Tukey Post-hoc Test for Dependent Variables Across Class Format Groups . 39 Table 8: One-way ANOVA for Dependent Variables Across housing situations ......... 39 Table 9: Multiple Regression Analysis of Variables Effects on Gemeinschaft and Gesellschaft .................................................................................................................... 40 Table 10: Multiple Regression Analysis of Variables Effects on Community .............. 40 Table 11: Comparison of a Semester with Most Online Classes and Most In-person Classes............................................................................................................................ 42 vi Introduction During the COVID-19 pandemic, many college and university students were required to attend school online, which caused changes in the contexts and types of interaction that they had with one another. Students in online classes, during a pandemic or not, tend to feel less connected to their school and to other students, which can negatively impact mental health and academic performance (Korpershoek et al. 2020; Shea et al. 2006; Moore 2014). Even during normal in-person school, a review of literature on schools and communities by Korpershoek et al. (2020) found that “having a sense of school belonging is positively related to student functioning such as students’ school motivation ... their social-emotional functioning such as their self-esteem … their classroom behavior … and their academic achievement” (Korpershoek et al. 2020:642). In an online learning context, a sense of community and belonging is much more difficult to create and maintain for students. In both online and in-person classes, frequent student interaction is strongly correlated with a sense of community and may even be necessary to create community (Dawson 2006). Using a survey of students at Whitman College, this paper addresses the question: For students during their first semester of college, how did starting school online, compared to in-person, affect their experience of community and connection with other students and professors? While much research has been done on learning communities both online and offline, the COVID-19 pandemic has provided a unique opportunity to compare interactions in an online and an in-person setting that share the same purpose and goals (learning), are comprised of similar groups of people, and have similar norms and expectations of members. At Whitman College, and likely many other colleges and universities, the online school experience was intended to closely match the experience students would have in-person. The classes and extracurricular activities offered by Whitman College remained largely the same in an online format, as did the groups of people participating in them. Whitman College’s course catalog remained very similar to what was offered in-person, with many of the same professors teaching the same classes, though slightly altered for an online format. Many student clubs also retained the same members and tried to recreate the activities they would have done in-person. In contrast to many previous studies of online student communities that have mostly focused on one learning environment (e.g., Shea et al. 2006; Fox and Roberts 1999), the transition to online school during the pandemic provides an opportunity to compare the effects of online and in-person learning environments on the same student community. This can help determine how being online affects students' experiences with school and community, separately from other factors that would normally influence whether students attend school online or in-person. This information could help colleges encourage community and improve student experiences both online and in-person. Setting Whitman College is a small liberal arts college in Walla Walla, Washington. During a normal, pre-pandemic year, students were required to live in dorms on campus their first two years at Whitman (with some exceptions), and juniors and seniors could choose to live in designated on-campus dorms or off campus (Whitman 2 College 2022). All classes would normally be held in-person in classrooms on campus. Partway through the Spring 2020 semester, as COVID-19 was spreading throughout the US, Whitman switched all classes to a fully online format, and encouraged all students who were able to leave campus to do so. Classes remained fully online through the end of the Fall 2020 semester, and in Spring 2021, some classes returned to in-person with required masks and social distancing. Since Fall 2022, Whitman College has had mostly in-person classes, with many classes having an option for students to attend online. During Fall 2020, the new first year class had a unique experience of starting college in a fully online format, where most students were not able to physically be on campus or meet their classmates in-person. This thesis investigates how beginning college in this online setting impacted the way that students felt connected with each other and with the college. Defining Community In order to study community online, an operational definition of community is required. Many sociologists have studied community, but historically there has been little consensus among sociologists on how it should be defined. In a survey of past research on community, Hillery (1955) found 94 different, and sometimes conflicting, ways that sociologists have defined community. The only commonality between them was that “...all of the definitions deal with people. Beyond this common basis, there is no agreement” (Hillery 1955:117). Despite this lack of agreement, Hillery (1955) found a set of 16 concepts that were used in different combinations in all descriptions, and three in particular that were used in the majority of definitions. While there are no aspects of community that all authors agreed on, “a majority of the definitions include 3 the following as important elements of community: area, common ties, and social interaction” (Hillery 1955:118). Multiple more recent sociologists have used definitions of community similar to, and based on, Hillery’s three elements (Sengstock 2001; Driskell and Lyon 2002). For example, Driskell and Lyon (2002) directly take inspiration from Hillery (1995) and define community as “identification with a specific geographic area, common ties through an identification by the residents with one another and with that area, and, finally, significant social interaction among the residents” (Driskell and Lyon 2002:375). Another example is Rovai (2002), who uses a definition of classroom community (which is different from overall student community), that focuses on what Hillery (1995) calls common ties: “sense of community, regardless of setting, include feelings of connectedness, cohesion, spirit, trust, and interdependence among members” (Rovai 2002:201). Because Rovai (2002) focuses on online classes, geographic area is not included in his definition. The same three core components of community are used by Sengstock (2001) to define community, though she views them as three distinct types of communities, as opposed to elements of a community. Sengstock (2001) argues that communities can be categorized as geographic, identificational, or purposive. Geographic refers to communities formed because its members live in the same place, and some amount of interaction and cooperation is unavoidable (Sengstock 2001). This is similar to Hillery’s (1995) concept of area. In the case of this thesis, that place would be the physical college campus where all students meet for classes, and where many students live. Identificational community is one defined by “a sense of identity or belonging” 4 (Sengstock 2001:153). Identificational community is similar to Hillery’s (1995) concept of common ties. In a college setting, identificational community can be measured by how connected a student feels to the college and other students, and how much they identify with their role as a student. Finally, a purposive community is one based on frequent social interaction, whose “members are drawn together for common objectives or purposes” (Sengstock 2001:153). Purposive community corresponds to Hillery’s (1995) social interaction and could describe a school or workplace where people interact frequently but do not necessarily feel a strong sense of connection with each other. To define community in this project, I use Sengstock’s (2001) three aspects of community: geographic, purposive, and identificational, though like Hillery (1995) and Driskell and Lyon (2002), I define them as different aspects of a community, as opposed to three distinct types of community. In an online learning context, class format (online or in-person) often alters or even removes the geographic aspect of community. My research question addresses how a change in the geographic aspect of community can cause changes to the purposive and identificational aspects of community. Types of Community and Community Spaces To define and measure community, it is also useful to consider Ferdinand Tönnies concepts of gemeinschaft and gesellschaft, which are often translated as community and society, respectively (Tönnies 2001). While Sengstock’s (2001) three aspects of community are useful to measure the strength of a community and how 5 connected members are to it, gemeinschaft and gesellschaft describe the types of interaction that occur within the community. The idea of gemeinschaft, usually translated as community, is arguably the most traditional conception of community, based around the idea of a village, and describes a community based on trust and mutual reliance (Tönnies 2001). In a gemeinschaft-type community, “trust and familiarity with others and viewing people as total persons who are significant in their own right provide a basis for core relationships in the community” (Bradshaw 2009:6). Gemeinschaft is more common in what Sengstock describes as identificational communities, where members have a strong sense of identification and belonging in the community. (Sengstock 2001). Gesellschaft, usually translated as society, describes a more utility-based approach to social relationships, which is associated with cities and workplaces that can be more alienating (Tönnies 2001). In a gesellschaft-like community, ties are less personal and social, and are instead based on official roles and a person's contributions to society. An example of a gesellschaft-like community would be a workplace, where most interactions are related to individuals' jobs, and their status in the community is heavily influenced by their job title. Although people can also develop gemeinschaft-like relationships with coworkers, the overall workplace community is often more gesellschaft-like. While these ties can still be friendly, they are often less personal, and provide less support outside of the roles that define the relationship (Bradshaw 2009). Gesellschaft is more common in what Sengstock (2001) calls purposive communities, which are based on frequent interaction around shared goals. To continue with the example of a workplace: a workplace community 6 would be purposive because it is based on frequent interaction around a shared objective. It is not identificational because employees do not necessarily feel identification or belonging in the community, and it is not necessarily based in geographical place. The lack of identificational community will discourage gemeinschaft-like relationships, and because the community is formed around a common objective (work), gesellschaft-like relationships will be more helpful in achieving this objective. At work, it is more useful to have a relationship with someone who provides knowledge and skills to successfully perform the job than to have a relationship with someone who provides emotional support. However, people who only have gesellschaft-like ties at work will likely find gemeinschaft-like ties in other areas of life. It can be helpful to think of gemeinschaft and gesellschaft as what individuals get out of being in a community: in a gemeinschaft-like community they get friendship and mutual support, while in a gesellschaft-like community they would get knowledge or resources that help them reach a goal, such as education in school or performing one's job. Scholars have argued that as community becomes less based around a geographic place, it shifts from being gemeinschaft-like to more gesellschaft-like, which results in more professional and impersonal ties (Driskell and Lyon 2002). Driskell and Lyon (2002) describe this process through three environments that community can occur in: local place, shared space, and cyberspace. Traditionally, community has been defined as being based around a “local place,” where people live near each other and depend on each other, which Driskell and Lyon (2002) argue encourages gemeinschaft more than any other community space. Because members 7 are more likely to consistently interact with each other, this type of place is likely to create the strongest community ties, with the least amount of intentional effort from members. As communities based around “local place” have become less common due to urbanization and more people living in cities where interaction with neighbors is not necessary, they have been replaced by what Driskell and Lyon (2002) call “shared space.” This includes spaces people voluntarily travel to and create community around, such as a school, a church, or a workplace. Because these are more voluntary, and community ties take more work beyond the primary focus of the organization, Driskell and Lyon (2002) argue that it is slightly more difficult to create a strong community in a shared space compared to a local place. This shared space encourages a more gesellschaft-like community, where ties are often based more on individuals’ roles in an organization than on personal ties (Driskell and Lyon 2002). The final type of space a community can occur in is what Driskell and Lyon (2002) refer to as “cyberspace”, or interactions that occur online without a physical space. This is the environment where it is the most difficult to create strong, meaningful, gemeinschaftlike ties, or a coherent community that trusts and relies on one another (Driskell and Lyon 2002). Cyberspace has many characteristics that are different from physical space, which are likely to encourage more gesellschaft and less gemeinschaft. Drawing on past research, Driskell and Lyon (2002) developed a list of characteristics that would describe a gemeinschaft-like social relationship: “the ideal gemeinschaft-type interactions reflect relationships that would be: Close and intimate, holistic and allencompassing, emotional and supportive, long term and enduring, based on common 8 values, [and] associated with regular social interaction” (Driskell and Lyon 2002:377). Driskell and Lyon (2002) argue that this kind of gemeinschaft-like community may not be possible in cyberspace for multiple reasons. First, the types of ties that occur online are socially ramified, meaning that people have a larger number of weaker ties, which are less likely to create community or mutual support and dependence (Driskell and Lyon 2002). While strong ties or friendships likely encourage gemeinschaft, a large number of weaker ties can help information travel faster through a community, which is useful for gesellschaft-like communities (Granovetter 1973). Online communities tend to be topically fused, with communication based around a narrow topic, as opposed to “holistic and allencompassing” relationships that would be more likely in physical space (Driskell and Lyon 2002). In an online school context, this could mean that most social interaction is focused on academics, and less on informal personal relationships. Driskell and Lyon (2002) argue that online communities are psychologically detached, allowing people to be members without any trust, commitment, or connection to other members. The final limitation of online communities that Driskell and Lyon (2002) describe is limited liability, which describes how people can leave an online community by simply clicking a button, where this would be a more difficult process in a place-based community. In a school context students are not likely to permanently leave, but by turning their camera off they can avoid being present or seen in the same way they would be in-person. Limited liability also means individuals feel less of a sense of responsibility to the group, which in online classes could translate to feeling less of a need to actively participate in class. 9 The idea of topical fusion in online communities was demonstrated by Fox and Roberts (1999), who studied how a community was formed around a professional email list. The partial anonymity of online interaction, as well as the fact that all messages were visible to all members of the community, influenced the types of communication that occurred. An interesting phenomenon they observed was heavy regulation of casual vs. professional communication on the email list. Because everyone can see all messages, casual or friendly conversations are often discouraged in favor of content that is directly useful to their profession. This topical fusion leads ties to be more gesellschaft-like, and less supportive of friendly personal relationships outside of the purpose of the community. When people only interact online for a specific professional or academic purpose, it is less likely that they will have opportunities for the type of informal social interactions that would lead to expressive ties, or a gemeinschaft-like community. If, as Driskell and Lyon (2002) argue, online communities have a large number of weak ties, are focused around a narrow topic, lack trust or commitment, and allow members to join or leave easily, then they will allow little opportunity for individuals to form meaningful personal relationships with other members of the community, or for a member to feel a deep sense of connection or belonging in the group. This means that the overall community will be more gesellschaft-like, as all interactions are based around members' roles as students and their contributions to the class. It also likely leads to less of what Sengstock (2001) calls identificational community, based on belonging and identification with the community. 10 Bradshaw (2009), in contrast, argues that place may no longer be necessary for community. While increasing urbanization and distance has created more gesellschaft-like communities, new technologies allowing easier travel and communication still allow gemeinschaft-like social relations. Some communities based around local place, such as suburbs and gated communities, may actually lead to more gesellschaft-like ties than many online communities (Bradshaw 2009). Bradshaw (2009) therefore argues for separating place from the concept of gemeinschaft, and instead focusing on the qualities of community, such as feelings of support and belonging, with or without place. A consequence of this is that communities not based around place become “variable rather than either-or ... The question is not whether you are in a community but how much community you have” (Bradshaw 2009:13). In communities defined by boundaries in physical space, a person is typically either in the community or outside of it. In communities not defined by physical space, such as online communities, it is more appropriate to talk about how much a person is a part of the community, and how connected they feel to other members (Bradshaw 2009). My research question uses Bradshaw’s (2009) approach of defining community separately from geographic place, and measures that community on a scale of how connected individuals feel to the community, rather than measuring it as a binary present or not. Similarly to gemeinschaft and gesellschaft, the idea of instrumental and expressive ties have been used to distinguish between ties that are formal or personal, though on a more individual level. Instrumental ties are those used for “tangible resources, knowledge, and information for instrumental purposes”, while expressive 11 ties are more personal and involve “social support and identity-conferral (e.g., friendship)” (Kim and Rhee 2010:481). Kim and Rhee (2010) use instrumental and expressive ties to classify social ties in a professional setting. Kim and Rhee (2010) found that in a large technology company, instrumental social networks significantly increased organizational commitment, or likelihood of an individual to stay at their job. While they did not find a significant effect of expressive ties on organizational commitment, they found in a review of past literature that expressive networks tend to increase “job satisfaction and social integration in the organization” (Kim and Rhee 2010:484). While an employee's organizational commitment and social integration in a technology company is likely different from that of a student at a college, instrumental and expressive ties likely have similar effects. Instrumental ties may aid students' experience in the classroom and their professional relationships with classmates and professors, while expressive ties can help develop a sense of community and connection, especially outside of class. Measuring the types of individual ties that make up a community can help determine the characteristics of the overall community. In a gesellschaft-like community, formed for the purpose of achieving a goal and based around formal roles, ties between members are more likely to be instrumental, since those ties will be more useful to be a successful member of the community (Granovetter 1973). A gemeinschaft-like community, which is more based on trust and mutual support, will be more encouraging of friendly expressive ties (Driskell and Lyon 2002). Gemeinschaft, gesellschaft, and purposive and identificational community can all be used together to determine how connected individuals are to a community, as 12 well as what they get out of being in the community (either emotional support, or knowledge and resources). While Sengstock’s (2001) three aspects of community are useful to measure the strength of a community and how connected members are to it, gemeinschaft and gesellschaft describe the types of interaction that occur within the community. Two examples can help illustrate the difference: at a very large university, a student likely does not have personal relationships with the majority of other students, though they interact with others often for academic purposes. Despite experiencing more gesellschaft than gemeinschaft in this community, the student may feel a strong sense of identification and belonging at the school. This student would likely have friendships and other gemeinschaft-like relationships outside of the school. In this situation, there is a high level of identificational and purposive community, and a lower level of gemeinschaft (meaning more gesellschaft). It is more difficult to find an example of an opposite situation, because for either gemeinschaft or gesellschaft to be present, there needs to be some level of community (purposive, identificational, and/or geographic). However, one example that comes close would be an individual being invited to dinner with a friend's family. A family is typically very gemeinschaft-like, with trusting and supportive ties between members. Although these gemeinschaft-like ties likely extend to the guest, that individual may not interact with the others frequently, or feel identification or belonging in that group. This means that there is a high level of gemeinschaft, but a low level of purposive and identificational community, at least for that individual. Although purposive communities tend to encourage gesellschaft and identificational communities tend to 13 encourage gemeinschaft, both gemeinschaft and gesellschaft can be present in any purposive or identificational community. Place Attachment The idea of place attachment can help explain why a physical space is important for creating a sense of community. Chow and Healey (2008) use place attachment to explain the process first year college students typically experience when leaving home to go to college. They argue that home is an important part of one’s identity, which can make the transition to a new location for college very difficult (Chow and Healey 2008). While leaving a parent’s home can make that home seem more significant, college students often come to see college as a second home (Chow and Healey 2008). As students spend more time at college, they grow more attached to it, and it can become a part of their identity. While place attachment partially depends on physical aspects of the location, it is also heavily dependent on the social relationships that occur in that place (Chow and Healey 2008). Chow and Healey (2008) also argue that for college students, place attachment and place identity are dynamic and changing, and “Significant disruption to the people, activities or processes, and places involved in place attachment(s) and place identity will ultimately threaten to undermine the very foundations of [place attachment]” (Chow and Healey 2008:370). Transitioning to online school during the pandemic had significant effects on all of these factors, which likely impacted how people feel attached to place and to the community that would normally occur in it. This reinforces the idea that physical space is important for identificational and purposive community. 14 When discussing place, it is important to note the difference in how sociologists use the terms space and place. While a space is simply a physical location and the things in that location, a place is a space that people have attached meaning to, either by naming it, living in it, or using it for various activities (Cresswell 2004). These terms provide more context for Driskell and Lyon’s local place and shared space. They argue that “The shared space has a location, but unlike the local place, the specific location is not important for common ties or social interactions to exist (e.g., the church might move from the downtown location to the suburbs, but the community remains)” (Driskell and Lyon 2002:374). This means that place attachment is less likely in a shared space than in a local place, because in a shared space, people are more attached to the people and activities than the physical space. In cyberspace, there is no place or space to be attached to, though there can still be connection to other people or a community. When people do grow attached to a physical place, the social interactions attached to the place are often more important to place attachment than the physical characteristics of the place. A study by Reese, Oettler, and Katz (2019) asked participants to imagine the place where they lived if it lost a physical feature, such as buildings, parks, or natural features, or a social feature, such as personal friends and family, events, or resident attitudes like "friendly" and "welcoming" (Reese et al. 2019:2). They found that an imagined loss of either a social or physical element of a place resulted in a reduced feeling of place attachment, and that a loss of a social element caused a larger reduction in place attachment than a loss of a physical element (Reese et al. 2019). When school moved online for Whitman College, there 15 was a loss of physical place for most students, which likely led to reduced place attachment. Although some of the social interaction that would have occurred at the physical campus continued online, there was likely less, especially outside of class, and it was not tied to a physical place. Even students who stayed on campus while classes were online likely experienced less place attachment to the college because of a reduction in social interaction attached to the place. Importance of Community for College Students Students’ sense of community in school is important for their mental health, academic performance, and satisfaction with their overall school experience (Korpershoek et al. 2020; Shea et al. 2006; Branand et al. 2015). For example, Branand et al. (2015) found that “inclusion of the college community in the self at the end of the junior year … predicted increases in satisfaction with the college experience upon graduation” (Branand et al. 2015:837). College is not only a place for students to get an education, but it is also an important time when students start to be more independent and develop their identities (Branand et al. 2015). This process of identity formation is heavily influenced by an individual's environment and their relationships with the people around them (Branand et al. 2015). Being heavily involved with a community and having strong ties to other students will improve students' satisfaction with their college experience and aid the process of their own identity development (Branand et al. 2015). Besides the mental and emotional benefits, a sense of community has also been shown to improve students' academic performance (Korpershoek et al. 2020; Moore 2014). In a book on the determinants of positive outcomes from college, 16 Christopherson (2021) discusses why people choose to attend college, rather than reading and studying the same materials independently. He argues that the main benefit in attending a college over independent learning lies in the community and relationships they form with other students and professors. This is important both in class, to directly benefit academics, but also outside of class to create a general sense of community and support for learning (Christopherson 2021). Moore (2014) similarly argues that “A sense of community allows students to feel connected not only to their instructors and classmates but also to the content itself” (Moore 2014:20). Although community is often seen as only important for students’ emotional wellbeing, it also has a significant effect on academic performance. Online Student Community Student interaction in class has been found by multiple past studies to significantly increase students’ feeling of community (Young and Bruce 2011; Shea et al. 2006; Dawson 2006). In online college courses, Young and Bruce (2011) found that interaction and collaboration between students in class was likely to increase feelings of community, as students in certain types of programs that encouraged more collaboration, including education and health sciences, tended to feel a stronger sense of student community than students in other programs (Young and Bruce 2011). To successfully build community, Young and Bruce (2011) argue that professors teaching online classes should intentionally create a balance between “task-driven” and “socio-emotional” interactions, which encourages collaboration and exchange of ideas (Young and Bruce 2011:226). This is further illustrated by what Shea et al. (2006) describe as “teaching presence,” or the way a professor structures a course to 17 encourage discussion and active participation. When a class is structured in a way that encourages classroom community, (which is more gemeinschaft-like), it has a positive impact on student learning and motivation (Shea et al. 2006). While it is still possible, this type of teaching presence is much more difficult to create in an online setting. In a study of a college Spanish class, Moore (2014) similarly finds that students in a hybrid class have less sense of community, and less satisfaction with their learning, compared with students in a fully in-person class. In order to create student community, these in class interactions have to be of a certain type. When instructors assign group activities based on “knowledge acquisition” that could be done independently, it “may actually create feelings of anxiety and conflict rather than feelings of a connected learning community” (Young and Bruce 2011:225). It is important that in class collaboration is also helpful and relevant to what students are learning, and not simply an attempt to build community while learning (Young and Bruce 2011). To be conducive of community, Dawson argues that in class interactions need to be synchronous, as “the mere act of accessing online content or reading forum postings does not assist in the establishment of a social identity or presence among fellow members, and therefore, does not contribute to community building” (Dawson 2006:155). To successfully create community in an online class, Young and Bruce (2011) argue that instructors need to be intentional about creating learning environments that encourage active participation and collaboration to achieve shared learning goals. Though the structure of classes is important to student community, interactions outside of class are equally important (Royal and Rossi 1996; Dawson 18 2006). For undergraduate and postgraduate students, Dawson (2006) found that frequent communication between students significantly increased their sense of community. Extracurricular activities also increase student feelings of community because they not only increase the amount of interaction between students outside of class, but they also often provide opportunities for students to interact with professors or other school staff in a less formal context, which can increase feelings of connection to the school (Royal and Rossi 1996). Besides interaction in and out of class, factors that previous studies have found to affect student sense of community include engagement with learning, grade level, and school size (Royal and Rossi 1996; Young and Bruce 2011). Grade level is important because as students spend more time at a school, they will have more opportunities to form ties with other students and become involved in the school community (Royal and Rossi 1996). In this thesis, grade level is not a factor as all questions ask respondents about their experiences during their first semester at Whitman College. Engagement with learning is important in relation to students' interactions in class. Students' interaction and collaboration in class will be more conducive to community if they are more engaged with the class material (Bruce and Young 2011). In a review of literature, Royal and Rossi (1996) find that “previous studies have suggested that school size is inversely related to sense of community at school” (Royal and Rossi 1996:938). Although this thesis is limited to one school and cannot investigate the effect of school size, it is likely that Whitman College students felt higher levels of community than students would at a larger school. 19 When school is moved online, there is more topical fusion, where most interaction is centered around academics and occurs in a formal context, and there is less opportunity for informal personal relationships. While this is theoretically true of in-person classes as well, there are many more opportunities for casual conversations before, after, or even during an in-person class. In an online class, nearly all communication is visible to the professor and other students, which makes monitoring and regulating interaction much easier. This visibility of communication reduces the opportunities for casual interactions that are not related to academics, which encourages a more gesellschaft-like community. While a gesellschaft-like community based specifically around learning is beneficial to students' learning and motivation, a gemeinschaft-like community is also important for students to create a sense of belonging, connection, and identification with a community (Shea et al. 2006; Christopherson 2021). Gesellschaft-like community is seen more frequently online and is very likely to still be present in an online learning context. There is a debate, however, around whether a gemeinschaftlike community is possible online, where relationships are based on trust, mutual reliance, and multiple shared interests. To describe community, I use five aspects from past literature: geographic, identificational, purposive, gemeinschaft, and gesellschaft. Geographic is how tied to a physical place the community is. Identificational measures whether members feel identification and belonging in the community. Purposive measures how often community members interact with each other, and how satisfied they are with the amount of interaction they have. Gemeinschaft and gesellschaft describe whether the 20 community provides members with friendship and support (gemeinschaft), or knowledge and resources that are useful for achieving a goal (gemeinschaft). Identificational and purposive measure how individuals are connected to the community, both though personal identification and belonging, and interaction with other members. Gemeinschaft and gesellschaft describe the type of support or resources that members receive as a result of membership in the community, as well as motivations for membership in the community. The geographic aspect of community, which is diminished or removed in an online learning context, has been shown by past research to affect gemeinschaft and gesellschaft, as well as identificational community (Driskell and Lyon 2002; Bradshaw 2009; Sengstock 2001). When communities move away from local place and into cyberspace, they tend to become more gesellschaft-like (Driskell and Lyon 2002). A loss of physical place prevents place attachment, which makes identificational community less likely (Reese, Oettler, and Katz 2019). My research question asks in what ways online learning, or the reduction of the geographic aspect of community, influences the level of purposive and identificational community students experience at school, as well as how gemeinschaft-like or gesellschaft-like that community is. Methods To collect data, I used an online Qualtrics survey, distributed to Whitman College students. 84 people responded to the survey. All responses with less than 80% of questions answered were removed, resulting in a sample size of 69. Most of these removed responses had less than 50% of questions answered, and many were 21 missing answers to questions central to the analyses. All participants were current students at Whitman College who began attending before 2022 and were over 18 years old. The survey was posted on February 7, 2022 and remained available to take until February 21, 2022. Participants were recruited through various college email lists and posts on social media. Upon completing the survey, participants had the option to be entered into a raffle to win two $10 prizes. All responses were anonymous. The survey consisted of 30 questions, with some questions using skip logic to not be displayed to respondents for whom they did not apply. The survey was limited to Whitman College largely out of convenience, because as a student I had access to students who could take the survey, and access to information about the college's COVID-19 and online learning policies. Limiting the sample to a single college also limited the number of other variables that could influence community, such as different school sizes and environments, and differing responses to the COVID-19 pandemic. In in-person classes, differing policies about masking and social distancing, and different student attitudes about COVID-19 could influence feelings of community. In online classes, the technology and preparedness of teachers, and the resources made available to students could have a large impact on feelings of community. General school and student attitudes toward COVID-19 and online learning could also have an impact. By limiting the sample to one school, students’ experiences in both types of classes, and the resources available to them, remain relatively similar. 22 Most of the questions on the survey ask students about their first semester of college (for transfer students, their first semester at Whitman College). This is intended to reduce the differences caused by upperclassmen having time to build friendships and community before having to move to online classes. By asking respondents about their first semesters, I can compare how students experienced community online compared to in-person, with no (or less) pre-existing ties with the college or other students. Limitations The survey had a small number of respondents, which resulted in a small sample size, particularly for the groups whose first semester classes were all online (n=7), and some online (n=6). This makes it more difficult to infer results to a larger Whitman student population, or to college students in general. The sample size was also too small for a more detailed analysis of different types of online classes, including synchronous and asynchronous. For questions that ask about respondents’ first semester of college, respondents’ memory may not be accurate, especially for upperclassmen who have had multiple years of college since their first semester. A request for respondents to review their transcript from their first semester and report their GPA for that semester, as well as questions about housing their first semester, aimed to improve respondents’ memory of their first year. While Whitman students in the same class year tended to have the same class format, there are other significant variables besides class format that differ between years. For example, some students may have had online classes during the pandemic 23 in high school, and therefore have more experience and skills for creating community online. Other world events including the pandemic could also create differences in community across different years. Professors became better equipped to teach online classes over time, so students who had online classes later likely had better experiences. Because no respondents who started attending Whitman in 2018 or 2019 did all online classes their first semester, and all respondents who began attending Whitman in Fall 2020 had all online classes their first semester, class year only differed from class format for students who began attending Whitman after Fall 2020, with the exception of some outliers who took a small number of online classes in earlier years. Because of this close correlation, the influences of class year and class format on community could not be separated for most respondents. Although the unique circumstances of the pandemic allow for research that can help to better understand online communities, the results of this study may also be less applicable to online communities in other times. The online student community created by the COVID-19 pandemic was relatively unique in that members did not choose to participate online rather than in-person. Because students were required to participate in school online, they likely had different experiences and perceptions of the online community than individuals who willingly chose to participate in school (or any other community) online. Ethics All responses to the survey were anonymous, and names and IP addresses were not associated with responses. At the end of the survey, respondents had the 24 option to enter their name and email for a raffle to win one of two $10 prizes. This information was collected in a separate survey and was not associated with responses from the main survey. Data was kept on a password-protected computer, and analyzed using Qualtrics, Excel, and Jamovi. The original data was only accessible to the main researcher and advisors and will be deleted by everyone who has access to it within 10 years of data collection. This paper only reports aggregate data, and not individual responses to the survey. This research did not pose any risks to the researcher or participants that were greater than those encountered in daily life. Sample My sample consisted of 69 Whitman college students. 14.5% of respondents were 18 years old, 23.2% were 19, 7.2% were 20, 30.4% were 21, 21.7% were 22, and 2.9% were 23. Respondents were 73.9% female, 21.7% male, and 4.3% were non-binary, third gender, or preferred not to say. In 2021, 56.0% of Whitman students were women, and 43.9% were men (Whitman College 2021). People who selected female or non-binary, third gender, or prefer not to say were overrepresented in my sample. Respondents could select one or more races they identified as. 94.2% selected white, 1.5% selected American Indian, Native American, or Alaska Native, 1.5% selected Asian or Asian American, 1.5% selected Middle Eastern or North African, and 4.4% selected “something not listed here,” with write-in answers including “mixed” and “none.” 15.9% of respondents indicated that they were Hispanic or Latino. Of the overall Whitman student population in 2021, 61.7% were white, 0.7% were Native American, Native Hawaiian, or Pacific Islander, 6.1% were Asian, and 25 2.4% were Black or African American (Whitman College 2021). 9.0% of Whitman students were Hispanic or Latino (Whitman College 2021). White and Hispanic or Latino people were overrepresented in my sample, and Asian and Black or African American people were underrepresented, compared to the overall Whitman College population. The largest portion of respondents began attending Whitman College in Fall 2018 (42.0%), followed by Fall 2021 (27.5%). The remaining respondents’ first semesters were dispersed fairly evenly from “before Fall 2018” to Fall 2021, with most students starting in Fall semesters. Whitman College has a roughly equal number of students who are sophomores, juniors, and seniors, and the first-year class is around 60% larger than the other classes. Data Analytic Technique Data was cleaned by removing all responses with less than 80% of questions answered. Analyses were done using Qualtrics, Excel and Jamovi. Univariate analysis was performed on all composite variables to find general trends. Bivariate analysis was done using ANOVA to measure the effects of online classes on gemeinschaft, gesellschaft, and the purposive and identificational aspects of community. Tukey's post-hoc tests provided more detailed information about the results of the ANOVA tests. Multiple regression analyses were used to determine whether extracurricular activities and class preferences, in addition to class format, had a significant effect on gemeinschaft/gesellschaft and community. A second ANOVA analysis was done to determine the effects of housing on the same dependent variables. The effect of place attachment on gemeinschaft, gesellschaft, and community was measured with a 26 Pearson's correlation test. To determine effects of online classes beyond the first semester, univariate analysis was performed on the class comparison variable. Survey Measures Independent Variable: Class Format. My main independent variable is class format, measuring what proportion of respondents’ classes were online or in-person. The class format variable is compiled from a set of six questions that ask respondents to indicate the number of classes they took in various formats, outlined in Table 1. Classes that respondents indicated were “fully in-person” or “hybrid- attended mostly in-person” were grouped into a “mostly in-person” category, and the remaining class formats were grouped into a “mostly online” category (see Table 1). Responses to the class format questions were used to calculate the percentage of respondents’ total number of classes that they took mostly or fully online during their first semester. Table 1 outlines which class formats were grouped as “mostly online” or “mostly inperson.” Using these percentages, respondents were divided into three groups based on class format: all online, some online, and all in-person. Independent Variable: Housing. Housing measured whether respondents lived on campus or off campus their first semester. This helps to determine the effects of the geographic aspect of community on purposive and identificational community, and on gemeinschaft and gesellschaft (Sengstock 2001). Although my research question focuses on online classes as an independent variable, physical separation is an important factor in the changes to community online. Independent Variable: Place Attachment. The place attachment variable is based on Chow and Healey’s (2008) discussion of place attachment during the 27 transition from a parental home to college and helps to understand how feeling attached to the physical place of the college influences community. To measure place attachment, a set of questions asked respondents to indicate how much they thought of various places as home during their first semester of college, on a scale of “a lot”, “a moderate amount”, or “not at all.” Respondents who selected “does not apply” for any of these questions were not included in the composite variable including that question. These questions were combined to make a college place attachment variable consisting of “the building I lived in while attending college,” “Whitman College campus,” and “Walla Walla.” A second variable measuring parental home place attachment consisted of “parent/guardians’ home” and “the city or town my parents/guardians live in.” Independent Variable: Class Preferences. To control for respondents’ introversion or extroversion, and individual preferences for particular types of community experiences, a class preferences variable was included to measure whether respondents preferred a gemeinschaft-like or gesellschaft-like student community. A set of two ranked choice questions asked respondents to rank various items in order of how important they are to them in both online classes and in-person classes (see Table 2). The rankings of each of these items were added together from both questions, with gesellschaft-like items (e.g., getting a good grade, learning the material well) reverse coded, and gemeinschaft-like items (e.g., feeling supported by classmates, interacting and collaborating with classmates) normally coded. This created a class preferences score, showing whether respondents’ preferences for classes were closer to gemeinschaft or gesellschaft. The class preferences score was internally consistent, with a Cronbach’s alpha of 0.7. 28 Independent Variable: Extracurriculars. The extracurriculars variable describes the number of extracurricular activities, such as clubs or team sports, that respondents participated in during their first semester of college. This was included as an independent variable, based on Royal and Rossi’s (1996) discussion of how extracurricular activities help students build stronger ties with each other and with professors. Because other independent variables focus on in-class interaction, it is useful to also account for structured interaction outside of class. Dependent Variables: Gemeinschaft and Gesellschaft. Several Likert scale questions were combined to construct variables measuring respondents’ perceptions of gemeinschaft and gesellschaft. These questions included statements about respondents’ first semesters and asked them to choose “describes me extremely well,” “describes me moderately well,” or “does not describe me at all”. Responses corresponded to a score, with “does not describe me at all” scored as 0, “describes me moderately well” scored as 1, and “describes me extremely well” scored as 2. These scores of individual questions were added together in groups corresponding to gemeinschaft and gesellschaft. Gemeinschaft, based on descriptions from Tönnies (2001) and Driskell and Lyon (2002), included “I had close friends at school”, and “I felt I could rely on my classmates”, and gesellschaft included “most of my interactions with classmates felt formal or professional.” (Driskell and Lyon 2002). The gesellschaft score was subtracted from the gemeinschaft score to get a variable I will refer to as GG, which measures gemeinschaft and gesellschaft on a scale with a higher score being more gemeinschaft, and a lower score being more gesellschaft. The gemeinschaft and gesellschaft variables were constructed by adding together 29 items listed in Table 3. While amounts of gemeinschaft and gesellschaft can be measured separately, it is more useful in this study to measure them on a scale from one to the other, to measure which type of community was more present. Because the amount of community is already being measured by the purposive and identification variables, measuring the amount of gemeinschaft and gesellschaft would be somewhat redundant. Measuring gemeinschaft and gesellschaft on the same scale allows an easier comparison of types of community across differences in independent variables. Gemeinschaft, gesellschaft, and the combined GG variable were all internally consistent, with a Cronbach's alpha of 0.9 for gemeinschaft, 0.8 for gesellschaft, and 0.9 for GG. Dependent Variables: Identificational and Purposive Community. Variables measuring the identificational and purposive aspects of community, as defined by Sengstock (2001), were constructed using Items from Likert scale questions similar to the gemeinschaft and gesellschaft variables. The purposive community variable also includes items from a set of questions that asked respondents to indicate how often they interacted with other students in various contexts, on a scale from “frequently” to “never”. Purposive community, based on Sengstock’s (2001) definition, included “I had as much social interaction with other students as I wanted to have,” as well as questions about frequency of interaction with other students. Identificational community included “I felt I belonged at Whitman,” and “Being a Whitman student was a significant part of my identity” (Sengstock 2001). The individual measures that make up these variables are described in Table 4. The identificational community variable was created by adding together the scores of the corresponding sub measures. 30 The purposive community variable was constructed by adding together the Z-scores of each sub variable, to standardize responses to different types of questions. Both variables were internally consistent, with both identificational and purposive community having a Cronbach's alphas of 0.9. The identificational and purposive variables were also combined to create an overall community variable, with Cronbach’s alpha 0.9. Dependent Variable: Recent GG. The recent GG variable measured students’ experiences of gemeinschaft during their most recent semester. A subset of the gemeinschaft and gesellschaft Likert scale questions were asked again, though about the most recent semester (Spring 2022, when the survey was published), instead of the respondents’ first semester. Items in this variable included “I feel like I am part of a student community,” and “most of my interactions with other students are directly related to school” (reverse coded). This variable is intended to measure whether taking online classes during the first semester of college has lasting effects on students’ experiences of community beyond the first semester. This recent gemeinschaft-gesellschaft variable is internally consistent, with a Cronbach's alpha of 0.8. Findings Univariate Analysis The majority of respondents took four or more classes their first semester and took no online classes. Table 1 shows the percentage of respondents that took each number of classes in each class format. For example, 85.5% of respondents indicated 31 that they took 4 or more fully in-person classes their first semester. The percentage of respondents who took all online classes their first semester was 10.4%, 9.0% of respondents took some online classes, and 80.6% took all in-person classes their first semester. Table 1. Respondents class formats during their first semester at Whitman (%) Class format Mostly in person: Fully in-person Hybrid, attended inperson Mostly online: Hybrid, attended online Online, synchronous Online, asynchronous Number of classes 0 1 2 3 4+ 10.1 — 1.4 2.9 85.5 92.6 2.9 — 1.5 2.9 97.1 85.5 94.1 1.4 — 4.4 1.4 1.4 — — 2.9 1.5 — 10.1 — For the class preferences variable, respondents tended to prioritize gemeinschaft-like items (e.g., support from classmates and professors, interaction and collaboration) over gesellschaft-like items (e.g., getting a good grade, learning the material well) in all types of classes (see Table 2). Among the highest ranked items for both types of classes were “feeling supported by other students” and “interacting and collaborating with other students.” The mean ranking for all gemeinschaft-like items was higher for online classes than in-person classes, and the mean ranking for all gesellschaft-like items was higher for in-person classes than online classes. Students prioritized gemeinschaft-like qualities over gesellschaft-like qualities in all classes, though the preference for gemeinschaft-like qualities was stronger in online classes. 32 Table 2. Rankings of Aspects of Online and In-Person Classes (%) Ranking place Measure Online Classes: Gesellschaft: Getting a good grade Learning the material well Feeling interested and engaged by the class material Gemeinschaft: Interacting and collaborating with other students Feeling supported by other students Feeling supported by the professor In-person classes: Gesellschaft: Getting a good grade Learning the material well Feeling interested and engaged by the class material Gemeinschaft: Interacting and collaborating with other students Feeling supported by other students Feeling supported by the professor 1st 2nd 3rd 4th 5th 6th Mean ranking 2.6 3.0 2.7 30.8 21.5 12.3 30.8 20.0 18.5 12.3 18.5 10.8 9.2 13.9 1.5 33.8 32.3 21.5 9.2 3.1 0 2.2 4.4 4.6 4.6 15.4 15.4 36.9 23.1 4.5 0 4.6 0 9.2 26.2 60.0 5.4 9.2 15.4 24.6 35.4 13.8 1.5 3.3 16.7 22.7 18.2 28.8 10.6 19.7 21.2 22.7 18.2 3.0 15.2 3.0 2.8 3.5 2.6 40.9 30.3 15.2 6.1 4.5 3.0 2.1 4.2 7.6 7. 24.2 19.7 27.3 13.6 3.9 1.5 1.5 7.6 10.6 18.2 60.6 5.2 10.6 13.6 22.7 19.7 28.8 4.5 3.6 In general, respondents tended to experience higher levels of gemeinschaft than gesellschaft. Table 3 shows responses to individual measures that make up the gemeinschaft and gesellschaft variables. 33 Table 3. Perceptions and Experience of Gemeinschaft and Gesellschaft (%) Measure Gemeinschaft: I trusted my classmates I felt that I could rely on my classmates I felt that my classmates relied on me I had close friends at school I had a lot of friends at school I felt that a student community existed at Whitman I felt that I was a part of a student community I knew many of my classmates personally, outside of class I voluntarily spent time with my classmates outside of class I talked with classmates outside of class often I had as much social interaction with other students as I wanted to have Gesellschaft: Most of my interactions with classmates felt formal or professional Most of my interactions with other students were directly related to school I cared about my classmates outside of class* I felt that my classmates cared about me outside of my role in class* I felt disconnected from my classmates Note: *reverse coded “How well did each of the following describe you during your first semester at Whitman?” Moderately Extremely Not at all well well 16.4 20.9 22.4 17.4 29.0 49.3 62.7 56.7 37.7 47.8 34.3 16.4 20.9 44.9 23.2 14.5 26.1 42.0 37.7 43.5 36.2 30.4 37.7 31.9 29.0 27.5 31.9 44.9 39.1 27.5 20.3 39.1 40.6 52.2 34.8 13.0 31.9 18.8 34.8 40.6 33.3 40.6 34.8 32.8 36.2 44.8 29.0 22.4 Respondents also tended to report relatively high levels of both identificational and purposive community. Table 4 shows responses to each of the questions that make up the purposive and identificational community variables. For the majority of these questions measuring community, the percentage of respondents who indicated that the statement described them “extremely well” was larger than the percentage of respondents who indicated it described them “not at all.” 34 Table 4. Perception and Experience of Community (%) “How well did each of the following describe you during your first semester at Whitman?” Moderately Extremely Not at all well well Measure Identificational community: I felt that I belonged at Whitman 24.6 39.1 36.2 I felt that I had many things in common with my classmates 26.1 43.5 30.4 I felt that I was a part of a student community 26.1 37.7 36.2 Being a Whitman student was a significant part of my identity 21.7 42.0 36.2 My academic major was a significant part of my identity 47.8 29.0 23.2 I had a lot of friends at school 29.0 47.8 23.2 I had close Friends at school 17.4 37.7 44.9 I felt disconnected from my classmates* 32.8 44.8 22.4 Purposive community (perceptions): I voluntarily spent time with my classmates outside of class 29.0 31.9 39.1 I talked with classmates in class often 10.1 52.2 37.7 I talked with classmates outside of class often 27.5 44.9 27.5 I had as much social interaction with other students as I wanted to have 20.3 39.1 40.6 Frequency of interaction (%) Purposive community (actions): Never Rarely Occasionally Frequently In class 1.6 7.8 29.7 60.9 Talking with classmates or professors before or after class 1.5 15.2 42.4 40.9 Professors’ office hours 16.7 40.9 37.9 4.5 Studying with other students 13.6 24.2 31.8 30.3 Clubs or team sports 31.8 18.2 18.2 31.8 Other people who lived with me in my dorm or house 13.6 6.1 15.2 65.2 Doing things with friends, not related to an official school activity 7.6 7.6 10.6 74.2 A paid job, internship, or volunteer work 40.0 15.4 18.5 26.2 College events or activities 20.0 21.5 36.9 21.5 Note: *reverse coded 35 Using the combined GG variable (gemeinschaft - gesellschaft) and the community variable (purposive + identificational), we can see that overall, respondents reported relatively high levels of both community and gesellschaft. Because the community variables used Z-scores to standardize responses to different questions, the mean score is zero. The median scores show that the distributions of both community and GG are negatively skewed, indicating that students in their first semesters at Whitman tended to experience relatively high amounts of both identificational and purposive community, and perceived community to be more gemeinschaft like than gesellschaft-like (see Table 5). Table 5. Perceptions of Community, Gemeinschaft and Gesellschaft Variable GG Community N 69 69 Mean 9.5 0.0 Median 12.0 0.7 SD 7.0 1.9 Minimum -5.00 -4.3 Maximum 20.0 3.0 Bivariate Correlations Purposive and identificational community were positively correlated with r = 0.68 and p < 0.001, indicating that if an individual has a high level of one type of community, they are likely to also have a high level of the other type of community. Gemeinschaft and gesellschaft were negatively correlated, with r = -0.72 and p < 0.001, indicating that as gemeinschaft increases, gesellschaft decreases, and vice versa. This negative correlation supports my method of measuring gemeinschaft and gesellschaft on a scale from one to the other, instead of measuring amounts of each separately. The GG variable was correlated with the combined community variable, with r = 0.89 and p < 0.001. If students perceive community to be more gemeinschaft36 like than gesellschaft-like, they are likely to also experience high levels of purposive and identificational community. Housing (on or off campus, and in or out of Walla Walla) was highly correlated with class format, with r = 0.88 and p < .001. Most respondents who took all in-person classes lived on campus, and all respondents who took all online classes lived outside of Walla Walla. There was a strong correlation between community and place attachment with the school. School place attachment was correlated with GG with r = 0.59 and p < 0.001 and was correlated with community with r = 0.40 and r < 0.001. Place attachment with a parent or guardian's home had no significant effect on student community or gemeinschaft and gesellschaft. Students who felt high levels of connection to community and viewed the community as gemeinschaft-like were also likely to think of the college campus, surrounding area, and residential buildings as home. Effects of Class Format and Housing on Dependent Variables To analyze the effects of class format on community and gemeinschaft and gesellschaft, I conducted an ANOVA analysis across groups of respondents who took all in-person classes, some online classes, and all online classes. The ANOVA test shows that proportion of online classes has a significant negative effect on gemeinschaft/gesellschaft, with a P-value of less than .05 (see Table 6). Respondents who took all online classes their first semester were likely to perceive community as more gesellschaft-like than gemeinschaft-like. Perceptions of overall community (constructed from the identificational community and purposive community sub 37 variables) was also lower for respondents who took all online classes their first semester at Whitman than respondents who took all in-person classes. Respondents who took all online classes were likely to experience less community and perceive that community to be more gemeinschaft-like than gesellschaft-like. Table 6. One-way ANOVA for Dependent Variables Across Class Format Groups Gemeinschaft Gesellschaft GG Identificational community Purposive community Overall community Recent GG F 6.75** 2.06 4.82* 7.82** 16.96** 17.07*** 2.42 df (class format) 2 2 2 2 2 2 2 df (DV) 8.79 8.26 8.48 8.72 8.65 8.78 10.44 Note: *p < .05. **p < .01. ***p < .001 The recent GG variable, measuring gemeinschaft and gesellschaft during the most recent semester (Spring 2022, when the survey was published) was not significantly affected by proportion of online classes during their first semester, nor by class preferences. Student experiences of gemeinschaft and gesellschaft were only significantly influenced by the formats of the classes they were taking that semester, and not by the format of classes taken previous semesters. This suggests that online classes only have an effect on community while students are taking them, and the characteristics of the community can likely change as class format changes. A Tukey post hoc test reveals that for all dependent variables affected by class format, there are only significant differences between the all online group and the all in-person group (see Table 7). This is likely because the some online group was smaller than the others (n=6), which makes significant results less likely. 38 Table 7. Tukey Post-hoc Test for Dependent Variables Across Class Format Groups (Mean Differences) Some online All online 2.11 6.61** — 4.51 3.34 — 9.2** 5.85 2.02 — 4.63** 2.60 5.26 — 14.31*** 9.05* 1.22 — 3.40*** 2.18 Gemeinschaft: All inperson Some online GG: All in-person Some online Identificational community: All in-person Some online Purposive community: All in-person Some online Overall community: All in-person Some online Note: *p < .05 **p < .01 ***p < .001 A second ANOVA test reveals that housing (living on or off campus) during the first semester of college had a similar effect on gemeinschaft, gesellschaft, and community as class format (see Table 8). Because housing and class format are highly correlated, with r = 0.84 and p < 0.001, both variables could not be analyzed together in a multiple regression. Table 8. One-way ANOVA for Dependent Variables Across housing situations Housing F GG 16.1** Community 61.4*** Note: *p < .05. **p < .01. ***p < .001 df (housing) 1 1 df (DV) 11.5 13.1 Multiple Regression Analyses To determine whether extracurriculars and class preferences affected community and gemeinschaft and gesellschaft, I used a multiple regression analysis, 39 with the class format variable as a percentage of the respondents’ total number of classes that were taken online. Class preferences helps to control for individual preferences for either gemeinschaft or gesellschaft like interaction, to see if being online had an effect separately from individual attempts to create certain types of community. Including extracurriculars as an independent variable similarly helps account for student interactions outside of class. The analysis revealed that neither of these variables had significant impacts on either gemeinschaft, gesellschaft or community when compared to the effect of class format (see Tables 9 and 10). Students’ preferences for the amount and type of in-class interaction they had with other students, and their number of extracurricular activities, did not have a significant effect on students’ experiences of community when compared to class format. Table 9. Multiple Regression Analysis of Variables Effects on Gemeinschaft and Gesellschaft Predictor Estimate Intercept 15.00** Class Format -0.08** Class Preferences -0.13 Extracurriculars 0.85 2 R = 0.23 Note: *p < .05. **p < .01. ***p < .001 SE 4.96 0.03 0.10 0.65 t 3.03 -2.86 -1.37 1.30 Table 10. Multiple Regression Analysis of Variables Effects on Community Predictor Intercept Class format Class preferences Extracurriculars R2 = 0.39 Estimate 1.12 -0.03*** -0.03 0.23 SE 1.12 0.01 0.02 1.16 Note: *p < .05. **p < .01. ***p < .001 40 t 0.94 -4.59 -1.09 0.14 Class year and coresidents (family, friends, students) are both factors that could influence community. A Pearson's r test shows that living with family is correlated with taking online classes, with r = 0.60 and p < 0.001, and living with other students was negatively correlated with taking online classes, with r = -0.74, and p < 0.001. Class year was not significantly correlated with class format, because in Fall 2021, most Whitman classes returned to being in-person (Whitman College 2021). However, for respondents who began attending Whitman before Fall 2021, class year is highly correlated with class format, with r = 0.70 and p < 0.001. Because all respondents who took all online classes their first semester started in Fall 2020, and all respondents who started in Fall 2020 took all online classes their first semester, an analysis of the effects of class year separately from class format does not provide any useful information. However, because of these correlations, I can assume that both class year and coresidents influenced students’ experiences of community. Effects of Online Classes Beyond the First Semester The Likert scale questions asking respondents to compare the semester they did the most online classes with the semester when they did the least online classes reveal that the same trends in online classes during the first semester likely apply similarly to later semesters (see Table 11). Respondents reported that they experienced gemeinschaft-like items, including feeling that most interactions were formal, professional, or directly related to school, more online than in-person. They also reported that they experienced gesellschaft-like items, including “I felt that I was part of a student community”, and an item indicating an identificational community (“I felt I belonged at Whitman”), higher in in-person classes than in online classes. An 41 item indicating purposive community, “I had as much interaction with other students as I wanted to have” was lower in online classes. These results suggest that even after the first semester of college, students in online classes tend to feel lower levels of purposive and identificational community and feel that the community that is present is more gesellschaft-like than gemeinschaft-like. Table 11. Comparison of a Semester with Most Online Classes and Most In-person Classes. I felt disconnected from other Whitman students Most of my interaction with classmates felt formal or professional Most of my interactions with other students were directly related to school I had as much interaction with other students as I wanted to have I felt that I belonged at Whitman I felt that I was a part of a student community N More inperson Equally inperson and online 51 2.0 11.8 86.3 50 12.0 24.0 64.0 52 0.0 34.6 65.4 52 49 82.7 55.1 11.5 40.8 5.8 4.1 52 73.1 23.1 3.8 More online Discussion Students who took all online classes during their first semester at Whitman were likely to experience lower levels of community and perceive that community to be more gesellschaft like, compared to students who took all in-person classes their first semesters. Because class format and housing are highly correlated, students who lived off-campus their first semester also likely experienced lower levels of community and more gesellschaft compared to students who lived on campus. 42 For students who took all online classes, those that stayed on campus to do online classes likely had more in-person interactions with other students compared with students who were off campus, which led to more community and gemeinschaft. The effect of living space on student community reflects Driskell and Lyon’s (2002) argument that gesellschaft is more likely in shared space than local place, and even more likely in cyberspace. My finding that online classes lead to more gesellschaft is consistent with past research arguing that online communities tend to be focused around a single topic, and discourage communication about other things (Driskell and Lyon 2002; Fox and Roberts 1999). In the class preferences questions, students prioritized gemeinschaft-like qualities over gesellschaft-like qualities in all classes, though this preference was stronger for online classes. This is likely because students experienced less gemeinschaft online, and therefore wanted online classes to be more gemeinschaftlike. This provides support for Christopherson’s (2021) argument that one of the reasons students choose to attend college, as opposed to studying independently, is for the social aspects that come with it. The higher prioritization of gemeinschaft in an online learning context shows that this type of community is less present, and more needed by students, in an online format. The findings reveal that a reduction in geographical community (physical separation in online classes) led to lower levels of identificational and purposive community. Although there was still some amount of both purposive and identificational community in all learning formats, the presence of a geographic space 43 and in-person interactions leads to students experiencing higher levels of identificational and purposive community. Being online directly impacts purposive community by limiting the contexts in which students interact. In an in-person class context, where the geographic aspect of community is present, students would interact with other students frequently simply by walking to class and being on campus, without having to put in much effort. As Dawson (2006) argues, this type of frequent synchronous interaction is important and possibly even necessary for creating a sense of student community. In an online learning context, these unplanned interactions don’t happen often, and any interactions outside of formal class meetings require conscious planning and effort. While purposive community is still very possible in an online format, it happens to a lesser degree because it requires conscious effort, and not all students consistently make an effort to maintain this type of community. Identificational community also decreases in an online format, though it does not decrease to the same degree as purposive community. This is likely because interaction is a significant part of purposive community, and interaction was more directly limited by an online class format than more subjective feelings of identification and belonging (Sengstock 2001). People can theoretically have high levels of identificational community without being physically near each other or interacting frequently. However, my results show that when people are physically near each other and have more in-person interaction, it does lead to higher levels of identificational community. When people are physically near each other and are able to have frequent in-person interaction, it is easier to feel a sense of connection with 44 others, and to identify with the community and place. These findings fit with those of Branand et al. (2015) who found that college students’ identity development is heavily influenced by their physical and social environments. The findings reveal that students tend to feel more gemeinschaft in-person, and more gesellschaft online. Students who lived on campus also felt more gemeinschaft than students who lived off campus. This can be explained by Driskell and Lyon’s (2002) local place, shared space, and cyberspace. When students live on campus, it is a local place for them, which encourages the most amount of gemeinschaft. When students live off campus, but still attend in-person classes on campus, the campus becomes a shared space for them, as they do not live there but travel there to participate in shared activities. This creates some gemeinschaft, though less than in a local place. When students participate in online classes, particularly if they are far from campus, they are participating in community through cyberspace, which encourages more gesellschaft. My findings support this idea that as community moves from local place to shared space to cyberspace, community is likely to become more gesellschaft-like and less gemeinschaft-like. The increased amount of gesellschaft online can also be explained by Driskell and Lyon’s (2002) idea of “topical fusion”, and Fox and Roberts’ (1999) findings about online content regulation. The gesellschaft measure included students feeling that interactions were formal, professional, or directly related to academics, and occurred more in an online context. In an online class, all interactions are typically visible to everyone in the class, and any side conversations that are not directly related to the class content are discouraged. In a normal in-person class, students 45 often have short conversations before, after, or even during class that can help students get to know each other better and form expressive ties. In many online classes, there is less opportunity for these kinds of casual one-on-one conversations, which will create a more gesellschaft-like community. Practical Implications: Supporting Student Communities To support a student community that is gemeinschaft-like (based in mutual trust and support), it is important that members have opportunities to interact frequently, both in and out of class (Dawson 2006). For in-person communities, local place and shared space provide a space for regular social interaction that is not necessarily connected to academics or defined by individuals’ roles as students. This provides opportunities for individuals to form expressive ties, which can help in developing a sense of connection and belonging. In an online learning context, most interaction happens around class meetings, or for a specific purpose usually related to academics. This environment causes most interactions to be instrumental, or focused on achieving a specific goal, which limits the amount or strength of personal expressive ties that are formed. Students’ feelings of community and connection online could likely be improved by a shared “space” online, where students can interact in a more casual non-academic context. At Whitman College in 2020, some examples of non-academic digital shared spaces included student clubs, college organized social events, and an official Whitman College server on Discord (a messaging app). 46 To encourage more gemeinschaft in class, it can be helpful to allow students to interact and collaborate in ways that are less structured and supervised. Topical fusion, one of the characteristics that Driskell and Lyon (2002) argue leads to more gesellschaft in online contexts, is often caused by a feeling of being watched or judged, which is more common in online contexts (Fox and Roberts 1999). Providing opportunities for students to collaborate without feeling that they are being watched may help encourage gemeinschaft. For example, Zoom breakout rooms with only a few students collaborating on a project would likely lead to more gemeinschaft-like interaction than would occur in a larger class or in a lecture. This works best with activities that require multiple people to complete, as activities that could be completed independently but are done as a group solely for the purpose of building community could actually lead to increased tension and discomfort in the group (Young and Bruce 2011). To summarize, gemeinschaft can be encouraged in classes where students can collaborate in ways that feel productive and not solely for community building purposes, and don’t feel that they and their interactions are constantly visible to everyone in the class. These findings can be applied to building community in physical space as well. In a normal in-person context, space and student interaction are important to build community. When students live on campus, this makes the school a local place for those students, which encourages a gemeinschaft-like community. For students that do not live on campus, shared spaces where students can choose to interact with others in an informal setting can also help form a gemeinschaft-like community. If schools want to encourage a more friendly and supportive student community, it can 47 help to pay attention to how physical spaces are used by students for both academic and social activities. Limitations A limitation of this study is that it mostly focuses on in-class interaction and does not do as much to address other student interactions outside of class. Though student interactions will likely be in the same format (online or in-person) as their classes, there are different ways students can choose to interact outside of class that likely have different effects on community. This study is also only limited to one college and studied a relatively unique year during the pandemic. While this does provide interesting information about online communities, the results act more like a case study and may not be generalizable to other colleges or online communities. Directions for Further Research Future research on this topic might focus on how students chose to interact with each other outside of class. For example, texting, video calls, or social media may each have different effects on students’ sense of community. Students’ motivation for connecting outside of class and factors that influence their likelihood to do this could also be useful to study. Different ways of structuring online classes could also be valuable to study. For example, reading and writing forum posts, and different kinds of group activities or discussions, could lead to different amounts or kinds of community. My survey aimed to measure differences between synchronous and asynchronous online classes, as well as hybrid classes, though my sample size was too small for a meaningful 48 analysis at this level of detail. Although my analysis treated all online classes as equal in terms of how they impact community, there are likely interesting differences in the details of how online classes function. Conclusion An online class format makes creating and maintaining a community based on belonging, trust, and support much more difficult. In an online class format, students tend to experience a lower level of community (both purposive and identificational) and perceive that community to be more gesellschaft-like than gemeinschaft-like. This means that students do not have frequent interaction with each other, do not feel a sense of identification or belonging in a student community, and feel that most of their interactions in schools are formal and directly related to academics. In order to maintain a strong gemeinschaft-like community, a conscious effort is required by both students and professors to create the types of environments and interactions that can build community. Professors can strengthen community in class by finding engaging ways for students to learn collaboratively in class. Students can build community by engaging in class, and by maintaining frequent contact with other students outside of class. The school itself can also influence online student community through the resources it provides to students and Professors. If students have easy ways to interact with other students and professors in an informal context, this will encourage the formation of expressive ties which will lead to a stronger gemeinschaft-like community. While my findings show that community ties become less close and more gesellschaft-like in an online class format, online spaces can be structured in ways 49 that make expressive ties and a supportive gemeinschaft-like community easier to maintain. 50 Appendix. Full Survey Thank you for participating in this survey on students' perceptions of community in online and in person learning environments. Please only complete this survey if you are over 18 years old, and are a current Whitman College student who began attending Whitman College before 2022. Participants who complete the survey will have the option to be entered into a raffle to win one of two prizes. The survey is not expected to pose risks to you, but does include some questions that may be uncomfortable to answer. Participation in this survey is voluntary, and you may skip any question or quit the survey at any time with no penalty. If you choose to participate, you will be asked questions about how you attended classes, your relationships with other students and faculty, your academic performance during your first semester of college, and where you lived during your first semester of college. This survey does not ask any questions that could identify you, and all responses are anonymous. If you have any questions about this survey, please contact Julian Landau (landaujr@whitman.edu), thesis chair Michelle Janning (janninmy@whitman.edu), or the Whitman College IRB (irb@whitman.edu). By clicking “next”, you agree that you are over 18 years old, are a current Whitman college student who began attending Whitman College before 2022, and you agree to participate in this survey. The next few questions are intended to determine who is taking this survey, to ensure responses are representative of a larger student population How old are you? o 18 o19 o 20 o 21 o 22 o 23 or older 51 Which race(s) do you identify as? (select all that apply) ▢ American Indian, Native American, or Alaska Native ▢ Asian or Asian American ▢ Black or African American ▢ Middle Eastern or Northern African ▢ Native Hawaiian or Pacific Islander ▢ White ▢ Something not listed here (please specify) ________________________________________________ Are you Hispanic or Latino? o Yes o No What is your gender? o Male o Female o Non-binary / third gender o Something not listed here ________________________________________________ o Prefer not to say 52 When did you begin attending Whitman College? o Before Fall 2018 o Fall 2018 o Spring 2019 o Fall 2019 o Spring 2020 o Fall 2020 o Spring 2021 o Fall 2021 Are you a transfer student? (attended another college before Whitman College) o Yes o No When was the first semester you attended any college or university? o Before Fall 2018 o Fall 2018 o Spring 2019 o Fall 2019 o Spring 2020 o Fall 2020 o Spring 2021 o Fall 2021 Where did you live during your first semester at Whitman? o On campus (residence hall, fraternity, interest house) o Off campus, in Walla Walla (this includes off-campus college-owned housing) o Outside of Walla Walla 53 During your first semester at Whitman, which of the following groups of people lived in the same building as you? (Select all that apply) If your living situation changed partway through the semester, select who you lived with for the majority of the semester. ▢ ▢ ▢ ▢ ▢ Parents/guardians Family other than parents/guardians Other college students Friends who are not college students Someone not included in any of the other categories The next few questions are about classes you took your first semester at Whitman. It may be helpful to refer to your transcript for information you do not remember. During your first semester at Whitman, how many of your classes were fully in-person? o 0 o 1 o 2 o 3 o 4+ During your first semester at Whitman, how many of your classes were hybrid (option to attend either in-person or online), which you attended mostly or fully in-person? o 0 o 1 o 2 o 3 o 4+ During your first semester at Whitman, how many of your classes were hybrid (option to attend either in-person or online), which you attended mostly or fully online? o 0 o 1 o 2 o 3 o 4+ During your first semester at Whitman, how many of your classes were fully online, and mostly or fully synchronous? o 0 o 1 o 2 o 3 o 4+ 54 During your first semester at Whitman, how many of your classes were fully online, and mostly or fully asynchronous? o 0 o 1 o 2 o 3 o 4+ During your first semester at Whitman, how many of your classes were fully online, and a fairly even combination of synchronous and asynchronous? o 0 o 1 o 2 o 3 o 4+ During your first semester at Whitman, how many co-curricular activities connected to the college did you participate in? (for example: clubs, team sports, etc.) this does not include internships, volunteer, or paid work. o 0 o 1 o 2 o 3 o 4 o 5 o 6+ During your first semester at Whitman, did you have a paid job, an internship, or do volunteer work? o Yes o No During your first semester at Whitman College, how much did you think of each of the following places as home? A moderate Does not A lot Not at all amount apply The building I lived in while attending college Whitman College campus Walla Walla Parent/guardians home The city or town my parent/guardians live in Somewhere not listed here o o o o o o 55 o o o o o o o o o o o o o o o o o o What was your GPA your first semester at Whitman College? Please enter only a number with two decimal places (example: 3.50). If this question does not apply to you, write "N/A". Please refer to your transcript to ensure this is accurate. ________________________________________________________________ The next several questions ask you about your first semester at Whitman. Please remember to think about your experiences then, which may be different from your experiences now. Please indicate how well each of the following statements describes you during your first semester at Whitman. ("classmate" refers only to students you had classes with during your first semester) Describes Does not me Describes me describe me at extremely moderately well all well I cared about my classmates academic success I felt that my classmates cared about my academic success o o o o o o I felt that my classmates cared about me outside of my role in class o o o o o o o o o o o o o o o I cared about my classmates outside of class I knew many of my classmates personally, outside of class I voluntarily spent time with my classmates outside of class I felt that I had many things in common with my classmates 56 Please indicate how well each of the following statements describes you during your first semester at Whitman. ("classmate" refers only to students you had classes with during your first semester) Does not Describes me Describes me describe me extremely well moderately well at all or does not apply I felt that I could rely on my classmates I felt interested and engaged in class o o o o o o o o o o o o o o o o o o I often worked on homework or other things during classes o o o I felt that my classmates relied on me I trusted my classmates I felt disconnected from my classmates I feel that I did well in most of my classes Please indicate how well each of the following statements describes you during your first semester at Whitman. ("classmate" refers only to students you had classes with during your first semester) Describes me Does not Describes me moderately describe me extremely well well at all I talked with classmates in class often I talked with classmates outside of class often o o o o o o Most of my interactions with classmates felt formal or professional o o o 57 Please indicate how well each of the following statements describes you during your first semester at Whitman. (These questions are about other Whitman students in general, not necessarily classmates) Describes me Describes me Does not extremely moderately describe me at well well all Most of my interactions with other students were directly related to school o o o I had as much social interaction with other students as I wanted to have o o o o o o o o o I had a lot of friends at school I had close friends at school 58 Please indicate how well each of the following statements describes you during your first semester at Whitman. Does not Describes me Describes me describe me extremely well moderately well at all Being a Whitman student was a significant part of my identity o o o My academic major was a significant part of my identity o o o I felt that a student community existed at Whitman o o o o o o o o o I felt that I was a part of a student community I felt that I belonged at Whitman 59 Please indicate how well each of the following statements describes you now Describes me Describes me Does not extremely moderately describe me well well at all Being a Whitman student is a significant part of my identity o o o o o o o o o o o o o o o I think in person interactions are more meaningful than online interactions o o o Most of my interactions with other students are directly related to school o o o My academic major is a significant part of my identity I feel that a student community exists at Whitman I feel that I am part of a student community I feel that I belong at Whitman 60 Please compare the semester at Whitman when you did the most online classes with a semester when you did the most in-person classes, and indicate when you experienced each of the following items more. (If multiple semesters were the same, choose the most If you have not taken both online and in person classes at Whitman, choose "does not apply" for all items. Experienced Experienced this about Experienced Does not this more the same in this more apply in-person person as online online I felt disconnected from other Whitman students o o o o I had as much interaction with other students as I wanted to have o o o o Most of my interactions with classmates (people I had classes with) felt formal or professional o o o o Most of my interactions with other students were directly related to school o o o o o o o o o o o o I felt that I belonged at Whitman I felt that I was a part of a student community Please rank the following in order of how important they are to you in an online class. (1 is most important). ______ Getting a good grade ______ Learning the material well ______ Interacting and collaborating with other students ______ Feeling supported by other students ______ Feeling supported by the professor ______ Feeling interested and engaged by the class material Please rank the following in order of how important they are to you in an in-person class. (1 is most important). ______ Getting a good grade ______ Learning the material well 61 ______ Interacting and collaborating with other students ______ Feeling supported by other students ______ Feeling supported by the professor ______ Feeling interested and engaged by the class material Please indicate how often you interacted with other Whitman students in each of the following contexts during your first semester at Whitman. Frequently Occasionally Rarely Never A paid job, internship, or volunteer work o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o College events or activities (for example an on-campus concert, study break, etc.) o o o o Other (please specify) o o o o In class Talking with classmates or professors before or after classes Professors office hours Studying with other students Clubs or team sports Other people who live with me in my dorm or house (if applicable) Doing things with friends, not related to an official school activity Would you like to be entered into a raffle to win one of two prizes? Your responses on this survey will remain anonymous. o Yes o No 62 References Bradshaw, Ted K. 2009. “The Post-Place Community: Contributions to the Debate about the Definition of Community.” Community Development 39(1):5-16. Branand, Brittany, Debra Mashek, Laura Wray-Lake, and John K. Coffey. 2015. "Inclusion of College Community in the Self: A Longitudinal Study of the Role of Self-Expansion in Students' Satisfaction." Journal of College Student Development 56(8):829-844. Chow, Kenny, and Mick Healey. 2008. “Place attachment and place identity: First-year undergraduates making the transition from home to university.” Journal of Environmental Psychology 28:362-372. Christopherson, Neal. 2021. Transformative Experiences in College: Connection and Community. London: Lexington Books. Cresswell, Tim. 2015. Place: An Introduction. UK: John Wiley & Sons Ltd. Dawson, Shane. 2006. “A Study of the Relationship Between Student Communication Interaction and Sense of Community.” The Internet and Higher Education 9:153-162. Driskell, Robyn Bateman and Larry Lyon. 2002. “Are Virtual Communities True communities? Examining the Environments and Elements of Community.” City and Community 1(4): 373-390. Fox, Nick, and Chris Roberts. 1999. “Gps in Cyberspace: The Sociology of a ‘Virtual Community’.” The Sociological Review 47(4):643-671. Granovetter, Mark S. 1973. “The Strength of Weak Ties.” American Journal of Sociology 78(6):1360-1380. 63 Hillery, George A., Jr. 1955. “Definitions of Community: Areas of Agreement.” Rural Sociology 20(2):111-123. Kim, Young-Choon and Mooweon Rhee. 2010. "The Contingent Effect of Social Networks on Organizational Commitment: A Comparison of Instrumental and Expressive Ties in a Multinational High-Technology Company" Sociological Perspectives 53(4):479-502. Korpershoek, H., E. T. Canrinus, M. Fokkens-Bruinsma, and H. de Boer. 2019. “The Relationships Between School Belonging and Students’ Motivational, SocialEmotional, Behavioural, and Academic Outcomes in Secondary Education: A Meta-Analytic Review.” Research Papers in Education 35(6): 641-680. Moore, Robert L. 2014. “Importance of Developing Community in Distance Education Courses.” TechTrends 58(2):20-24. Reese, Gerhard, Leonie M.S. Oettler, and Laura C. Katz. 2019. “Imagining the Loss of Social and Physical Place Characteristics Reduces Place Attachment.” Journal of Environmental Psychology 65:1-6. Rovai, Alfred P. 2002. “Development of an instrument to measure classroom community.” The Internet and Higher Education 5:197-211. Royal, Mark A. and Robert J. Rossi. 1996. “Individual-Level Correlates of Sense of Community: Findings from Workplace and School.” Journal of Community Psychology 24(4):395-416. Sengstock, Mary C. 2001. “Sociological Work in Communities.” Pp. 151-168 in Handbook of Clinical Sociology, edited by Howard M. Rebach and John G. Bruhn. New York: Kluwer Academic. 64 Shea, Peter, Chun Sau Li, and Alexandra Pickett. 2006. “A study of teaching presence and student sense of learning community in fully online and webenhanced college courses.” The Internet and Higher Education 9:175-190. Tönnies, Ferdinand. 2001. Community and Civil Society. edited by Jose Harris. Cambridge: Cambridge University Press. Whitman College, 2021. “April 23, 2021 - COVID-19 and Plans for the Fall.” Accessed March 29, 2022. https://www.whitman.edu/coviddashboard/updates/april-23-2021-%E2%80%93-covid-19-and-plans-for-thefall Young, Suzanne, and Mary Alice Bruce. 2011. “Classroom Community and Student Engagement in Online Courses.” MERLOT Journal of Online Learning and Teaching 7(2):219-230. 65