Interpersonal Relationships Drive Successful Team Science: An Exemplary Case-Based Study

Scientists are increasingly charged with solving complex and large-scale societal, health, and environmental challenges. These problems require interdisciplinary teams to collaborate, coordinate, create shared terminology, and engage in complex social and intellectual processes. To successfully approach complex research questions, teams must synthesise knowledge across disciplines, create shared language, and engage a diverse research community. Despite substantial investment in collaboration, little research has examined the impact of scientific team support measures (training, facilitation, team building, performance metrics). Early work used bibliometrics to identify predictors of success (team diversity, size, proximity, inter-university and repeat collaborations). More recent studies focus on team processes, finding that effective framing of scientific problems involves emotional engagement and effective interaction, and that collaboration must consider processes, collaborators, and human capital. Network studies have linked emotional intelligence to team outcomes via team processes. Researchers have called for longitudinal, mixed-methods studies of project teams to move beyond bibliometric outcomes and explore complex, interacting features in real-world teams. There remains a paucity of research on how teams develop both contributory expertise (disciplinary knowledge required to contribute) and interactional expertise (socialised knowledge of expert group practices), which are often tacit and learned by doing. Developmental, process, and outcome evaluations are needed, ideally using mixed methods (SNA, observation, surveys, interviews), but these are not widely employed. Prior work highlights psychological safety, dependable members, clear roles, meaningful goals, and interaction qualities (even turn-taking, social sensitivity, gender balance) as keys to team success, and recommends building knowledge banks on contributory and interactional expertise. This article reports a longitudinal case-based study of an exemplary interdisciplinary scientific team successful in awards, publications, and training. It examines how scientific productivity, advice, and mentoring networks intertwine to promote success and how team processes to train scientists propel productivity and fulfill the land grant mission. Team dynamics were evaluated via social network surveys, participant observation, focus groups, interviews, and historical SNA over 15 years to develop theory and evaluate complex relationships contributing to success.

Prior literature initially relied on bibliometric indicators to predict team success (diversity, size, proximity, inter-university, repeat collaborations). Subsequent research emphasizes team processes and socio-emotional dynamics, showing that successful teams engage emotionally, develop shared language, and interact effectively. Studies identify psychological safety, dependable members, clear roles, meaningful goals, even turn-taking, social sensitivity, and balanced gender representation as associated with success. The distinction between contributory expertise (disciplinary knowledge) and interactional expertise (socialised, tacit knowledge) is crucial; both are needed to solve complex problems, yet interactional expertise is difficult to measure and under-documented. Calls in the SciTS field advocate for longitudinal, mixed-methods approaches (combining SNA, observation, surveys, interviews) to capture developmental, process, and outcome evaluations beyond archival outputs. Recommendations include creating knowledge banks to strengthen understanding of contributory and interactional expertise and documenting effective mentoring structures that relate to productivity.

Design: Longitudinal, exemplary case-based study of one interdisciplinary scientific team, selected from 25 teams monitored by a SciTS evaluation group over 5 years. The case was designated exemplary based on outcomes, interdisciplinarity, longevity, fulfillment of land grant mission elements (research/discovery; education/training; outreach/engagement), integration of members, and use of external reviewers. Participants: Team included PIs, postdocs, graduate and undergraduate students, and external collaborators. Data collection and instruments:

• Social network surveys (2015–2019): Annual roster-based surveys queried extent and type of collaboration (publications, presentations, grants, serving on student committees) and relationship types (learning, leadership, mentoring, advice, friendship, fun). 2015 captured prior collaborations; later years captured additional interactions. Response rates: 94% (2015), 83% (2016), 95% (2017), 81% (2018). Tool: Organisational Network Analysis Surveys (online). IRB protocol #19-8622H; informed consent obtained. Software: RStudio, UCINET for analysis; Visone for visualization.
• Constructed networks: (a) Scientific productivity network combined research/consulting, grants, publications, and student committee service (proxy for contributory expertise). (b) Mentoring network from "who is your mentor?" (c) Advice network from "who do you go to for advice?" (mentoring and advice as proxies for interactional expertise). Network metrics included average degree, indegree, and outdegree; stratified by role (postdocs, graduate students, faculty).
• Network comparisons: Pearson correlations between adjacency matrices for mentoring, advice, and productivity networks. Statistical significance via Quadratic Assignment Procedure (QAP), p<0.05.
• Historical social network data: Compiled from PI interviews, a PI-authored historical narrative of team formation, and team rosters spanning 81 members since inception to map how connections formed and evolved across projects.
• Retrospective team survey (2018): Sent to 22 members from the 2018 roster via Qualtrics; 86% response rate. Assessed skills developed, personal/professional support, and favorite aspects of team membership.
• Interviews: Two semi-structured one-hour interviews with two PIs (2018) on team history; recorded and transcribed.
• Participant observation (2015–2019): Observed four annual three-day off-campus retreats and 1–2 additional meetings per year, including students, PIs, external collaborators, and families. Analytic field notes documented interdisciplinary interactions, problem-solving, and cross-career engagement. Analysis: Descriptive SNA (average degree, indegree/outdegree by role), temporal comparison of network structure (2015–2018), matrix correlations with QAP for significance testing, thematic analysis of qualitative data (surveys, interviews, observations) to interpret processes supporting expertise development and productivity.

• Team outcomes and scope (2004–2018): 33 extramural awards totaling >$5.6M (including two large federal awards totaling >$4.5M); 58 peer-reviewed publications with collaborators from 39 universities, 13 state agencies, and 11 other organizations; 141 presentations; 21 graduate students and 15 postdocs trained; institutional interdisciplinary team award; one PI elected to the National Academy of Sciences. Global co-authorship on all continents except Antarctica.
• Team growth and composition: Team grew from 4 members in 2004 to 43 by 2018; 81 individuals participated over the observation period. Disciplines included ecology, wildlife biology, evolutionary biology, genetics, and veterinary medicine; PIs spanned Colorado State University, University of Wyoming, University of Minnesota, UC Davis, and University of Tasmania, with extensive agency and community collaborators.
• Onboarding and training structures: 15 students held co-advised positions, facilitating cross-lab collaboration and broader training. Fourteen members advanced roles within the team (e.g., undergraduate to PhD, PhD to postdoc); one postdoc became a PI in 2012. Students were included in leadership-level conversations, enhancing exposure to tacit team processes.
• Scientific productivity network dynamics: Average degree prior to 2016: 8.8; 2016: 6.2 (integration of new members and role reformation around a new grant); 2017 peak: 9.7 during intensive collaboration on a 5-year NSF award; 2018: postdocs increasingly central, with two overlapping the faculty core.
• Mentoring network: Team members reported an average of 2.4–3.1 mentors (outdegree) per year. By role, average mentors reported: graduate students 6.0–7.7; postdocs 2.4–3.5; faculty 2.2–4.3 (highest in 2018). Lead PI had the highest indegree, 13–14 per year (i.e., being named as mentor by 13–14 members). Average degree by year: 3.1 (2015), 2.4 (2016), 3.6 (2017), 2.5 (2018).
• Advice network: Average advice ties per member: 5.1–6.4. Faculty clustered centrally in 2015–2017; in 2018, postdocs and graduate students joined the center, indicating broader integration of junior scientists in core advice structures.
• Interrelations among networks (QAP correlations, p-values): Mentoring-to-collaboration correlation: 0.21 (p<0.004) in 2015; 0.93 (p<0.001) in 2016; 0.59 (p<0.001) in 2017; 0.35 (p<0.001) in 2018. Advice-to-collaboration correlation: 0.29 (p<0.001) in 2015; 0.72 (p<0.001) in 2016; 0.57 (p<0.001) in 2017; 0.35 (p<0.001) in 2018. These significant correlations indicate that interpersonal mentoring and advice ties are tightly intertwined with scientific collaboration.
• Qualitative evidence: External reviewers praised the team as a “dream team” with rare scope and training opportunities. Survey and interview excerpts emphasized gains in tacit skills (communication across disciplines, leadership, conflict management), mentorship, and networking, alongside tangible outcomes (publications, grants).

This longitudinal mixed-methods case demonstrates that interpersonal relationships are central drivers of scientific productivity in an exemplary interdisciplinary team. Four major implications emerge: (1) Interactional and contributory expertise are intertwined. Mentoring and advice networks created shared language, vision, and practices that fueled collaboration, supporting the view that successful teams require cognitive, emotional, and interactional platforms. (2) Team size, accumulated tacit knowledge across projects, and interdisciplinary breadth effectively and efficiently train scientists. Co-advising, cross-lab experiences, and inclusive exposure of students to leadership dialogues cultivated tacit capabilities that are typically hard to codify. (3) Interpersonal relationships increase productivity. Long-standing ties built cohesion, reliability, and reciprocity; statistically significant correlations between mentoring/advice networks and collaboration networks across years show a symbiotic relationship where social interactions structure the network and network structure reinforces productive interactions. (4) Fulfilling the land grant mission. The team integrated research/discovery, education/training, and extension/engagement, leveraging a core–periphery structure with external collaborators to span boundaries, influence policy and practice, and tailor research to societal needs. Collectively, the findings move beyond archival indicators to show how real-time team processes and networks cultivate both the tacit and explicit expertise necessary for sustained interdisciplinary success.

This study provides evidence that in a successful interdisciplinary team, interpersonal mentoring and advice relationships are tightly intertwined with scientific collaboration and productivity. Purposeful structures—co-advising, cross-lab engagement, inclusive retreats, and role progression—cultivate both contributory and interactional expertise, effectively training scientists while advancing research outcomes and community impact. Future research directions identified include: (1) systematically identifying best practices that support or inhibit team effectiveness; (2) documenting how large teams generate new knowledge, complementing evidence that small teams often disrupt; (3) developing and evaluating training concepts for graduate students and postdoctoral researchers; (4) investigating the role of graduate students as network bridges that connect disciplines and reduce clustering; and (5) improving recognition and rewards for scientists leading integration and implementation efforts.