Early Years of the Journal of Learning for Development: A Combination of Bibliometrics and Thematic Analysis

The paper analyses the contributions to Journal of Learning for Development (JL4D) from volume one to five using bibliometrics and content analysis techniques. Analysing the 91 papers in various categories of JL4D, the paper identifies authorship pattern, topics covered, research methods used, types of documents used in citations, core journals and the median age of citations to JL4D. The finding of the study reveals that JL4D has created a niche for itself as a specialised research journal focusing on innovations in learning contributing to development.


Introduction
The Journal of Learning for Development (JL4D) was conceived in 2013 as an open access journal to be published on a rolling basis (Tait, 2014) and the first issue was published in 2014.In the last five years, JL4D has published three issues regularly in March, July and November.While introducing the Journal, President and CEO of Commonwealth of Learning (COL) ̶ publisher of the Journal ̶ emphasised that there is a growing recognition that learning must lead to development, and there is a need to share the "vast body of knowledge and experience" in this field.Therefore, JL4D was started with an explicit aim to "showcase the practical dimension of how learning for development works" (Kanwar, 2014).The Journal website states that it is a "forum for the publication of research with a focus on innovation in learning, in particular but not exclusively open and distance learning, and its contribution to development.Content includes interventions that change social and/or economic relations, especially in terms of improving equity" (JL4D, n.d).So, JL4D was not envisaged to be another journal of distance and online learning but intended to provide a platform for "all innovation in learning that had as an aim to contribute to social and economic development" (Tait, 2016).It also intended to be a vehicle for engaging "a broad audience of researchers, scholars and practitioners"early careers as well as established scholars from the Commonwealth and beyond (Kanwar, 2014).Latchem (2014), appreciating the launch of JL4D, stated that: • it affirms COL's commitment to learning for development; • the scope of the Journal covers sectors other than formal education; • the Journal provides the opportunity to apply the "broader principles of educational technology" and provides evidence of success; and • the Journal provides a blank slate to practitioners and researchers to express their thoughts, ideas, experiences and findings to support lifelong learning.Baggaley (2014) while analysing the paucity of journals covering 'learning' and 'development' emphasised that "JL4D can anticipate the problems of current and future educational trends via its applied focus on the access and equity problems of learning and development in the developed and developing nations".
Journals are the life blood of research in any discipline, disseminating new ideas and reporting the latest developments in the field (Mishra, 1997).Those, being 'peer reviewed' publications, enjoyed a higher reputation (Royal Society, 1981), but the process eventually delayed publication of research findings (Cronin, 1984).The emergence of electronic publishing and subsequently open access journals made it possible to reduce the time-lag in publication of results.JL4D adopted both an online mode and open access principle to publish the Journal.As it completed five years of its existence, it is important to analyse and reflect how the Journal is progressing and trying to create a niche in its identified field.The objectives of the study were to examine: • authorship pattern, including contribution segregated by gender and region; • topics covered, and research methods used; and • citation pattern, including types of documents cited, age of citations, core journals cited, etc.

Methods
Considering the objectives of the study, a combination of bibliometrics method and content analysis was used in the study to analyse the articles published in JL4D and the citation appended to them as references.Since the term bibliometrics was coined by Pritchard (1969), it has been used for analysing research trends in a variety of fields: the history of science, social sciences, library and documentation studies and science policy (Okubu, 1997).In the field of educational technology and distance education a number of single journal studies have been reported in the literature.Some of these include: British Journal of Educational Technology (Bond, Zawacki-Richter & Nichols, 2019), Computers & Education (Zawacki-Richter, & Latchem, 2018); Distance Education (Zawacki-Richter, & Naidu, 2016); Indian Journal of Open Learning (Mishra, 2002;Tripathi, & Kanungo, 2010); The International Review of Research in Open and Distributed Learning (Zawacki-Richter, Alturki, & Aldraiweesh, 2017).A series of bibliometric studies were published in the Educational Technology magazine from 2011 to 2015 covering 22 journals (West, 2016).Bibliometrics methods have emerged as an important field to study the properties and behaviour of recorded knowledge.Various statistical methods are applied to "measure authorship, citation and publication pattern, and the relationship within scientific domains" (Patra, Bhattacharya, & Verma, 2006).Data about the research papers in JL4D was collected from the journal website and manually coded in Microsoft Excel to create tables and charts for analysis.
Amongst the bibliometrics tools, study of citations is a common practice since the emergence of citation indexes (Garfield, 1970).The earliest study to use citation data to evaluate the importance of scientific journals was by Gross and Gross (1927) for the study of the Journal of the American Chemical Society.The premise is "bibliographic citation is an expression of a relationship between two documents, the citing and the cited" (Cronin, 1984).The JL4D is indexed in Google Scholar, Directory of Open Access Journals (DOAJ), Education Resources Information Centre (ERIC), WorldCat and Bielefeld Academic Search Engine (BASE).While the JL4D has so far avoided being listed in citation indexes due to the proprietary nature of the indexes, Zawacki-Richter, Anderson and Tuncay (2010) stated that the proprietary owner of the Citation Indexes, such as the Social Science Citation index, has "little interest in indexing and calculating impact factors for journals in relatively small disciplines and especially those that compete with it as open access publications."It may be noted that the citation analysis also has limitations due to the underlaying motive of the authors while citing a document, which may be serious or frivolous (Cronin, 1984), questioning the validity and reliability of citation data (Rice et al, 1989).Despite its inherent limitations, Zawacki-Richter and Anderson (2011) consider citation data as "the best objective measure describing the relationships between journals and the flow of information in a research discipline" (p.445).References from all the papers in the JL4D were collected and added to a Microsoft Excel sheet for further analysis.For the citation of articles published in JL4D in other journals, Publish or Perish software (Harzing, 2007) was used with Google Scholar as the source.This is an acceptable method (Zaugg et al., 2011) to analyse citations, particularly when the journal is not indexed in major commercial indexing sources.
While analysis of a citation network built around publications allows us to have a better grasp of how a scholarly community has evolved in a field (Jo, Jeung, Park, & Yoon, 2009), content analysis provides understanding of the trends and issues covered in the discipline.According to Lee, Driscoll, and Nelson (2004) "understanding trends and issues in terms of topics and methods is pivotal in the advancement of research on distance education" (p.225).Content analysis of journal articles has been done by several researches (Mishra, 1997(Mishra, , 1998;;Bond, Zawacki-Richter & Mark Nichols, 2019;Lee et al., 2004;Marín, Duart, Galvis, & Zawacki-Richter, 2018;Zawacki-Richter, Bäcker, & Vogt, 2009;Zawacki-Richter et al., 2017;Zawacki-Richter & Naidu, 2016;Zawacki-Richter & Latchem, 2018).Most of these studies used computer-assisted content analysis as an appropriate method for content analysis.As in the study by Bond et al. (2019), in this study Leximancer TM was used to analyse the title and abstract of the papers published to identify significant concepts and themes covered in the journal.

Data Source
JL4D publishes items under different sections: (i) Editorial/ Foreword, (ii) invited articles, (iii) research articles, (iv) special feature, (v) report from the field, (vi) case study, (vii) commentary, and (viii) book reviews.Except the editorial and book reviews all other sections in the Journal are peer reviewed.The invited articles are normally solicited papers form senior experts in the field but still undergo the review process.The special feature section is used to cover topics of special interest such as sustainable development, and a whole series on leaders in open and distance learning was published under this section in 2017 and 2018.From 2014 to 2018, JL4D published 118 items under different categories (Table 1).For analysis in this paper, editorial and book reviews were not used, and thus, 91 items were taken for coding and analysis.A publication count of The International Review of Research in Open and Distributed Learning, a peer reviewed journal, revealed that it published 81 research articles in the issues of the initial five years.Similarly, the only comparable published data is from Indian Journal of Open Learning that published 68 items in its initial five years (Mishra, 1997).

Results and Discussion
Authorship Pattern Collaboration Authorship pattern is a measure of research collaboration in any field of study (Subramanyam, 1983).
It is also widely assumed that collaboration in research is a 'good thing' and "for decades multipleauthor publication, frequently referred to as a co-authored publication, has been used as a basic counting unit to measure collaborative activity" (Katz & Martin, 1997).Research also shows that collaboration is associated with increased scientific productivity (Parish, Boyack & Ioannidis, 2018).
The average number of co-authors per paper published by individual scientists has steadily increased in all fields over the past century (Fanelli & Larivière, 2016).In the case of JL4D, except in the first volume, the number of multiple author papers were more than single author papers.In the five-year period, multiple author papers counted for over 55% of papers (Table 2).Using a formula suggested by Subramanyam (1983), the calculated score (0.56) shows a higher degree of collaboration.In the field of distance education, which is related to the scope of JL4D, Mishra (1997) (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013), the percentage of co-authored papers was 84% (Christensen et al., 2015) and in Journal of Computing in Higher Education (2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012), the percentage of co-authored paper was 68% (Langton et al., 2015).Therefore, though the multi-authored publications are not high in JL4D as in the case of many major educational technology journals, the current degree of collaboration (0.56) could be seen as good for a relatively young journal.

Gender
Not much information is available on gender of the authors contributing to research in the focus area of JL4D.However, it is an important consideration for COL to encourage women researchers to support learning for development.Of the authors, whose gender could be identified (through appropriate searches), we found that 61.20% were men (Table 3).The study on distance education research by Zawacki-Richter et al. ( 2009) found that 55.4% of first authors were male.In comparison the study of Indian Journal of Open Learning showed the male contributors at 72.83%.While the available data is not sufficient to make any generalisation on gender of researchers in the field covered by JL4D, there is a need to target female researchers to encourage contributions from them.

Regional Distribution of Contributors
As indicated before one of the objectives of JL4D is to encourage research contributions from the Commonwealth and beyond (Kanwar, 2014).A regional analysis (Table 4) of the 186 contributors to JL4D revealed that the highest percentage of contributors were from Africa (34.41%) followed by Asia and the Americas with 18.28% and 16.67% respectively.There is variation in distribution of contributions across the regions, but more contributions from Africa shows that the Journal has been able to provide researchers in Africa with a good platform to contribute.However, of the 37 countries of the contributors to JL4D, Canada and the United Kingdom came out on top (Table 5), with over 50% coming from only five countries (Canada, the United Kingdom, India, South Africa and Tanzania).Interestingly, looking beyond the Commonwealth, JL4D has also reached to non-Commonwealth countries (37.84%) indicating that the majority of contributions came from the Commonwealth with 86% of the contributors (Table 6).

Leading Contributors
Of the 186 authors within the period of five years, sixteen authors contributed at least two papers.However, there were only three who contributed more than three papers (Table 7) -all three of them from Africa, indicating the space created by JL4D to support researchers in the region.

Research Methods
To analyse the research methods used in the papers published in JL4D, we broadly used the categorisation followed by Zawacki-Richter et al. ( 2009) covering quantitative, qualitative, triangulation/mixed method, or others (Grant, Ward & Rong, 1987).While the first three categories are self-explanatory, the category 'others' included papers that were either conceptual, descriptive or opinion/commentary in nature.Table 8 indicates that the highest percentage (48.35%) of contributions in JL4D were in the category of 'others' followed by qualitative (23.08%), quantitative (15.38%) and mixed method (13.19%).The percentage of 'others' category or descriptive papers in JL4D are similar to an analysis of papers published between 1991 to 1996 in four key journals of distance education by Mishra (1997) who reported the percentage of descriptive papers as 47.6.However, Berge and Mrozowski (2001) classified 75.9% of the articles published in four key journals of distance education as descriptive between 1990 and 1999.Interestingly, the Zawacki-Richter et al. ( 2009) study reported 38.1% of all articles as descriptive, with 12.9% in the category of mixed-method design (triangulation), which is similar to this study with 13.19% for papers with mixed method research design.In contrast, Bozkurt et al. (2015) analysed seven key journals of distance education between 2009-2013 and reported that DE researchers adopted mostly qualitative (47%) and quantitative (37%) studies, and just a few employed mixed-method (16%) designs.The second highest category of research methods used in JL4D is aligned towards the findings of this study, though it is far behind as a comparison Nevertheless, it indicates that so far JL4D has attracted more descriptive and qualitative studies.It may also be noted that the findings are not in line with the majority of bibliometrics studies conducted on journals in the field of educational technology, as West (2016) reported that "the field skews strongly towards quantitative methods.Also, most mixed method studies were not balanced in how they honored the different research paradigms" (p.44).

Content Analysis
JL4D has a clear focus on innovations in learning leading to development-change in social and/or economic relations, especially in terms of improving equity as a result of learning.However, the historical focus of COL as the sponsoring agency of the Journal may influence its content largely towards open education and distance learning.As indicated before, content analysis of the abstracts and the title of the articles were carried out using Leximancer TM software to understand the coverage and focus of the Journal.The thematic summary reveals that education has the most direct mentions within the text with 113 (100% relative count), followed by learning (88% connectivity), teachers (35%), need (34%) and research (31%).The concept map (Figure 1) shows linkages of key terms within the five clusters of themes.The longest links can be traced amongst three sub-themes spanning education-need-teachers (see education-policy-quality-access-need-challenges-online-educational-teachersschool).The number of papers covered in this analysis was only 85 (six papers had no abstract).Therefore, the thematic clustering is not very strong, though it indicates that the papers published were not just focused on open and distance learning, and the editors and peer reviewers have done the right level of gatekeeping to focus on the Journal's core area of innovations in learning.Thus, while the focus of the papers was by and large on issues related to policy, quality, access, openness, learners and support in specific contexts, learning was the second core issue discussed with a focus on impact, courses, and factors.This was followed by a focus on teachers and teaching, learning and open educational resources.The other foci were on need (focusing purpose and appropriateness to the context) and research.The latter may be attributed to the importance attributed to research methods in any peer-reviewed publication.

Citation Characteristics
Bibliographic references in a publication symbolises a metaphoric relation between the cited document and the citing document (McInnis, 1982).References are also indicators of the scholarliness of a journal (Cline, 1982).Table 9 shows that there were only four items without references in JL4D, and the average references per article in the Journal was 27.67 (excluding the items without references).This is similar to the mean score of 29 references for the journals covered in the study by Zawacki-Richter et al. (2009).While references appended to articles may not be treated as a criterion for quality articles, Price (1970) and Avramescu (1980) suggest that an article with 16±6 reference indicate scholarliness.According to Price (1970) articles providing fewer than 10 references imply that "scholarship does not exist but is irrelevant or exists relevantly but is unknown" (p.8).In the light of emphasis placed on review of previous research as indicator of quality (Moore, 1985), we can assume that JL4D has maintained standards from the beginning by ensuring authors' focus on previous research in their works.

Types of Citation
Table 10 shows the types of documents cited in the papers published in JL4D: 36.54% of the citations were from Journal sources, followed by 32.77% in the 'others' category, which included government reports, weblinks, blogs, personal communication, etc. References to books, including chapters in books were 23.8% and conference papers accounted for only 6.27%.This information use pattern by contributors of JL4D indicates that they are highly dependent on journals and grey literature.Out of the total of 2,408 references in all the articles, 1,093 (45.39%) had a Uniform Resource Locator (URL) attached to it indicating increased use of electronic resources by scholars publishing in JL4D.This trend is comparable to the analysis of types of document cited in Indian Journal of Open Leaning (Tripathi & Kanungo, 2010) which showed 34.07%journal citations.

Core Journals Cited
Of the 880 citations from journal articles, analysis was carried out to study scattering of journals in the field of learning for development.There were 406 journals cited, of which 289 (32.84%) were with only one citation and 25 journals accounted for over 40% of the citations (Table 11).While The International Review of Research in Open and Distributed Learning is the leading journal cited with 8.41% of citations, the dispersion of journals cited by scholars contributing to JL4D indicates that the field draws from several fields of studies and is not necessarily influenced by open and distance learning alone.

Most Cited Authors
Most cited authors indicate their significant impact to the field.However, in the case of JL4D, the current analysis does not indicate high citation of any single individual or a group of researchers.However, Table 12 indicates that the top cited authors in the articles published in JL4D.Interestingly, most of the authors listed in this table are from the field of open and distance learning.In other words, this shows that JL4D contributors are largely influenced by scholars in open and distance learning, though the thematic analysis and scattering of journals do not support this assertion strongly.

Most Cited Works
An author could write many works, but not all of his/her works will receive similar attention from peer researchers.While the dispersal of references in JL4D was large, covering a variety of areas, Table 13 shows the list of documents cited more than four times.Martin Weller, whose name appears in the most cited author list also has two publications in the list of most cited works.

Conclusion
The bibliometric and thematic analysis of the contributions published in JL4D from volume one to five revealed the early years of an open access journal in the field of social science.Especially considering the focus and scope to the Journal, the number of items published is more than similar to comparable journals in their early days.JL4D has maintained the balance of different types of contributions from both Commonwealth and non-Commonwealth countries.It has been publishing relatively more items form Africa, which indicates that the JL4D has found a niche amongst researchers there.While only about 40% of the contributors were female during the period, there is a satisfactory degree of collaboration amongst authors.In future studies it may be useful to analyse the relationship between gender and collaboration and types of research methods used (Zawacki-Richter & von Prümmer, 2010).The content analysis of the abstracts and titles revealed that JL4D is focusing on educational issues in general, with a major focus on 'student learning', 'teachers and teaching' and 'contextual needs'.The sample covered in this study was too small to make a more critical analysis on the linkages of the themes covered, but the use of the computer assisted content analysis could be used in future to gain a deeper understanding of how JL4D is contributing to various sub-domains within the 'learning for development' field as a unique journal.Contrary to the thematic content analysis, the citation analysis revealed that the contributions are by and large influenced from the field of educational technology in general and experts in the field of open and distance learning.However, these researchers extensively used literature from a variety of journals and fields indicating that the Journal's focus was still on innovations in learning, and not open and distance learning per se.The halflife of the references to the papers published shows that contributors of JL4D use information within eight years, which is more than the 6.5 years estimated by Davis and Cochran (2015), indicating that the field needs to use more of the latest research to reflect on innovations in learning.Five years is a small period in the life of an academic journal, but the study provides a bird's eye view of the Journal's progress and contribution to the literature.

Figure 1 :
Figure 1: Concept map of themes in JL4D

Table 13 : Most Cited Works in JL4D
Battle for Open: How openness won and why it doesn't feel like victory.London: Ubiquity Press.