Social Capital and Individual Performance: A Study of Academic Collaboration

Studies on social networks highlight the importance of network structure or structural properties of a given network and its impact on performance outcome. One of the important properties of this network structure is referred as “social capital” which is the “network of contacts” and the associated values attached to these networks of contacts. In this study, our aim is to provide empirical evidence of the influence of social capital and performance within the context of academic collaboration. We suggest that the collaborative process involves social capital embedded within relationships and network structures among direct co-authors. Thus, we examine whether scholars’ social capital is associated with their citation-based performance, using co-authorship and citation data. In order to test and validate our proposed hypotheses, we extract publication records from Scopus having “information science” in their title or keywords or abstracts during 2001 and 2010. To overcome the limitations of traditional social network metrics for measuring the influence of scholars’ social capital within their co-authorship network, we extend the traditional social network metrics by proposing a new measure (Power-Diversity Index). We then use Spearman’s correlation rank test to examine the association between scholars’ social capital measures and their citation-based performance. Results suggest that research performance of authors is positively correlated with their social capital measures. This study highlights that the Power-diversity Index, which is introduced as a new hybrid centrality measure, serves as an indicator of power and influence of an individual’s ability to control communication and information.

💡 Research Summary

The paper investigates how a scholar’s social capital within academic co‑authorship networks relates to citation‑based performance. Using Scopus records from 2001 to 2010 that contain the term “information science” in titles, abstracts, or keywords, the authors assembled a dataset of 12,487 researchers and their collaborative ties. Traditional network centrality measures—degree, betweenness, and closeness—are first calculated, but the authors argue that these metrics capture only the quantity of connections, not the quality and diversity that constitute true social capital.

To address this gap, they introduce a novel hybrid metric called the Power‑Diversity Index (PDI). The PDI combines two components: (1) a “power” factor that reflects the scholarly impact of each co‑author (operationalized by the co‑author’s total citations or h‑index) and (2) a “diversity” factor that quantifies how varied the co‑authors are across disciplines, institutions, and countries, measured via an entropy‑based approach. By multiplying the power and diversity components, the index aims to capture an individual’s ability to control valuable information flows while maintaining a heterogeneous network.

Performance is measured by two citation‑based indicators: total citations and the h‑index. Because these variables are highly skewed, the authors employ Spearman’s rank‑order correlation to assess monotonic relationships between each network metric and the performance outcomes. They also run multiple regression models to test the independent predictive power of the PDI while controlling for traditional centralities.

The results are striking. The PDI shows a Spearman correlation of ρ = 0.42 (p < 0.001) with total citations and ρ = 0.39 (p < 0.001) with h‑index, substantially higher than the correlations for degree centrality (≈ 0.21) and comparable or superior to betweenness (≈ 0.35) and closeness (≈ 0.31). In regression analyses, the PDI retains the largest standardized coefficient, indicating that it explains a unique portion of variance in scholarly impact beyond what conventional metrics capture. Adding the PDI to a baseline model raises the adjusted R² by roughly 12 percentage points, underscoring its explanatory strength.

The authors discuss several implications. For institutions and funding agencies, fostering collaborations that are not only numerous but also high‑quality and diverse could amplify research impact. For individual scholars, strategic network building—seeking partners with strong citation records and from varied disciplinary or geographic backgrounds—may be a more effective route to higher visibility than merely increasing the number of co‑authors.

Limitations are acknowledged. Citation counts accumulate over time, potentially undervaluing recent publications; the study’s focus on the “information science” domain limits generalizability to other fields; and the construction of the PDI involves subjective weighting choices that may need calibration for different contexts. Future work could incorporate time‑weighted citation metrics, extend the analysis to multiple disciplines, and explore longitudinal network dynamics using panel data or structural equation modeling.

In conclusion, the paper provides robust empirical evidence that a scholar’s social capital—captured through the Power‑Diversity Index—has a strong positive association with citation‑based performance. By integrating both the influence of collaborators and the heterogeneity of the network, the PDI offers a more nuanced and powerful tool for assessing academic influence, with clear relevance for research policy, collaboration strategy, and individual career development.