Without proper support for these challenges, many students survive rather than thrive in academia (Dixon, 1973; Sumprer & Walfish, 2001). Beyond coursework and research, success in graduate programs is influenced by institutional ‘politics’, which Sumprer and Walfish (2001) define as “intangible factors and the interpersonal likes and dislikes” (p. 78). While students differ in skills, experiences, and personalities, there is limited research on how organizational politics shape their academic and social experiences in graduate school.
Graduate programs, like corporations, follow hierarchical structures that expose students to power dynamics, resource allocation challenges, and competitive ambition. Often perceived negatively, Faiz et al. (2022) suggest that organizational politics can produce both positive and negative outcomes. Understanding how political dynamics work within graduate programs is crucial for academic and career success. This study examines how graduate students and faculty perceive organizational politics and its influence on the graduate program experience.
For graduate programs, organizational politics involve institutional culture and interpersonal relationships, influencing students’ academic and career development while obtaining an advanced degree (Sumprer & Walfish, 2001). Success extends beyond academic performance, where networking and social integration are crucial. Central to organizational politics is power, defined as the “exercise of influence over others” (Ferris & Treadway, 2012, p. 3). Power is significant because students have no institutional power other than control of their interests and research involvement. Student involvement varies throughout a graduate program, with those more engaged, especially those nearing completion, receiving more encouragement from peers and faculty than their less involved counterparts or newcomers (Gardner & Barnes, 2007). Although students are responsible for seeking out opportunities, their level of involvement can shape their experience with access to mentors, resources, and professional advancement.
To continue the idea that students have limited power, Foucault’s (1984) concept of disciplinary power reinforces the hierarchical structure of faculty-student relationships. Foucault (1984) explains that disciplinary power maintains control through surveillance and normalization to shape behavior that internalizes institutional expectations through observation and evaluation. Faculty have authority and serve as advisors and instructors who exude power in the environment of the graduate program, where organizational politics can influence the student’s academic experience (Gibson, 2006). The inherent power dynamics between students and faculty could be damaging through organizational politics, where students could face positive or negative consequences from the relationships they build with faculty (Hess & Sauser, 2001).
The primary reason students stop engaging in a PhD program is due to an incompatibility between them and their advisors (Lee, 2008). From prior studies, approximately 50% of all PhD students quit before completing their degrees (Lovitts, 2001), attributed to a lack of satisfaction with the degree programs they attended (Barnes & Randall, 2012). Faculty are models for conducting research and teaching. For some, this role expands into mentorship, where they guide their mentees through the social and political dynamics of the program (Deale & Lee, 2017). However, faculty should be responsible for socializing all students into the program or profession by offering the guidance needed to navigate the associated politics and build professional networks (Deale & Lee, 2017; Fiaz et al., 2022).
The game of politics is never played alone. While faculty have significant influence, students promote political dynamics by forming close relationships with colleagues, competing for opportunities, and shaping the hierarchies within the program. Students must navigate social structures that can improve or hinder their academic progress. Fiaz et al. (2022) identify social networks as a positive outcome of organizational politics that build and maintain relationships that are beneficial to career advancement, with more access to information and resources. Mauldin et al. (2022) address the importance of cohort-based education and student-peer relationships for social integration by providing academic collaboration, professional development, and emotional support. Students can thrive by building strong peer networks. However, those who do not build these relationships early on can face exclusion from informal and formal interactions.
While forming relationships based on shared characteristics and values is natural (Mauldin et al., 2022), organizational politics can also reinforce exclusion of individuals and allow certain groups more access to advisement, mentorship, resources, and opportunities. Those who experience social isolation are more likely to be withdrawn from the program and collaboration opportunities and adopt negative perceptions of peers, faculty, and the program (Fiaz et al., 2022). Students who are able to navigate organizational politics are likely to express some influence within its power dynamics, thus increasing their involvement, which often leads to additional opportunities for academic and professional success. Students who do not integrate into these peer networks risk missing out on co-authorship, funding, and career advice. Given the significance of peer and faculty relationships, it is crucial to include the examination of student perceptions of the political dynamics within the graduate programs through the following research questions.
How do students and faculty perceive organizational politics within their current graduate programs?
What differences exist between students’ and faculty’s perceptions of organizational politics in their programs?
How do students and faculty perceive the impact of organizational politics on academic and professional opportunities in graduate programs?
This study was designed as an explanatory mixed-methods design with sequential phases of quantitative and qualitative data collection and analysis. The study’s ideal population consisted of actively enrolled graduate students and actively teaching or advising faculty members, regardless of program, at a Midwest university. Approval was given from the university’s Institutional Review Board (IRB) for the following methods outlined in Figure 1. From the institutional resource services, a request of 500 random emails of both active graduate students and faculty was approved for email invitations to participate in this study’s purpose and a link to complete a self-reporting quantitative survey.
The survey was adapted from Ferris and Kacmar’s (1992) Perceptions of Organizational Politics Scale (POPs) to align with a graduate program’s context and relevant concepts. The survey consists of three screening questions to confirm whether participants are active graduate students or faculty and how long they have been in their role. Participants needed to complete the participation form embedded in the survey before proceeding to the POPs questions, which consist of 35 items.
Figure 1
Research Method Flowchart
The breakdown of the POPs survey is six demographic questions and 29 questions related to organizational politics. The POPs survey is categorized into five content areas. The first is general political content pertaining to a graduate program’s status quo or normalized power dynamics, while the second, “go along to get along,” content covers self-reflection in those power dynamics. The third is student-student relationships, for how students interact and influence each other. The fourth is the interaction of student-faculty relationships within the graduate program. Finally, the satisfaction content covers perceptions of the program’s capacity to foster opportunities for success and professional development. The analysis will consist of comparing perceptions between graduate students and faculty through the mean scores of each category between both groups through independent sample t-tests and a regression analysis of gender and number of years in the program. From this analysis, we could have addressed the first and second research questions to inform the questions of the qualitative interviews.
The end of the survey prompts participants with an invitation to participate in follow-up interviews through a separate link to provide contact information. The separate link allows the researchers to maintain confidentiality and privacy of those who completed the survey while recruiting for qualitative interviews to understand their perceptions on the impact of organizational politics. A smaller sample pool of those who completed the survey and then participated in interviews, either in-person or online, would provide a deeper exploration of participants’ perceptions to assist in the construction of themes and narratives to address the third research question. Lastly, a triangulation of survey results and themes constructed from the interview transcripts would provide a robust understanding of the student and faculty perceptions of organizational politics in graduate programs.
Despite the efforts of a random sample group of graduate students and faculty, the survey response rate was less than 3%, and no survey was completed. The recruitment challenges could be attributed to the limited response rate associated with email recruitment or the niche topic of organizational politics. The limited response for this study resulted in no data to address the research questions adequately.
Future studies should investigate alternative recruitment methods to improve data collection. First is considering simplifying recruitment by utilizing snowball sampling or word-of-mouth, and/or expanding the participant pool to include undergraduate students and non-faculty staff. Alternatively, we could expand our sample to maintain broader generalizability by utilizing social platforms like Reddit or Twitter. Each approach could enhance the response rates and improve the study’s application across a larger population.