In labyrinths of digital text: use of Web literature by faculty from two Croatian universities

Introduction

Faculty members now have large amounts of online texts available for their research and teaching duties. Many of these texts are not accessible without payment. However, today it is also possible to freely access a large amount of online scientific literature. A faculty member could access literature through open access databases, repositories or pirate digital libraries. Researchers and lecturers have always been active in exchanging literature, so that is another way to acquire literature. In this study we will explore the extent to which Croatian faculty from universities of Split and Rijeka use different Web sources of scientific literature, as well as their practices and attitudes related to availability and use of Web-based scientific papers.

Web sources of literature

In the year 2006 the total number of reviewed papers published in scientific journals was about 1.35 million. Of this number, 19.4% of the works were freely available on Web (Björk, Roos and Lauri, 2009). Therefore, it is not surprising that work practices of many researchers and lecturers have been greatly changed. Important elements of the digital information environment are more comprehensive access to scientific information, easier publication of scientific research results and simpler data exchange (Vrana, 2010). It is not unusual that many researchers are exploring these changes in the digital information environment. One important aspect are changes in scholars information behaviour brought by Web information sources. There are a number of studies about that topic (Abels, Liebscher and Denman, 1996; Centre for Information Behaviour... 2008; Connaway, 2007; Connaway and Dickey, 2010; Eason, Richardson and Yu, 2000; Prabha, Connaway, Olszewski and Jenkins, 2007; Talja, 2002; Talja and Maula, 2003; Tenopir, 2003). Information sources are represented in various important models of information seeking. Information seeking is described as a 'process of either discovering patterns or filling in gaps in patterns previously recognized' (Case, 2007, p. 80). Information seeking is a term encompassed by broader term information behaviour, which is defined by Tom Wilson as:

the totality of human behaviour in relation to sources and channels of information, including both active and passive information seeking, and information use. Thus, it includes face-to-face communication with others, as well as the passive reception of information as in, for example, watching television advertisements, without any intention to act on the information given (as cited in Case, 2007, p. 81).

Information sources are very important aspect of information seeking and behaviour, as we can see from Wilson’s definition. Therefore, information sources are also represented in various models of information seeking, which Case presented in his book (2007). In his model of information seeking, Krikelas divided information sources into main types: internal and external. Internal information sources are information seeker’s memory and direct observations. External information sources are the information seeker’s direct, interpersonal contact and recorded literature. In Wilson’s first model of information seeking from 1981 barriers for access to information sources are represented. In a digital information environment, these barriers could be high prices of the subscription to scientific databases or complicated interfaces of information systems. Wilson’s second model from 1996 invoked explicit theories to explain questions, such as, why some sources of information are used more than others. Leckie, Pettigrew and Sylvain’s model of information seeking also included information sources as one of its main components (Case, 2007). In the next section, more detailed data from some studies about scholars’ use of Web information sources will be presented.

In a study about Slovenian scientists it was established that they use Web search engines more than official sources, they prefer e-content and digital tools, and they demand direct access to digital texts (Vilar, Juznic and Bartol, 2015). Web-based scientific literature used for writing PhD dissertations is a topic researched by a group of Lithuanian scientists. Fifty-seven percent of texts cited in these dissertations are freely available on Web (Grigas, Juzeniene and Velickaite, 2017). According to a study written by Orduña-Malea, Ayllón, Martín-Martín and López-Cózar, Google scholar enables searching of about 160 million documents related to scientific work (2017). Many of these documents are freely available on the Web, so it is not unusual that for many scientists Google scholar is the first place to search for literature. For example, laboratory bio-scientists and nano-scientists from South Korea primarily use Google, journal websites and databases such as PubMed and SciFinder for gathering literature (Kwon, 2017). The researchers from two Finnish research units in molecular medicine intensively use Web portals, especially the portal PubMed for searching scientific facts. However, to locate general Web pages, research groups, methods and tools, they use primarily Google (Roos, Kumpulainen and Järvelin, 2008). In a study about Israeli LIS students’ preferences for sources of information, a combination of two data collection methods was used: study participants wrote a personal diary about their search activities and answered a questionnaire (Bronstein, 2010). Participants mainly used networked sources followed by human sources. Another important finding was that the ‘main preference criterion for networked sources was accessibility which was perceived as easiness of use, time saving, language and physical proximity’. Bronstein contextualized this finding with various ‘studies that have found that information sources that are easier to use are perceived as more accessible and will be used more frequently’. In a study about the use of data from OAPEN Library, another author came to similar conclusions. The majority of book downloads in this library came from users who do not use the Website of OAPEN Library. It seems that the highest use of OAPEN books is ensured by ‘making the metadata available in the user's systems, the infrastructure used on a daily basis’ (Snijder, 2014). In a study about the use of The Finnish National Electronic Library it was found that keyword searching was the most important access method. The roles of browsing, chaining or obtaining material from colleagues were diminishing (Vakkari and Talja, 2006). One important aspect of access to information sources is exchange of information between scientists. Many important Web portals with scientific literature are based on the exchange of information between scientists, for example, ResearchGate or the pirate digital library SciHub. Namely, the creation of SciHub is supported by scientists around the world who gave to the creators of this portal their credentials for access to commercial databases. In that way many journal papers were acquired by SciHub (Karaganis, 2018). This cooperation in illegal sharing of information between scientists and SciHub creators is an example of a very risky type of information sharing. Wilson wrote a paper about the concept of information sharing and related variables: trust, risk, reward and organizational proximity (2010). Various relationships of these variables can influence the probability of information sharing. Wilson’s analysis could be helpful for the creation of research studies related to information sharing in organizations. Another study also focused on information sharing and trust. In this study, strategies of design scholars were identified for ‘assessing and creating trust in relation to information sharing’. Participants talked about criteria for the assessment of people's trustworthiness: friendship, impressions from previous interactions, rank and experience (Pilerot, 2013). In a previous study written by Duić, Konjevod and Grzunov the research goal was to explore the extent to which faculty at the University of Zagreb (Croatia) use various Web sources of scientific literature (2017). Research findings clearly indicate that Google scholar is their most popular Web source. Many Croatian scientists also use DOAJ (Directory of open access journals), ResearchGate, Academia.edu, Science Direct and Hrčak (Croatian portal). About 70% of participants strongly agree with the statements that subscriptions to databases with international scientific journals are too expensive and that all scientific papers should be freely available. Many scientists from various parts of world also share this opinion. For example, sixty major German research institutions have cancelled subscriptions for journals of publishing house Elsevier protesting against the high cost of subscriptions (Göttingen State and University Library, 2016). Elsevier is one of the scientific mega publishers that owns a large number of scientific journals. In 2013 Elsevier and other mega publishers (Springer, Taylor & Francis, Wiley-Blackwell) published more than half of all scientific papers in the world. These publishers use successfully their influence to significantly increase subscription prices to scientific journals and databases (Larivière, Haustein and Mongeon, 2015). Because of this, many countries have problems in securing the availability of scientific literature. One of them is Croatia. In a study written by A. Zubac and A. Tominac it is emphasised that Croatian higher education institutions do not have enough financial resources to pay for expensive subscriptions for printed and electronic scientific journals (2012). However, it needs to be emphasised that in Croatia there is a very good system for securing availability of Croatian scientific literature. The main backbone of this system is Web portal Hrčak that was created in 2006. The significance of portal Hrčak could be confirmed by the fact that 461 Croatian journals and 179451 scientific and professional papers were freely available on this portal (Hrčak, March 11, 2018). Open access movement and open access policy are in a great measure developed in Croatia precisely because of portal Hrčak. Finally, scientists can also access literature through scientific social networks such as ResearchGate. This social network allows opening of personal profiles of scientists, as well as putting their scientific papers on Web, so that they are freely available to anyone interested.

Pirate or shadow digital libraries

Pirate digital libraries, such as Library genesis and Sci-Hub, are another way of accessing scientific literature. These libraries contain a huge amount of digital texts which are freely available, often without consent of the copyright owners. They have many users from every part of the world (Bohannon, 2016). Pirate digital libraries with enormous numbers of books and journals has become popular in the last few years (Karaganis, 2018). These pirate digital libraries are also called shadow libraries, which Balázs Bodó defines as ‘piratical text collections which have now amassed electronic copies of millions of copyrighted works and provide access to them usually free of charge to anyone around the globe’ (Bodó, 2015, p. 75). Bodó described a development of one large and popular, pirate digital library: Library Genesis (also known as LibGen). He wrote that this library was made by Russian scientists in 2008 to integrate mostly Russian-language text collections (Bodó, 2018a). In the year 2008 the LibGen catalogue contained only 34000 items and in 2014, it contained almost 1.2 million items (Bodó, 2018b). In the year 2011 there was enormous growth of LibGen library when it integrated collection of another large pirate library: Library.nu. With this integration LibGen acquired many scientific texts in English and other world languages. In that way LibGen became interesting to scientists from every part of the world. It is a specific pirate library, because it offers its text collection for anyone to download and it also offers for download its own library infrastructure (source code of library software and catalogue). This means that anyone can take this data, duplicate it or adapt it to a new version of LibGen library with different names, different users and functionalities. Therefore, today there are many versions of LibGen around the Web. These mirror sites greatly increase the likelihood of long-term preservation of access to documents collected by LibGen (Bodó, 2018a). There are many users of LibGen from Russia and post-Soviet countries, and from developing countries such as India, and Iran. These countries have many characteristics which favour the use of pirate libraries, such as low per-capita GDP and a rapidly growing university sector. For example, in India only few universities can pay subscriptions to the major academic databases (Liang, 2018). In the year 2009, the largest academic library, the Indian Institute of Science (IISc), subscribed to more than nine thousand journals. This represents only a small fraction of the number of journals typically received by elite U.S. universities. Columbia University, for example, received 133,831 serials in 2007. The government-supported CAPES Journals Portal is a very important source of online materials for Brazilian researchers and graduate students. In the year 2015, CAPES included about 38 thousand full-text journals, available to 436 university partners (Mizukami and Reia, 2018). However, as the cost of journal access has risen, the CAPES journal budget went from about $21 million in 2004 to about $99 million in 2014. These high expenses are opening a possibility of the demise of the program or a drastic reduction in the number of subscriptions. There are also many LibGen users from developed countries (United States, Germany, UK, Netherlands, Sweden, etc.). Bodó emphasized following reasons for using of LibGen in these countries with strong economies, university and library infrastructure: high pricing of access to literature, unavailability of scholarly works in digital formats, and users who are collecting documents without the intention of reading them (Bodó, 2018b). One interesting fact is that highest number of LibGen users, in relation to population size, are found in ‘relatively poor countries at the edges of the European Union’ (Lithuania, Estonia, Greece, Latvia, Slovenia, Croatia, Macedonia, Hungary, Bulgaria, Armenia, Czech Republic, Romania, Montenegro). Many of these countries have common characteristics which favour a use of pirate libraries: highly educated populations, much weaker resources and educational infrastructure in relation to the core European countries, and obligations to implement EU educational standards. The problem is that their faculty and students often cannot find needed literature in legal libraries. That is a fertile ground for pirate libraries (Bodó, 2018b). For example, in the Polish academic community there is a high demand for books from foreign publishers, but because of financial reasons, there is not sufficient numbers of these books in libraries (Filiciak and Tarkowski, 2018). Except LibGen, another large and popular pirate library is SciHub. It was launched in 2011 by a Russian neuroscience student named Aleksandra Elbakyan (Karaganis, 2018). In 2016 this library contained fifty million papers from scientific journals and it had over 28 million downloads from all around the world over a six-month period. Faculty members from around the world supported the creation of this library by giving Elbakyan their online credentials for access to the commercial, scientific journal databases available at their universities.

Methodology

The research goal of this paper was to explore the extent to which faculty use different Web sources of scientific literature, as well as their practices and attitudes related to availability and use of scientific literature.

Research questions were:

• Which Web sources of scientific literature do faculty use and how often?
• What are the faculty's practices and attitudes related to availability and use of scientific literature?

We did not make any formal hypothesis. However, the expectations were that research results would be similar to research results that were acquired in a previous study with the same research goals, but with faculty members from other Croatian universities (Duić et al., 2017).

To acquire the insights, data were gathered through a survey by questionnaire among teaching and research employees at the faculties in two Croatian cities: Split and Rijeka. At the University of Split (http://eng.unist.hr/) the following faculties were included: Faculty of humanities and social sciences, and Faculty of science. At the University of Rijeka (https://www.uniri.hr/index.php?lang=en) the following faculties were included: Faculty of humanities and social sciences, Faculty of teacher education, Department of biotechnology, Department of physics, Department of informatics, and Department of mathematics. The questionnaire was made available on Web at the beginning of July 2017. The e-mails with the request to participate in survey were sent to 642 faculty employees with teaching and research responsibilities: 361 invitation e-mails for employees in Rijeka and 281 invitation e-mails for employees in Split. The questionnaire was completed by 108 faculty members of faculty during July and August 2017. The survey method is a very popular research method in information behaviour studies (Case, 2007). Case describes the strengths and weaknesses of this method based on the Zhang's paper from the year 2000. Questionnaires can be easily distributed by e-mail to individuals or groups with known interests. Also, an advantage of survey method is that responding can be easy. However, responses could suffer from bias favouring those who have access to the Internet and interest in computers. Also, questionnaires cannot easily capture the complexity and context of information seeking. Other research methods, such as interviews and participant observation could be better for these purposes (Case, 2007).

The questionnaire in this study had nineteen open and closed questions. Some of them contained additional questions. The questions were based on the literature review. They were also based on personal insights and experiences. Namely, the author of this paper is a lecturer and researcher at a Croatian university. This position allows him to acquire insights and experiences about the research topic.

The majority of survey questions were related to: a) the frequency of use of a certain Web source; b) participants’ attitudes related to Web sources. In the majority of questions, participants were asked to choose a grade that most precisely indicated their frequency of use of a certain Web source or their attitudes related to the use of Web sources. Participants had to choose one grade from the following range: one, two, three, four, five.

The survey was anonymous, which allowed participants to be more sincere about their practices and opinions. In this paper only some of the findings are presented. In another paper, data were analysed in relation to four variables: age and sex of study participants; field of science in which they work; and, their practices of exchanging scientific literature (Duić, 2018).

Sample

The questionnaire was completed by a total of 108 faculty employees. The total number of employees who teach and conduct research at the faculties included in research is estimated at about 640. Therefore, the response rate of about 17% was achieved. It is hard to tell if this is a high or low response rate. Some studies indicate that e-mail response rates are decreasing; for example, in one study it was established that while in the year 1986 the mean response rate was 61.5%, in the year 2000 the mean response rate was 24% (Sheehan, 2001). One of the reasons for low response rates could be a large number of invitations for participation in e-mail surveys that scientists are constantly receiving.

The majority of study participants were women (65.7%). The majority of study participants were 30 to 39 years old (42.6%). However, the survey was completed by faculty of various ages. It was completed by 35.2% of participants who work at the University of Split and 57.4% of participants who work at the University of Rijeka. The majority of study participants work in the fields of social sciences (37%), humanities (30.6%) and natural sciences (22.2%).

Results

In Table 1 we see that majority of study participants download and read scientific papers several times a week (38%). There are 23.2% participants who do it on a daily basis. This means that Web sources of scientific information are very important for faculty.

Table 1: Frequency of downloading and reading scientific papers on the Web

In Table 2 we see how often study participants send requests to text authors so that they send them their scientific papers. Almost 60% of participants never or almost never ask authors to send them their scientific papers (grades one and two). There are about 20% of participants who often and very often ask authors to send them their papers (grades four and five). About 63% of participants never or almost never ask persons from Croatia to send them scientific papers (grades one and two).

In Table 3 we see the frequency with which study participants access scientific papers through various Web portals. The most popular Web portal for accessing scientific papers is Hrčak, which is often and very often used by 50% of participants. This portal is followed by Google scholar with 44% of participants (grades four and five). A considerable number of participants are members of social networks for scientists, i.e., the Web portals ResearchGate and Academia.edu, which enable those scientists to upload their papers and share them with other members. ResearchGate is often and very often used by 42% of study participants, and Academia.edu by 27% of study participants (grades four and five).

The portal of library at the University of Rijeka (Library at UNIRI) is often and very often used by 26% of participants, while the portal of the library at the University of Split (Library at UNIST) is often and very often used by 9% of participants. Portal NSK is the portal of National and university library in Zagreb (Croatia). Only 6% of study participants often and very often use this portal (grades four and five). A similar percentage of participants often and very often use the portal DOAJ (Directory of open access journals). The Croatian portal PERO is often and very often used by 1% of participants.

Estimate on a scale from 1 to 5 to what extent do you access scientific papers through the specific Web portal (1 = never; 5 = very often; n. r. = no resp.)

In Table 4 we see how often study participants are using pirate digital libraries. Sci-Hub is often and very often used by about 20% of participants, and Library genesis by 15% of participants (grades four and five). These two information sources are the most used pirate digital libraries.

In Table 5 we see the frequency of downloading and reading scientific papers using pirate digital libraries. There are about 16% of study participants who daily or several times a week use these libraries. About 8% of participants use them several times a month and about 20% of participants use these libraries several times a year.

In Table 6 we see use frequency of three types of Web information sources: Croatian portals (Hrčak, PERO, portal NSK, etc.), international legal portals (Google scholar, ResearchGate, DOAJ, etc.), and pirate digital libraries (Sci-Hub, Library genesis, etc.). Legal international Web portals are mostly used: about 68% of study participants use these portals often or very often (grades four and five). They are followed by Croatian Web portals with about 46% of participants. Pirate digital libraries are the least used with about 24% of participants who often or very often use them.

In Table 7 we see levels of study participants' agreement with various statements. Especially interesting are findings from the last four statements. Regarding the Statement 4, we found out that only about 10% of participants almost completely or completely agree that they are content with the availability of international scientific literature available through subscription (grades four and five). The fifth statement was that teaching and research in Croatia could be significantly improved if the availability of papers is increased through subscription by the Ministry of science and education and other institutions. About 85% of participants almost completely or completely agree with this statement. About 75% to 80% of participants almost completely or completely agree with the statements that subscriptions for databases with international scientific journals are too expensive and that all scientific papers should be freely available to all interested.

Estimate on a scale from 1 to 5 to what extend do you agree with the following statements (1 = completely disagree; 5 = completely agree; n. r. = no response)

Discussion

Among many findings acquired through this research, one important finding is that many study participants very frequently use Croatian Web portals. This could be a somewhat surprising finding considering the fact that there is a amount of literature in international Web portals. Hrčak is largest portal with Croatian journals and it is the most used portal, even in comparison with the very popular Google scholar. This is strong evidence that an open access policy, accepted by journals available at Hrčak, is an extremely successful approach to dissemination of scientific papers. Financial support from the Croatian Ministry of science and education is one of the important elements which have enabled creation and development of this portal. As a model of good practice, Hrčak could be an inspiration for similar projects outside Croatia and it can serve as a model for the free dissemination of scientific literature. Also, research findings indicate that study participants from two Croatian universities are often the members of international social networks for scientists. ResearchGate is a social network which has the largest number of members from the whole world. According to this research, it has also the largest number of users in Croatia, in relation to another large social network, Academia.edu. The huge popularity of social networks is a phenomenon which would be valuable to explore more thoroughly. This type of digital information sharing between scientists is a very convenient form for exchange of ideas and literature. It would be interesting to explore the experiences of Croatian scientists regarding the use of social networks: what are their benefits, deficiencies and other characteristics.

Research findings have also indicated that pirate digital libraries are used by a considerable number of study participants. About one fifth of participants frequently uses two pirate digital libraries: Sci-Hub and Library genesis. Possible reasons for their use are indicated by the following research results: almost two thirds of study participants are not content with the availability of international scientific journals and papers in Croatia; about 80% of participants strongly agree with the statement that subscriptions for databases with international scientific journals are too expensive and that all scientific papers should be freely available to all who are interested, without subscription payment to publishers. These opinions are a clear sign that a huge majority of study participants are not supporting the system in which it is necessary to pay inappropriately high prices for subscriptions to databases with scientific journals. The payment is financed by money which was collected through taxes paid by these same researchers and lecturers. This situation has characteristics which incite some scientists to call it unfair. We found confirmation of this interpretation in sentences that some study participants wrote as their personal opinion about this situation. Some of them wrote that they use pirate digital libraries as a form of protest and that publishers have monopolies through which they exploit the free work of authors and reviewers of scientific literature.

Conclusion

Web portals in Croatia, especially the portal Hrčak, have a very important place in finding and accessing scientific literature. Therefore, these portals should continue to be financially supported by the state and various scientific institutions. The opinions of the researchers and lecturers who participated in survey indicate that there is a problem of access to scientific literature and these respondents disagree with the commercialization of scientific literature. These important problems should be addressed by the Croatian university community.

Further research could help to identify and explain in more detail the reasons and circumstances for popularity of certain Web sources. The practices and attitudes of researchers and lecturers from other Croatian universities should also be explored. Also, findings from this study could be helpful to build new and enhance older Web portals. If researchers and lecturers frequently use Google scholar and the portal Hrčak, then it would be valuable to connect some elements of these portals. For example, when somebody is searching literature in the portal Hrčak, relevant results from Google scholar could be presented in some part of the Hrčak interface. Similar integration could be made between other sources of literature (commercial and open access databases with scientific journals). A meta-search portal could be very valuable, so that users can enter search keywords in one simple interface, and get the results from various information sources (Google scholar, Hrčak, databases with journals, etc.).

In this paper there is also a literature review in which many studies were presented about researchers' and university lecturers' use of Web information sources. This review could be helpful for future studies on related topics. The paper also contains a presentation of new and controversial information sources: pirate digital libraries. Creation of pirate digital libraries has been supported by information sharing by many individuals around the world. This aspect of information sharing could be additionally explored.

Acknowledgements

This research was completed with the support of a bursary from the Information Literacy Group (ILG) of the Chartered Institute of Library and Information Professionals (CILIP). The studies were conceived and conducted in partnership with colleagues from the Scottish Information Literacy Community of Practice and Edinburgh City Libraries and Schools. The authors would like to thank Katie Swann, Cleo Jones, Ian McCracken and David Hastings for their help with the project and the pupils and headmaster, Tom Rae, at Craigmount High School in Edinburgh.

About the author

Mirko Duić is a postdoc at the University of Zadar (Croatia), Department of Information Sciences. He received his PhD in 2015 from the University of Zadar. The topic of his doctoral dissertation was The Film collections in Croatian public libraries. His research interests are in digital literacy, film in information organizations, film history, development of library collections, digital libraries, people's cultural interests (books, film, music...), and cultural diversity. He can be contacted at Department of Information Sciences, University of Zadar, Ul. Dr. Franje Tuđmana 24i, 23000 Zadar, Croatia. E-mail: miduic@unizd.hr.

How to cite this paper