Abstract
When a scientific discipline is maturing, scholars usually review the literature produced by the scientific community and assess the general state of development of the focal discipline (Ramos-Rodríguez & Ruíz-Navarro, 2004). U.S. News is a popular resource for world university rankings that examines the most influential scholars in key academic areas by using various bibliometric measures (Morse & Castonguav, 2021). Bibliometric analysis is a statistical technique that is commonly employed to present an overview of the growth or evolution of literature within a specified discipline over a certain period (Leung et al., 2017). This method utilizes several techniques, including citation-based analysis, co-word analysis, keyword co-occurrence analysis, and co-authorship analysis, with co-word analysis being the most commonly used technique (van Eck & Waltman, 2014). Previous bibliometric analyses have mostly adopted bibliometric indicators, including but not limited to article counts (Baker, 1989; Jayaratne, 1979; Thyer & Polk, 1997), impact factors (Lacasse et al., 2011), h-index (Hodge et al., 2021; Holosko & Barner, 2016; Thyer et al., 2019) and g-index (Holosko, 2021), to evaluate the quantity and impact of the work of scholars, institutions and journals.
Among these indicators, journal impact factors have been used to measure the impact of a scholar's work. However, these factors have been criticized due to their time-specific nature (Lacasse et al., 2011; Smith et al., 2018). Meanwhile, the h-index proposed by Hirsch (2005) is defined as “A scientist has index h if h of his/her Np papers have at least h citations each, and the other (Np – h) papers have no more than h citations each”. Relative to impact factors, the h-index is more reliable for rating a scholar's contributions throughout his/her career (Holosko et al., 2017). The g-index is an improved version of the h-index that takes into account the weight of the citations of the most frequently cited papers of a scholar (Egghe, 2006). The above bibliometric indicators shed light on who has the highest productivity and influence in a specific period. Bibliometric analysis has been widely used in many disciplines, such as business (Leung et al., 2017; Ramos-Rodríguez & Ruíz-Navarro, 2004), medical science (Thompson & Walker, 2015), sociology (Phelan, 2000) and psychology (Tur-Porcar et al., 2018).
Social work has charitable roots with a history that can be traced back to more than 100 years ago (Cnaan & Dichter, 2008). Since then, this field has developed into a profession that constantly creates, accumulates, and transmits disciplinary knowledge (Hodge et al., 2012). Over the past four decades, many researchers have examined the knowledge base of social work (e.g., academic papers) and the contributions of relevant entities (e.g., faculty members). For instance, Holden et al. (2005) performed bibliometric analysis to assess the productivity and impact of individual social work researchers and academic institutions. Lacasse et al. (2011) computed an estimated h-index value for social work faculty, whereas Holosko et al. (2017) examined the citation impact factors of accredited social work programs in Canada. In examining the evolution of social work, Martínez et al. (2014) detected ten thematic areas using science mapping technology. They analyzed research articles published from 1930 to 2013 across 25 main social work journals, and their findings shed light on the scientific evolution of social work. However, publishing articles in journals of a different discipline is not unusual for scholars. For example, social work scholars often publish articles in the American Journal of Orthopsychiatry (Al-Krenawi et al., 2007; Benbenishty et al., 2018), Journal of Behavioral Medicine (Ai et al., 2009; Ai et al., 2011; Levy et al., 2010), and Journal of Nursing Care Quality (Galambos et al., 2021), whereas scholars of nursing (Chen et al., 2020; Julion et al., 2021), law (Braun, 2021) and education (Chacón-Cuberos et al., 2021) publish articles in Research Social Work Practice. Therefore, merely focusing on papers published in discipline-restricted journals (Ho, 2013; Holden et al., 2005; Holden et al., 2009; Holosko et al., 2017; Lacasse et al., 2011; Martínez et al., 2014) or non-discipline-restricted journals (Klein & Bloom, 1992; Lacasse et al., 2011) can only paint half the picture of the major research activities and the dynamic nature of social work.
A promising solution to this problem is combining the most representative articles written by the most influential social work scholars. Baas et al. (2021) compiled a list of the top 2% of most-cited scholars from 22 scientific fields. Social work was founded in the “Economic & Social Science” domain and under the “Social Science” field, which may provide the necessary data pool for achieving the abovementioned aim. To compile this list, the most-cited scientists in each field were systematically ranked (Ioannidis et al., 2019). The composite indicator applied in this analysis managed to overcome the drawbacks of using a single measurement indicator and can capture the citation influence and take into account the relative credit in multiauthored papers (Ioannidis et al., 2016).
This study aims to advance this area by using the bibliometric software CiteSpace 5.8.R3 (Chen, 2006) to conduct a co-word analysis of selected literature published by top-cited social work scholars. To highlight social work scholars, the bibliometric analysis in this study was restricted to the first-authored articles written by scholars holding a doctoral degree in social work. According to Shneider (2009), the evolution of a scientific discipline follows four stages, namely, (I) conceptualization, (II) instrument development, (III) question investigation, and (IV) knowledge transfer. Non-social work scholars may play active roles at stages III and IV as they bring new techniques and tacit knowledge from their respective disciplines to investigate problems and synthesize knowledge in social work. Meanwhile, social work scholars are more likely to be visioners and inventors at stages I and II who establish the object of the research field and develop instruments for studying the underlying phenomena. In other words, social work scholars decide the goal of the discipline and lay the foundation for its subsequent development. The relationships among the concepts, problems, and practices introduced in the articles written by social work scholars can clarify the background and trends of research in this discipline. By identifying the research hotspots and representative scholars, the individual contributions of social worker scholars can be understood along with how their works facilitate the scientific evolution of the discipline. These stages equip social work practitioners with up-to-date knowledge and improve their understanding of the frontiers of social work knowledge and technology.
Method
Participants and Procedures
To explore the subject areas and their linkages, the input data in the analysis should combine the most representative literature with the most influential social work scholars. Baas et al. (2021) compiled a list of the top 2% most-cited scholars from 22 scientific fields (https://elsevier.digitalcommonsdata.com/datasets/btchxktzyw/3), of whom 243 were specializing in social work. Information about the highest education degree of these scholars was collected from the public websites of their affiliated organizations or retrieved directly from these scholars by sending them individual emails. The email enquiry lasted from February 22 to March 10, 2022. The educational backgrounds of 175 scholars were collected from public websites, and 53 of these scholars were holding a Ph.D. degree in social work. Meanwhile, the educational backgrounds of 68 scholars could not be found. Accordingly, individual emails were sent to these scholars on March 2, 2022 to confirm whether they received a Ph.D. in social work or just perceived themselves to have successfully completed all requirements for the conferment of a Ph.D. in social work upon receiving their highest educational degree. These scholars were given until March 10, 2022 to respond to these emails. Only 10 of these scholars responded to the email, of whom 7 confirmed that they have doctoral degrees in social work. In sum, 125 scholars were identified as nonsocial work scholars, 58 scholars had missing information, and 60 scholars holding a doctoral degree in social work were shortlisted for the next step (see Appendix). The first-authored articles and reviews of these scholars published from 2000 to 2022 were identified from the Web of Science Core Collection. Although this database contains articles published as early as 1900, the research themes and hotspots in the new century are more in line with the development of the current society. In addition, conversations about the results can be carried out using other articles that utilize data published after 2000 (e.g., Thyer et al. (2019)). In sum, the inclusion criteria of this study are (a) first-authored articles published by the 60 scholars holding social work doctoral degrees and listed among the top 2% most-cited scholars; (b) articles published from January 1, 2000 to March 4, 2022; (c) articles and reviews; and (d) articles indexed by the Web of Science Core Collection. The final sample for the analysis included 1,007 publications.
Data Analysis Plan
Co-word analysis was conducted using CiteSpace 5.8.R3 (Chen, 2006). This content analysis technique (Callon et al., 1983) relies on the co-occurrence of keywords in textual data to extract themes from texts and then analyzes their relationships (He, 1999). The analytical process was divided into several steps. First, the key terms standing as noun phrases were extracted from four text sources (i.e., titles, abstracts, author keywords, and keywords plus). In co-word clusters, each node represents a key term. A node whose betweenness centrality exceeds 0.1 has the potential for boundary spanning and plays an important role in linking other nodes (Chen, 2017; Chen et al., 2010). The primary theme for each cluster is manually identified as computer-assigned labels that may not fully capture the theme for clusters with multiple research foci. Second, the clusters of key terms were identified using default parameters (e.g., time slicing = 1 and modified g-index = 25). In cluster analysis, the modularity (Q-value) and silhouette values were calculated. The cluster structure is clearly defined when Q > 0.3, whereas the clustering effects are deemed reasonable and credible when Silhouette > 0.5 and > 0.7, respectively (Chen, 2004; Chen et al., 2010). Third, the primary themes for major clusters were identified. By default, CiteSpace has the largest connected components (LCC) of the underlying network. Only those clusters classified as LCC can be considered major clusters. A computer-generated label was assigned to each major cluster, but this label may not fully capture the theme for clusters with multiple research foci. Therefore, the primary themes of major clusters were manually identified based on the abstracts of their representative literature. Fourth, the key terms with bursts of at least 1 year were identified and grouped based on their associations with the major clusters.
Results
A total of 778 key terms were extracted from 1,007 publications (from 275 journals) to form the synthesized network. A total of 44 clusters were identified, of which 19 were major clusters on LCCs. The LCCs included 720 nodes (i.e., key terms), which accounted for 92% of the entire network. The network had a highly significant modularity (Q = 0.722), and the obtained clusters were deemed credible (Silhouette = 0.861). Burstness detection yielded 32 key terms, of which 21 had bursts of at least 3 years.
Research Hotspots and Representative Literature
A total of 44 clusters were formed, of which 19 were identified as major clusters on LCCs. Following Chen (2017), this study only focused on the largest major clusters. The clusters in CiteSpace were numbered in descending order based on their size. The largest clusters whose accumulated size would reach half of the network were initially selected. Clusters #0 to #7 included 412 nodes accounting for 52.96% of the entire network. The abstracts of the representative literary works of clusters #0 to #9 were then read, and no new research hotspot could be summarized starting from cluster #7. Therefore, the final analysis only focused on clusters #0 to #6 (Tables 1 and 2), which included 373 nodes and accounted for 47.94% of the entire network.
Properties of Seven Largest Clusters.
Major key Terms of Seven Largest Clusters.
Cluster #0: Professional Discourse and Discipline Development of Social Work
Cluster #0 is the largest cluster containing 98 key terms. The major key terms in this cluster include evidence-based practice (EBP), religion and social work. EBP first appeared in a social work journal more than 20 years ago (Gambrill, 2018). Since then, many researchers have focused on the EBP movement, including highlighting its importance (Gambrill, 2011; Gray & Schubert, 2012; Thyer, 2007, 2008), addressing the misunderstandings (Cnaan & Dichter, 2008; Gambrill, 2007; Gray et al., 2015b; Thyer & Myers, 2011; Thyer & Pignotti, 2011), identifying its problems (Gambrill, 2019; Rubin, 2011; Rubin & Parrish, 2007) and investigating its acceptance in education and practice (Beddoe, 2011; Bride et al., 2012; Gray et al., 2014; Gray et al., 2015a; James et al., 2019; Rubin & Yu, 2017b).
Both the EBP and spirituality literature shows high correlations with the key terms of cluster #0. Despite the lack of an agreed-upon measurement, the applications of spirituality in social work expanded over the previous decades (Crisp, 2020). The growing acceptance of spirituality may be attributed to the advocacy of religious freedom and social justice (Hodge, 2007a, 2007b), the applications of EBP in faith-based interventions (Ferguson et al., 2007a; Ferguson et al., 2007b; Hodge, 2006, 2010) and the improvement of spiritual competency through education (Crisp & Dinham, 2019; Hodge et al., 2006).
The third focus in cluster #0 points out the development directions for social work education and practice. Major issues include continuing education (Beddoe, 2015), student competencies (Regehr et al., 2012), cooperation with nonsocial workers (Beddoe, 2019; Weiss-Gal & Caduri, 2015), engagement in policy (Weiss-Gal, 2016, 2017; Weiss-Gal & Gal, 2008, 2020) and new challenges for social work (Nurius, 2017; Nurius et al., 2017a; Nurius et al., 2017b).
Cluster #1: Child Welfare System
Cluster #1 is the second largest cluster with 54 key terms. Although the primary theme of this cluster is the child welfare system, which provides both child protective services and out-of-home care (Berger & Slack, 2020), considerable attention was paid to children in out-of-home placement. Researchers cared about not only the experiences and reactions of children following their entry into out-of-home care (James et al., 2008; James et al., 2006; Wulczyn, 2020) but also the impacts of out-of-home placement on their wellbeing (e.g., cognitive skill) and behavioral problems (e.g., sexual risk behaviors) (Berger et al., 2009; Berger et al., 2015; James et al., 2009). Two problems were of particular interest, namely, delinquency and substance abuse. Efforts were made to compare the delinquent behaviors of adolescents receiving care across different settings (Ryan, 2012; Ryan et al., 2008; Ryan et al., 2016), to identify the risk and protective factors associated with delinquency (Cheng & Li, 2017; Ryan et al., 2007) and to study substance abuse among adolescents involved in child welfare (Cheng & Lo, 2010, 2011, 2012a).
To enhance the family reunification rate, which is a positive outcome of out-of-home care, scholars have studied its associated factors (Cheng, 2010; Cheng & Li, 2012) and called for integrated services targeting the specific needs of individual families (Marsh et al., 2006; Marsh et al., 2011). In addition, organizational climate (Glisson et al., 2006; Glisson & Green, 2006, 2011) and consideration of racial disparities (Cheng, 2009; Cheng & Lo, 2012b) may improve the child welfare system in general by optimizing service utilization and outcomes.
Cluster #2: Bullying Perpetration and Victimization Among Urban Adolescents
Cluster #2 contains 51 key terms indicating that adverse behaviors increase the risk for adolescents to engage in delinquency and risk-taking behavior (Hong et al., 2021a; Hong et al., 2021d; Hong et al., 2019). The most representative publications for this cluster were written by Jun Sung Hong, who mainly collected data from African-American adolescents in Chicago's southside. His works focus on the risk and protective factors for bullying perpetration and victimization at multiple levels (Hong et al., 2021b; Hong et al., 2021e; Hong et al., 2020a; Hong et al., 2020b; Hong et al., 2020c) and the factors that buffer the association between victimization and its detrimental impacts (Hong et al., 2021c; Hong et al., 2021d; Hong et al., 2021f; Hong et al., 2021g).
Cluster #3: Mental Health
Cluster #3 contains 47 key terms. The representative literature for this cluster covers three aspects of mental health, namely, the within-group effect-size benchmarks for therapies targeting mental health problems (e.g., depression) among children and adults (Rubin et al., 2017; Rubin & Yu, 2017a, 2017b), the associations between mental health problems and physical conditions for people suffering from a certain disease (Ai & Carrigan, 2007; Ai et al., 2007; Levy et al., 2006) and the risk and protective factors for mental health problems among Latin and Asian Americans (Ai et al., 2014a; Ai et al., 2013; Ai et al., 2015a; Ai et al., 2015b; Ai et al., 2014b).
Cluster #4: Ecological System Analysis
Cluster #4 contains 42 key terms. Apart from three articles on the bilingualism and wellbeing of children (Han, 2010; Han & Huang, 2010; Han et al., 2012), the majority of the representative literature for this cluster was written by Jun Sung Hong, who applied Bronfenbrenner's ecological systems theory to investigate the risk and protective factors for health risk behaviors among the minority youth (Hong et al., 2011a; Hong et al., 2011b), domestic violence (Hong et al., 2010b; Hong et al., 2012; Hong et al., 2011c) and school bullying (Hong & Espelage, 2012; Hong et al., 2011c) as well as the factors that operate within five system levels for school safety and campus shooting (Hong et al., 2011a; Hong et al., 2010a; Hong & Eamon, 2012).
Cluster #5: Intimate Partner Violence
Cluster #5 contains 41 key terms. Although some representative publications for this cluster coincide with the themes of the literature for the former clusters, a prominent theme for this cluster was intimate partner violence. The literature for this cluster focused on those factors associated with intimate partner violence (Cheng & Lo, 2016a) and its impacts on health problems (Cheng & Lo, 2019a), help-seeking behaviors (Cheng, 2013; Cheng & Lo, 2019b) and the wellbeing (Cheng & Lo, 2013, 2016b) and behaviors of children (Hong et al., 2021h; Hong et al., 2021i).
Cluster #6: Spiritual Needs and Assessment
Cluster #6 contains 40 key terms. Except for three articles on resilience (Ungar, 2013a, 2013b; Ungar et al., 2013), the other literature for this cluster was written by David R. Hodge, who focused on spirituality in healthcare settings. Mainstream health services can be ineffective for minorities and indigenous peoples whose perspectives on wellness are influenced by their spiritual and cultural beliefs (Hodge et al., 2009). Therefore, identifying and addressing spiritual needs may improve overall satisfaction with the hospital service provision for US minorities (Hodge et al., 2014a; Hodge et al., 2014b; Hodge & Wolosin, 2014a; ; 2015a, 2015b). To help practitioners provide culturally competent services and assessments, experts were invited to modify and validate the tools and frameworks for spiritual assessment and to suggest a culturally valid question protocol for operation (Hodge & Limb, 2009a, 2009b, 2010a, 2010b, 2010c).
Research Frontiers
Table 3 shows the burst key terms in the research conducted by the selected top-cited authors. The surge duration of a key term is indicated by its beginning and end years. A larger strength value corresponds to a larger increase in the amplitude of key terms per unit of time (Chen, 2017). The burst detection yielded 32 key terms, with the highest intensity reaching 7.459.
32 key Terms With Bursts of at Least 1 Year.
The 32 key terms can be semantically divided into four groups (Table 4) based on what they stand for and their associations with the seven largest clusters.
32 key Terms With Bursts of at Least 1 Year in Four Groups.
Group 1 included 4 key terms associated with clusters 0, 3, and 6 that indicate how spirituality evolves into an embodiment of religion, indigeneity, and multicultural competence. These terms also propose that such development is triggered by the need to improve the healthcare services provided to people of different cultures, races, and ethnicities.
Group 2 included 15 key terms associated with clusters 1, 2, 4, and 5. 2014 was defined as a dividing year. Specifically, before 2013, researchers focused on children's wellbeing, the impacts of domestic violence on children, whether those families who are involved in the child welfare system receive proper services and the service outcomes for children. However, since 2014, researchers shifted their focus to youth violence, including bullying perpetration, victimization and juvenile delinquency, particularly among African Americans in Chicago's southside.
Group 3 included 7 key terms associated with cluster 0 and other clusters whose representative literature focused on service provision. The representative literature for this group proposes three stages of social work practice and development, including encouragement of EBP (2007–2009), emphasis on service provision with high ethical standards (2013–2016) and the impacts of changing policies on social work education and practice (2016–2020). The gap between 2009 and 2013 may be supplemented with group 4, which included 6 key terms associated with clusters 1–4. The representative literature for this group highlighted the preference of social work scholars to adopt ecological system theory in studying the risk and protective factors for certain problems (e.g., victimization) based on national samples.
Discussion
To explore the subject areas and evolution of social work discipline in the past two decades, this study performed a co-word analysis on 1,007 selected publications written by the most influential social work scholars. They hold doctoral degrees in social work and are among the top 2% of most-cited authors worldwide (Baas et al., 2021). Seven research hotspots emerged from their first-authored articles and reviews: professional discourse and discipline development of social work, child welfare system, bullying perpetration and victimization, ecological system analysis, mental health, intimate partner violence, and spiritual needs and assessment. In total, 32 keywords were detected to have bursts for at least 1 year. They correspond to four essential elements of social work, namely, spiritual and cultural needs, children's well-being and youth violence, social work practices and development, and social work research.
Previous bibliometric analysis has a long-lasting interest in the impacts of journals and programs, and the productivity of scholars and faculties (Jayaratne, 1979; Rosen, 1979; Rosenberg et al., 2005). It serves as an important information source for those making academic decisions (e.g., for which graduate program to apply) (Baker, 1989; Holosko, 2021; Holosko et al., 2017; Smith et al., 2018; Thyer & Bentley, 1986; Thyer et al., 2019). A few studies have analyzed selected literature from the core social work journals (Holden et al., 2009; Rosenberg et al., 2005), but the present study is among the first to focus on the conceptual structures and scientific evolution hidden in the publications. The most recent counterpart to our study was conducted by Martínez et al. (2014), who performed science mapping analysis on publications from 1930 to 2012 in 25 main social work journals. They identified six classical thematic areas (i.e., children, social services, health care, women, social workers, and education) and four emerging thematic areas (i.e., violence, human immunodeficiency virus [HIV]/acquired immunodeficiency syndrome [AIDS], lesbian, gay, bisexual, and transgender [LGBT], and grief).
The seven research hotspots echoed the six classical and one emerging thematic areas in Martínez et al. (2014)'s study. Our hotspot of professional discourse and discipline development corresponds to Martínez et al. (2014)'s areas of social work and education, but with a greater emphasis on EBP and spirituality. It suggests that the importance of the fine-defined culture of research is widely acknowledged (Barner et al., 2015; Holosko & Barner, 2016; Thyer & Polk, 1997) and social work scholars have made efforts to enhance the evidence base for traditional practice. The hotspot of the child welfare system is related to both areas of children and social services, but the focus is on out-of-home placement. The hotspots of bullying and ecological system are related to areas of children and violence. It indicates that youth problems are bringing more concerns and scholars tend to explore them from a comprehensive perspective. The hotspot of intimate partner violence matches the area of woman. The hotspots of mental health and spiritual assessment are related to the area of healthcare.
We didn’t find research hotspots related to HIV/AIDS, LGBT, and grief (i.e., the rest three emerging thematic areas) despite that these keywords appeared in a few articles in our bibliographic records. There are three possible explanations. First, Martínez et al. (2014) discovered that the thematic areas of LGBT and GRIEF were not growing consistently like the rest eight areas ten years ago. Our result is coherent with their finding and provides further implication that the topics of LGBT and grief may appear under the umbrella of more general topics (e.g., social services and healthcare) concerning social worker scholars. Second, Martínez et al. (2014) included the literature coming from core social work journals while the present study included the articles of social work scholars published in journals covering all disciplines (275 journals in total). Such differences in research scope may lead to different results for the main thematic areas. Third, a simple guess is that there were not enough publications to form clusters highlighting these issues. Since we only analyzed first-authored articles, the publications of senior scholars may outnumber those of their junior counterparts as they have accumulated more in their careers. However, this should not be interpreted as a devaluation of the importance of these three topics.
Although we didn’t divide the time frame (2000–2022) into consecutive periods of time, the 32 key terms with bursts gave us a hint on the research frontiers and development tracks over the past two decades. Cautiously interpreting these key terms could provide requisite knowledge to both social work researchers and practitioners. The bursts were mainly about spirituality, children, and social work services. The concept of spirituality has witnessed a constant expansion from religious belief to culturally adapted coping. It indicates the social workers’ proactive response to individual spiritual and cultural needs and their devotion to diversity and difference. The shifted concerns from children's well-being to youth violence manifests how the scientific community responded to rapid economic and social changes. Given that the bursts came in groups and last long, the frontiers of vulnerability and violence among children may last for a long time and therefore warrants attention from both researchers and practitioners of social work. Bursts about social work services suggest two directions for development: maintaining high ethical standards under changing policies and implementing scientific practices with the aid of EBP, systematic frameworks, and national samples. Worth noting is that, although the burst of EBP only lasted for one year in 2007, it has laid the foundation for social work practice. The growing application of scientific methods, including EBP, ecological system theory, and representative samples has directed the development of social work research and facilitated the social work practices to deliver practical issues encountered at different times.
Despite that co-word analysis yields inspiring results when examining scientific evolution, it has two major limitations, and we call for caution when interpreting results. First, as co-word analysis requires extensive literature to increase its accuracy, it may well extract research hot spots and frontiers in the past decade but may not make full use of its advantages when frequent updates are appreciated. The present study analyzed publications of 60 scholars over a time span of 20 years. Is it possible that an annual report may update the research hotspots timely? In that case, researchers and practitioners may trace or even predict an upcoming hotspot at the very beginning from the annual reports in the past 2–3 years. Second, as co-word analysis is based on co-occurrence of keywords, the results can be difficult to interpret. A single key term may appear in multiple clusters of different research foci. One solution is to read through abstracts of the highly representative literature and configure manual decisions as what we did in the present study. This could be time-consuming and bring research bias. Another solution is to conduct document co-citation analysis, which can be more precise than co-word analysis as the unit analysed is a paper rather than a word.
Besides, the present paper has certain limitations. First, the bibliographic records used in this study solely came from the Web of Science Core Collection. Although this database is of significance, some other important works exclusive to other databases or published earlier than 2000 may have been missed. Nevertheless, similar studies such as Holden et al. (2009); Martínez et al. (2014); Thompson and Walker (2015) used the same database to conduct the same analysis in other disciplines. Second, in order to highlight social work scholars, this study only considered the first-authored articles and reviews written by scholars with doctoral degree in social work. These works represent the most elemental practice and fundamental development for social work as well as serve as the intellectual base for cross-disciplinary social work research. However, scholars from disciplines other than social work have also made significant contributions to this field (Hodge & Turner, 2022). Future comparative studies should be conducted to understand the differences in the contributions of these two types of scholars across the four stages of the evolution of a scientific discipline (Shneider, 2009). Third, 58 scholars had missing data on their education and were therefore excluded from this work. However, such exclusion may introduce a bias on part of the representative 60 scholars eventually selected in this study (Hodge & Turner, 2022). Fourth, the seven research hotspots identified in this work did not cover all themes. For example, 15 papers on associations between neo-liberalism and social work failed to form a research hotspot as they were scattered across different clusters.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Appendix
60 scholars with a doctoral degree in social work from the list of top 2% most-cited scholars worldwide
