Social media and impression management: Veterinary Medicine students' and faculty members' attitudes toward the acceptability of social media posts.

INTRODUCTION: While social media has the potential to be used to make professional and personal connections, it can also be used inappropriately, with detrimental ramifications for the individual in terms of their professional reputation and even hiring decisions. This research explored students' and faculty members' perceptions of the acceptability of various social media postings.
METHODS: This cross-sectional study was conducted in 2015. All students and faculty members at the College of Veterinary Medicine were invited to participate. The sample size included 140 students and 69 faculty members who completed the Social Media Scale (SMS), a 7-point semantic differential scale. The SMS consisted of 12 items that measured the extent to which a variety of behaviors, using social media, constituted acceptable and unacceptable behaviors. Items appearing on the SMS were an amalgamation of modified items previously presented by Coe, Weijs, Muise et al. (2012) and new items generated specifically for this study. The data were collected during the spring semester of 2015 using Qualtrics online survey software and analyzed using t-tests and ANOVA.
RESULTS: The results showed that statistically significant differences existed between the students' and faculty members' ratings of acceptable behavior, as well as gender differences and differences across class years.
CONCLUSION: These findings have implications for the development of policy and educational initiatives around professional identity management in the social sphere.

Introduction

An estimated 1.2 billion people participate in social media, with Facebook being the most common (1). While social media has the potential to be used to make professional and personal connections, it can also be used inappropriately, with detrimental ramifications for the individual in terms of their professional reputation and even hiring decisions. Beyond that, student use of social media has the potential to reflect poorly on the institution and the students’ future profession. One need not look far to see the consequences of posting pictures of oneself to Facebook. The image we convey via social media matters, even if we have high privacy settings and are discerning when accepting friend requests. Potential breaches to privacy can occur due to Facebook and other social media platforms’ interfaces. Friends can like, comment, and share information that is posted by others that then becomes visible to their friends, friends of friends, and so on.

Social Media and Veterinary Medicine

Students and early career veterinarians have Facebook profiles with varying degrees of privacy, though the majority have medium exposure/privacy (2). Of concern are the students and early career veterinarians with high exposure/low privacy profiles. This group posts information that could be potentially damaging to the individual, the institution/practice, and the profession including indications of substance abuse, breaches of confidentiality, demeaning others, and venting (2,3). These posts occur despite students’ attitudes regarding varying levels of acceptability for such behaviors. For example, a majority of veterinary students reported that it is acceptable to post comments about coursework and vent about school, but fewer than 10% of these students felt it was acceptable to comment about clinical cases, clients, or patients (4). It seems, then, that students’ attitudes toward acceptability of various postings could adversely impact them as they apply for internships, residencies, and jobs. In fact, Kogan, Hellyer, Stewart, & Dowers recommend that students be educated regarding management of their electronic professional image, since many potential employers screen applicants through an exploration of social media presence (5). One study found that 76% of employers look at Facebook to gather information about potential employees (6), despite the fact that this platform is typically viewed as a personal site geared toward self-expression and connecting with friends. Of note, consistency in personal and professional identity becomes paramount when seeking employment or career opportunities (7).

The notion of e-professionalism-the communication, practices, and behaviors comprising one’s persona as displayed on social media-is a growing focus of research among healthcare professionals (8). In particular, research examining e-professionalism among pharmacy students found that while a majority of them felt the image presented on Facebook was representative of who they are as a person, only 65% felt the image accurately represents who they will be as a professional. Moreover, 57% thought it was unfair for prospective employers to use Facebook information and almost 40% have posted something they would not want potential employers to see (9). At a basic level, these findings point to a desire to keep private and public selves separate, despite the potential for breaches of confidentiality associated with posting information on social media sites.

Computer-Mediated Professional Identity

Fieseler, Meckel, and Ranzini introduced the term online personae to describe an individual’s social media identity (10). Computer mediated communication contributes to the increasingly blurred lines/overlap between private and public or personal and professional which can problematize the online personae a professional communicates through social media representations (11). As a result, individuals might resort to boundary regulation (12) to protect privacy and prevent access to the private, personal self through managing disclosure or access (13) or maintaining distinct personal and private online personas (14). This boundary regulation allows individuals to perform their identity for a specific function (expressive, communicative, and promotional) with a specific audience (friends, colleagues, employers, hiring managers, etc.) (7,15). Where we see less demarcation between the private and professional is under conditions of strong identification with one’s professional role (10,16) such that boundaries between private and professional (work/non-work) become more permeable.

Perhaps students underestimate the extent to which a negative posting can reflect on the entire profession. These postings not only reflect negatively on the individual who posted them, rather, they have the potential to mar the image of the entire profession and adversely impact public trust (17). In other words, the representation of private lives in the public sphere can have profound professional impacts. Given this, it is surprising that individuals would choose to publically share specific aspects of their private lives. Yet, perceived anonymity results in individual’s sharing information without concern for personal or professional consequence (18).

Most people assume that the social communication characterizing Facebook, for example, is a private space (19). Increasingly, however, we see a blurring of public and private, thus contributing to broader ethical questions surrounding the use of social media in general. Who decides the criteria on which to base an online persona, and to what extent does an online persona reflect professional ability and attitudes period? (19). Previous research with medical students, for example, found that online identity conflicts were common and that their online persona did not necessarily reflect who they wanted to be (20). Moreover, students felt constrained in terms of their ability to connect with others and express themselves fully online. In some cases, students understand the importance of online impression management, particularly where future employment is concerned, but they are unwilling to modify or limit social media participation. In other words, while they want privacy, they resist censoring their online communication and instead hold fast to the belief that their personal activities should not be used as a barometer for future professional activities. In fact, some students perceive employer reviews of social media profiles to be a breach of trust since the information posted on those sites “is not intended for them” (21).

Yet, as professionals, the public affords doctors of veterinary medicine and by extension, DVM students, a certain level of respect, credibility, and trust, and in return expects certain professional behaviors. Students’ professional identity develops through socialization processes, (22) notably participation in professional education, including the curriculum and interacting with role models (23,24). Through socialization, students learn the values, norms, behaviors and skills appropriate to their role as a veterinary professional, both through navigating the curriculum and through interactions with faculty and other role models (25,26). These interactions shape students’ professional self-image in terms of who they see themselves being in the future (22,24). Thus, exposure to faculty members’ attitudes surrounding social media use, as well as instruction related to professionalism is paramount to students’ developing professional identity.

Our objective was to explore students’ and faculty members’ perceptions of the acceptability of various social media postings to guide development of educational initiatives around professional identity management in the social sphere. Specifically, we hypothesized that students would find posting of all types of content to be more acceptable than faculty members. We were also interested in exploring differences in students’ attitudes across progression through professional school and suspect that the closer students are to graduating; the less acceptable they would find posting certain information to social media. As such, we explored the following research question: what differences exist in students’ attitudes about posting content across the four years of the Veterinary Medicine curriculum?

Methods

The Social Media Scale (SMS) consisted of 12 items and measured the extent to which a variety of behaviors, using social media, constituted acceptable and unacceptable behaviors. A 7-point semantic differential scale (1=Completely Acceptable to 7=Completely Unacceptable) was used to capture respondents’ perspectives. Items appearing on the SMS were an amalgamation of modified items previously presented by Coe, Weijs, Muise et al. and new items generated specifically for this study. The university’s Institutional Review Board declared the study exempt and the participants provided informed consent.

The SMS scale was administered via a census sample to all College of Veterinary Medicine (CVM) students and faculty at a large southeastern university. A total of 376 students were surveyed across all four years of the DVM program. A total of 140 students participated in the study, yielding a response rate of 37.2%. Table 1 and 2 provide a breakdown of the students’ demographic characteristics.

Table 1

Student participants by DVM program year

DVM Year	N	%
Seniors	27	19.3
Juniors	42	30.0
Sophomores	43	30.7
Freshmen	28	20.0
Total	140	100

Table 2

Student participants by gender

Sex	N	%
Male	28	20.0
Female	112	80.0
Total	140	100

To investigate the extent to which the sample of respondents was representative of the larger student population, chi-squared goodness-of-fit tests were performed. The results indicated the sample was not significantly different from the larger student population based on program year, χ2 (3, N = 140) = 3.11, p=0.375, or by gender, χ2 (1, N = 140) = 0.01, p=0.920.

A total of 194 veterinary faculty (including part-time and adjunct faculty) were administered the survey with 69 faculty responding. This resulted in a 35.6% response rate among all faculties. It should be noted, however, that the CVM has 154 full-time faculty; thus, it is possible the response rate for full-time faculty was greater than estimated because it is unknown how many associated faculty (e.g., part-time and adjunct) periodically check their campus email. Demographic data were not collected for faculty, as the research team believed potentially identifiable information would have greatly affected faculty participation due to disproportionate department sizes.

The data were collected during the spring semester of 2015 using Qualtrics online survey software. The survey window was open for approximately one month, with non-responders receiving weekly reminders to participate. The Institutional Review Board (IRB) provided permission to conduct the study (protocol #5471).

Data analysis involved calculating reliability estimates, descriptive statistics, and inferential statistical tests to investigate group differences. The data were evidenced to be normally distributed, thus independent samples t-tests were used. We used a Bonferroni adjustment to control for family-wise error by adjusting alpha to 0.004. To complement the analysis for statistical significance, we also calculated the “practical significance” by calculating Cohen’s d effect size estimates (27). Additionally, an ANOVA was performed to determine if students’ responses differed by class year and a t-test investigated differences by gender. A Tukey post-hoc procedure was used to discern statistical differences within the model.

Results

Reliability was calculated in three ways. First, Cronbach’s alpha estimates were produced for combined faculty and student data, then for individual student and faculty samples. Cronbach’s alpha estimates for combined groups on the SMS items was 0.853, 0.852 for students, and 0.803 for faculty. Each of these values indicate high-moderate to high levels of statistical reliability (28).

Table 3 presents each of the SMS items. Table 4 provides a statistical comparison of student and faculty responses. Results show that for all but five items, there are statistically significant differences between students’ and faculty members’ ratings of acceptable behavior, thus providing partial support for our hypothesis.

Table 3

Items appearing on the Social Media Scale

No.	Item
1	Posting inflammatory comments about coursework on social media sites
2	Posting inflammatory comments about instructors on social media sites
3	Posting inflammatory comments about classmates on social media sites
4	Posting pictures of other people without their permission
5	Posting pictures of class parties
6	Posting pictures of animals, anatomy specimens or class activities that are prohibited by the standards of academic conduct
7	Posting comments about clinical cases on social media sites
8	Discussing clients on social media sites
9	Posting information about clients on social media sites
10	Posting information about client animals (including shelter or rescue-group owned) on social media sites
11	Posting pictures of client animals (including shelter or rescue-group owned) on social media sites
12	Commenting on referring veterinarians on social media sites

Table 4

Comparison of student and faculty responses on the Social Media Scale

Item Number	Faculty Mean±SD	Student Mean±SD	p	Cohen’s d
1	5.67±1.74	4.81±1.88	0.002*	0.47
2	6.08±1.59	6.25±1.32	0.423	0.11
3	6.18±1.60	6.16±1.50	0.913	0.01
4	5.86±1.56	4.46±1.76	<0.001	0.84
5	3.88±1.93	2.69±1.61	<0.001	0.67
6	6.88±0.45	6.45±1.14	0.003*	0.50
7	6.60±0.92	5.60±1.66	<0.001*	0.75
8	6.89±0.47	6.69±0.83	0.029	0.29
9	6.94±0.39	6.81±0.65	0.089	0.24
10	6.65±0.85	5.69±1.72	<0.001*	0.71
11	6.50±0.95	5.52±1.72	<0.001*	0.70
12	6.62±1.17	6.31±1.19	0.084	0.47

p-values were flagged as statistically significant when α<0.004, the adjusted p-value after a Bonferroni correction for family-wise error. (1=Completely Acceptable to 7=Completely Unacceptable).

Faculty versus students’ responses revealed seven of the twelve items yielded a statistically significantly different response (see Table 4). Effect size estimates ranged from 0.47 to 0.84, indicating the practical significance of these differences were quite large.

The statistically significant item demonstrating the smallest effect was item #1, Posting inflammatory comments about coursework on social media sites, and the statistically significant item demonstrating the largest effect size was item #4, Posting pictures of other people without their permission.

When investigating students’ responses by class year, several statistically significant differences were discernable. Item #4, Posting pictures of other people without their permission, was rated as more acceptable by 3rd year students (3.83±1.80) than 1st year (4.92±1.96), 2nd year (4.63±1.61), and 4th year students (4.67±1.54) (p=0.049). Item #7, Posting comments about clinical cases on social media sites, was rated as more acceptable by 2nd year students (4.98±1.94) than 1st year (5.65±1.41), 3rd year (6.03±1.44), and 4th year students (5.89±1.53) (p=0.025). Item #10, Posting information about client animals (including shelter or rescue-group owned) on social media sites, was rated as more acceptable by 2nd year students (5.00±1.91) than 1st year (6.27±1.25), 3rd year (5.95±1.69), and 4th year students (5.78±1.58) (p=0.013).

When investigating students’ responses by gender, several statistically significant differences were discernable. Item #2, Posting inflammatory comments about instructors on social media sites, was rated as more acceptable by males (5.58±2.10) than females (6.41±1.00) (p=0.004). Item #3, Posting inflammatory comments about classmates on social media sites, was also rated more acceptable by males (5.50±2.23) than females (6.31±1.22) (p=0.012). Item #9, Posting information about clients on social media sites, was rated more acceptable by males (6.58±1.10) than females (6.87±0.48) (p=0.039).

Discussion

This study explored faculty and students’ perceptions about the acceptability of posting various types of information to social media. Despite several statistically significant differences between faculty members and students, as well as between groups of students (class year and gender), the data revealed similar trends in terms of what content is most and least acceptable to post on social media. For example, both faculty and students found it most unacceptable to post comments and information about clients, reinforcing results by Coe, et al. (4). Both faculty and students also found it unacceptable to post content that is prohibited by standards of academic conduct, thus pointing to the importance of rules and policies to enforcing professional standards. In contrast, both faculty and students found it most acceptable to post pictures of class parties, perhaps due to the personal nature of such posts and the implicit acknowledgement of protecting an individual’s private life. In spite of these generally similar trends among faculty and students, faculty members did find posting of pictures, information about clinical cases, and inflammatory comments about coursework to be more unacceptable than students. Of note, previous research found that students view comments about coursework and venting about school to be acceptable to post on social media (4), perhaps because aspects related to school experiences are indeed a part of students’ private lives, and therefore are posted for a specific audience-friends, in a private space-Facebook.

The data also revealed interesting gender differences. Specifically, male students found it more acceptable to post inflammatory comments about instructors and classmates, as well as information about clients. These results reflect similar trends in other research (29), showing that male students often view posting of various kinds of content to be more acceptable than female students. Perhaps this relates to gender socialization and communication such that males are typically more comfortable asserting themselves and even communicating aggressively. That is, traditional masculine styles of communication privilege dominance and aggression over connection and avoidance (30). Another potential explanation for this gender difference could be that male students are less concerned about the possible ramifications of posting what could be construed as less than professional information. Or, perhaps male students resist feeling constrained in terms of their ability to express themselves fully online. That is, they might be more likely to view their social media postings as private, social communication that should not be used as a barometer of their (future) professional activities (21).

Even more curious are the differences across class year. We speculated that the closer students were to graduation, the more conscious and careful they would be in terms of constructing their computer mediated professional identity. Yet the data show that fourth year students did not report the highest degree of unacceptability for any of the items. Curiously, second year students found it more acceptable to post comments and information about clinical cases and client animals than any other group of students and third year students found it more acceptable to post pictures of other people without their permission. While we can merely speculate as to the reasons for these differences, what is most interesting about these findings is how students’ perceptions are impacted by their progression through the degree program. Posting information and photos about clients, cases, and animals is prohibited and discussed as such during new student orientation. As a result, it is possible that students who are more removed from freshman orientation might forget this information, or have developed a more nuanced opinion. Or, perhaps second year students feel more removed from the clinical experience than third or fourth year students, resulting in the view that posting information about clients, cases, and patients is less unacceptable than students who are already participating in clinical rotations. While we cannot be certain, we can use these findings to make the case for greater integration of social media training, including the ethical and professional ramifications of social media communication.

More broadly, it seems that both faculty and students recognize the importance of respecting clients and their animals, as posting this type of information on social media was generally scored as more unacceptable than postings about coursework, as well as pictures of class parties and people without their explicit permission. These findings seem to support the notion of the bifurcation of the professional and personal; that is, posting information that relates to DVMs professional activities (clinical cases, client and patient interactions) is viewed as more unacceptable than postings related to personal aspects of students’ lives. Not only do these results illustrate students’ appreciation for respecting confidentiality-an important aspect of their professional role-they also demonstrate that students have an intuitive, if not explicit understanding of the notion of front stage v. backstage performances (15). Yet, we know that the representation of private lives in the public sphere can have professional impacts, especially in the age of increasingly blurred professional and personal identities. As a result, we see two key implications arise from these results.

First, social media is inherently collaborative and interactive and thus, never truly private and personal. Private lives become public when individuals relay information about themselves and post comments, images, links, and videos. Making identity available for public consumption has problematized the relationship between employers/employees and professionals/clients. Before social media, private lives remained, for the most part, private, out of the public eye. That is, back stage encounters were unavailable to anyone outside an individual’s personal realm. This is no longer the case, even for those with the highest privacy settings. As a result, a disconnect exists between how Facebook, for example, was intended to be used and how it is actually used by some to make professional judgments. This is especially concerning for DVM students who are somewhere in the liminal space between embodying their student role and their professional role.

Second, with the potential for private information to become available for public consumption, it is important to educate students on the importance of impression management and provide training in effective boundary regulation (12). That is, students need to understand how questionable or negative postings can adversely impact them, their institution, and the entire profession. Beyond that, training must be provided on how to protect privacy and prevent access through one of two primary ways: minimizing disclosure or maintaining a highly protected private persona as well as a professional persona. By maintaining a demarcated personal and professional identity, students would have some freedom in communication with particular audiences for specific purposes. While breaches of confidentiality are always a possibility, this form of boundary regulation affords students a degree of freedom in expressiveness and control over their online impression management.

Limitations of this study include the use of a sample from only one institution, though previous research leads us to suspect that similar trends would be reported at other institutions. Additionally, we had a limited number of faculty respondents and have no way to discern if the faculty who did respond are clinical or tenure-track, full or part-time, or adjunct. A more representative faculty sample might yield different results. Finally, this research explored faculty and students’ self-report of social media posting acceptability. Future research should objectively explore students’ postings on various social media platforms to reveal more detailed information about how they construct their electronic persona.

Conclusion

This study provided insight into DVM students’ and faculty members’ perceptions of acceptable social media behavior. While faculty members and students appear to be in alignment in terms of the unacceptability of posting client information on social media, faculty members find it more unacceptable than students to post information about clinical cases, pictures of any kind, and inflammatory comments about coursework. The data also illuminated interesting gender differences in that male students found it more acceptable than females to post inflammatory comments about instructors and classmates, as well as information about clients. Perhaps most perplexing are the differences across class year. Second year students found it more acceptable to post comments and information about clinical cases and client animals than any other group of students, while third year students found it more acceptable to post pictures of other people without their permission. Taken collectively, these results illustrate differences between groups-faculty and students, males and females, and first through fourth year students. Opportunities exist to further educate students on the importance of actively and purposefully constructing an electronic persona, particularly where private and public personas intersect. By providing training and reinforcing the importance of online identity management throughout the curriculum and extracurricular activities, as well as faculty mentorship, we can better prepare students to manage their electronic presence and showcase their professional identity appropriately.