Publishing in library and information science journals: The success of less experienced researchers

1. Introduction

The aphorism publish or perish describes the fact, that in order to succeed in academia, researchers need to publish academic works. Thus, less experienced researchers such as early career researchers are under great pressure to publish [1] and they seem to have responded to the pressure by increasing their productivity measured as the number of published articles or reviews [2]. Productivity in the early phases of the research career is important as early career publishing is correlated with future research excellence. A high number of early career publications have been found to be positively correlated with later research productivity [3–9]. The importance of selecting a target journal is described recently as follows:

Researchers attempt to choose the most appropriate platform to highlight their research work so that their work gets published in good order, read, and referred. Hence, selecting an appropriate journal is the most vital task for them. [10]

Advice on how to get published is available in plentiful to help less experienced researchers to ‘publish more papers in top academic journals’ [11]. Researchers are recommended to select the target journal early in the writing process [12] and generally how to publish before even beginning the project [13]. However, publication strategies are often not explicitly included in doctoral student training taught at the university level, and doctoral students often do not discuss this with their supervisor [14]. Furthermore, the lack of formal training on the various aspects of the publishing process is increasingly being acknowledged and addressed in dedicated programmes [15].

Recent studies confirm that choosing an appropriate journal is a complex decision and that many factors are considered when deciding on a publication outlet [16–18]. However, selecting a target journal is often based on incomplete information regarding the journal’s ability to support the authors in their pursuit of disseminating their research and advancing their academic career [19]. One factor is particularly difficult for the less experienced researcher to assess. Authors take their prior experience with the journal into consideration [20–22], and indeed, authors tend to publish in a relatively small number of journals [23,24]. However, for less experienced researchers, past submission experience is per definition limited. They may, of course, ask more experienced colleagues for advice. However, the journals within a field develop over time and the experience of older colleagues may not be transferable. It may therefore be valuable for less experienced researcher to gain insights into publishing success for other less experienced researchers in the journals within their field.

This study aims to explore the publication patterns of less experienced researchers in core library and information science journals. More specifically, the goal is to investigate the share of less experienced researchers in each journal to analyse changes over time as well as differences across journals. Providing such an overlook of the journals in our field will hopefully assist authors with limited past submission experience choosing the optimal journal for publishing their research.

In the next section, we discuss the concept of journal hierarchies and how it may relate to publication success. Then, we define the entities of the study and the methods applied. Results are presented separately, and the implications of our findings are finally discussed.

2. Choosing a publication outlet

When selecting where to publish ones research, a researcher will typically take the issue of journal quality into consideration. This issue is closely tied to the so-called order hierarchy, defining a value-based ranking of a set of elements under investigation [25]. Although order hierarchies are always value-based, in academia, the basic values of order hierarchies are far from always explicated. Instead, they are often tacitly implied. Kuhn [26] was among the first to address the concept of tacit knowledge in academia. According to him, tacit knowledge covers the common rules that create the implicit basis for the paradigms, which define scientific standards. They are typically acquired through apprenticeship, and tradition is passed down through learning by doing, by submitting to authority or by trusting the experienced practitioner. In every scientific discipline, there is an order hierarchy of journals [27]: ‘Those at the top are the most prestigious, are perceived as having the highest quality standards and are difficult to publish in’.

Although several national and international rankings of journals exist, the order hierarchy of academic journals is primarily tacit knowledge. Researchers simply come to know the structure of this hierarchy over time: ‘Interestingly, as time goes on, most authors discover their ecological niche in this hierarchy and seldom venture far away from it’ [27].

The order hierarchy of journals may consequently play a role when authors determine where to publish. The common assumption (e.g. Garvey [27] and Heintzelman and Nocetti [28]) is that authors will try to maximise professional rewards, and will therefore aim at publishing in the most prestigious journals. If rejected, they will continue down the order hierarchy until acceptance is reached. Based on a study of British biochemists, Gordon [29] found that authors selected journals primarily on the basis of the journal audiences, rather than by potential reward. He therefore concluded that the reward-seeking model of selection behaviour found little or no support in his sample. Luukkonen [30] came to the opposite conclusion. Based on interviews with researchers from four university departments in the fields of zoology, biomedicine, and automation and control technology, he concluded that researchers placed emphasis on both the reward and communication functions of publishing. Yet, there were differences between fields. The publishing behaviour of biomedical scientists fitted the maximisation of professional rewards strategy best. Calcagno et al. [31] presented results from a large-scale survey of the submission process, covering 923 scientific journals from the biological sciences (2006–2008) showing among other things that resubmissions tend to flow from journals with higher impact factors to journals with lower impact factors. Other studies [32–34] have also shown that generally, authors are attracted by journals with higher impact factors when deciding where to submit their manuscripts.

Summing up, the journal hierarchies are generally taken into account when authors determine where to publish. However, inexperienced and early career researchers may not necessarily fit the general patterns. Existing studies confirm that the publication strategies of early career researchers focus on publishing the highest quantity of journal articles in the best quality journals as possible being aware of the importance of metrics such as the journal impact factor (JIF) in evaluations [35,36]. Many early career researchers use the JIF to identify the best quality journals within their field [35,36].

There are studies indicating that less experienced researchers are using journal selection criteria differently from their more experienced colleagues. Researchers engaged with questionable journals tend to be young and inexperienced when they publish in these journals [37,38], review for them [39] and cite them [40,41]. Consequently, these studies suggest that experience makes a difference when deciding for a publication outlet. Furthermore, the study by Søreide and Winter finds that age correlated with valued importance of the journal reputation but was inversely correlated with fast-track decision, time from acceptance to paper in e-press, journal previously published on the topic, author previously published in the journal and option to suggest or exclude reviewers. Consequently, increasing age means that there is a decline in the percentage of responders holding the impact factor (IF) as the most important factor when deciding on a journal. However, increasing age means relying more on past publishing success [20]. Consequently, the less experienced researchers value the importance of the criteria differently from more experienced researchers. However, contrary to the findings by Søreide and Winter, Cheung [21] finds that the stage of a researcher’s career does not affect the importance attached to a particular factor by an author whereas having editorial experience or being affiliated to a higher ranked university have an impact on the appraisal of each factor.

3. Methods

For the purpose of this study, we need to define a less experienced researcher. Studies of publication activity and citation impact of specific researcher groups may include chronological age, career age, professional categories or academic experience.

Less experienced researchers may be defined based on career or chronological age. Early career researchers represent the transition stage between PhD and senior academic positions and can be defined as being less than 10 years since PhD completion [42]. A definition used in other studies as well (e.g. Wanelik et al. [43]). A similar approach is recently used, although a shorter time frame is being used as the definition of early career researchers is up to 7 years after completing the PhD [44]. Furthermore, early career researchers can be defined as academic researchers up to 5 years after completing the PhD while allowing for extension where significant life factors have limited the research conducted due to, for example, illness or parental leave [45]. Chronological age may also be part of the definition. Combining chronological age with professional categories, a recent study defines early career researchers as researchers less than 36 years old who are currently doing a doctorate or have received their doctorate but are not in established or tenured positions [46].

Professional categories may also be used to define less experienced researchers. A recent study examines three categories of early career researchers: doctoral students, postdoctoral fellows and research staff on fixed-term contracts [47]. Similarly, analyses have been made on doctoral candidates and postdocs [48]. An early career researcher can also be defined as anyone engaged in research who is not recognised as an independent leader of a research group which includes undergraduates, graduates, postdoctoral researchers and junior research assistants [49]. Combining professional categories with career age, early career researchers can be defined as researchers who earned a doctoral degree less than 10 years prior to applying or were employed in colleges or universities as assistant professors for less than 5 years [50].

Finally, less experienced researchers may be defined based on publication records. Studies of publication activity and citation impact may be defined through academic experience of specific researcher groups. Publication records can also form the basis of the definition [51]. In this study, a similar approach is used as it characterises the less experienced researchers: early career researchers, practitioners in the field that rarely publish as well as unproductive researchers. The next step is to define the less experienced researcher bibliometrically and a threshold needs to be set, however, turning to the existing literature we see that these definitions are typically made ad hoc. In a large-scale study of research careers, four career stages are defined as follows: junior (<5 years since first publication), early career (≥5 and <15 years since first publication), mid-career (≥15 and <30 years since first publication) and late-career (≥30 years since first publication) [52]. The authors note that the time periods were selected for convenience and that any other division could have been selected. Another example is a study of stratification in the global higher education research community the author defines full-timers as researchers who authored or co-authored at least five papers in a set of elite journals and argues that they are defined arbitrarily [53].

To allow for analyses of the consequences of defining the less experienced researchers in this study, we work with two different definitions. Consequently, we define less experienced researcher using two different definitions:

In this study, the order of authors is not considered. The order of the authors of a scholarly publication can take place using several different approaches: by amount of contribution, alphabetical order, multiple first author or multiple last author, by seniority or reverse seniority, by raffling or lottery system, by negotiation or mutual understanding [54]. Many of these approaches indicate that author order can be tied to experience which is also confirmed empirically [52,55]. The order author in a publication co-authored by a less experienced researcher would therefore be tied to this. Consequently, we measure the number of publications and co-authors, but we are not able to determine the role of each co-author or the relative sizes of their contributions.

Next, we determine the publications of library and information science journals to be included. For the purpose of this study, we draw on an existing list including library and information science journals categorised as information science journals and library automation journals [56]. The list or modified versions of it has formed the basis of several studies of library and information science (recent examples include Zhao and Strotmann [57], Hoiu et al. [58] and Nicolaisen and Frandsen [59]). The list is adjusted according to title changes. Discontinued journal titles are excluded, and we therefore have 10 journals included in this study. From these journals, we include all articles and reviews from 2001 to 2020 with author information in the data set (Table 1) – a total 14,612 of articles and reviews. Due to co-authorships, the total number of authors analysed is more than twice the amount of publications (36,417). Table 1 shows that the library automation journals (Data Technologies and Applications, Electronic Library, Information Technology and Libraries and Library Resources and Technical Services) published fewer articles from 2001 to 2020 than the information science journals (Information Processing & Management, Journal of Documentation, Journal of Information Science, Journal of the Association for Information Science and Technology, Library and Information Science Research and Scientometrics) did.

OPEN IN VIEWER

Table 1.

Sample.

Journal	Articles and reviews	Authors
Library automation journals	2551	5206
Data Technologies and Applications	564	1257
Electronic Library	1242	2614
Information Technology and Libraries	422	774
Library Resources and Technical Services	323	561
Information science journals	12,061	31,211
Information Processing & Management	1656	5040
Journal of Documentation	1182	2148
Journal of Information Science	1103	2808
Journal of the Association for Information Science and Technology	3144	8173
Library and Information Science Research	606	1248
Scientometrics	4370	11,794

Finally, we define the variables of the study in the following way.

Author experience is defined as the total number of articles and reviews published by every author in the years leading up to but not including the publication year of the specific publication in our data set. We define two indicators of author experience as follows:

We use Scopus to identify all articles and reviews published by a researcher and more specifically, Scopus Author Identifier forms the basis of collecting data on author experience. Scopus Author Identifier is reliable for bibliometric analyses although it is not without limitations [60,61]. It should be noted that the publication record of each author is determined on the basis of all their publications and is thus not restricted to the included 10 journals in this study. An author of a single publication in our data set may thus be highly experienced due to many publications in other journals. The experience of an author is based on the number of articles and reviews they have published in the years preceding the publication in our data set. An author may appear several times in our data set with different experience levels if they have published articles or reviews in the 10 journals with different publication years. Consequently, the author experience is determined for each author of every paper which in this study is the publication history of 36,417 authors. An example is an article published in 2015 in Journal of Documentation by three authors. These authors have published a total of 10, 6 and 20 articles and reviews, respectively, up until 2014. Absolute author experience is then determined as 10 + 6 + 20 = 36, whereas normalised author experience is determined as (10 + 6 + 20)/3 = 12. We use Scopus for our data collection and the data are analysed using graphic overviews showing the development over time using moving averages.

4. Results

Generally, the number of previously published articles and reviews by the authors of papers in the 10 journals in this study has increased over time. Figure 1 provides an overview of the average number of previous articles and reviews by the authors of publications in our data set. Library automation journals are marked with a dotted line. The experience of authors in Scientometrics can serve as an example to illustrate how to read the figure: in 2002, the articles and reviews published in Scientometrics were authored by one or more authors that in total had an average number of previously published articles and reviews of approximately 40. The unit of analysis is the paper and the experience is thus measured as an average per paper to avoid bias from different patterns of authorship. In 2020, the absolute experience of authors of articles and reviews in Scientometrics had increased to approximately 80. Consequently, in 2020, each article and review in Scientometrics was authored by an author or a team of authors with twice as many previously published articles and reviews as in 2001. However, not all the journals in the data set show the same tendency towards increased author experience. The four library automation journals publish papers by authors with little increase in absolute publication experience. However, the six information science journals are characterised by noticeable increase over time in absolute author experience. However, this development is to some extent caused by an increase in the average number of authors per publication. To further examine this, the author experience needs to be normalised by the number of authors per publication.

OPEN IN VIEWER

Figure 1.

Absolute author experience per article from 2001 to 2020 in 10 library and information science journals. The four library automation journals are marked with a dotted line.

Figure 2 shows the development over time in author experience having normalised to correct for differences in the average number of authors across time and journals. Again, Scientometrics can serve as an example as to how the figure is read. In 2001, the average publication experience for authors in Scientometrics was about 20 articles and reviews. This means that articles and reviews in Scientometrics were published by authors who on average had published 20 articles or reviews each in the previous years. In 2020, the normalised author experience had increased to about 30. After normalisation, the considerable increases in author experience over time are now smaller and in most cases negligible. In addition, from the figure, we observe that the library automation journals in this study generally publish articles and reviews by authors with less experience than the information science journals.

OPEN IN VIEWER

Figure 2.

Normalised author experience per article from 2001 to 2020 in 10 library and information science journals. The four library automation journals are marked with a dotted line.

We now look at the share of less experienced researchers in each journal. Figure 3 presents the share of less experienced researchers over time. Again, the individual article or review is the unit of analysis, and we analyse the experience of the authors of the paper having controlled for the number of authors. Here, a paper by less experienced researchers is defined as a paper published by one or more authors having controlled for the number of authors. A less experienced author is an author who has published less than five articles and reviews in the years before but not including the publication year of the publication in our data set. We normalise by the number of authors, and therefore, we define a less experienced researcher publication as one being published by one or more authors with less than five publications on average. Using the definition of a less experienced author as an author who has published less than three articles and reviews similar although more unstable results due to the smaller numbers, and therefore, we only present the results of the analysis using the definition of a less experienced author as an author who has published less than five articles and reviews in the years before but not including the publication year of the publication in our data set.

OPEN IN VIEWER

Figure 3.

Share of less experienced researchers (<5 previously published articles and reviews) from 2001 to 2020 in 10 library and information science journals. The four library automation journals are marked with a dotted line.

Figure 3 depicts a picture of great variety when it comes to the past publishing success of the authors of articles and journals in library and information science journals. In 2020, the shares range from 10% to 75%. Figure 3 also shows that the library journals are publishing considerably more publications by less experienced researchers than the information science journals do. In addition, the figure shows that the share of publications by less experienced researchers has decreased over time. In the information science journals, the shares of less experienced researcher publications were between 30% and 40% in 2001. In 2020, the shares of less experienced researcher papers have decreased and only two journals are above 20%. Due to the small number of publications each year, it is more difficult to see the decrease in the library automation journals. However, in general, the shares of less experienced researcher papers are also decreasing in the library automation journals.

5. Discussion and conclusion

First, in this study we have defined the less experienced researcher using two different definitions. LER-3 is defined as a researcher who has authored or co-authored less than three articles and reviews, whereas LER-5 is defined as a researcher who has authored or co-authored less than five articles and reviews. These definitions are typically made ad hoc in bibliometric studies and alternative definitions are not necessarily tested but, in this case, we can see that the latter LER-5 produces more stable results as that definition includes more researchers and the results are reported using this definition.

The findings of this study show that there are considerable differences between library and information science journals when it comes to the average past publication success by the authors publishing in these journals. On average, the library automation journals are characterised by publishing articles and reviews by authors with less experience than the information science journals. The library automation journals publish considerably more publications by less experienced researchers than the information science journals. This marked difference between the information science and library automation journals has also been found in other bibliometric studies. There seems to be a strong relationship between research topics and technological developments [62] which may have led to the decrease in research interest over the years. Library automation appeared in the 1960s and was of increasing interest in the 1970s but is now somewhat out-of-date in the research literature [63]. Consequently, these four journals may represent a more applied approach and therefore attract practitioners from the field with less experience. Furthermore, these four journals seem to be characterised by a lower degree of specialisation measured by bibliographic coupling, and therefore possibly easier to enter for an unexperienced author [64].

Furthermore, the results show that the share of publications by less experienced researchers has generally decreased over time in the 10 journals examined here. These results need to be considered in the light of the publish or perish paradigm. Since the production of publications is key to a research career and academic advancement, the production of publications during doctoral studies is increasingly expected and constitutes a part of the socialisation process [65]. In pursuit of getting published, the focus is on traditional, formal publications as they remain indispensable for institutional reputation setting [36,66] although early career researchers are more probably to be able to reach out and increase their influence [67].

The results of this study are relevant to the modelling of the choice of publication outlet. The existing frameworks for explaining the choice of publication outlet generally do not include career stage as a parameter for the choice of journal. However, we find age and career stage to be included in several studies modelling the choice of publication outlet. In the model developed by Björk and Holmström, the authors argue that the factors that influence the decision are weighted according to individual preferences which may include career stage although this is not explained in detail [68]. Wong et al. [69] use a multi-objective optimisation analysis to determine the set of journal submission pathways that balance the objectives for three stages of a researcher’s career and argues that the trade-offs between objectives even for co-authors of the same manuscript may not always be aligned due to their different career stages. They find that optimal solutions may be partitioned into different classes of strategies, and these are attractive at different career stages.

Research careers and academic advancements are tied to publishing, and, as a consequence, less experienced researchers may struggle to publish the highest quantity using the best publication outlets for their work. Identifying and selecting the optimal journals is a complex decision, and many factors play a role. One factor is particularly difficult for the less experienced researcher: experience with publishing in the field’s journals. For less experienced researchers, past submission experience is per definition limited. This study shows that generally publications in these 10 library and information science journals are being published by authors with an increasing publishing experience and fewer papers are being published by author teams with little experience in total.