E ﬀ ect of Communication Practices on Volunteer Organization Identiﬁcation and Retention

: Volunteering has taken on growing signiﬁcance as a beneﬁt to society and in initiatives to promote sustainability; it is therefore important to understand the factors driving its success. One increasingly studied variable with a positive e ﬀ ect on volunteer behavior and retention is organization identiﬁcation. The antecedents inﬂuencing the organization identiﬁcation variable, however, have not yet been explored in the volunteer literature. We address this gap by implementing a survey among volunteers at the OUR HOUSE Grief Support Center in Los Angeles and analyzing results via simple and multivariate linear regression analyses. Speciﬁcally, we investigate whether or not communication factors a ﬀ ect both organization identiﬁcation and volunteer intention to continue. We ﬁnd that speciﬁc communication factors, including a relationship with one’s supervisor, internal communication, and external social media postings signiﬁcantly increase the level of organization identiﬁcation and retention. Our ﬁndings are consistent with the theories of leader-member exchange and absorption capacity. Practitioners and nonproﬁts can improve the organizational environment of volunteers by optimizing these communication practices.


Introduction
Volunteering has become an increasingly important force in society, benefiting from both growing trends toward social consciousness and the desire for individual connection [1]. It is estimated that 25% of the United States population volunteers each year [2]. Volunteers are furthermore the backbones of nonprofit organizations, since 85% of these entities have few to no paid employees [3]. Volunteering builds people's capacity, social inclusion, and responsiveness to societal initiatives [4]. In addition, volunteers play an important role in initiatives to promote sustainability, or development which fulfils the needs of the present generation without hurting needs of future ones [5][6][7]. Given the prevalence and value of volunteering, understanding volunteer dynamics has accordingly become more critical.
One variable which has received growing attention in the literature for its relevance to a positive volunteering experience is organization identification. A recent review of the volunteer literature by Alfes et al. [8] describes organization identification as a promising route for capturing the depth of the volunteering concept. Organization identification is particularly critical for volunteer retention; the more connected volunteers feel to their organizations, the more likely they will choose to remain in their roles [9,10]. However, the factors influencing the organization identification variable have not yet been investigated in the volunteer literature. It is thus valuable to gain an understanding of the determining factors of organization identification. Our study seeks to expand on the literature by identifying specific communication factors which build volunteer organization identification and retention. By specifying these factors, we can assist nonprofits to build organization identification and retention among their volunteers.
In this study, we focus on identifying which communication practices within nonprofits most effectively build volunteer organization identification and retention. We hypothesize that the relationship with one's supervisor; internal communications such as an organization's newsletter; external communications such as social media postings on sites like Facebook, Instagram, and Twitter; mission statement; and relationships with other volunteers have a positive effect on organization identification. We also hypothesize that the combined impact of these factors positively affects retention. We test our hypotheses using an online survey at the OUR HOUSE Grief Support Center in Los Angeles.
This study contributes to the existing literature in a few ways. First, by showing a significant statistical association between organization identification and the variables of relationship with supervisor, internal communication, and external social media postings, our study adds to previous studies by identifying specific antecedents of organization identification. Second, by providing the combined effect of these variables on retention, our findings provide additional support for the theories of leader-member exchange [11] and absorptive capacity [12]. Third, our study is to the best of our knowledge the first research in the volunteer literature to use sampled data from nonprofits to evaluate the effect of discrete communication factors on organization identification. Some studies in the employee literature have investigated general communication practices between workers and organizations [13], but not the specific factors influencing organization identification.
This paper is organized as follows. We first describe the importance of organization identification among volunteers, the factors affecting organization identification, and the potential drivers of volunteer organization identification and retention. We turn to building hypotheses regarding the effect of specific communication practices on volunteer organization identification and intention to continue. This is followed by the research methodology and measures. We then present our results, followed by a discussion of their consequences for theory and practice. We end with a conclusion and ideas for future directions.

Importance of Organization Identification
Organization identification is defined as a person's identification with organizational values, mission, and culture [14]. This concept stems from the social identity theory holding that people categorize themselves by memberships, affiliation, age, gender, culture, or other self-identification [15,16]. The process of social identity allows individuals to organize their surrounding environments and relations with others. As social identity evolves, people develop a sense of oneness with elements of their environment, such as the organizations in which they volunteer or work. Their identities become intertwined with these salient groups in their lives, and they become willing to engage in behaviors which benefit others as well as themselves. Ashforth et al. [9] state that this concept of organization identification is critical in understanding why employees join and remain in organizations.
Research on the impact of organization identification has been extensive in the employee literature, with this variable shown to have a positive effect on a wide variety of factors which benefit organizations [13]. Employees who feel a strong connection and sense of oneness with their organizations, as embodied in organization identification, are more likely to be intrinsically motivated [17], engage in beneficial decision-making [18], display a customer orientation [19], and show organizational citizenship [20]. Moreover, they express less intention to leave the organization [10].
Although research in the employee literature is relevant for our discussion, volunteers do have a structurally different organizational relationship due to their unpaid status [21]. Earlier studies show that volunteers are more independent in how they work, demonstrate greater flexibility in how they interpret their roles, and often bring different motivations to the job such as being more focused on an organization's mission and social interaction with others [21]. Despite the differences between employees and volunteers, however, Kang [22] finds organization identification still to have a positive impact on volunteers. In her study of volunteers in a faith-based charity organization, organization identification is associated with engagement. She defines engagement as a psychological state of volunteers marked by affective commitment, positive affectivity, and empowerment. Among volunteers, engagement is shown to affect variables such as self-efficacy [23], satisfaction with volunteer experiences [24], psychological well-being [25], and the intention to remain in organizations [26].

Factors Affecting Organization Identification
Numerous studies in the employee literature show that different communication channels have a positive effect on organization identification among employees [13]. They reveal that the content of communication to employees enables identification because it cites the organization's goals, values, and achievements [27]. Exposure to an organization's identity via communication is key to identification among workers [28].
Atouba et al. [13] demonstrate the importance of organizational communication adequacy as a driver. This concept captures the degree to which an organization informs its employees via communication methods such as objectives, policies, job descriptions, procedures, intranet content, and leader communications. By ensuring that employees are well-informed regarding an organization's goals, values, and norms, they develop a greater understanding over time. This understanding then becomes the basis for the development of organization identification. The results indicate that providing employees with useful and relevant information leads to a sense of organization identification.
Vertical communication, or messaging which travels up and down the hierarchy within an organization, has been studied as an additional driver of organization identification among employees by Bartels et al. [29]. The authors point out that vertical communication helps employees comprehend what the organization represents and reduces uncertainty regarding the organization's stance on issues. It further allows employees to distinguish their organization from others and establish their positions within the organization. They measure vertical communication using twelve communication climate items divided into four dimensions: Participation in decision-making; information adequacy; organizational support; and the trust, confidence, and credibility of senior management. The authors conclude that organization identification depends on an appreciation of communication from upper management. A study by Karanjes et al. [30] also provides support for the effect of the vertical communication variable on organization identification.
Smidts et al. [31] investigate the effect of communication climate on employee organization identification. Communication climate refers to openness in communication, perceived participation in decision-making, and general supportiveness. The authors argue that a positive communication climate builds organization identification because it is rewarding and inviting, allowing employees to categorize themselves more easily as part of the organization. They measure communication climate using a scale testing the dimensions of trust and openness in communication, participation in decision-making, and supportiveness. Nakra [32] also finds an explicit positive relationship between satisfaction with the communication climate and organization identification among employees in India.
Despite the important studies implemented in the employee literature on the relationship between communication and organization identification, similar work has not yet been conducted in the volunteer literature. It is important to establish this relationship among volunteers within nonprofits due to the different dynamics shown within such organizations. Since volunteers are not as affected by the power and reward structures influencing paid employees [21], their reactions to communication practices can be different. Furthermore, work is needed among volunteers to test specific communication factors, e.g., internal communication such as a company newsletter and external social media postings, rather than the broad communication constructs evaluated in the employee literature. These constructs are general in nature and do not identify specific factors which nonprofits can use to improve organization identification and retention. By testing these types of actionable factors, we can potentially make our study more helpful to nonprofit practitioners.

Potential Drivers of Volunteer Organization Identification and Retention
Although there has not yet been work on the antecedents of organization identification either within the volunteer literature or on specific factors within the employee literature, studies in both have evaluated the effect of discrete antecedents on related organizational metrics [8,33]. These studies are relevant to the framing of our hypotheses. The antecedents encompass relationship with supervisor, a type of hierarchical communication; internal communication, such as an organization newsletter; external social media postings; the mission statement; and the influence of other volunteers in the organization.

Relationship with Supervisor (Hierarchical Communication) and Organization Identification
One potential antecedent shown to influence a wide variety of organizational metrics is the relationship with one's supervisor [8,33], a type of hierarchical communication. Leader-member exchange theory [11] explains the importance of this variable. It posits that a leader who forms a relationship with members of the organization based on trust, liking, and respect promotes higher member job satisfaction, organizational commitment, job performance, and job retention. In the volunteer literature, Dwyer et al. [34] find a positive association between supervisor support in the form of transformational leadership and satisfaction. This kind of leadership aims to inspire followers to move past their own interests in order to achieve more than originally deemed possible. Allen and Bartle [35] demonstrate that managers who support supervisee autonomy generate greater engagement among UK sports event volunteers. Lo Presti [36] finds that a greater level of social support, task support, appreciation, and information all positively affect organizational commitment, satisfaction, and intention to remain among volunteers in Italy. In India, Yagi [37] shows that frequent and regular supervision causes more motivated volunteers at literacy centers. These findings lead to the hypothesis for our first variable of relationship with supervisor, a type of hierarchical communication.

Hypothesis 1a (H1a).
A better relationship between volunteers and their supervisor is associated with a higher level of organization identification among volunteers.

Internal Communication and Organization Identification
A second variable which serves as an antecedent is internal communication, such as an organization newsletter. The process of internal communication is postulated as the sharing of information in order to create a sense of community in an organization [38]. It serves to develop what researchers describe as a second important relationship other than that with one's supervisor: Connection with the organization as a whole [39]. The value of internal communication can be understood in the context of the theory of absorptive capacity [12]. The theory states that absorbing new knowledge can assist individuals in an organization to achieve higher levels of performance. Internal communication transferred between individuals in an organization builds knowledge and connections. This leads to our second hypothesis: Hypothesis 1b (H1b). A better internal communication is associated with a higher level of organization identification among volunteers.

External Social Media Postings and Organization Identification
External communication channels, such as social media advanced by the organization on sites like Facebook, Instagram, and Twitter, can be a third antecedent of organization identification. The idea that external communication channels may have a positive effect on organizations is supported by Schlagwein and Hu [40]. They relate the use of external social media postings within Australian organizations to increased learning. The researchers explain the results within the theory of absorptive capacity [12], citing a beneficial effect on organizations of processes which create and use knowledge. Even though the social media postings are externally oriented, they have a beneficial impact on members of an organization due to their educational and connective value. We thus form the third hypothesis as follows: Hypothesis 1c (H1c). Better external social media postings made by an organization are associated with a higher level of organization identification among volunteers.

Mission Statement and Organization Identification
A fourth antecedent of organization identification we consider is mission statements. These statements capture an entity's "reason of being" [41] and are a key step in defining its strategy [42]. They can accordingly be a useful factor for building organization identification. Hustinx and Lammertyn [43] find a positive association between identification with the organization's mission and commitment among Flemish volunteers. Thus, we offer our fourth hypothesis as follows: Hypothesis 1d (H1d). The mission statement is associated with a higher level of organization identification among volunteers.

Relationship with Other Volunteers and Organization Identification
An additional antecedent of organization identification we study is the influence of other volunteers within the organization. The positive effect of these relationships can be explained in terms of the building of a sense of community [44], in which people have a feeling of belonging to others and the organization. Several studies using volunteers within the U.S. demonstrate the positive impact of other volunteers on satisfaction [45,46]. Peer influence is further shown to affect retention positively in a study of Flemish volunteers [47]. In research with volunteers within the U.S., Dwyer et al. [34] find that relationships with other volunteers influence satisfaction but not work contribution. Nencini et al. [48] show that these relationships have a positive effect on satisfaction but not retention among Italian volunteers. These results lead to our fifth hypothesis: Hypothesis 1e (H1e). A better relationship with other volunteers is associated with a higher level of organization identification among volunteers.

Drivers of Volunteer Organization Identification and Their Impact on Retention
We note that our hypothesized antecedents of organization identification have been shown to affect retention [36,47]. The leader-management exchange [11] and absorptive capacity [12] theories which we have used to explain the communication practices hypotheses also support their beneficial impact on retention. We accordingly test a model incorporating all of these antecedents in Hypotheses 1a, 1b, 1c, 1d, and 1e as a predictor of a volunteer's intention to continue with an organization. This leads to our next hypothesis as follows: Hypothesis 2 (H2). Volunteers' intention to continue is associated with the combined effect of connection with one's supervisor, internal communication, external social media postings, the mission statement, and connection with other volunteers.

Sampled Data-Survey
We implement an online survey among 141 volunteers at the OUR HOUSE Grief Support Center, in Los Angeles. This nonprofit is very suited for the study because it has been serving the community for 25 years and considers volunteers to be critical to its operation. The response rate among all volunteers sent the survey is 56% (79 individuals). The volunteers are trained to lead groups of adults, teens, and children with the objective of helping the participants manage recent experiences of grief.

Variable Measurement and Control Variables
We measure organization identification by adapting Mael and Ashforth's scale [10], using ratings on a five-point scale of the following three statements: (1) "When someone criticizes OUR HOUSE, it feels like a personal insult," (2) "When I talk about OUR HOUSE, I usually say 'we' rather than 'they'," and (3) "OUR HOUSE's successes are my successes." This adapted scale shows a Cronbach's Alpha of 0.74. Other variables are not tested using the Mael and Ashforth scale, but rather with questions developed together with management at the nonprofit. As these variables are very clear and specific, single item measures are judged to be appropriate for them.
The way we assess the relationship with supervisor is similar to the method of Lawthom et al. [49]. Response format ranges from strongly agree (5) to strongly disagree (1). We measure four variables-internal communication, external social media postings, the mission statement, and relationship with other volunteers-based on a similar method to Smith et al. [50]. The variables are assessed by a five-point scale, and the response format ranges from strongly agree (5) to strongly disagree (1). To measure volunteers' intention to continue, we use a method similar to Raina et al. [51]. The variable is assessed by a five-point scale, and the response format ranges from definitely will continue (5) to definitely will not continue (1). These questions were developed jointly with the management at OUR HOUSE Grief Support Center in Los Angeles.
The antecedents in H1a through H1e are tested via ratings on a five-point scale of the following statements: (1) "I feel connected to OUR HOUSE due to my relationship with my supervisor," (2) "I feel connected to OUR HOUSE due to communications I receive from OUR HOUSE such as the newsletter," (3) "I feel connected to OUR HOUSE due to the postings by OUR HOUSE on social media sites like Facebook, Instagram, and Twitter," (4) "I feel connected to OUR HOUE due to the mission," and (5) "I feel connected to OUR HOUSE due to my relationship with other volunteers." The above format is used for testing all communication factors because we, together with management at the nonprofit, judged it as an effective and consistent way of evaluating the importance of the variables. Intention to continue is evaluated via ratings on a five-point scale of the following question: "Do you intend to continue your volunteer experience at OUR HOUSE?" We also measure volunteer engagement via ratings on a five-point scale of the following statement: "I am enthusiastic about my volunteer experience at OUR HOUSE." The survey also includes open end questions on the most satisfying and dissatisfying parts of the volunteer experience. In addition, we have collected additional data, such as volunteer work and living locations, preferred social media channels, current employment status, occupation, etc. This collected information is available upon request.
In order to control other potential factors affecting both dependent and independent variables, we consider gender, age, and the number of years with the organization. Information on gender, age, and the number of years volunteered at the nonprofit are also collected via the survey.
The number of years volunteered is as follows: less than one year: 20%, one year: 13%, two years: 15%, three years: 8%, four years: 12%, five-nine years: 20%, and over ten years: 12%, reflecting a good mix of volunteer experience at the nonprofit.

Results
We test Hypotheses 1a to 1e using simple linear regressions. Panel A of Table 2 presents the results. Our findings support Hypothesis 1a with a significantly positive relationship for the relationship between a volunteer's supervisor and organization identification. Hypothesis 1b is also confirmed, with a positive relationship between internal communication and organization identification. We also find a significantly positive relationship between external social media postings and organization identification, supporting Hypothesis 1c. However, we fail to find strong support for Hypothesis 1d and Hypothesis 1e, i.e., we find only weak evidence showing the association between the mission and organization identification, and no significant evidence showing the association between the relationship with other volunteers and organization identification. As a further test of these five antecedent variables, we run simple linear regression analyses with "intention to continue" as a dependent variable (see Panel B of Table 2). This test is relevant since earlier studies report that several antecedents, e.g., social and technical support, internal communication, and feedback, affect volunteer retention [36,47]. The three variables showing significant association with organization identification as the dependent variable-relationship with supervisor, internal communication, and external social media postings-are also significant in these analyses. The two variables not showing a strong relationship in the first analyses (see Panel A of Table 2)-mission and relationship with other volunteers-are still insignificant in the analyses using "intention to continue" as the dependent variable.
To further understand the considered variables, we examine the correlations between relationship with supervisor, internal communication, external social media postings, mission statement, relationship with other volunteers, organization identification, intention to continue, engagement, gender, age, and the number of years worked at the organization. Table 3 reports both the Pearson and Spearman correlation test results. Consistent with our prediction, the simple correlations in Table 3 indicate a significantly positive association between relationship with supervisor, internal communication, and external social media postings with both organization identification and intention to continue. On the other hand, we find that the simple correlations between the mission and relationship with other volunteers with intention to continue are insignificant. However, the correlation results show that other factors such as engagement and age are associated with organization identification, intention to continue, relationship with supervisor, and internal communication. In order to control for these other potentially affecting factors, we accordingly perform our second hypothesis test using a multivariate linear regression analysis.
Since the number of years volunteered at the nonprofit could affect the degree of organization identification and retention, we implement analyses using this factor as a control variable. Prior research shows that retention is higher among people who devote more time to volunteering [49]. Multivariate linear regression analyses with years volunteered as the control variable show that the three variables of relationship with supervisor, internal communication, and external social media postings are each still significant for both organization identification (see Panel A of Table 4) and retention (see Panel B of Table 4). Variance inflation factor (VIF) analyses show no evidence of multicollinearity.
We utilize the following multivariate linear regression model in order to test the relation between "intention to continue" and communication practices: We evaluate Hypothesis 2 postulating that the combined impact of these five antecedents-relationship with supervisor, internal communication, external social media postings, mission statement, and relationship with other volunteers-affects "intention to continue" as a composite variable created using an average of the five antecedents. Since organization identification is shown to affect worker retention [9,10], this factor is added to the analysis as a control variable. Engagement is similarly included as a control variable, since this factor has been found to affect retention [26]. As years worked as a volunteer [52], age [53], and gender [54] are shown to affect the volunteer experience, these factors are also included as control variables.  Our multivariate regression results are reported in Table 4. When the multivariate linear regression analyses are run with all of these variables, the composite independent variable is not significant (see Model G in Panel B of Table 4). However, a second analysis is implemented using a composite independent variable which includes only the three antecedents which are significant in simple linear regression analyses: relationship with supervisor, internal communication, and external social media postings. In this case, the composite independent variable is significant (see Model H, Panel B of Table 4). The R 2 value in this analysis reaches a level of 0.30 (see Model H, Panel B of Table 4), far higher than the value obtained from the simple linear regressions of relationship with supervisor on intention to continue (0.10 as shown in Model D, Panel B of Table 4), internal communication on intention to continue (0.14 as shown in Model E, Panel B of Table 4), and external social media postings on intention to continue (0.11 as shown in Model F, Panel B of Table 4). VIF tests show no sign of multicollinearity. The importance of the relationship with supervisor, internal communication, and external social media postings variables is therefore further demonstrated by our finding that their combined effect has a significant positive effect on intention to continue at an organization.

Discussion and Implications
Our results advance the literature in the volunteer area by identifying specific antecedents of organization identification. Previous work in the volunteer literature has shown the importance of these variables in leading to engagement [23] and retention [26]. However, studies have not investigated discrete factors which lead to organization identification. In addition, work on the antecedents of organization identification in the employee literature has focused on general communication constructs rather than specific communication factors [13].
These findings clearly point to the importance of relationship with supervisor in building organization identification among volunteers, in line with leader-member exchange theory [11]. The theory posits that by liking, helping and respecting people who report to them, leaders build high quality relationships which have a wide variety of positive effects. The relationship between supervisors and volunteers has previously been found to influence satisfaction [34]; engagement [32]; commitment, satisfaction, and retention [36]; and motivation [37]. The relationship with supervisor seems to be a fundamental one with a major impact on volunteers.
Organization identification among volunteers is also shown to be enhanced by internal communications such as an organization's newsletter. This finding supports absorptive capacity theory, which describes the positive effects on organization members of assimilating useful information. Internal communications help volunteers learn about and feel more connected to their organizations. Our results are also consistent with work in the employee literature pointing to vertical communication as a general construct influencing organization identification [29,30].
The study also demonstrates that external social media postings on sites such as Facebook, Instagram, and Twitter influence organization identification among volunteers. While external social media postings are primarily aimed at customers, suppliers, and business partners [55], our research shows that internal members of an organization are also affected by these communications. The postings seem to build bonds between volunteers and the organizations they are helping. Schlagwein and Hu [40] have cited absorptive capacity theory to explain the positive impact of social media postings. They discuss how external broadcasting and dialogue via social media help support exploratory learning among organization members.
The value of the three variables found to be significant with organization identification-relationship with supervisor, internal communication, and external social media postings-is further demonstrated by findings that they have a significant positive effect on the intention to continue at an organization. These results add to the literature in the retention area showing that the relationship with one's supervisor affects the intention to remain at an organization [36]. They are also in line with leader-member exchange theory, which postulates that high quality supervisor-member relationships lead to lower turnover [11].
Our research does not provide support for the mission statement as an antecedent of organization identification. This result may seem surprising since the mission statement is the clearest statement of the organization's purpose [41], and a main motivation of volunteering is a desire to help others [56]. An explanation for this lack of significant results is that the mission is perhaps not internalized sufficiently by volunteers. Defining and communicating a mission is not the same as achieving its internalization among organization members. Feedback from the tested organization suggests that volunteers grasp the essence of the mission statement but could benefit from a more granular understanding of it. Marimon et al. [57] identify five dimensions of mission internalization: (1) leadership as the feeling that a member of an organization has regarding managerial engagement with the mission, (2) importance as the extent to which a member believes that the mission is important in fulfilling their societal contributions, (3) knowledge as the degree of a member's knowledge regarding the mission, (4) co-worker engagement as the belief that one's colleagues are engaged with the mission, and (5) implication as the extent to which member opinions are requested during the process of updating the mission over time. The absence of significant results shows a need for strengthening the connection with the mission among volunteers along one or more of these dimensions.
Our results also do not support the hypothesis that a relationship with other volunteers leads to organization identification. Several studies [45,46,57] show a positive impact for relationships with other volunteers, although other research produces less clear results. Dwyer et al. [34] demonstrate that this variable leads to greater satisfaction but not contribution; Nencini et al. [48] find that relationships with other volunteers affect satisfaction but not retention. The lack of significant results in our study can be explained due to lower contact with other volunteers within the tested organization. A number of open-end responses in the research survey to a question on the most dissatisfying part of the volunteer experience suggests that the degree of exposure to other volunteers is an issue: (1) "I believe there is a lack of a more cohesive volunteer community, and would like more opportunities to meet/form relationships with other volunteers through recreational events and advanced training workshops", and (2) "I would like to meet other volunteers outside of supervision". These responses are not uncommon given the unique status of volunteers versus paid employees in an organization. Since volunteers are not physically present at organizations as much as paid employees, they tend to have less contact with others working within the organization [21].
Our findings point to several steps practitioners can take in order to increase organization identification and retention among volunteers. First, managers must do everything possible to ensure a strong relationship between supervisors and volunteers. The study shows that the impact of a solid supervisor-volunteer connection goes beyond the personal relationship and affects bonds with the organization itself. If volunteers have positive feelings regarding their supervisors, they will feel better about the organization and will be more likely to continue in their roles. Second, our research demonstrates that internal communication is a critical tool which managers can use to build organization identification. The survey results suggest that an organizational newsletter is a particularly useful device because it can easily be distributed in a digital format. In fact, the nonprofit included in this study emails a newsletter to all stakeholders monthly, and our results suggest that this effort has a positive impact on a volunteer's bond with the organization as well as intention to remain. Third, we show that external social media postings are important not only for stakeholders outside an organization, but also for individuals within it. Our focal organization regularly posts information on Facebook, Instagram, and Twitter; these outreach initiatives seem to build identification and retention among volunteers. Practitioners should understand the value of these postings and encourage their volunteers to follow their organizations on social sites. Chung et al. [51] postulate a model of organizational social media which progresses from (1) experimentation and learning to (2) rapid growth to (3) formalization to (4) consolidation and integration to (5) institutional absorption. Managers can follow this model's path by better understanding the value of social media both internally and externally.
A further implication of this study for practitioners concerns the lack of significant findings on the mission statement. Although a mission statement is a critical organizational document, it is only valuable if it is truly internalized by volunteers. Managers should follow the model suggested by Marimon et al. [57] by showing volunteers that they are personally engaged in the mission, help volunteers understand how the mission is important in fulfilling their contributions to society, make sure that volunteers understand the mission thoroughly, aid volunteers in understanding how colleagues are engaged with the mission, and allow volunteers to help update the mission over time. As one of the main motivations of volunteers is to perform activities important for society [1,[5][6][7], they should be very receptive to such actions on the part of managers.
A final implication concerns the lack of significant findings between the relationship with other volunteers and organization identification. This variable should accordingly be addressed by managers via initiatives to build bonds between volunteers. Such initiatives would most likely be well received, as one of the key motivations for volunteering is social contact [21].

Limitations and Future Research Opportunities
A limitation of our study is that it is conducted with only one nonprofit organization. In the future, it would be advantageous to implement the survey using additional nonprofit organizations. It would also be particularly important to test the hypotheses in organizations devoted to sustainable initiatives, as numerous studies have evaluated volunteers in these entities [5][6][7]. Another limitation is that we did not evaluate different types of supervisor/supervisee relationships [58]. This area could be addressed in future studies. In the same vein, a limitation is that various types of internal/external communication tools such as blogs, internal forums, and mandatory reads, are not analyzed. Similarly, the merits of postings on specific social media communication channels, e.g., one's own website, Facebook, Instagram, and Twitter, are not tested for their effects on organization and worker retention. These areas could be pursued in additional research.
Another direction for future work could be to explore additional influencing variables. For example, one could examine the effect of relationships between volunteers and paid employees other than supervisors on organization identification and retention. Finally, one could evaluate the influence of organization identification and retention on the degree to which organizations enable effective role mastery among volunteers. Saksida et al. [59] show that role mastery builds commitment and can be stimulated through effective manager training; McCormick and Donohue [60] find that lower role ambiguity leads to greater affective commitment; and Slattery et al. [61] note that role clarity is related to higher organizational commitment. Since role ambiguity is one of the key problems in the volunteer experience [21], this area could be particularly promising.

Conclusions
Understanding and optimizing the experience of volunteers is critical due to their increased importance in contributing to the betterment of society. This study has expanded our insight into the factors driving organization identification and retention among volunteers. We have added depth to the volunteer literature by identifying specific antecedents which positively affect organization identification, such as relationship with supervisor, internal communication, and external social media postings. In addition, we have shown that these variables have a positive influence on volunteer intention to continue at an organization. Results provide support for leader-member exchange and absorption capacity theories.
Nonprofits can build greater connection with their volunteers and increase the likelihood of continuing by enhancing the relationship between supervisors and volunteers, advancing internal communications such as the organization newsletter, and implementing external social media postings on sites such as Facebook, Instagram, and Twitter. We suggest that organizations make greater efforts to internalize their mission statements among volunteers as well as build relationships between volunteers throughout the organization.
Appendix A   Table A1. Definition of key variables.

Supervisor
Variable indicating the impact of the relationship between a volunteer and his/her supervisor scaled between 1 and 5 (5 high, 1 low).

Int_Com
Variable indicating the impact of internal communications such as the organization newsletter scaled between 1 and 5 (5 high, 1 low).

Social_Media
Variable indicating the impact of organization posting on sites such as Facebook, Instagram, or Twitter scaled between 1 and 5 (5 high, 1 low).

Mission
Variable indicating the impact of the mission statement scaled between 1 and 5 (5 high, 1 low).

Volunteers
Variable indicating the impact of the relationship between volunteers scaled between 1 and 5 (5 high, 1 low).

Org_Identif
Variable indicating the level of organization identification scaled between 1 and 5 (5 high, 1 low).

Intention
Variable indicating the intention to continue at the organization scaled between 1 and 5 (5 high, 1 low).

Engagement
Variable indicating the level of engagement scaled between 1 and 5 (5 high, 1 low).

Gender
Variable, 1 for female and 0 for male.