A Citation Network Analysis of Social Work Scholarship: How the Grand Challenges for Social Work Aid Our Understanding of Interrelated Social Problems

Background and purpose: The American Academy of Social Work & Social Welfare’s (AASWSW) Grand Challenges for Social Work (GCSW) framework organizes the major goals of social work (SW) practice and research. However, the social problems (i.e., poor health, income inequality, and domestic violence) addressed by the GCSW are interconnected, such that addressing one challenge in isolation is unlikely to create lasting change. Using the GCSW framework, we examine how SW research is interrelated by using 1) article-level and 2) GCSW-level citation network analysis (CNA) to illustrate research clusters.

Methods: Using Google Scholar’s H-5-index and Hodge et al.’s (2020) quality and prestige journal rankings, we selected eight prominent SW journals. We collected all peer-reviewed articles published in these eight journals in 2022 (n=1,211), conducting forward and backward reference searches on each article to develop a citation network. Using supervised machine learning (SML), articles were classified into 19 categories: a GCSW or one of five additional categories developed by the authors. Citations that did not fit into one of these categories or citations with low prediction confidence (i.e., <.40) were removed (n=8,671; 17.21%). This left 41,715 nodes and 46,814 edges in the complete article-level directed network to answer the following research questions: what GCSW categories are most cohesive (i.e., form distinct clusters) and what categories bridge gaps between clusters? We also created a weighted GCSW-level network, composed of our 19 categories/nodes and 296 edges. Using this network, we aimed to answer: what categories are most closely related? CNA and visualizations were created in Gephi using modularity class and centrality metrics (e.g., betweenness, degree).

Results: In the article-level network, three dominant clusters emerged: ensure healthy development for youth (26.83% of nodes), build healthy relationships to end violence (17.56%), and SW practice (4.97%), with a smaller close the health gap (17.82%) cluster nested within youth development. Using betweenness centrality, some of the largest nodes were methodological papers (7.57%), indicating that research design and methods cross-cut research clusters. To investigate other clusters, we removed the edges that connected to 2022 articles in the three largest categories. Articles on SW practice and economic (in)equality formed the two largest clusters, with smaller clusters on violence prevention and homelessness emerging. Articles on (in)equality and social justice functioned as boundary spanners between clusters.

In the weighted GCSW network, we detected 5 research community groups, grouped by modularity class: (a) technology and isolation, (b) violence prevention, (c) poverty and inequality, (d) wellbeing of vulnerable groups, and (e) SW education, practice, and advocacy.

Conclusions and Implications: Our findings demonstrate how SML and CNA can reveal structural patterns in SW scholarship. The pervasiveness of research on (in)equality and social justice throughout the network suggests that these are threads throughout SW research. Additionally, we found that most articles tend to cite articles within the same category, which inhibits innovation, as bringing together diverse research areas often yields more profound advancements. Given our findings, future SW research may benefit from scholarship that integrates multiple or novel combinations of Grand Challenges to advance SW science.