Introduction

Academic advancement has been shown to progress unequally for men and women in science, technology, engineering, math, and medicine as well as in veterinary medicine—a profession that has been predominantly female in the US since 2009.^1–5 Several hypotheses have been described to help explain these discrepancies. One proposed mechanism by Carnes et al⁶ anchors on the concept of bias, suggesting that gender-linked assumptions persist throughout organizations, fundamentally devaluing work performed by women and functionally constraining promotions. Alternatively, the “Matilda effect” proposed by Rossiter⁷ posits that differences in resource availability between men and women faculty create an unequal playing field, disadvantaging women’s career advancement. Carnes highlights the importance of research and Principal Investigator status as a pathway to leadership and influence, pointing to established evidence for gender differences throughout these areas.^8,9 Considered together, these hypotheses generate a singular argument that academic research, and all the levers that affect it, may be an important component of the overall advancement gap.

Scholarly productivity is a powerful driver for advancement to associate and full professorship and toward leadership roles. This holds true for many nontenured clinical faculty—positions common in veterinary academia—as publication is encouraged and often required for promotional dossiers and for establishing a national reputation. Across various science, technology, engineering, math, and medicine fields, publication rates and publication impact (as measured by the h-index) of women faculty are lower compared with male faculty.¹⁰ Giuffrida et al¹¹ showed that fewer senior authors on veterinary surgical publications are women, and Wang reports that women are comparatively underrepresented on journal editorial review boards.¹² The h-index, a continuous measure of research productivity, is positively associated with increasing academic rank among surgical faculty in veterinary medicine.³ Given the relative predominance of men in associate and full professor roles,^1–3,13 it follows that there is a greater overall volume of published and cited literature from senior male faculty. Individual success in research not only propels personal career development but is an indirect form of currency that can be used to influence the success and progress of others.

Academic research is increasingly a collaborative activity that engages the sharing of ideas, influence, experience, and resources to increase the quality and quantity of output. To a junior faculty member, collaboration with an established researcher can be one of the most important contributions to the establishment of their research program and pace of their career progression. Collaboration may provide access to resources including mentorship, financial support, access to equipment or personnel, and an increase in the credibility of an unknown investigator on grants and manuscripts. Despite its importance and in contrast to more structured processes to resource new scholars, collaborations occur largely through informal networking and social linkages. While networking is critical for career success,¹⁴ studies show that gender differences exist in the quality and size of networks, creating less powerful or effective grids for women than for men.^15–17 Gaining a better understanding of how these networks manifest differentially for men and women in veterinary medicine through research relationships may provide useful insight to understanding how the advancement gap develops.

The objective of this study was to examine patterns of collaborative research behavior among academic faculty in veterinary medical colleges to determine whether differences exist in how men and women, particularly in their early career, access resources that may ultimately contribute to their career progression. We utilized data collected from a national cross-sectional survey sample of the experiences of academic faculty in accredited veterinary colleges to compare academic rank and experiences in research collaborations between male and female faculty. We hypothesized that gender dependencies in research partnerships would exist and coincide with a gender-based advancement gap.

Materials and Methods

Results

Collaborative research behaviors

To assess collaborative behaviors, junior faculty were defined as unpromoted faculty, either at the assistant or instructor rank. Senior faculty were defined as promoted faculty, including associate and full professors. No differentiation was made by tenure versus nontenure status. Respondents were asked whether, during their time as a junior or unpromoted faculty, they received an offer from a senior faculty to collaborate on a project. Data were first analyzed using all respondents both currently and previously holding roles as junior faculty. There was no difference by gender in the proportion of men (68.7% [184/268]) or women (64.5% [269/411]) who had been invited to collaborate (P = .4). Men were asked to collaborate more frequently (median, 3 times; IQR, 2 to 6) compared with women (median, 3 times; IQR, 2 to 4; P = .0006). Gender discrepancies were observed in the patterns of collaboration among junior faculty, in which male respondents were asked more frequently by male senior faculty (median, 2 times; IQR, 1 to 3) compared with women (median, 1 time; IQR, 1 to 2; P < .0001). Overall, junior faculty were asked more frequently to collaborate by senior male faculty than senior female faculty (P < .0001). There was no difference in the number of times male and female junior faculty were asked to collaborate by senior female faculty (median of 0 times [IQR, 0 to 2] vs median of 1 time [IQR, 1 to 1.25]; P = .34).

Analyses were repeated among only those currently holding junior faculty positions. As before, there was no difference in the incidence of being asked to collaborate (63.8% [125/196] of women and 66.2% [51/77] of men; P = .81). Women were asked to collaborate less frequently overall than men (median of 2 times [IQR, 2 to 3] vs 3 times [IQR, 2 to 5]; P = .013). All previously identified patterns of gender dependencies remained significant at a level less than P ≤ .003. As before, there were no gender differences in how junior male and female faculty received requests for collaboration from female senior faculty (P = .38).

Respondents were asked whether, as a junior faculty member, they had asked for collaborative assistance and/or sharing of resources from a senior faculty member on a project they were undertaking. Responses were first analyzed among those both currently and previously holding junior faculty roles. A majority of both male and female respondents indicated that they had asked for collaborative assistance and/or sharing of resources from a senior faculty member, and this proportion was not significantly different (73.3% [193/263] of men vs 73.4% [298/406] of women; P = 1.0). Male junior faculty indicated that they had requested collaborations a greater number of times than female junior faculty (median of 3 times [IQR, 2 to 5] vs 2 times [IQR, 2 to 4], respectively; P = .0003). A greater number of requests were made to male senior faculty overall (median, 2 requests; IQR, 1 to 3) as compared with female senior faculty (median, 1; IQR, 0 to 2; P < .0001). Requests to male senior faculty were more frequently from male junior faculty than female (median of 2 times [IQR, 1 to 3] vs 1 time [IQR, 1 to 2]; P < .0001). There was no difference in the number of requests made to female senior faculty by men or women (median of 0 times [IQR, 0 to 2] vs 1 time [IQR, 0 to 2]; P = .38).

Data were again reanalyzed to assess responses only from current junior faculty. There was no difference in the proportions of current male and female junior faculty who requested collaborative assistance from senior faculty members (71% [54/76] of men and 70.3% [137/195] of women; P = 1]. Women asked less frequently than men (median of 2 times [IQR, 1 to 3.35] vs 3 times [IQR, 2 to 5]), which approached a level of significance (P = .073). All previously identified patterns of gender dependencies remained significant at a level less than P ≤ .02. As before, there were no differences in requests made to female senior faculty by male or female junior faculty (P = .54).

Respondents currently classified as senior faculty were asked whether they had asked an early-career (unpromoted) faculty member to collaborate on a project. Among senior faculty, there was no difference by gender in the rate of senior faculty offering collaborative assistance to junior faculty (79.5% [136/171] vs 79.4% [139/175]; P = 1.0). Male senior faculty reported offering assistance more times than female senior faculty overall (median of 4 times [IQR, 3 to 6] vs 3 times [IQR, 2 to 5]; P = .002). Senior faculty overall reported asking junior female faculty more times (median, 2 times; IQR, 1 to 3) than junior male faculty (median, 1 time; IQR, 0 to 3; P < .0001). There was no difference by gender among senior faculty in how often female junior faculty (P = .33) were asked to collaborate. However, male senior faculty reported asking junior male faculty to collaborate more times (median, 2 times; IQR, 0 to 3) than was reported by female senior faculty (median, 1; IQR, 0 to 2; P = .006). A schematic summarizing the collaborative relationships described above is provided (Figure 1).

View Full Size

Schematic depicting the collaborative relationships reported between male and female junior and senior veterinary faculty respondents in an online survey conducted between February 20 and March 26, 2018, to gather information regarding patterns of collaborative research behavior by gender among academic faculty in veterinary medical colleges. Arrows point in the direction of the request to collaborate; arrow size is associated with collaboration frequency. Numbers at the base of the arrow represent median number of requests made as reported by respondents currently and previously holding junior faculty roles (first or only number) and by respondents currently holding senior faculty roles (second number, italicized).

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.23.02.0111

• Download Figure
• Download figure as PowerPoint slide

Objective and subjective experiences with resources

Respondents were asked if they received a startup offer at the time of hire into an assistant professorship role; there was no difference by gender in a positive response (64.5% [165/256] of men vs 67.4% [254/377] of women; P = .5). There was also no difference in whether the startup packages were deemed sufficient to conduct a research project, either in full or to generate pilot data (65% [105/161] of men vs 70.1% [175/249] of women; P = .33). Men and women reported equivalent values for their startup packages (median of midpoint average, $85,747 [IQR, $12,500 to $150,000] for men vs $80,627 [IQR, $12,500 to $150,000] for women; P = .91).

Finally, respondents were asked about their perceptions of access to ongoing resources or opportunities to succeed based on their personal identities. There was a marginally significant difference among male and female respondents in these perceptions (P = .048), with a greater proportion of women than men responding “Yes” that identity had affected access (36.4% [149/431] vs 26.2% [73/279]) and a smaller proportion responding “No” (51.7% [223/431] vs 60.2% [168/279]; Figure 2). Among those responding “Yes,” there was clear difference in what specific effect men and women felt gender had on access (P < .0001), with 87.6% (127/145) of women experiencing a disadvantage or significant disadvantage compared with 21.9% of men (16/73; Figure 3). Conversely, a high proportion of men felt their gender had provided them with some level of advantage during their career (46.6% [34/73] vs 4.1% [6/145] of women).

View Full Size

Percentage column chart representing how women (n = 431) and men (279) responded to the question, “Do you feel that your race or ethnicity, age, ability status, nationality, or gender has affected your access to resources and opportunities?”

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.23.02.0111

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Percentage column chart representing how women (n = 145) and men (73) who responded “Yes” to the question, “Do you feel that your race or ethnicity, age, ability status, nationality, or gender has affected your access to resources and opportunities?” rated the relative effects of their own gender.

Citation: Journal of the American Veterinary Medical Association 261, 8; 10.2460/javma.23.02.0111

• Download Figure
• Download figure as PowerPoint slide

Discussion

This was the first study, to the authors’ knowledge, evaluating gender dependencies in patterns and frequency of research collaboration among veterinary college faculty. Though high rates of collaboration were reported among both men and women, collaborative outreach was more frequent for men at both the junior and senior faculty levels. Gender concordance appeared to benefit men at the junior faculty level, resulting in greater numbers of collaborative opportunities offered and requested. Conversely, gender concordance was not advantageous for women, who received and requested more opportunities from senior male faculty. This trend appears unrelated to historic demographic imbalances that were previously weighted more heavily toward men, as the patterns generally persist in the current environment where demographics are more balanced in senior ranks and even skew female at the junior faculty levels. Male and female respondents were both provided with startup resources to begin academic research programs by their hiring institutions, yet there was a stark difference in how each perceived gender equity in their ability to access resources or opportunities during the course of their academic careers. Within the same population, given the same number of years of experience, a male faculty was almost 50% more likely to be at a higher stage of academic rank compared with a female faculty (based on promotion and track). In other words, gender was associated with a higher academic rank, although neither gender in and of itself nor the reported collaborative experiences should be interpreted as causal for greater academic advancement.

The data used in this cross-sectional study were derived from a large population of faculty at accredited veterinary medical colleges across North America and the Caribbean. A true response rate could not be evaluated given the lack of available data for the total number of faculty receiving invitations to participate. The proportion of female respondents in the current study was higher than the proportion of female faculty reported in the 2018 Comparative Data report from the American Association of Veterinary Medical Colleges (61% vs 48%),²⁰ indicating the possibility for a gendered response bias. The proportion of respondents in nontenure track roles identifying as female (68% total) is similar to the proportion reported by the American Association of Veterinary Medical Colleges in clinical track positions in 2018 (62%),²¹ suggesting a comparable distribution of faculty members across different faculty tracks. No public data were available on the distribution of faculty at various promotional ranks either at the institutional or national level for comparison. Indeed, the information provided here may give a broad snapshot of the state of academic promotion in veterinary medical colleges.

Research collaborations can be complex, including a variable number of individuals and organizations. Collaboration, when analyzed in the literature, is often defined as coauthorship on a published manuscript.^22–24 Here, publications were not measured as an outcome of shared effort. Instead, collaboration was assessed as the incidence of dyadic relationships formed within the context of research. These data were therefore indirectly representative of personal networks in a research environment than outputs. Networks are increasingly being evaluated as influential factors in organizational structure and output, often using formalized techniques in network analysis (NA). Research in NA has included methods such as mapping organizational structures, employee email communications, teams or project groups, and observation of interpersonal behaviors in the workplace.^24,25 Dimensions of NA have been shown to correlate with research productivity and collaborations among academic faculty.^24,26,27 Network characteristics, such as levels of status and power, also vary between men and women within organizations, ultimately providing greater benefits to men.^17,28 The data presented here provide preliminary evidence that formal and informal research relationships may be an important lens through which to understand gender-based differences in faculty experiences in veterinary academia and that network analysis may be an innovative platform for future work.

Collaborations were retrospectively reported by faculty members, with faculty of all ranks invited to recall experiences as junior faculty. Data were first analyzed in this manner to understand the collaborative environment in which all faculty considered in this study had experienced early career and how early collaborations could influence career trajectory. It is certain that changing faculty demographics could account for variation in academic environments for junior faculty today versus those beginning their careers 20 years ago. Historical overrepresentation of men in academia previously might have created a scenario that would favor collaborative relationships forming among men. However, repeat analysis among current junior faculty only, a subset predominantly populated by women, revealed no differences in the gender-specific patterns of collaborative behaviors. Gender demographics within veterinary academia have been shifting for many years due to increases in women’s enrollment in veterinary colleges and biological training programs. In our data set and others,^1–3 even at the associate level men no longer outnumber women in veterinary academia. As such, our findings are likely to be less associated with the numbers or proportions of available men and women at various academic levels but rather may be influenced by other factors associated with gender.

Two important themes that emerged from our results were that (1) gender homophily produces more frequent collaborative efforts among men than among women and (2) senior male faculty serve as the strongest central node for collaborative efforts. The homophily effect has been shown in research collaborations in which, irrespective of the successes of individual authors in attracting collaborators, researchers tend to collaborate with those of the same gender.^25,29 Here, among junior faculty, women participated equally in gender homophilic and heterophilic collaborations, whereas minimal heterophilic collaboration was recalled by men. Some researchers have shown that women engage collaboratively more frequently than men, though exceptions to this pattern occur with international collaborations or collaborations with industry.^30,31 In general, researchers tend to collaborate repeatedly with the same coauthors, and the larger the coauthor network the more collaborations that are likely to occur.^22,32 These tendencies ultimately favor the larger networks created by men and may potentiate the influence of highly productive or well-resourced individuals.

In this study, senior female faculty participated in the fewest research collaborations, despite many women respondents in associate and full professorship roles. Research productivity was not measured; as collaboration has sometimes been associated with productivity, it is possible similar outcomes could be extrapolated from these findings.^27,33 It is beyond the scope of this paper to investigate the specific factors that may have influenced individuals to either perform research or seek out collaborations. However, consideration of either or both actions may be a basis for understanding the patterns observed in our data. Multiple possibilities exist for why fewer collaborations were observed between senior female faculty and junior colleagues compared with male senior faculty. Areas to consider or investigate further may include to what extent male or female faculty hold positions with limited or no research requirements, reduce research activity following promotion due to competing administrative responsibilities, have greater independency in research pursuits, receive fewer grants, pursue collaboration with nonfaculty (such as students or postgraduate trainees), have smaller collaborative networks, or to consider intrinsic motivation factors such as confidence, reward, or purpose, among others. Among women responding to this study holding associate or full professor roles, a greater number were in tenured than nontenured roles and therefore unlikely to be exempt from research. Interestingly, the subjective experience of women in this study, regarding their ability to access resources and opportunities, pointed toward experiential barriers in their careers that may have limited their research productivity or collaborative process. Further investigation into the needs and experiences of women researchers will be critical to supporting the most rapidly growing demographic in veterinary academia. And, as women increase in number and proportion in academia and in tenure-track roles, concentrating on the power of their networks may be an important lever to ensure that their academic success does not lag further.

Similar to previous reports,^1–3,13 gender was associated with academic rank, in line with observations from medicine, basic sciences, mathematics, and business.^5,34–37 Discussions about lagging career progression in women relative to men have evaluated bias, productivity, commitment, family, negotiation, and mentorship among other factors to consider areas for intervention.^38,39 Productivity differences are a possible driver of the advancement gap due to their clear implications for recognition and promotion.^40–42 We show here that dyadic research collaborations were consistently reported most frequently between men than between women or across genders. While not all collaborations lead to a successful publication, those with established senior faculty are likely to facilitate productivity. Additionally, the relationships built by such collaborations may also provide access to mentorship and sponsorship. Collaboration in this context may therefore contribute to the structural (resources) and cultural (relationships) environments in which individual academic careers are cultivated. While associations identified here are casual only, given the consistency of these patterns in a sample in which a gender-based advancement gap is detected, it is reasonable to consider that increased access to senior researchers can offer a professional advantage.

To better understand the experiences of faculty in accessing resources valuable for a successful academic career, we evaluated both objective and subjective domains of resource allocation. Objectively, men and women appeared to start from a level playing field, as equal proportions reported receiving startup offers of equivalent size, sufficient to begin research programs. This is an important finding given institutional initiatives aimed at reducing gender inequities surrounding pay and funding and the lack of transparency typically surrounding startup packages in academia due to the sensitive nature of negotiation and funding processes. However, subjectively women and men respondents felt disadvantaged and advantaged, respectively, in their ability to access resources, at significantly different rates. Experiences such as these are difficult to quantify, and harder still to substantiate, yet the repetition of these perceptions across many institutions suggests some common themes throughout academic environments. Women’s contributions to research are less-often acknowledged both systemically and through authorship as compared with men also participating in collaborative efforts.⁴³ Additionally, papers authored by women are less frequently cited than those authored by men,⁴⁴ a phenomenon which may be gender driven in itself.⁴⁵ Mentorship is a complex phenomenon to document and understand, yet differences have been shown in the resources and prestige among those who mentor women versus men,⁴⁶ indirectly affecting the access of the mentee. Consistent with our findings regarding gender disparities in research collaborations, access to interpersonal networks and collaborative patterns have been observed elsewhere in academia.^32,47 These findings, taken together, produce elements of a possible framework for understanding how women researchers experience a different environment of support during their academic journey.

There are many bright spots in the data presented here. First, the newest generations of veterinary college academics were predominantly female, creating a strong pipeline for advancement. While the proportion of junior faculty entering tenure-track roles was smaller than historically among full professors, women appeared evenly split among the 2 options in the early ranks, similar to other reports.^1,2 As such, the productivity and influence of women researchers is likely only to grow as young academics continue to build their dossiers. In line with this, we found that senior faculty as a whole reported offering more collaborative opportunities to junior female faculty as compared with junior male faculty. If reliable, the process of women growing their influence and centrality in the field of veterinary science is already underway.

Finally, based on the evidence collected here, collaboration is widespread, frequent, and continues throughout all stages of an academic career. To solidify this trend and promote opportunities for all faculty regardless of gender, race, or track, the authors suggest more structured institutional development programs to help connect early career faculty with more senior faculty representing a wide array of interests and disciplines to encourage collaboration and resource sharing. While the novelty of collaborative relationships was not examined, incentivizing network variation within and even across institutions will benefit individual faculty, schools, and scientific progress.

These data were collected in 2018 and therefore do not account for more recent demographic shifts or new veterinary medical colleges that have since been established. The retrospective and self-reported nature of the collaborative event data created the possibility for recall bias. The track of appointment held by each respondent was not evaluated due to the variability in research requirements for different tenure and nontenure roles across institutions and in requirements for promotion with tenure over time within certain institutions. The collaborative relationships described are also not limited to the sample (respondent) pool from which these data were generated. The differences in the number of requests from senior faculty described by junior faculty may not (and here, frequently did not; Figure 1) match the number of offers reported by senior faculty, on account of parties external to this sample population affecting the results. It was also possible that recall bias was driven by the length of time involved in remembering or by a social desirability bias in which respondents recalled generosity more frequently than it occurred.

The topic of demographics and identity in academia, science, and veterinary medicine continues to be relevant as organizations evolve in their thinking and commitment to diversity, equity, and inclusion. Even in such a large sample of veterinary college faculty, the number of people identifying a race or ethnicity other than white was too small to provide meaningful statistical analysis. However, observing the experiences of a single identity group (female) highlights the importance to organizations of understanding the experiences of all groups, even if those investigations cannot be performed with statistically powerful sampling. It is clear from these data that personal networks are powerful, and gaining deeper insight into how relationships are formed and utilized will be essential to support diversity through inclusion and equity and in general create environments where all have the resources to thrive. Cross-disciplinary collaboration to expand our knowledge about professional networks will be a valuable addition to both the science and management of our organizations.

Supplementary Materials

Supplementary materials are posted online at the journal website: avmajournals.avma.org