Introduction

There is a long history on social work research. In the early nineteenth century, hospital social work was presented (Beard 1923). At the same time, most research focused on psychiatric social work (Jarrett 1918; Neilson 1919). Theory and practice in social work and evaluation of social work were investigated in 1970s (Segal 1972; Sheldon 1978; Gibbons et al. 1978). Strong perspectives in social work practice were presented in latter last century (Weick et al. 1989; Saleebey 1996). In recent years, the short form of the Childhood Trauma Questionnaire (the CTQ-SF) (Bernstein et al. 2003) has been frequently applied in social work. Topics related to children have been paid more attention especially child sexual abuse (Wyatt 1985; Beitchman et al. 1992; Mullen et al. 1996).

Classic papers (Garfield 1976) also called most frequently cited articles (Garfield 1973), highly cited articles (Garfield 1974), top cited articles (Bentler 1992), or top publications (Korevaar and Moed 1996) in various fields have been studied, with focus on rehabilitation (Shadgan et al. 2010), environmental and occupational health (Smith 2009), ophthalmology (Ohba et al. 2007), water resources (Chuang et al. 2011), and library and information science (Blessinger and Hrycaj 2010). Bibliometric analysis of classic papers using Science Citation Index or Social Science Citation Index (SSCI) databases for a specific journal was also found including Social Biology (Osborne and Osborne 1999), Veterinary Quarterly (Elsinghorst and Sybesma 2000), British Journal of Nutrition (Trayhurn 2004), American Journal of Roentgenology (Bui-Mansfield 2005), Clinica Chimica Acta (Csako 2007), Australian Dental Journal (Smith 2008a), Australian Veterinary Journal (Smith 2008b), and Journal of Orthopaedic Trauma (Lefaivre et al. 2010). It has been reported that due to the presence of many most cited papers, there have been influential subsequent advances in molecular biology resulting in helping a great number of people (Picknett and Davis 1999). Highly cited articles nevertheless provide an interesting and useful insight into which authors, articles, and topics are influencing the research profession over time (Smith 2008b). Furthermore, using classic papers to teach physiology (Raff 2005), organometallic chemistry (Duncan and Johnson 2007), capillary filtration, and presentation skills development (McGeown 2006) were also presented.

Number of articles cited was often used as an indicator of scientific performance in subjects of anaesthesia and pain (Terajima and Åneman 2003), ophthalmology (Ohba et al. 2007), and obstetrics and gynecology (Brandt et al. 2010). Top cited articles were commonly listed to provide a basic source of information and were categorized according to publication year, journal, authors, countries, and institutions (Brandt et al. 2010). However, the number of citations of a paper is probably not sufficient to show the impact it had in the research field. Indicators such as numbers of authors cited, numbers of institutions cited, numbers of countries cited, numbers of subject areas cited, citations per year, and total number of citations of a paper to date have been applied to the high impact papers on the subject of water resources (Chuang et al. 2011). Citation life cycles of highly cited articles were also considered to be important (Aversa 1985). The citation histories of papers give more details of the impact characteristics of articles (Wang et al. 2011).

The articles analyzed in the present work were those in the field of social work in SSCI having at least 50 citations since 1956. This study identified and examined the characteristics covering publication year, journals, authors, institutions, countries, life citation cycles, and attributes of the top cited articles.

Methodology

The methodology used in this study was based on the SSCI database of Web of Science from Thomson Reuters (updated on 30 November 2012). According to Journal Citation Reports (JCR) of 2011, it indexes 2,966 journals in 56 Web of Science categories of social science edition. There were 41 journals listed in the category of social work in JCR in 2011. There were 6,452,237 documents from 1956 to 2011 found in SSCI. Results were refined by the Web of Science category of social work (73,506 documents). TC2011 denotes the total citations since publication of the article up to the end of 2011 (Wang et al. 2010; Chuang et al. 2011). TC2011 ≥50 was used as a filter to extract the classic cited documents (782 documents). Documents with citations ≥50 were downloaded showing the total annual citations for each article. C2011 denotes the total citations only in 2011. The advantage of TC2011 and C2011 is that they are invariable and ensure repeatability compared with the index of citation from Web of Science (Fu et al. 2012). Therefore, 1.1 % of the total documents published in the Web of Science category of social work are regarded as the classic papers including articles (721; 92 %), reviews (46; 5.9 %), proceedings papers (39; 5.0 %), editorial materials (8; 1.0 %), notes (6; 0.77 %), and one discussion. The 721 articles having TC2011 ≥50 were retrieved as classic articles for further analysis. All results were graphed using Microsoft Excel 2007 for visual representation. In the SSCI database, the corresponding author is designated as the “reprint” author; this study will hereby use the term “corresponding author”. In a single author article where authorship is unspecified, the single author is both first and corresponding author. Similarly, in a single institutional article, the institution is classified as the first as well as the corresponding author institution. In a multi-author article where authorship is unspecified, the first author is classified as the corresponding author. In subsequent analysis, articles originating from England, Scotland, Northern Ireland, and Wales were classified as being from the United Kingdom (UK) (Chiu and Ho 2005). Articles from Hong Kong before 1997 were included with China. The contributions from institutions and countries were identified by the appearance of at least one author in the publications. Collaboration type was determined from the addresses of the authors. The articles were classified into five types based on the country and institution: (1) TP: the number of “total articles” of an institution or a country; (2) SP: the number of “single country article”, if the researchers’ addresses were from the same country or “single institution article”, if the researchers’ addresses were from the same institution; (3) CP: the number of “internationally collaborative article”, if the articles were coauthored by researchers from multiple countries (Chiu and Ho 2005) or “inter-institutionally collaborative article”, if authors were from different institutions; (4) FP: the number of “first author article”, if the first author was from the country or institution for analysis; and, (5) RP: the number of “corresponding author article”, if the corresponding author was from the country or institution for analysis.

Results and discussion

Publication year

A total of 721 classic articles (TC2011 ≥50) had records in social work field in SSCI. These papers were published between 1957 and 2008. TC2011 ranged from 50 to 752 citations and 90 citations per publication (CPP) on average. Figure 1 illustrates the distribution of these 721 classic articles over the years, and their CPP. No classic articles have yet emerged in the most recent 3 years (2009–2011). Similar result in lack of recent classic articles was also found in the top cited articles in chemical engineering in SCI-Expanded (Ho 2012a). Obviously, newly published articles require time to accumulate citations (Picknett and Davis 1999). Only two of the classic articles were published prior to 1950s, while 47 % of the classic articles appeared in 1990s. It is accepted that with continual passing of time since publication there is an increasing chance of the paper being forgotten (Picknett and Davis 1999). In addition, as time goes by, even “true classics” are gradually less often cited, probably because their substance is being absorbed by the current knowledge, a phenomenon that has been termed “obliteration by incorporation” (Garfield 1987). To cite the original paper would be appropriate (Ho 2004). It not only respects authors who presented a novel idea in research but also facilitates colleagues to read the original work to better understand the idea in detail (Ho 2010). In particular, the decade of the 1950s with two articles and the 1960s with five articles had much higher CPPs, 202 and 120, respectively, which can be attributed to the article by Greenwood in 1957 with a TC2011 of 254 and Riessman in (1965) with a TC2011 of 214. The CPPs of the other four decades ranged from 74 to 103. Furthermore, the mean number of CPP for the top 100 classic citations was 211, which was lower than the value of 100 classic citations in some medical fields, for example 283 in anesthetic journals (Baltussen and Kindler 2004), 318 in ophthalmology journals (Ohba et al. 2007), 405 in general surgical journals (Paladugu et al. 2002), 447 in orthopaedic surgery (Kelly et al. 2010), and 629 in urology (Hennessey et al. 2009). One of the reasons might be that researchers in social work field are less frequently cited and use fewer references than other medical areas. The earliest classic article in social work field was published in Social Work in 1957 (Greenwood 1957). The most recent classic article was published in 2008 in American Journal of Community Psychology with TC2011 = 50 by Wandersman et al. (2008).

Fig. 1
figure 1

Number of articles and citation per publications by decade

Full size image

Journal

The classical articles were published by 32 journals in the Web of Science category of social work. Ten of the 32 journals were not listed in the category of social work in 2011 including Family Coordinator with 14 classic articles listed in Web of Science from 1968 to 1979, Social Casework with four articles (1956–1978), Child Care Quarterly with three articles (1972–1986), Social Casework-Journal of Contemporary Social Work with three articles (1979–1989), Social Work Research & Abstracts with three articles (1983–1993), Journal of Education for Social Work with two articles (1965–1984), and Public Welfare (1956–1996), Urban & Social Change Review (1969–1986), Journal of Gerontological Social Work (1980–2000), and Journal of Social Distress and the Homeless (1994–2002) with one article respectively. Forty-one journals were listed in the category of social work in JCR in 2011. The impact factor of a journal according to JCR for the year 2011 was determined. The classic articles in social work field were distributed in more journals than those classic articles in some medical fields, for example, the seven journals cited in the top 100 from classic papers of orthopaedic surgery (Kelly et al. 2010), ten journals cited in the top 100, from general surgical citations classics (Paladugu et al. 2002), eleven journals cited in the top 100 cited articles in obstetrics and gynecology (Brandt et al. 2010), 13 journals in 100 from ophthalmology class citations (Ohba et al. 2007), and 15 journals cited in the top 100 cited articles, from urology (Hennessey et al. 2009).

Of these 22 journals listed in the Web of Science category of social work in 2011, Trauma Violence & Abuse was the only journal with impact factor higher than 3. Child Abuse & Neglect published the most classic articles with 217 articles (30 %), followed by American Journal of Community Psychology with 194 articles, while their impact factors were 2.471 (rank 3 in 41 journals) and 1.736 (4/41) respectively (Table 1). As expected, the classic articles were published in journals with high impact factors, similar to the subject area of anesthetics (Baltussen and Kindler 2004). The leading journals attracted the classic publications, which in turn maintained the high impact factor of these journals (Schein and Fingerhut 2000). However, classic articles with TC2011 ≥50 in social work field could also be found in journals with lower impact factors such as Families in Society-The Journal of Contemporary Human Services (IF = 0.286; 32/41) with four articles and Administration in Social Work (IF = 0.360; 33/41) with one article.

Table 1 Characteristics of 21 journals in social work field with the classic articles
Full size table

Authors, institutions, and countries

Each person listed as an author of an article has made an independent material contribution to the manuscript (Coats 2009). The results of author analysis have identified those researchers who have made significant contributions. Among the 1,536 authors contributing to 721 classic articles, 1,303 authors (85 %) published one classic article; 150 authors (10 %) published two classic articles; 50 authors (3.3 %) published three classic articles; 20 authors (1.3 %) published four classic articles; and 13 authors published five or more of the classic articles. The top 18 most productive authors worked in USA. Table 2 lists the authors with five or more classic articles with three indicators such as total number of classic articles, first author articles, and corresponding author articles. It has been accepted convention among the experimental sciences that the most important authorship positions are the first and the last, who very often is the corresponding author (Costas and Bordons 2011). The first author is the person who contributed most to the work, including conducting research and writing of the manuscript (Riesenberg and Lundberg 1990). The corresponding author supervised the planning and execution of the study and also participates in writing of the paper (Burman 1982). D. Finkelhor at University of New Hampshire in the US published not only the most classic articles but also the most first author and corresponding author classic articles. The most cited article among Finkelhor’s ten classic articles was titled “Sexual abuse in a national survey of adult men and women: prevalence, characteristics, and risk factors” (Finkelhor et al. 1990) in Child Abuse & Neglect (TC2011 = 720). A. Wandersman at University of South Carolina of the US ranked second published nine classic articles and had one article as single author. Both M.A. Zimmerman at University of Michigan of the US and M.E. Courtney at University of Wisconsin of the US ranked second in first and corresponding author classic articles, respectively. J. Rappaport at University of Illinois of the US published four single author classic articles. The most frequently cited single author classic article (Rappaport 1987) was published by Rappaport in American Journal of Community Psychology with TC2011 = 410. Article entitled “Area, class and health: should we be focusing on places or people” (Macintyre et al. 1993) by authors in the UK in Journal of Social Policy with TC2011 = 458 was the most cited article by non-US authors.

Table 2 Thirteen highly productive authors of five or more classic articles
Full size table

The 710 classic articles with author address information in the Web of Science were further analyzed regarding institutions and countries. Altogether, 710 articles originated from 443 institutions in 17 countries. The top 38 most productive institutions were in US. Table 3 shows the top 11 institutions published more than 15 classic articles each with University of Illinois (29 articles) and University of California, Los Angeles (26 articles) leading the list. University of Illinois also published the most first and corresponding author articles while University of California, Los Angeles had the most inter-institutionally collaborative articles and Arizona State University had the most single institution articles. Of the total articles, 425 articles (60 %) came from single institutions themselves and 285 articles (40 %) from inter-institutional collaborations. The inter-institutional collaboration rate was observed to be large in science fields, such as 62 % of global climate change (Li et al. 2011), 53 % of atmospheric simulation (Li et al. 2009), and 44 % of solid waste research (Fu et al. 2010). But inter-institutional collaboration rate was smaller in classic articles of some medical fields, such as 12 % of 100 top cited articles in general surgical journals (Paladugu et al. 2002), and 8 % of 100 ophthalmology class citations (Ohba et al. 2007). It might be concluded that the classic articles had less collaboration than those of non-classic.

Table 3 Characteristics of the 11 most productive institutions (TP >15)
Full size table

Classic articles usually originated from a small field centered in a few countries, for example, 100 top cited articles in general surgical journals were produced by six countries (Paladugu et al. 2002); top 100 most frequently cited articles in anesthetic journals originated from nine countries (Baltussen and Kindler 2004); the top 100 ophthalmology classic citation articles originated from 10 countries (Ohba et al. 2007). In this study, only 20 articles (2.8 %) were international collaboration and they were contributed by nine countries, while the other 690 articles (97 %) were articles contributed by 15 countries. The percentage of collaboration in the classic articles was much lower than that of other areas in SCI-Expanded, such as 14 % biosorption technology for water treatment (Ho 2008), 16 % of desalination research (Tanaka and Ho 2011), 18 % of financial crisis research (Chang and Ho 2010), and 14 % of acupuncture research (Han and Ho 2011). In general, collaboration played an important role in enhancing the impact of articles (Narin et al. 1991; Leta and Chaimovich 2002). The highly cited papers typically involve more collaborative research than the general norm (Aksnes 2003). However, results indicated that higher percentage of single country’s classic articles was found in social work field. This phenomenon was also found in science (Fu et al. 2012) and engineering (Ho 2012a) fields. The characteristics of these 17 countries that published classic articles are illustrated in Table 4. The leading country was US (629 articles), accounting for 89 %, followed distantly by the UK (34 articles) and Canada (29 articles). Eight of 17 countries published only one classic article. Domination in classic articles by the US was not surprising since this pattern occurs in other scientific fields such as, obstetrics and gynecology (Brandt et al. 2010) and orthopaedic surgery (Kelly et al. 2010). As for the 20 internationally collaborative articles, 18 (90 %) involved contribution by the US; and as for non-collaborative articles, 574 articles (89 %) were published by the US. As US took the lead with an overwhelming majority, they had the highest percentage of classic articles in social work field than other research fields, such as articles of general surgical (78 %) (Paladugu et al. 2002), anesthetic (70 %) (Baltussen and Kindler 2004), ophthalmology (86 %) (Ohba et al. 2007), and urology and subspecialty (76 %) (Hennessey et al. 2009). The top cited publications originated from the United States are explained in part by the large number of the American scientific publications (Zhou and Leydesdorff 2008). There is some evidence that US authors tend to reference articles from US journals than from other countries (Campbell 1990). Reviewers from the United States and outside the United States who also evaluate non-US papers are likely to evaluate papers submitted by US authors more favorably (Link 1998). In particular, all of the classic articles from Finland, Italy, Switzerland, Norway, Ireland, China, and Netherlands were single country articles, while France and Germany published only one internationally collaborative article each.

Table 4 Characteristics of the 17 contributing countries
Full size table

Citation life cycles of articles with the highest TC2011

The articles with the highest TC2011 can be considered the most popular articles in the research field. Table 5 shows the four most frequently cited classic articles that were cited at least 500 times since publication of the articles up to the end of 2011. Two of the four articles were published in American Journal of Community Psychology. All four were published by authors from the US. A classic article, titled “Measures of perceived social support from friends and from family: three validation studies” (Procidano and Heller 1983), received the most citations (TC2011 = 752). It was published in American Journal of Community Psychology by M.E. Procidano at Fordham University of the US and K. Heller at Indiana University. A bias was found that authors’ affiliations were recorded as INRA (Kenneth Heller Institut National de la Recherche Agronomique), France and Fordham University, USA in Web of Science. The detailed figures of cited units, such as authors, institutions, and countries were listed to reveal more of the characteristics that influenced the classic articles in social work field. In general, more citations meant that more authors, institutions, and countries were influenced by the article. Further analysis of citations showed that the article was cited by 1,652 authors from 548 institutions in 34 countries, cited in 330 journals in 75 Web of Science categories, and with an average of 25 times per year. Article life shows impact history of articles (Chuang et al. 2011). Five typical citation frequency curves of individual articles, such as initially much praised articles, basic recognized work, scarcely reflected work, well-received but later erroneous qualified work, and genial work were reported (Avramescu 1979). The citation life cycles of highly cited articles were studied in following years (Chuang et al. 2011). The citation lives of the top four classic articles (TC2011 >500) are shown in Fig. 2. Three of the top cited articles were published in 1980s and one in 1990. Only one article entitled “Sense of community: a definition and theory” (McMillan and Chavis 1986) saw continually increasing trends in most years since published, while the others’ citation rates climbed initially to a plateau and then maintained steady for some years.

Table 5 Characteristics of top four articles (TC2011 >500)
Full size table
Fig. 2
figure 2

Citation life cycles of the top four articles (TC2011 ≥500)

Full size image

Citation life cycle of articles with the highest C2011

Total citations indicated impact of an article in a research field. Article life shows characteristics of an article impact after publication. About 74 % of all classic articles had no citations, 17 % articles had one citation, and 4.7 % articles had two citations during their year of publication. TC2011, an accumulative number may reach a large value as long as the time span is long enough. An article impact might not be always high (Fu et al. 2012). Although some recently published articles within the past few years had great potential, they did not have a high TC2011. Thus it would be interesting to investigate new publications that have shown high impact in 2011. Of the articles evaluated in 2011, 13 % of all classic articles had no citations, 10 % articles had one citation, and 10 % articles have two citations. Indicators including the number of authors of the article, the number of authors cited, institutions cited, countries cited, journals cited, citations per year, country, C2011, and TC2011 with C2011 greater than 40 are all shown in Table 6 (Chuang et al. 2011). In terms of citations per year, the article by Bernstein et al. (2003) had the highest value, 30 citations per year. “The long-term impact of the physical, emotional, and sexual abuse of children: a community study” (Mullen et al. 1996) was the only one internationally collaborative article by authors in New Zealand and Australia. Other five articles originated from US. The article by Bernstein et al. (2003) had the greatest total citations in 2011 (C2011 = 78), TC2011 = 271, ranked 18th and received the highest citations per year. The article by McMillan and Chavis (1986) had C2011 = 62, TC2011 = 536, ranked 3rd, was cited by the most authors (1,023), Web of Science categories (82), and journals (264) which showed evidently the article was more multidisciplinary. This article also had the highest citation in its publication year. However, it sometimes turns out that even if an article is cited very frequently, it does not influence as many institutions, countries, and categories as would have been expected, for example the article by Reichman et al. (2001). Of all the articles listed in Table 6, only McMillan and Chavis (1986) were also listed in Table 5, indicating strong performance both in quantity as well as in the impact of their research. Although Table 6 shows the classic articles with the highest citations, it does not provide any information on the history of citation impact. In addition, it was reported that very high percentage of the most frequently cited articles have changed their ranking with time (Bui-Mansfield 2005). A historical perspective on the classic papers in the literature that are still influential in social work was examined (Fig. 3). Six articles had more than 40 citations in 2011. Among them, three articles were published in Child Abuse & Neglect and one in Journal of Community Psychology, Children and Youth Services Review, and Child Maltreatment respectively. The article by Bernstein et al. (2003) is the only article with more than 10 authors. Its annual number of citations sharply increased from the time of its publication. The article by Straus et al. (1998) had the second highest number of citations in 2011. It had an article life pattern similar to the most frequently cited review paper by Bernstein et al. (2003). Overall, it appeared to be going strong and showed no sign of leveling off. An early article by McMillan and Chavis (1986) was the only one in both Figs. 2 and 3 with TC2011 >500 and C2011 >40. Three articles, Bernstein et al. in 2003, Reichman et al. (2001), Finkelhor et al. (2005), had high citations in 2011 (C2011 >40) but low total citations (TC2011 <500). In comparison with the six articles in Fig. 3, these latter three articles did not have enough time to accumulate citations, but all of them have sharply increased in citations since their publication.

Table 6 Characteristics of top six articles (C2011 >40)
Full size table
Fig. 3
figure 3

Citation life cycles of the top six articles (C2011 ≥40)

Full size image

Publication characteristics of authors and institutions

The first author is known to have made the most contribution and should receive a greater proportion of the credit (Riesenberg and Lundberg 1990; Yank and Rennie 1999; Marušić et al. 2004). It has been reported that the first author is the person who contributes most to the work and writing of the article (Gaeta 1999). The corresponding author responsibilities include supervision of the planning and execution of the study, along with writing the paper (Burman 1982). At the institutional level, the determined institution of the corresponding author might be a home base of the study or origin of the paper. A new Y-index (j, θ) is developed to evaluate publication characteristics of authors and institutions. It is related to numbers of first author publications (FP) and corresponding author publications (RP), defined as:

$$ j = {\text{FP}} + {\text{RP}} $$
(1)
$$ \theta = \tan^{ - 1} \left( {\frac{\text{RP}}{\text{FP}}} \right) $$
(2)

An author, institution, or country with a higher j indicates more publications as first or corresponding author, playing the leadership role. θ, differentiates the nature of the leadership role. When θ > 0.7854, indicates more corresponding author publications and θ < 0.7854, means more first author publications. When θ = 0, j = the number of first author publications and when θ = ∞, j = the number of corresponding author publications.

There is a limitation for analysis of publication characteristics of authors and institutions by Y-index. Only articles with both first author and corresponding author information could be considered. A total of 721 classic articles by 1,570 authors in social work field were analyzed. Only 575 of the 1,570 authors had both first author and corresponding author classic articles. Five of 575 authors had θ > 0.7854 and two authors had θ < 0.7854 while 568 (99 %) had the same numbers of first author and corresponding author articles. Percentage of author published classic articles with θ = 0.7854 in social work field was found to be much higher than authors published in the top cited articles in the SCI-Expanded (Ho 2012b) and in the chemical engineering field (Ho 2012a). Figure 4 displays the distribution of the top 84 authors with j ≥ 4, (j Cosθ and j Sinθ are chosen as the x and y coordinate axes). Each dot represents one value that could be one author or many authors. The authors who contributed the most classic article were D. Finkelhor (j = 16), followed by M.A. Zimmerman and M.E. Courtney with j = 12 respectively. Publication characteristics constant, θ, could help to obtain the different proportion of corresponding author articles to first author articles. It is very helpful especially when j of authors is the same to distinguish the different performance of authors. For example, the j of S.E. Hobfoll and M. Barrera were both the same of 7. However θ of Hobfoll was 0.9273 but θ of Barrera was 0.6435. Hobfoll had greater proportion of corresponding author articles to first author articles than Barrera. Within these 84 authors, only four authors with θ > 0.7854 such as S.E. Hobfoll (θ = 0.9273), and S.R. Dube, C.M. Sullivan, and G.M.H. Pretty with θ = 0.9828 respectively. These authors had more corresponding author articles than first author articles indicating that top cited authors contributing to the classic articles were more likely to be designated as the corresponding authors. The top cited authors probably contributed more to the initial conception and supervision of study (Wren et al. 2007). M. Barrera (θ <0.7854) was the only one who had more first author articles than corresponding author articles. Other 79 authors just on the boundary line (θ = 0.7854) having the same quantity first authors articles and corresponding author articles.

Fig. 4
figure 4

Top 84 authors with Y-index (j ≥ 4)

Full size image

Only 710 classic articles with both first author and corresponding author institution information were analyzed for publication characteristics by Y-index. There were 443 institutions contributing to the 710 articles. Within these institutions, 202 (46 %) institutions had no first author articles nor corresponding author articles; 60 (14 %) institutions had only corresponding author but no first author articles (θ = ∞); 35 (7.9 %) institutions had θ > 0.7854; 171 (39 %) institutions had θ = 0.7854; 36 (8.1 %) institutions had 0 < θ < 0.7854; and 52 (12 %) institutions had θ = 0. With respect to j, 90 (20 %) institutions had j = 0, with no first author articles nor corresponding author articles; 274 (62 %) institutions had 1 ≤ j < 5; 67 (15 %) institutions had 5 ≤ j < 20; and 12 (2.7 %) had institutions had j ≥ 20. The top 36 institutions were located in US. The top 26 institutions with j ≥ 10 are shown in Fig. 5. University of Illinois (49, 0.7650) and Arizona State University (43, 0.8086) took the leading positions. Eleven institutions including Suny Albany (10, 1.166), Rutgers State University (11, 1.052), University of California, Los Angeles (33, 0.9358), Johns Hopkins University (20, 0.8851), University of Southern California (22, 0.8761), Centers for Disease Control and Prevention (11, 0.9761), University of Wisconsin (30, 0.8520), Case Western Reserve University (17, 0.8442), Washington University (17, 0.8442), University of California, Berkeley (35, 0.8140), and Arizona State University (43, 0.8086), had more corresponding author articles (θ > 0.7854), while ten institutions including University of Washington (29, 0.5485), University of Virginia (10, 0.5880), University of South Carolina (19, 0.6288), University of Rochester (14, 0.6435), Columbia University (14, 0.6435), University of Maryland (15, 0.7188), Michigan State University (21, 0.7378), University of Chicago (23, 0.7419), University of Michigan (33, 0.7551), and University of Illinois (49, 0.7650) had more first author articles than corresponding author articles (θ < 0.7854). Other five institutions had the same number of first author articles and corresponding author articles (θ = 0.7854). The proportion of institutions had articles with θ > 0.7854, θ = 0.7854, and θ < 0.7854 were found to be 11:5:10 for top 26 articles and 35:170:36 for all articles with 0 < θ < ∞.

Fig. 5
figure 5

Top 26 institutions with Y-index (j ≥ 10)

Full size image

Conclusions

During the period from 1957 to 2008, 721 classic articles were published in 32 journals covered in social work field in SSCI from 17 countries. The results indicated that 57 % of the total classic articles in social work field were published in Child Abuse & Neglect and American Journal of Community Psychology. 97 % of the classic articles were single country publications and 3 % were internationally collaborative publications. US produced the most single, collaborative, first author, and corresponding author articles followed distantly by the UK and Canada. Information from citation life cycles of articles with the highest TC2011 and C2011 gives more details about impact of the classic articles. The classic articles might not always be highly cited. Results from evaluation by Y-index show that D. Finkelhor, M.A. Zimmerman, and M.E. Courtney were the three most contributing authors to the classic articles in social work. University of Illinois and Arizona State University were the main institutional contributors. University of Illinois published more first author articles and Arizona State University published more corresponding author articles.