Using network analyses to examine the extent to which and in what ways psychology is multidisciplinary

The study applies a science-of-science perspective to quantify how multidisciplinary Psychology is and how that extent has changed from 2008 to 2018. Motivated by calls for increased multidisciplinary research to address complex societal problems (e.g., climate change) and recognizing Psychology’s history of integrating concepts from Economics, Biology, and Physics, the authors use network science to provide quantitative baselines. They focus on journals categorized by Web of Science as "Psychology, Multidisciplinary" and analyze citation patterns and research themes to assess: (1) the proportion of citations within Psychology vs. outside disciplines; (2) longitudinal changes in multidisciplinary engagement; and (3) thematic areas driving multidisciplinary connections. The work aims to inform evidence-based decisions by researchers and administrators and illustrates methods applicable to other Humanities and Social Sciences.

Prior work demonstrates the utility of network analyses in studying scientific communication and structure, including Price (1965) on citation distributions, Xhignesse and Osgood (1967) on psychology’s journal network, and numerous applications of community detection and co-word analysis to map research areas (e.g., Ding et al., 2001; Newman, 2006; Blondel et al., 2008). Growth trends in science publishing (Fortunato et al., 2018) and journal proliferation (Mabe, 2003) provide context for interpreting changes in article and journal counts. Additional literature highlights the balance between consolidating knowledge and bridging disciplines (Foster et al., 2015) and the increasing use of network science across Psychology and the social sciences (Borsboom & Cramer, 2013; Vitevitch, 2019).

Design: Two network analyses were conducted covering 2008–2018. Data source: Web of Science (Social Sciences Citation Index), restricted to journals with the subject category "Psychology, Multidisciplinary." Citation networks: For each year (2008–2018), a directed citation network was constructed where nodes represent journals. An outward link from a node X (a Multidisciplinary Psychology journal) to node Y was added if an article in X cited an article in Y. Cited journals Y were categorized as: (a) Multidisciplinary Psychology (MultiDisc Psych), (b) other Psychology subfields (Other Psych), or (c) other disciplines (Other Disc). Only journals in the SSCI were analyzed; articles from journals not registered in SSCI were excluded. Self-citations at the author level and authors’ geographic locations were not distinguished, as journals served as proxies for disciplines. Multidisciplinary Psychology journals could have both outgoing and incoming links; non-Multidisciplinary journals only had incoming links from the focal set. Co-word networks: To explore thematic content within modules detected in the citation networks, co-word networks were built from author-provided keywords in articles belonging to each module. Nodes are keywords; undirected, weighted edges connect keyword pairs co-occurring within the same article’s keyword list, with weights equal to co-occurrence frequency. To focus on salient themes and account for size variation, a degree threshold was applied to display only the top 5% (or fewer) of nodes and links in each co-word network. Community detection and software: The Louvain method (as implemented in Gephi 0.9.2) was used to detect modules in the 2008 and 2018 citation networks, and modularity Q was recorded to assess community structure. Gephi was used for all network analyses and visualization.

- Publication growth: Articles in Multidisciplinary Psychology journals approximately doubled from 4,441 (2008) to 9,496 (2018). This growth mirrors the broad exponential increase in scientific publishing (doubling ~15 years), suggesting Multidisciplinary Psychology kept pace with overall science rather than outpacing it. - Journal counts: Multidisciplinary Psychology journals increased from 107 (2008) to 132 (2018) (~25% over the decade). However, overall journals in the networks increased from 1,776 to 2,657, averaging ~4% annual growth (consistent with Mabe, 2003). The Multidisciplinary Psychology category grew at a mean ~0.6% per year, indicating only modest expansion relative to general journal growth. - Citation composition (stable across years): Approximately 25% of citations were to Multidisciplinary Psychology journals, ~50% to other Psychology journals, and ~25% to journals in other disciplines. This 25:50:25 distribution remained fairly constant from 2008 to 2018, indicating a stable extent of multidisciplinary engagement. - Community structure: • 2008 citation network: Seven modules detected; modularity Q = 0.32 (significant community structure). • 2018 citation network: Seven modules detected; modularity Q = 0.20 (community structure present but less pronounced). - Thematic patterns from co-word networks: • 2008 modules featured themes including: computers/communication technologies; mental and physical health (e.g., depression, oncology); relationships/relational-cultural theory; socio-cultural/critical discourse analysis; body image/sexuality; learning and memory (e.g., hippocampus); forensic psychology (e.g., violence, recidivism). • 2018 modules featured themes including: broad core psychology topics (e.g., anxiety, motivation, personality); addiction (e.g., gambling, alcohol, impulsivity); domestic violence and victimization; happiness/well-being/life satisfaction; online communication and social media; mental and physical health; identity and discrimination (e.g., stigma, integration). - Overall interpretation: The stable 75% Psychology vs. 25% other disciplines citation pattern suggests work remains identifiable as Psychology while engaging with other fields. Topic areas evolve over time, with some persistent (e.g., mental-physical health) and others more transient.

Findings indicate that the extent of multidisciplinary engagement in Multidisciplinary Psychology remained stable from 2008–2018, as reflected by a consistent ~75:25 split between citations to Psychology (MultiDisc + other Psych) and other disciplines. This may reflect the need for research to remain identifiable within Psychology for publication and review, a balance between consolidating within-discipline knowledge and bridging to new areas, or the mainstreaming over time of initially external methods and ideas. The longitudinal module and co-word analyses reveal how thematic foci shift: some areas (e.g., mental and physical health) persist, while others emerge or fade, potentially responding to societal challenges or becoming integrated into disciplinary mainstreams. Implications: Rather than solely urging more multidisciplinary work, administrators might focus on creating sustained opportunities to build new cross-disciplinary bridges as existing ones are absorbed or fade. Suggested mechanisms include recurring, jargon-free lightning talk series to catalyze collaborations and institutional strategies to train future researchers for multidisciplinary work (e.g., multiple mentors across departments). For individual researchers, co-word patterns highlight gaps and fruitful avenues (e.g., connections between time pressure and happiness; adolescent depression intersecting with social media addiction). The approaches generalize to other Humanities and Social Sciences to monitor thematic evolution and identify emerging opportunities.

This study quantifies Psychology’s multidisciplinary engagement using citation and co-word network analyses over 2008–2018. Despite growth in publications and modest increases in Multidisciplinary Psychology journals, the proportion of cross-disciplinary citations remained stable (~25% to other disciplines), suggesting steady, not increasing, multidisciplinarity. Community detection and co-word analyses reveal both enduring and evolving research themes. The work provides a methodological template and baseline for administrators and researchers to monitor and strategically foster multidisciplinary research, and it can be extended to other social science and humanities domains. Future research should leverage richer network models (weighted edges, node-attributed networks), expanded measures (e.g., centralities), alternative units (e.g., co-author networks), and longer time spans to better distinguish among hypothesized mechanisms and capture broader dynamics.

- The analyses cannot adjudicate among alternative explanations for stable citation proportions (discipline identity, consolidation vs. bridging, mainstreaming of external ideas). - Networks were relatively simple (unweighted citation links, limited set of measures). Richer models (weighted edges, node-attributed networks with impact factors or h-indices) and additional metrics (centralities) could yield deeper insights. - Unit of analysis focused on journals and author keywords; other perspectives (e.g., co-author networks) might reveal different structures and individual-level dynamics. - The time window (2008–2018) may be too short to capture larger shifts; longer horizons could reveal different trends. - Author-provided keywords may be constrained by publisher lists, potentially biasing co-word themes.