How plural ist ic is the research f ield on adult education ? Dominating bibl iometrical trends , 2005-2012

What the field of adult education research is and how it can be described has been a debated issue over the decades. Several scholars argue that the field today is heterogeneous, borrowing theories and methods from a range of disciplines. In this article, we take such statements as a starting point for empirical analysis. In what ways could it be argued that the field is pluralistic rather than monolithic; heterogeneous rather than homogenous? Drawing on bibliographic data of the top cited articles in three main adult education journals between 2005 and 2012, we illustrate how the citation patterns have tendencies of homogeneity when it comes to the geographical country of authorship, since the USA, UK, Australia and Canada dominate, as well as the research methods adopted, since qualitative approaches have near total dominance. Furthermore, there is a tendency to adopt similar theoretical approaches, since sociocultural perspectives, critical pedagogy and post-structuralism represent more than half of the articles in our sample. At the same time, the results of our analysis indicate signs of scholarly pluralism, for instance, in terms of authorship, since both early career researchers and established researchers are represented among the top cited publications. We conclude the article by arguing that empirical analysis of publication and citation patterns is important to further the development of reflexivity within the field, not least for early career researchers, who might benefit from knowledge about what has been recognized among peers as worth citing in recent times.


Introduction
What the "field" of adult education research is and how it can be described has been a debated issue over the decades.In the 1970s and 1980s the debates were to a large extent of a philosophical character, focusing on issues of epistemology, i.e. the status of adult education as a discipline or as a unique field of knowledge (cf.Hirst, 1974;Bright, 1989).Such debates seem to have declined in the 1990s and, entering the 2000s, these debates had almost disappeared (Rubenson, 2000).Today, there seems to be a consensus concerning the epistemological status of adult education, since scholars construe this "field" as inherently interdisciplinary, borrowing theories and methodologies from a range of disciplines (cf.Fejes & Salling Olesen, 2010;Hake, 1992;Larsson, 2010;Rubenson, 2000).Openness to the inclusion of scholars from diverse disciplines with different methodological and theoretical inclinations is thus arguably an important part of the self-image of scholars active in the field.However, whether or not adult education research is characterized by methodological and theoretical pluralism, could also be treated as a research question, open to inquiry and contestation.In what ways is the field characterized by scientific practices that are heterogeneous and pluralistic?Is this interdisciplinary and institutionalized pluralism reflected in the most cited contributions in some of the main adult education journals in the field?
To study the publication outlets of adult education is a feasible way to map out how adult education as a field of research is construed.Studying this field by analysing its leading scientific journals has become a popular strategy in recent years (Hayes, 1992;Blunt & Lee, 1994;Rachal & Sargent, 1995;Taylor, 2001;Fejes, 2008;Harris & Morrison, 2011;St. Clair, 2011;Larsson, 2010;Fejes & Nylander, 2014), and the early examples of bibliometric studies and content analysis of adult education research date back to the 1970s (Dickinson & Rusnell, 1971;Long & Agyekum, 1974;Boshier & Pickard, 1979).In this article, we study the top-cited articles in three of the leading English-speaking adult education journals between the years 2005-2012 in order to examine if and in what aspects the field is shaped as heterogeneous and pluralistic.Thus, our analysis aims to provide a description of the field based on what scholars have recognised as worthy to cite.
The analysis adopts a two-step procedure: firstly, we map the content of the topcited contributions in terms of theoretical and methodological perspectives, common research themes and ways to construct the object, i.e. the what and how of the most cited articles.Secondly, we deepen the empirical account of who publishes in the adult education research field.This is done by tracing the institutional affiliations, academic positions and geographic site of enunciation of the most cited authors as well as by providing some demographic characteristics.Our findings show how the most cited articles have tendencies of homogeneity when it comes to the geographical country of authorship, since the USA, UK, Australia and Canada dominate, as well as the research methods adopted, since qualitative approaches have near total dominance.At the same time, the citation pattern shows some signs of pluralism for instance, in terms of authorship, since both early career researchers and established researchers are represented in our sample.
Before presenting, in greater detail, the research questions and method that have led us to unravel the dynamics of contemporary adult educational research, we briefly sketch some historical trends in the research field.This sketch will act as a basis for the discussion of our results.
A sketch of the history of adult education research Adult education as a field of research has emerged largely out of a concern for the practice of adult education.In North America, such concern was partly connected to the emergence of university programs in adult education.The focus was on developing knowledge in order to define and develop the field (Jensen, Liveright & Hallenbeck, 1964) and much research was directed at developing programs and instructional methods.This is illustrated, for example, by Long and Agyekum's (1974) analysis of articles that was published in the AEQ (then named Adult Education).They found that during the period 1964-1973 more than half of the articles (55%) concerned one of the three following areas: program planning and administration, instructional material and methods, and adult learning.Rubenson (1982, p. 62) notes that empiricism and psychology heavily influenced adult education research in the US before the 1980s, and argues that that this was partly due to the dominating idea that instruction 'derives linearly from research'.There was thus a lack of research on sociological aspects of adult education, and questions of power were not a common issue dealt with.
In Europe, the history of adult education research is more diverse than in the US, not least because the US is one nation compared to the multiple ones in Europe, and there is one official language in the US, whereas Europe has many.Such diversity in Europe could be illustrated, for example, by the different ways professorships in the field have been labelled: professor of andragogy in some locations such as the former Yugoslavia or the Netherlands (Bron, 2006), or professor of adult education in Finland, Germany and Sweden.In some countries, there has never been a professorship in adult education.A second illustration of diversity is the way adult education is organised.In some countries, e.g. in Scandinavia, there are long historical traditions of institutionalised adult education (cf. Laginder, Nordvall & Crowther, 2013), while in other locations such history is much shorter.A third illustration of diversity, connected to the two above, is how well developed the adult education research field is within each country.This could partly be expected to depend on the history of the institutionalisation of adult education and based on different political priorities (social democratic, liberal, conservative).However, one trend in Europe during the 20 th Century has been how adult education in many locations became integrated into social and economic policy, making research funding available (Rubenson, 2000).Such developments can partly be related to trends within the social sciences and educational science more generally.
In the 1950s, there was an increase in the public spending on education in countries such as Germany, Sweden, the USA and the UK.One of the effects of these political reforms and growth of public expenditure on education was a growing need to find means to evaluate and administer the population in the educational realm.Partly connected to these developments, many countries established governmental authorities and research institutes that were meant to monitor and survey what was happening in the educational field.The increase in large-scale surveys and statistical data analyses based on micro-and census data after the World wars can, in part, be seen as a means for politicians to have researchers provide results deemed useful for policy-making and steering (Husén, 1983).The ascend of a quantitative research paradigm in educational research was driven largely by the expectations that funding of such research would increase efficiency and productivity of education.In Sweden, this was seen, for example, in the vast number of studies exploring the phenomena of 'reserve of talent' (begåvningsreserven), which refers to those who have not been given the opportunity to reach the level of education which their talent foresaw (Härnqvist, 1958;Husén, 1956).Such studies paved the way for an increase in public spending on adult education in Sweden as it became apparent that many citizens lacked the educational opportunities that the measurements of 'talent' predicted they were suited for.Thus, research on adult education in Europe partly came to focus on policy issues and the state, topics seldom treated in the US despite empiricism and statistical data dominating the field at this time.According to Rubenson (2000, p. 5) the lack of focus on policy issues and the state in the US could partly be explained due to a 'decentralized political and economic system and individual emphasis on social mobility, [which] promotes a research focus on the individual'.
However, the dominance of empiricism and the quantitative research paradigm would come to change in the later part of the 20 th Century.In the 1960s and 1970s, alongside the rapid growth of the whole university sector and the political radicalization of students, the domination of empirically-oriented research was questioned by the increased use of hermeneutic, phenomenological and critical perspectives.This development also paved the way for qualitative research methods and that did not have much legitimacy and scientific status from the start (cf.Husén, 1983Husén, , 1988;;Larsson, 2006).These more general trends within the social sciences and educational science had repercussions for the field of adult education research.We will mention three such repercussions here.
Firstly, there was an increase in qualitative studies.Even though research in the US still, to a large extent, lingers on a psychological research tradition, previous research indicates that there has been a decline in empirically oriented research in the US (Rubenson, 2000) as well as quantitative studies (Taylor, 2001).Secondly, theorisations in the field of adult education that came to focus on issues of power developed.For example, the work of Paulo Freire (1972) became influential for many adult education scholars during the 1970s, not only in Latin America but also in Europe among the former colonizers (cf.Kane, 2013).Freire was also picked up in the US, most notably through the work of Henry Giroux (1983).The writings of critical pedagogues such as Freire and Giroux provided critiques of oppressive relations and inspiration for how to design educational practices that might be mobilized to counter oppression.At the same time, intellectual movements that later became known under the label of 'poststructuralism' emerged.These had a strong emphasis on abandoning the quest for essence and causality, offering new ways to conceptualize power, class, gender, and the making of social scientific knowledge.With an interest in deconstructing and problematising practices of adult education, researchers in adult education now started to ask other questions and provide different answers than those offered by critical pedagogues (cf.Usher & Edwards, 1994).However, the risk of such a shift was in the US context, according to Rubenson (2000, p. 5) that 'the concern over the lack of a theoretical base in some quarters of adult education has resulted in a preoccupation with abstract theory building'.The third and final repercussion that we want to draw attention to is that the debates about the epistemological status of the field have almost disappeared, and instead the field has been construed by many scholars as interdisciplinary and pluralistic (cf.Fejes & Salling Olesen, 2010;Hake, 1992;Rubenson, 2000).

Bibliometric research on the who, what and how of adult education research
In this article we use bibliographic data to give a synthetic image of the research field where knowledge about adult education is produced.We believe that bibliometric measurement and content analysis might prove helpful in identifying what is promoted and valued within the field and, as such, may provide a pertinent basis for scientific reflexivity as the modus operandi of the field is outlined (Bourdieu, 1988, p. xii).
Similar approaches have previously provided fruitful knowledge in terms of the who, what and how questions in this particular field.By aggregating bibliographic data on what is being produced in adult educational journals, one could, for instance, specify the demographic backdrop such as the relative numbers of women and men among authors (Hayes, 1992;Taylor, 2001;Harris & Morrison, 2011) or the institutional affiliation of authors in terms of geography (Taylor, 2001;Harris & Morrison, 2011;St. Clair, 2011).Both Taylor's (2001) and Harris and Morrison's (2011) analyses indicate that the research field has gone through a change from a dominance of male authors to female authors.They have also confirmed that there is a clear dominance of anglophone authors in the Adult Education Quarterly and the Australian Journal for Adult Learning.On the same topic, Larsson (2010) has stressed that publication patterns in the research field of adult education and learning are clearly related to specific geographic sites of enunciation and different language regimes.Building on Larsson's (2010) arguments, Fejes and Nylander (2014) have shown how anglophone authors are not only more prone to contribute to the English-speaking journals in the field, they have also a much better revenue on their publications than non-anglophone scholars in terms of who is considered worth citing.
Apart from identifying who it is that contributes to the scientific journals in the field, this line of research has also provided some answers as to what kind of research is being published, i.e. it answers the what and how questions.For instance, Taylor (2001) and St. Clair (2011) have noted an increasing trend for qualitative studies in the field and a simultaneous decrease in the number of quantitative studies.Taylor (2001) found that the most common subject treated in the articles of AEQ between 1989 and 1999 was 'adults learning', a finding supported by Harris and Morrison (2011) who, from the analysis of the Australian Journal for Adult Learning, added 'teachers/educators/teaching' as common themes.Some of these results might appear self-evident in the eyes of someone whot has worked in the field for a long time, yet given the historical changes discussed above, it should not be taken for granted that these circumstances will not change further as time goes by.
One of the limitations of previous bibliometric studies and content analyses of journal publications in the field is that they tend to map out one journal at the time.As the different journals might uphold positions that are, hypothetically at least, complementary to one another, it seems more reasonable to study the research field with a sample that includes multiple journals published in different geographical locations.Therefore for our analysis we have chosen three different journals to represent the field published in three different geographical locations: Adult Education Quarterly (AEQ) (USA), International Journal of Lifelong Education (IJLE) (UK) and Studies in Continuing Education (SICE) (Australia).Analysing these journals will, based on those articles recognized by scholars in the field through the highest rates of citations, provide the basis for answering our questions about how this research field is structured in relation to bibliographics and content.

Categories and principle of classification
We have conducted an analysis of the content and authorship based on a sample of the most cited articles published in three leading journals in the field between 2005-2012.Our focus on academic journals is based on the premise that academic publishing provides one important knowledge base for any discipline or field of research, and represents vital parts of its history, social structure, norms and ways of communication.
Compared with books and book chapters, which are less likely to be subjected to review processes, journals include papers that have been deemed to have a certain level of rigour and quality.Arguably, peer-reviewed journals have gained increased importance in evaluating research in recent years, a tendency that has accelerated through a series of political reforms (Karpik, 2011;Nylander, Aman, Hallqvist, Malmquist & Sandberg, 2013;Fejes & Nylander, 2014).Although we acknowledge that entire research fields and several advanced scientific journals have emerged specializing in bibliometrics and scientometrics 1 , our paper engages with other adult education researchers who have conducted similar studies (cf.Taylor, 2001;Larsson, 2010).
Our analysis aims to provide a broad description of how the field has been shaped in terms of authorship and content in recent years.We have therefore focused our exercise on ten different categories that are of particular interest to us. 2 These categories correspond to a set of how and what questions -in terms of research method, theory, context and object.A second set of questions deal with the question of who -in terms of authorship.Under this heading we have included variables on department affiliation, geographic site of enunciation, academic title, gender, number of authors, as well as number of citations.In order to identify how the most cited articles relate to the questions above, we have used data available in the Scopus database.Besides being one of the main indexing databases in the academic community that applies certain quality procedures when including journals, the reason for choosing this database is mainly pragmatic.Scopus includes all three of the journals chosen for this analysis, that represent three different geographical locations, whereas only two of them, AEQ and SICE, are listed in the Web of Science, the other major database available for bibliometric measurements (cf.Archambault, Campbell, Gingras & Lariviere, 2009).

Classification based on dominance
In selecting and classifying the data for our analysis, we have adopted the principle of dominance.First, we narrow down the sample of our analysis to the top cited articles published in each of the three journals.The reason for focusing on the most highly cited articles is the collegial recognition ascribed to them by either relating, opposing or building on their arguments.Altogether, this means that we have carried out a detailed analysis of 57 articles, all of which appeared in the journals throughout the period of 2005 to 2012, which amounts to the 19 most cited articles in each journal.As our focus here is on recognition among peers, all self-citations have been excluded from the statistics.
Based on this sample, each author of this paper separately read the articles and categorised them in relation to the above-mentioned questions.Throughout these readings we sought to construct broader thematic categories that could allow for quantifiable descriptions, for example the object of study being either students, teachers, workers, or research.Since the quantitative approach requires a reduction of the number of possible themes or subcategories than it is possible to account for, we grouped the articles based on the dominance criteria.When these recurring themes had been classified individually, we compared our categorisations.Those subcategories that were sizeable enough to form their own category were listed as such, whereas smaller unconventional units were lumped together as broader categories or placed among the leftovers classified as 'Other'.In those instances that the two researchers categorized articles differently from each other -which amounted to approximately 20% of the total sample -we read the articles again and came to a joint conclusion about the most appropriate way to label them (cf.Taylor, 2001).This method can be seen as a strategy to increase the reliability of the categorisation process, as the nomenclature effects are controlled for by collegial means There are some obvious limitations to this study, in terms of our selection of only journal articles, journals published in English, as well as in certain geographical locations. 3However, we do not claim that our analysis is fully generalizable in its findings in relation to the field as a whole, especially since other modes of publication (that are not journals) and other publication languages (that are not English) are left out.However, by selecting journals positioned as key ones within the field from three different continents and picking out those articles that have been most cited between 2005-2012, we hope that our analysis will provide a description that, to some extent, could be generalizable to the field.

Findings: Content analysis and bibliographics based on citations
In the following section, we present the results from our analysis of the 57 most cited articles by drawing from the questions raised under the two overarching areas of interest sketched above.Firstly, we will address the issues pertaining to the content of these articles.Secondly, we will direct attention to the social and institutional structures of the contributing authors, i.e. the bibliographics of the most cited authors in the field.The findings will be divided into two different tables, after which we will embark on a discussion of what we see as our most interesting results.

Content of research
Table 2. Number and share of top-cited articles in three adult educational journals, categories related to content.
Table 2 provides data on how the top-cited research was carried out.As for the method deployed, qualitative research clearly dominates, representing 62 percent.If we also include purely conceptual papers and theoretically oriented research reviews the dominos of the qualitative paradigm becomes even more pronounced (88%).That means that only 12 percent (n=7) of the top-cited articles used quantitative methods or mixed methods to reach their conclusions.Four of these studies used quantitative approaches, while three of them drew on mixed-method approaches.Three out of the four quantitative articles were published in the AEQ, perhaps indicating that this publication outlet is the major one for quantitative research within the three journals studied.Looking closer at these three articles, we see that the first one is based on an attitudinal survey of traditional and non-traditional undergraduate students' motivation and interest (Bye, Pushkar & Conway, 2007), the second is a comparative survey of institutional barriers to participation in adult education in various countries (Rubenson & Desjardin, 2009) and the third is a study of the educational effects of online instruction for police officers (Donavant, 2009).The fourth quantitative study that made it into our sample was published in IJLE and focuses on older learners' motivation for participation in online learning (Mulenga & Liang, 2008).Taken together, these articles represent a small number of quantitative contributions to the field that have been given recognition based on citation measurements.Whether this is due to the relative lack of quantitative articles submitted, the profile of the journals, or the limited impact of those quantitative studies carried out, is a question beyond the scope of this study.
Looking more closely at the vast number of qualitative studies, we can see how interviews and interviews in combination with observations dominate, together representing almost half (46%) of the top-cited articles.For example, English (2006) published an interview study in AEQ focusing on learning within feminist organizations.Another interview study published in SICE used a combination of interviews and observations to study recognition of prior learning practices (Breier, 2005).Lastly, we can see how 27 percent of all top-cited articles are conceptual papers or review articles.As an example of this sub-genre we can mention an argumentative paper calling for a social transformation approach to lifelong learning published in IJLE where the author draws on a plenitude of scholars to make her point (Rogers, 2006).The research reviews were typically qualitative in nature, but we separated these from the other research methods as they amounted to a rather sizeable and distinguishable research strategy of their own. 4 Continuing with the how-question of research and what theories are mobilised in these articles, we can see how three theoretical perspectives dominate and represent more than half (54%) of all the top-cited articles: critical pedagogy, post-structuralism and socio-cultural theory.Our working definition of socio-cultural theory includes authors who focus on activity theory, socio-cognitive or situated perspectives on learning.For example, in an article in SICE, Mason (2006), draws on her own experiences as a tutor in online master's courses and on literature on adult learning when discussing three technologies used in online teaching (blogging, learning object and e-portfolios).References to Wenger and social learning theories are one part of the argument pursued.In another article, published in AEQ, O 'Donnell and Tobbel (2007) analyse adult students' transition to higher education.Drawing on Wenger's community of practice concept, in combination with other theories, they focus on adults' transition in terms of participation, learning and identity.
In mapping out the articles within the critical pedagogy paradigm, we have included contributions that draw from Marxism, critical theory, feminism, social movement learning, and post-colonial perspectives.Looking more closely at some of the articles, Tisdell and Thompson (2007) study adult educators' consumption of media, drawing on critical media literacy to problematize educational equity with the aim of creating transformative learning processes and alternative counter-hegemonic narratives.In another article, published in SICE, Forrester (2005) argues for a reshaping of learning within the trade unions in the UK, a movement he argues has to move from an 'employability framework' to one informed by virtues of 'democratic citizenship'.
As for the category labelled post-structuralism, we include a series of articles that draw on Foucault, actor-network theory, science and technology studies and post-structuralist feminism.Although these articles sometimes share certain features with critical pedagogy, they also differ in important regards.Normative claims are here much less conspicuously outlined.Often the author avoids them altogether as any search for essence, and causality and foundational norms are looked upon with suspicion.One example is an article in AEQ where St. Pierre (2006) critically addresses the concept of scientific research and how it has emerged as a policy term as of late.The author draws from various other authors associated with post-structuralism such as Butler, Spivak and Foucault in order to conclude that it is impossible to separate methodology from epistemology, which is why, she says, adult education researchers need to engage with epistemologies that are not their own.
Focusing instead on the where question, the context of research, we find that schools and universities represent the most common empirical location (30%), while work and workplaces are the second most common (17.5%).ICT (16%) as well as nonformal education locations (14%) are also quite common.
To sum up our findings so far, we have found that only a small share of the topcited articles in recent years have been using quantitative research methods.Instead, the methodological strategies adopted by leading adult education scholars seem heavily tilted towards qualitative research approaches in general, and interview studies, in particular.Theoretically, we have found that the field is dominated by three overarching approaches which share certain internal family-resemblances.These were grouped as socio-cultural theory, critical pedagogy and post-structuralism.Schools and universities as well as workplaces and ICT are the most common empirical locations.
In order to deepen our understanding of who produces this knowledge and who has managed to reach our sample of the most cited scholars in the three journals, we will turn to the bibliographic data on institutional affiliations, academic position, as well as demographic variables such as gender and country of origin.

Authorship
Table 3. Number and share of top-cited articles in three adult education journals, categories related to authorship.
Table 3 provides bibliographic and institutional data on the authors that contributed to the 57 most cited articles within the rather limited time-frame of our study.The results show a distinctive pattern when it comes to the geographical aspects, which is a topic that has been dealt with in one of our previous publications (Fejes & Nylander, 2014).Suffice to say here, among the top-cited scholars the anglophone world dominates within the research field, representing 88% of all top-cited articles (authors from the USA, UK, Canada and Australia), while the rest of the world represents as few as 12%.Between the major anglophone countries we find that United Kingdom has most of the well-cited contributions, with 30% of all top-citations, followed by researchers with institutional affiliations with Australia, the USA and Canada, which each assembled around 20% of the total share.The rest of the world has a modest share in this 'leaguetable' of adult education citations, amounting to no more than 12%, largely due to some well-cited contributions in IJE. 5  Beyond geopolitical patterns of dominance, we find that the well-cited authors in this period are placed predominantly in departments and research units specializing in adult education and lifelong learning (42%).Apart from departments that could be immediately connected to adult education specialties, there were another 37 percent of researchers listed in education departments of various sorts.Among these, most had unspecified positions with regards to education departments, whereas others belonged to units focusing on research in higher education or educational technology.This illustrates that those who publish in the field, but who are not in an adult education department, are affiliated with departments or units quite close to adult education.Some of these departments might have research groups in adult education, but without these being described as such.Only one out of five (21%) of the top-cited contributions had a first author from another scientific discipline (social scientific fields, 12.3%, and more rarely the 'hard sciences', 5.3%).Thus, it seems that even though the adult education research field is commonly viewed as particularly heterodox and pluralistic, the majority of first authors belong to an adult education department or a department with close proximity to adult education.
Furthermore, we find that although most of the top-cited authors were established researchers at the time of their research some of these contributions were made by lecturers and PhD-students.In fact, within our sample, only one third (33%) of the topcited contributions were written by full professors, while a substantial number of papers were written by associate professors (19.2%), assistant professors (12.3%), senior lecturers (12.3%), lecturers (10.5%) and other researchers (7%).Five percent of the topcited papers (n=3) had even been written by PhD students and other authors without a PhD.
Yet, looking at the most cited papers in our sample, the authors are mainly more established researchers.The most cited paper in our sample is a review of transformative learning theory, single-authored by the US male scholar Edward Taylor (2007), associate professor in adult education, published in IJLE and cited 81 times.Taylor's contribution has an outstanding citation frequency as compared to all the other publications.The second most cited paper is one on the future of e-learning, singleauthored by a male Australian scholar John Hedberg (2006), professor in educational technology, published in SICE and cited 28 times.Top cited papers three, four and five are cited 23 times each.In an article published in AEQ, female Canadian scholar Dorothea Bye (Bye et al., 2007), PhD student in psychology (co-authored with professors), analyses motivation among non-traditional students in higher education.Male Australian scholar Stephen Billet (Billett & Pavlova, 2005), professor in vocational and adult education, with a colleague, analyses workers' development of the notion of self.In another article published in IJLE by male UK scholar Mark Olssen (2006), professor in political theory and education policy, the focus is on discourses on neoliberalism and lifelong learning.
On the overall gender patterns of the authors contributing to all the top-cited articles, we see a clear female dominance.Taken together, the female authors represent 65 percent of all top-cited articles as compared to 35 percent for their male counterparts.
In terms of number of authors, we see a relatively equal distribution between single and multiple authorship, with 51% of the articles single-authored and 49% with multiple authors.

Bibliometric trends in the field of adult education research
The aim of this article has been to analyse to what extent the field of adult education can be characterized as heterogeneous or homogeneous based on the most cited publications in some of the leading journals in the field.By means of analysing the publication patterns of content and bibliographics of the most cited authors between 2005-2012 we have provided some empirical evidence on the dominating theoretical traditions and methods deployed for attaining collegial recognition in the field, and we have given a collective portrayal of those authors who have succeeded in generating the most citations in recent years.
As our results show, the field can be seen as pluralistic or heterogeneous only in certain aspects, while being homogenous and monolithic in others.In terms of content, a clear pattern relating to method emerged, which was almost entirely qualitative in nature.Within the near total dominance of qualitative research among the top-cited articles, research based on interviews was the most popular method chosen.The field is theoretically construed as partly heterogeneous in terms of the wide array of theories and concepts used.At the same time, however, three theoretical perspectives dominate, representing more than half of the articles in the sample.Focusing on authorship, the field seems rather homogeneous on the basis of measurements of geographic site of enunciation, with a clear dominance of anglophone authors prevail.Heterogeneity, on the other hand, is visible in relation to the academic position of the authors, with a mixture of early career researchers and professors represented in our sample.In the following, we will discuss our main findings further.We will consider the dominance of qualitative studies, the dominance of three theoretical perspectives, as well as questions about authorship.

A qualitative research paradigm
With regards the research methods used, our results illustrate how there has been a near total dominance of qualitative research methods within the top-cited contributions in adult education research journals in recent years.Only four articles in our sample can be categorised as solely drawing on quantitative methods.Such results are partly in concordance with earlier research (Harris & Morrison, 2011;St. Clair, 2011;Taylor, 2001) insofar as these previous studies found that qualitative research has become more common and quantitative research less common over the years.However, despite the decrease in quantitative research, previous studies indicate that it is still quite common.Taylor (2001) for example, illustrates how quantitative and qualitative research were equally common at the end of the 1990s (see also argument by Groen & Kawalilak, 2013).A focus on the top-cited articles, however, indicates that quantitative research methods are much more endangered than previous review descriptions have been able to convey.
The difference in results might be due to our focus on a wider range and on slightly different journals compared to previous studies, or because we considered a later timeperiod than in Taylor's (2001) study.Another possible reason for the decline in quantitative papers may be that our focus has been on those articles that have been picked up and cited by others, whereas previous reviews in the field have grasped the full research output in specific journals (Taylor, 2001;Harris & Morrison, 2011).It might be that numerous quantitative studies have been published with low citation rates, and that it is the quality of these papers or the numerical (il)literacy of scholars in the field that prevents them from being cited.Regardless of the reason, our results show how qualitative research has gained a dominant role in the field.In fact, our findings suggest that not only should research be framed within a qualitative paradigm in order to be picked up and cited extensively by peers, it should also preferably focus on individuals and their narrations elaborated through interviews (46%), sometimes in combination with observations.Having all possible research strategies in mind, this is quite a remarkable outcome that calls for further discussion.
One possible explanation regarding the dominance of qualitative research might be found in the historical trends outlined above.Firstly, those adult education scholars who currently hold positions as professors have, to a large extent, shaped their academic careers during a time when qualitative research was emerging and establishing itself as a dominating trend within the field.It can be argued that there is a risk that these leading professors, who often fought hard to make qualitative research legitimate, had focused on providing more doctoral courses and supervision within the frames of a qualitative research paradigm.Thus, it could be hard for doctoral students interested in conducting quantitative studies to find suitable supervision at their institutions or receive meaningful feedback at adult educational conferences.
Secondly, quantitative methods have been important, and previously dominated adult education research in the US.Our results seem to indicate that, even though the volume of quantitative research seems to have drastically declined in the last few decades, such research is still to be found, but predominantly so in the adult education journal in the US.Three out of four of the quantitative articles in our sample are published in AEQ, and three out of four of the main authors of these articles are from North American institutions.The fourth of the main authors is from an institution in Zambia, but is affiliated with a North American university.Thus, one could argue that adult education scholars who draw on quantitative methods seem to have benefited from being located and published in North America.
Another explanation of our results might be found in the question of recruitment to doctoral studies in the field.There is a long-standing tradition in the field of recruiting students who themselves have been engaged in practices of adult education, as teachers, activists or community workers.From a life-course perspective, it is logical to assume that these students bring with them interests that relate to prior experiences of their vocational and political practices.The propensity to use qualitative research methods and to construct research objects that are intuitively recognizable (students' motivations, transformative learning experiences, pedagogic and political strategies, etc.) might therefore, at least in part, depend on the recruitment to the doctoral level.It could also be connected to a wish to produce the kind of knowledge demanded in educational and teaching programs which, according to Taylor's (2001, p. 336) diplomatic observation, has led to 'less debate about its [qualitative methods] validity as a research method'.
Another important lead in explaining the qualitative dominance might be found by reading about the aims and scopes of the three investigated journals.As the sample builds on research journals within adult education that stress the relation between theory and practice as especially important, ambitions to publish research that is deemed useful to practitioners, might also render the journals more inclined towards qualitative approaches, in which the connection to those working in the field appears more straightforward.Provided that one would like to contribute to an increase in quantitative studies, there would be a need to develop the competence to carry out such studies, either by fostering the appropriate skills within the field or by bringing in such competence from other departments or disciplines.This latter dimension seems more visible in our material, where a researcher in psychology, and another one from a department specializing in criminal justice, conducted two of the four quantitative studies in our sample.
Teamwork between scholars who are competent within different areas of conducting social scientific research is not rare and almost half of the top-cited articles in our sample were collaborative projects that involved more than one researcher.Yet, in order to be able to carry out large-scale quantitative studies empirically, there is a need for researchers active in the field to attract larger research grants.As our results clearly illustrate, very few studies are quantitative, and the qualitative studies are, to a large extent, small-scale interview studies in terms of empirical material.One possible explanation for this finding is that adult education researchers are not very successful in attracting funding for large-scale research projects in which mixed method approaches and quantitative research form an integral part.However, the lack of quantitative studies might also indicate that adult education research is not highly esteemed among research funders, or that adult education, which is often a very minor part of the educational or public system, has been ignored during this particular time period, i.e. the early 2000's Three theoretical perspectives Our results illustrate that, even though a wide array of theoretical perspectives are used among the top-cited articles, three perspectives dominate the field: socio-cultural perspectives, critical pedagogy and post-structural perspectives.This pattern indicates that broad conceptual pathways are open for the conduct of adult education research.So how come these three perspectives dominate the field?
One explanation could be related to the methodological observation that almost all articles in the sample draw on qualitative research methods.These theories are often mobilized to help explain and problematize qualitative data of various sorts, and thus authors deem these theories appropriate in relation to the choice of method.Secondly, as the bibliographic variable on institutional affiliations of the authors illustrates, the dominance of these three theoretical perspectives should probably be seen as tied to specific sites of enunciation, e.g.adult education research as it is practiced in the USA, UK, Canada and Australia.But what happens outside this anglophone world of publications, say in Korea, Germany, China, or France?Such a question can also be raised in relation to our own sample and its internal relationship.For example, the three dominating perspectives are most clearly represented in IJLE (n=10) and SICE (n=14), while they are less represented in AEQ (n=7).This could indicate that adult education research in North America, to a large extent, is still influenced by psychology, with a stronger focus on the individual rather than on sociological questions and issues of power (cf.Rubenson, 2000).Such an explanation is partly supported by our sample, since three of the articles in AEQ drew on transformative learning theory, while only one of the papers in the other two journals focused on this theory (authored by a US scholar).Transformative learning theory was developed in the US, and is very much focused on the individual and her/his cognitive dispositions.
Thirdly, the dominance of these perspectives could be related back to the history of the field.With the emergence of critical pedagogy in the 1970s, not least through the writing of Freire, issues of power came to the fore in much of the adult education research.Critical pedagogues are brought together by a 'preoccupation…with social injustice and how to transform inequitable, undemocratic, or oppressive institutions and social relations' (Burbules & Berk, 1999, p. 47).The critical inclination typically comes from researchers identifying themselves with a social cause or movement, which leads them to take on the role of spokespersons or judges who unveil the destructive disparity between the ideal and reality, between how the world really 'is' and how it ought to be (Boltanski, 2011).Critical pedagogy could thus be expected to appeal to adult education scholars who themselves come from the adult education field, bringing along a wish to conduct research that might help improve practice by focusing on issues of power.As already argued, previous practitioners and activists have been a common source of recruitment to PhD programs in adult education.Similarly, post-structuralism, being part of a critical tradition as well, offers a different way to understand power, and thus adult education.
Socio-cultural perspectives, rather than being critical or focusing on issues of power, could be viewed as descriptive.Generally, their focus is on describing how learning occurs in relation to and within socio-cultural practices, through the appropriation of language, rules, tools, etc. (cf.Wenger, 1998;Vygotsky, 1978).Such perspectives were most commonly used in our sample.Firstly, the popularity of sociocultural perspectives could be explained by wider tendencies within educational research, since socio-cultural perspectives on learning have taken an dominant position in terms of how to understand learning, at least in a European context (often through rather simple comparisons to 'cognitive perspectives' or 'behaviouristic perspectives' on learning) (cf.Säljö, 2013).Secondly, by looking in a cross-sectional way at our results, the dominance of socio-cultural perspectives can be related to the research contexts that are most common in these studies.Out of thirteen articles that we grouped as belonging to the socio-cultural perspective, five focused on the workplace, four on elearning and four on a specific educational group of people.Thus, some of the top-cited articles focused on two contexts that have emerged as important in policymaking on lifelong learning during the last decade (e-learning and workplaces).Arguably, the continuing success of socio-cultural perspectives has benefited from, and contributed to, a change in emphasis from education to learning.As argued by Fejes (2006) and Nicoll and Fejes (2011), there seems to have been a shift in adult education research from lifelong education to lifelong learning -where the focus on lifelong learning account for other learning processes than those associated with educational institutions.Thus, the increasing emphasis on workplace learning within adult education seems to have resulted in well-cited articles that extend the domains in which knowledge is seen to be acquired, reflected for instance in workplaces amounting to no less than 21% of the most cited articles as compared to the 10% that focused on the transformations of entire educational systems.With the spread of research on workplace learning, socio-cultural perspectives have been able to gain or sustain momentum during the last decade (cf.Fenwick, 2010).Yet another explanation to the dominant pattern described might be that sociocultural perspectives are very generic and broad, framed within a social constructive theoretical terrain, thus making it attractive for being taken up in a range of different versions, and contexts, offering tools for many adult education researchers in the mainstream.

Authorship
In terms of authorship, the top cited contributions to the field are homogeneous with regard to the dominance of anglophone authors.There is also a majority of authors in the sample that come from an adult education department or from a department closely related to such expertise.Furthermore, established researchers have authored most of the highly cited articles.However, there is also diversity in terms of academic titles among first authors, since both early career researchers as well as well-established researchers are represented among the sample.How can we explain such results?The first issue regarding the anglophone dominance and inherent provincialism in the field has been illustrated by several contributions in previous research (Larsson, 2010;Taylor, 2001;Harris & Morrison, 2011), and we have in another publication (Fejes & Nylander, 2014) ourselves extensively elaborated on possible explanations.In short, the three journals included in our sample, especially AEQ and SICE, are edited by anglophone scholars, and also have a majority of anglophone scholars on the editorial boards.Furthermore, the aim and scope of the journals do not indicate that they are 'international', although one of them has the word 'International' in its title (IJLE).Thus, even though scholars in many locations, through the economy of publication and citations, are forced to publish in journals indexed in the Web of Science and Scopus (cf.Fejes & Nylander, 2014), the journals in our sample (especially AEQ and SICE) are not clearly defined as international journals, i.e. they are rather national/regional journals and thus might not encourage submissions from other locations.Furthermore, as the journals publish in English, those with English as a native tongue are at a great advantage, as non-native speakers have to reconceptualise their research in another language, as well as to another audience.
The result that most authors are located in adult education departments or departments closely related to adult education might not be considered as very surprising, as we have focused primarily on the contributions from three adult education journals that have been picked up and cited by peers.However, it might be important to stress this finding anyhow.Firstly, this result indicates that the field is rather homogeneous as for its institutional embeddedness, i.e. the majority of contributions come from within the field, rather than from altogether other departments.This might cast doubt on the claim that the field is inherently transdisciplinary or pluralistic.Secondly, the finding suggests that the most cited authors often rely on a university infrastructure in which the topic of adult education and lifelong learning has become an institutionalized research speciality.To a fairly large extent the field is dominated by scholars who are embedded in research units and departments where seminars, graduate schools and teacher training programs all point towards proficiency in adult education or lifelong learning.
There is wide representation of authors in different stages of their career among the sample, thus indicating that the top cited contributions to this field are not clearly correlated to the researchers' hierarchical position within academia.One could even speculate that the chances of attracting high number of citations are more dependent on the authors' site of enunciation rather than academic title.However, turning to those five papers that have attracted the most citations within our sample, only one of them is authored by a doctoral student (together with established authors), and all of them are authored by anglophone scholars, indicating that there might be a rather intricate correlation between site of enunciation, title and citations.Another common denominator among these five articles is that they were all published in the early phase of our study (2005)(2006)(2007) which is hardly surprising as the aggregated cited half-life within education more broadly, is 8.3 years (Larsson, 2009).Thus, it takes several years before a paper can have been expected to attract a large number of citations.
One of the five papers (Taylor, 2007) stands out, with an outstanding citation frequency consisting of 81 citations (the second most cited paper has "only" 28 citations).Such a citation frequency could probably be explained by several factors.Firstly, the article is an literature review and discussion about the development of transformative learning theory, a kind of text that could be argued to increase the chances of receiving citations.Secondly, transformative learning theory is one of the more successful theories developed within the field of adult education as such.So even if Larsson (2010) illustrates that there is a low citation relation between different adult education journals, in the case of transformative learning theory, a community has emerged within the field.The theory is especially common in articles published in AEQ, probably because the theory was originally developed in the US.Thirdly, the paper is published by a US scholar in the UK journal IJLE, thus providing an opportunity to offer an adult education theory developed in the US to a wider international audience.Limiting the publication on transformative learning theory to AEQ would probably only keep the debate within North America since AEQ is construed as a national/regional journal rather than an international one (Fejes & Nylander, 2014).As citation rates between adult education journals are generally low (Larsson, 2010), IJLE thus emerges as a strategic output for an overview of an adult education theory developed within the US.

Concluding notes
In sum, our findings question all too generalized statements about the field of research on adult education and learning.Statements about the dynamics of this field need to be refined and nuanced through empirical investigations such as the one we have just carried out.There is a risk that we, as adult education scholars who publish in adult education journals as well as read them, take our own set of assumptions of the field to be true.Even though such assumptions are important and inevitable, they need to be complemented with systematic empirical inquiries in order to further the discussions on what the field is and how it might develop.
The main concern that we have raised in this paper relates to the questions of who and what are allowed to enter and are worth citing in three leading academic journals.Our result partly concurs with Rubenson's (2000, p. 5) statement concerning the field before 2000, in that it suggests a 'preoccupation with abstract theory building'.One quarter of the articles in our sample was written in a form that is either purely conceptual or aimed at providing research reviews.Among the empirical contributions, many of them draw on a rather limited amount of data, often in the form of a few interviews.
Speculating further from our results, could it be that the chances of being published and well-cited increase if one is a professor, or at least co-writes with a professor, and/or is located at a university in either the US, UK, Canada or Australia, and/or conducts qualitative studies, preferably interviews, and/or uses socio-cultural perspectives, critical pedagogy or post-structuralism as theoretical perspectives?If so, it raises several questions.To what extent do the three dominating theoretical traditions and the qualitative paradigm in the field enable or hinder the emergence of new knowledge?To what extent are established networks of adult education research based on proximity and familiarity with these theoretical approaches?To what extent do relations with certain networks and/or certain scholars in the field enable or hinder entrance into these publication venues and/or affect the chances of high citation rates?These questions are important, not only in terms of reflecting upon where the field 'is', how it might develop, what is valued as worthy of citing and what might be excluded, but also for doctoral students and early career researchers in order to help them reflect on their own position in the field and on the choices they may need to make to increase their chances to enter these publication venues.Read from a more heretical point of view, our study may also help researchers who strive to renew this research field.Arguably there is no better way to do that than to know the history and dominating relations of the field to date.
In line with our findings, a series of questions for future studies might be raised.Firstly, are there any correlations between authorship and content as we have described?For example, both our own result and previous studies (e.g. Taylor, 2001) indicate a dominance of female authorship in the field.Does this in any way correlate with the dominance of qualitative studies and approaches that take the views and narratives of people as their starting point?Are there ways to conduct adult education research critically, while still building on statistical methods?Secondly, what would our results be if we drew on data from other geographical sites and included altogether different language regimes?Would the image of the field perhaps look entirely different if we included other sources of data in our analysis, such as books and book chapters, or conference proceedings?And to what extent do field-specific assets that authors have accumulated in their previous track records of articles, books, keynotes and editorial position, affect the propensity of other adult educational scholars to cite and make reference to their work?Notes 1 See, for instance "A bibliometric view of Scientometrics" by Olle Persson, retrieved 2014-12-18: http://www8.umu.se/inforsk/scientometrics/ 2 A more detailed analysis focusing on the geography of authorship is available in Fejes and Nylander (2014). 3See Fejes and Nylander (2014) for a more elaborate discussion on the limitations of bibliometrics for adult education research. 4Other qualitative methods represent 16% of all articles.In this category we have sorted qualitative studies where the method does not fall under the conventional procedures such as statistical analysis, interviews or ethnographic observations.For example, in a study of recognition of prior learning within higher education, teacher voices are used to exemplify the authors' argument, but without it being clear from where these voices come from methodologically (Armsby, Costley & Garnett, 2006). 5For a full list of top-cited articles, see Appendix 1.

Table 1 .
The bibliometric analysis divided up by content and authorship