Introduction

The open access movement arose in the early 21st century in response to the emergence of digital publishing, the eruption of a crisis in journal publishing, the relentless increase in the price of subscriptions, the monopoly of the largest publishing houses and the paradoxical model of traditional scientific communication whereby the producer of information is also the consumer. This movement advocates free online access to scientific literature in order to disseminate knowledge rapidly and widely, and has received support from governments and institutions funding research. For example, research projects funded via the European Commission’s Horizon 2020 Framework Programme must provide open access to the results obtained. Similarly, Article 37 of the Spanish Law on Science, Technology and Innovation (Law 14/2011) advises researchers to deposit publications resulting from funded research projects in open access repositories.

The open access initiative proposes two routes to achieve open access. The first is the "gold" open access route, whereby researchers publish their studies in open access journals to ensure that these are immediately available to everyone. In this model, most publishers charge publication costs or an article processing charge paid by the authors via the bodies funding their research. The second is the "green" open access route, whereby digital copies of pre- or post-print versions of publications are deposited in subject-based or institutional open repositories, where they can be consulted freely, albeit in some cases after a waiting period established by the publishers (Francke, et al., 2017).

According to one of the best-known definitions, these repositories are not mere document deposits but rather “a set of services that a university offers to the members of its community for the management and dissemination of digital materials created by the institution and its community members” (Lynch, 2003). Therefore, they should be considered complex information systems that require a technological infrastructure, qualified staff and a number of services to be useful and efficient. Some of the proposed definitions of the institutional repository (Lynch, 2003; Lynch and Lippincott, 2005; Bailey Jr., 2006; Harnard, 2005; Markey, et al., 2007; McDowell, 2007; Suber, 2007) evidence the importance of ensuring that these provide researchers with dynamic, quality services to facilitate the management, evaluation, dissemination and preservation of their publications and thus extend their use and impact.

Despite the recent increase in the number of repositories in academic institutions (Pinfield, et al., 2014; Asadi, et al., 2019) and their usefulness as a means to disseminate academic publications, numerous studies have reported that institutional repositories are under-utilised. The reasons cited include their low perceived usefulness (David and Connolly, 2007; Nicholas, et al., 2012; Tmava and Miksa, 2017) and the existence of more attractive alternatives for disseminating scientific publications, such as academic social networking sites (Borrego, 2017; Rovira, et al., 2019) or thematic repositories (Robinson-García et al., 2020).

Considering that institutional repositories are complex information systems that require attractive and useful services and that the use of them could be improved if they were to provide value-added services that researchers and academics found useful and attractive, this work is focused on value-added services provided by institutional repositories. Specifically, the aim of the present study is to analyse the presence and characteristics of value-added services offered by institutional repositories in Spanish public universities.

Related work

Under-utilisation of institutional repositories

Numerous studies have asked authors about the use they make of repositories or have compared authors’ scientific production with the publications deposited in their institutions’ repositories, and have found that few university and research centre staff regularly deposit their publications in institutional repositories. In a study on use of the University of Cornell repository, David and Connolly (2007) found that few teaching staff participated in the development of the institutional repository, and that many displayed scant knowledge of how to self-archive their publications and were poorly motivated to do so. In a survey of more than 3,000 researchers from various European countries, Creaser, et al. (2010) found that only half of them had deposited a scientific journal article in their institutional repository in the previous five years.

As regards the percentage of publications deposited in institutional repositories, most studies have focused exclusively on articles in scientific journals and their figures vary considerably. Björk, et al. (2009) studied the percentage of journal articles published worldwide in 2006 that were available via Green Open Access. Their estimates showed that open access copies were only available for 11.3% of articles and only 5% of articles were deposited in institutional repositories. A similar study analysing data on publications indexed in Web of Science and Scopus in 2009 found that 11.9% articles were available via 'Green' open access, of which 24% were deposited in institutional repositories, 43% were deposited in subject-based repositories and the remaining 33% in another type of webpage (Björk, et al., 2010). Subsequent studies have reported higher rates of deposit in repositories, but in general these continue to be relatively low. A study conducted in Sweden found that only 9.6% of scientific articles published in that country in 2011 had been deposited in institutional repositories (Fathli et al., 2014). A study conducted within the framework of the European research project, PASTEUR4OA, reported that only 15.5% of articles published in Europe and indexed in Web of Science during the period 2011-2013 had been deposited in an institutional repository (Swan, et al., 2015).

This under-utilisation of institutional repositories has been analysed in numerous studies in an attempt to determine the reasons for the low rate of deposit by researchers and to propose measures to increase repository use. The most frequent reasons given for not depositing work in institutional repositories are copyright infringement worries (Foster and Gibbons, 2005; Nicholas, et al., 2012; Cullen and Chawner, 2011; Björk, et al., 2014; David and Connolly, 2007), lack of time to perform a task considered tedious and time-consuming (Foster and Gibbons, 2005; Nicholas, et al., 2012; Cullen and Chawner, 2011; Björk, et al., 2014; Abrizah, et al., 2015), lack of knowledge about the deposit procedure (Cullen and Chawner, 2011; Abrizah, et al., 2015), lack of motivation since repositories are not considered useful (Nicholas, et al., 2012; Joo, et al. 2019; David and Connolly, 2007; Ten Holter, 2020) and fear of plagiarism (Nicholas, et al., 2012; David and Connolly, 2007).

Another of the primary reasons for the scant use of institutional repositories as a means to disseminate scientific publications is the existence of other alternatives, such as personal webpages, digital subject-based archives and above all, academic social networking sites. In recent years, the latter have positioned themselves as the preferred means for teaching and research staff to raise the visibility of their work because they are perceived as being more user-friendly with better features, offering more rapid dissemination of information, greater visibility and a wider range of resources (Borrego, 2017; Rovira, et al., 2019). Thus, some studies have found that researchers disseminate more publications via academic social networking sites such as ResearchGate than via institutional repositories (Eva and Wiebe, 2019; Scott and Swanepoel, 2018; Borrego, 2017).

Institutional repositories in Spain

The number of open access institutional repositories in Spain has increased steadily since the creation of the country’s first repository in 2001, and they now exist in many universities, research centres and other academic institutions. Various studies have analysed this upwards trend in institutional repositories and open access in Spain (Melero, 2008; Melero, et al., 2009; Barrueco and García Testal, 2009; Ferreras Fernández, 2018).

An associated institutional policy, whether in terms of a directive or a recommendation, has been one of the keys to success for these repositories, since it implies a commitment on the part of the university or research centre to facilitate and ensure self-archiving, access and preservation of scientific documents. According to the report entitled The state of open access in Spanish universities published by the Spanish University Library Network Open Access Subgroup (REBIUN... 2017), the number of Spanish universities with open access policies increased by nearly 30% between 2012 and 2017, and more than half of higher education institutions in Spain now have a recommendation model. Besides institutional declarations, there is also a legal and regulatory framework that guarantees the existence of open access digital archives because authors whose research receives public funding are required to deposit their resulting publications in open access repositories. Institutional policies and regulations alike must observe publishers’ intellectual property and self-archiving policies, promote the integrity of metadata entered by the authors or delegated third parties and increase the visibility of scientific production by including repositories in search engines and directories.

In Spain, the report on the state of open access in Spanish universities (Casal Reyes, et al., 2013) indicated that in general, few documents were deposited in repositories. Only 21% of repositories housed more than 10,000 records and the total number of documents deposited in the 50 existing Spanish university repositories was 482,126. A 2016 report by the FECYT (Spanish Foundation for Science and Technology) revealed that in Spain, only 9% of articles resulting from Spanish Research, Development & innovation projects and indexed in Web of Science were deposited in open access institutional repositories. Similarly, Borrego (2016) found that only 14.4% of articles published in 2012 resulting from publicly funded research in Spain and indexed in Web of Science were available in institutional repositories. Meanwhile, in an analysis of publication data from 2012 to 2014, Melero, et al. (2018) found evidence of wide variability in Spanish universities as regards rates of deposit in institutional repositories, which ranged between 1% and 61%.

One of the objectives pursued in order to develop and consolidate university and research centre repositories is to ensure they meet the principles of quality that characterise robust services. Several studies have analysed the criteria for certification, the different evaluation frameworks and the various factors of success for institutional repositories (Shearer, 2003; Westell, 2006; McKay, 2007; OCLC, 2007; Swan, 2007; Thibodeau, 2007; Deng and Li, 2008; Kim and Kim, 2008; Yong and Hyun, 2008; Markey, et al., 2009; Barrueco Cruz, et al., 2010; Casella, 2010; Cullen and Chawner, 2010; Tripathi and Jeevan, 2011; Azorín Millaruelo, et al., 2015; Lagzian and Mee-Chin Wee, 2015; Serrano Vicente, 2017). Their results can be summarised as follows:

• The existence of institutional directives.
• Use of suitable technological resources and software platform for the repository.
• Accessibility and usability of the system interface.
• Integration of the service in the institution’s general plan and with the surrounding community.
• The existence of strategies to promote, measure, fund and preserve the repository.
• The provision of a good service in terms of availability, reliability, ease of use and user guidance tailored to their needs. Promotion of user engagement and awareness.
• Substantial incorporation of content, taking into account the percentage of researchers who deposit documents and the different departments and research groups they belong to.
• Availability of the full texts of stored documents.
• The existence of value-added services such as CVS files, statistics, bibliography services, author identifiers and profiles, guides, FAQs, controlled vocabularies and notifications.
• The existence of specialist, full-time staff responsible for management and coordination, including delegated third-party archiving as one of their functions.
• Careful development of metadata standards that facilitate interoperability and visibility in national and international directories, search engines and service providers.

Other initiatives include parameters for evaluating institutional repositories that have served as a model for other projects. The DINI (Deutsche Initiative für Netzwerkinformation E.V.), developed by German universities, has designed a quality assurance certificate for repositories that assesses basic criteria that must be met now and others to meet in the future. The DINI recommendations for repositories include raising the visibility of the service through statistics, providing advice to authors and ensuring the security, integrity and long-term availability of documents. The DINI certificate acts as a seal of quality of institutional repositories and constitutes an instrument for their evaluation and improvement (Deutsche Initiative für Netzwerkinformation, 2011). Meanwhile, the DRIVER (Digital Repository Vision for European Research) project is another international initiative that proposes a series of criteria that must be met by open access research archives in order to be considered quality services. These criteria primarily concern text resources and metadata, particularly in terms of content interoperability and visibility (DRIVER, 2008).

As indicated in the literature, the inclusion of value-added services in institutional repositories is one of the factors that substantially contribute to improving this type of academic resource. Baessa, et al. (2015) have emphasised the importance of incorporating ORCID author identifiers in repositories as a key element to ensure the quality of data on researchers’ scientific production and facilitate visualisation of their profiles in the repository’s author lists or indexes. They also suggested the inclusion of metrics and altmetrics for use by researchers and integration with the CRIS (Current Research Information System) or institutional research system as a means to obtain indicators to inform decisions and improve open access repository quality. Scherer (2016) has highlighted the need to include other value-added services in institutional repositories, focusing on three aspects. The first of these concerns tools for providing guidance to authors and researchers on copyright and the agreements they have signed with the journals publishing their work, and for depositing permitted versions in the repository. The second concerns the mechanisms established by the institution for the deposit of documents, which is usually delegated third-party archiving by the library. Although the percentage of self-archiving is very low and third-party archiving the norm, it would be useful if instructions were included somewhere visible on the repository website describing a simple procedure for authors to self-archive their work. The third refers to marketing services provided by open access repositories to promote widespread dissemination of content. Scherer (2016) places particular emphasis on the use of social networking sites to obtain low-cost support. In a 2017 report, COAR (Confederation of Open Access Repositories) defined the new functions of next-generation open access repositories, which included new technologies and value-added services for users such as the inclusion of content reviews and comments, modern applications for social networking sites, controlled vocabularies and better workflows to facilitate interoperability with other types of repository and open access aggregators.

Method

To meet the study objectives, we first identified the study population, which consisted of all institutional repositories in Spanish public universities. According to data supplied by the Spanish Ministry of Education, Culture and Sport, the Spanish university system comprises 84 universities, of which 34 are private and 50 are public. Two of these latter belong to a special category: the Menendez Pelayo International University (UIMP) and the International University of Andalusia (UNIA), which only offer postgraduate and doctoral programmes and specialisation courses. We searched the websites of each of the 50 public universities for evidence of an institutional repository, and found that 48 of them had one and two did not. These two universities are the UIMP, a special category university, and the University of La Rioja, which does not have a repository as such (this university is the producer of Dialnet, a well-known database in Spain, but it cannot be considered an institutional repository).

Then, from September to October 2017, we analysed the 48 institutional repositories in Spanish public universities (Appendix I), collecting data on their main characteristics and their value-added services. A value-added service in a repository was defined as a service that was not absolutely necessary for repository operation but which served to render it more useful and attractive to users.

Data collection instrument: the checklist

Information on the repository

This dimension referred to the provision of information about the repository itself for internal users (the people who deposited documents) and external users (anyone who wished to view the documents deposited) regarding its purpose and deposit policies and offering an overview of its content. Our rationale for including this dimension was that better knowledge of the repository would favour its use, especially by internal users. The items comprising this dimension were as follows:

• Description of the repository. Basic information describing the scope, objectives and institutional affiliation of the repository.
• Information on open access and copyright. General information intended to guide and educate users about the open access movement and copyright.
• Institutional policy. Specification of official institutional policies and regulations regarding open access and the deposit of documents in the institutional repository.
• Repository policy. Specification of repository regulations, indicating who can deposit, what kinds of material can be deposited and under what conditions and explaining the approval period, deadlines and modifications.
• Contact details. Possibility for users to contact repository staff to clarify any doubts or questions. For this item, information was collected on the different channels offered by the repository: telephone, email, chat, a form or others.
• General repository statistics. General statistical information on the documents deposited in the repository. For this item, specific data were collected on the statistics provided: number of documents deposited, downloads, views and most frequently downloaded or visited documents.
• General statistics by section. Statistical information on the documents deposited in different repository communities (subjects, faculties, etc.). For this item, specific data were collected on the statistics provided: number of documents deposited, downloads, views and most frequently downloaded or visited documents.
• Most recently added documents. Section of the repository that allows users to see the most recently deposited documents in the repository.

Information on records

This dimension encompassed information additional to the descriptive metadata (Dublin Core or similar) that comprised the document records. We considered that this information could encourage repository use by researchers, since it would allow them and their institution to monitor the impact of the deposited documents. The items included in this dimension were: download statistics; view statistics; provenance of downloads; provenance of views; link to author profiles (in the repository itself, in ORCID, Google Scholar or on academic social networking sites); information on citations; information from other metrics (PlumX, Altmetric, etc.); and Creative Commons licences for the deposited documents.

Instructions for use and dissemination

This dimension was related to value-added services aimed at facilitating repository use by internal and external users alike, as well as promoting and disseminating the deposited documents. The items included were as follows:

• Instructions on how to deposit documents. Detailed information on the procedure for depositing documents in the repository, enabling users to do this independently.
• Delegated third-party archiving service. On request, repository staff deposit documents on behalf of the authors.
• Query instructions. Information for users on how to search and use repository documents.
• Notification system. Information sent to users about new documents that have been deposited in the repository, based on past queries.
• Export of records. Possibility of exporting information on bibliographic records to other formats so that they can be used by users according to their possibilities (reference managers, CSV files, etc.). Information was collected on export formats.
• Repository profile on social networking sites. Presence of a repository profile on social networking sites to publicise its existence. We recorded the social networking sites on which the repository had a profile.
• Listed on aggregators. Information was collected on whether the documents deposited in the repository were listed on an aggregator, thus increasing the likelihood of being viewed and cited, and also whether, as a means to incentivise deposit, this information was publicised.
• Indexed in Google Scholar. Information was collected on whether the documents deposited in the repository were indexed in Google Scholar, thus increasing the likelihood of being viewed and cited, and also whether, as a means to incentivise deposit, this information was publicised.
• Possibility of sharing records on social networking sites. Possibility, based on an icon, of sharing and disseminating records on social networking sites. We recorded the social networking sites on which records could be shared.

Data coding and analysis

The checklist was completed on an Excel spreadsheet, first coding the existence or not of each of the value-added services, its location within the repository (on the home page, in a dedicated section or in a more general section), and in the case of some items, as described in the previous section and in Appendix II, any other, more specific characteristic or modality of the value-added service.

Once all the information had been collected for each of the repositories included in the present study, we conducted a descriptive analysis of frequencies and percentages at global level and for each of the dimensions comprising the checklist. In order to identify possible factors that could explain the number of value-added services offered in each repository, we gathered information about the size of the university in terms of number of faculty, budget and staff of each university library and analysed the correlations of these variables and the number of value-added services. These data are available from the website of the Spanish Network of University Libraries: REBIUN.

In addition, we generated a graphical representation of the similarity between the 48 repositories analysed, according to the various value-added services they offered, assuming that two repositories were very similar, and therefore very close graphically, when they offered the same value-added services. To this end, we reduced the dimensions by means of multidimensional scaling using the PROXSCAL program in SPSS v. 24, based on a matrix of the Euclidean distance between variables. This yielded the coordinates to graphically represent the repositories analysed in two dimensions.

Results

Our results show that there was wide variability between the repositories as regards the number of value-added services they offered, which ranged between 5 and 23 out of a possible 25. However, in general, the repositories analysed provided a high number of value-added services, with a mean of 15.7 per repository and a median of 16. It should be noted that 35 (77%) of the 48 repositories included in our study offered at least half the value-added services analysed. Figure 1 shows the number of value-added services provided by each repository, together with the contribution of each dimension to the total number. As can be seen, the dimensions related to information on the repository and to instructions for use and dissemination contributed most value-added services.

Surprisingly, the number of value-added services offered by the repositories analysed is not related with the three variables included in the study. The Pearson’s correlation analysis showed positive values, but very low, close to zero and therefore very insignificant in the three cases: in the case of the number of library staff, the value of the Pearson correlation coefficient was 0.03, it was 0.04 regarding the budget and in the case of the size of the university it was 0.05. Therefore, from the results of the correlation analysis, it cannot be inferred that in these university repositories there is a direct relationship between the number of value-added services and the library budget, size or number of staff.

Figure 2 shows the similarity of the repositories analysed according to the value-added services they shared in common. The size of the bubbles corresponds to the number of value-added services provided by each repository. As can be seen, no two repositories were 100% similar, and those which presented most similarity were the repositories with the highest number of value-added services (for example, UAM, UC3M and UPV), which differed by two or three. The remaining repositories primarily differed from each other in the presence or absence of a series of value-added services related to general statistics on the repository and the records, which presented the greatest variability. As we can see in the right part of the graph, it is the repositories with the least value-added services that are more dispersed and those that have the least in common. A general reading of the graph would indicate that there are no clear patterns when it comes to offering value-added services and that each repository offers those it considers most appropiate. In other words, there are no clearly defined groups of repositories that offer the same or very similar value-added servicies.

The value-added services also presented wide variability: some showed a high frequency, being offered by most of the repositories analysed, whereas others showed a much lower frequency and were only provided by a few. As can be seen in Figure 3, almost all the repositories were indexed in Google Scholar or listed on aggregators and provided a repository description, whereas only a few gave information on the provenance of downloads or had a profile on social networking sites. In general, we found that most of the value-added services (18 out of 25) were offered by over half of the repositories, but the services corresponding to the dimension of information on records were less frequent than those from the other two dimensions. Below, we present a detailed analysis of each of the dimensions considered.

Information on the repository

Descriptive information about the repository itself was the dimension which obtained the best results, with a mean of 5.65 value-added services out of a possible eight per repository (70.3%) and with seven services offered by over half of the repositories. Figure 4 shows the percentage of repositories offering the value-added services analysed.

A description of the objectives, scope and institutional affiliation of the repository was the most frequent value-added service from this dimension, being provided by 47 of the 48 repositories analysed. Although this could be interpreted as a good result, we believe that such information is fundamental and should be included in all repositories, since this is their mission statement, indicating their function and purpose. In addition, this information should be clear and visible, enabling any user to locate it rapidly and intuitively. In 74% of the repositories analysed, this information appeared on the home page, while in another 21.3% it was given in a dedicated section entitled “Description of the repository”, “What is it?” or something similar and in 4.3% of cases it was necessary to navigate within the repository to find it.

Some 79.2% of the repositories analysed provided information about open access and copyright. In 10.53% of cases, this appeared on the home page, while in another 48.94% it was located in a dedicated section and in 19.15% of cases it was contained in a more general section such as FAQs or Help. While this is not an essential service, it provides an important added value since many people are still uncertain about open access and legal aspects related to the deposit of documents in repositories (Nicholas, et al., 2012; Cullen and Chawner, 2011; Björk, et al., 2014).

Only 52.1% of the repositories gave information on university policies and regulations concerning open access and the deposit of documents in the repository: 12% appeared on the home page, 68% in a dedicated section and 20% in another more general section. This information is of particular interest to researchers, since they must observe these guidelines when disseminating their research results. Regardless of whether each university provides this information elsewhere, we believe that it should also be noted or signposted by the repository given its relevance to the deposit of documents.

Surprisingly, information on the repository’s policies and regulations was only available in 35.4% of cases, appearing in a dedicated section in 47% of cases and in another more general section in 41%. This was the least frequent value-added service from this dimension, despite the fact that inclusion of this information would endow a repository with transparency and help users understand how it worked, which we believe would increase its use and perceived value by research and teaching staff.

Another surprising result was that 16.7 per cent of the repositories did not provide contact details. This is common practice on any webpage and is highly recommended in an information service that seeks user participation and interaction, since this type of support enables users to request clarification. The 40 repositories that offered this service did so via email, and some also offered alternative channels such as a web form (62.5%) or a telephone number (32.5%).

In relation to the statistics provided by repositories on their content and use, 64.6% gave general statistics and a 58.3% gave statistics on the main sections or communities. Statistics thus varied between general repository statistics and statistics on the main sections. The most common among the former were: most frequently downloaded documents (96.8%); visits (58.1%); downloads (35.5%); and total number of documents (22.6%). Among the latter, the most frequent were: visits (89.3%); number of documents (46.4%); downloads (17.9%); and most frequently downloaded documents (3.6%). In addition, 91.7% of the repositories indicated the most recently added documents.

Information on records

The dimension related to information on each record was the one that obtained the worst results. The mean number of value-added services from this dimension was 4.15 out of a possible 8 (51.9%), and some repositories only offered one of these services. Figure 5 shows that half of the services included in this dimension were offered by less than 40% of the repositories analysed, while those most frequently offered were provided by barely 80% of the repositories.

Most of the repositories provided document statistics, although they did not all offer the same kind. Thus, statistics on downloads (81.3%), views (79.2%) and provenance of views (77.1%) were available in most repositories, whereas data on the provenance of downloads were only offered by 8.3%. In general, our results can be considered positive, since most repositories allowed authors to monitor use of their publications. This value-added service is available on platforms such as Academia.edu or ResearchGate and would encourage authors to deposit their documents because it allows them to see that their publications gain visibility from being in the repository.

The other value-added service corresponding to this dimension which most repositories offered was the inclusion of Creative Commons licences for the deposited documents, observed in 81.3% of the repositories. This function allows authors to specify the conditions under which they wish to disseminate their publications and informs readers of the use they can make of the documents.

The three remaining services, offered by very few repositories, were highlighted by Baessa, et al. (2015) as being especially useful to enable the institution to monitor its researchers’ scientific production and its impact. The first of these is a link to authors’ academic profiles on other platforms hosting their scientific production, and this was provided by 20.8% of the repositories. The author profiles most frequently linked to by the 10 repositories that offered this service were from ORCID (8), Dialnet (3), ResearchGate (1), Google Scholar (1) and Scopus (1). The second is information on document citations, which was available in 18 (37.5%) repositories, and the most frequently used sources were Google Scholar (16), Microsoft Academic (4), Scopus (4) and Web of Science (1). The third of these value-added services is a link to other metrics that quantify the impact of publications, and this was available in 29.2% of the repositories. All the repositories that provided this service gave a link to Altmetric and two of them also gave a link to PlumX.

Instructions for use and dissemination

The mean number of value-added services corresponding to this dimension was 5.96 out of a possible 9 (66.2%), and there was marked variability in the frequency of these services in the repositories, as can be seen in Figure 6.

The two value-added services related to the deposit of documents differed in frequency in the repositories. While instructions for deposit (self-archiving) were given in 77.1% of cases, usually in a dedicated section or in the form of help in the data entry application, only 39.6% of the repositories provided information on or mentioned the possibility of depositing a document via a delegated third party. It should be noted that 16.7% of the repositories did not give either kind of information on how to deposit documents. Although it is highly likely that information on how to deposit can be obtained elsewhere or the repository could be contacted for instructions, we believe that this should be provided by the repository itself, especially given that the specialist literature indicates that many authors do not deposit their publications because they are not sure how to do so (Cullen and Chawner, 2011; Abrizah, et al., 2015).

As regards value-added services for users seeking information, 56.3% of the repositories gave instructions on how to perform a query, usually in a section entitled “Help” or “FAQs”; meanwhile, 72.9% provided a notifications service, mainly by email or RSS, 64.6% permitted users to share records on social networking sites (100% of these on Twitter, 96.8% on Facebook and 87% on other sites) and 72.9% permitted users to export records to reference managers such as Mendeley (57.1%), Refworks (40%) and Endnote (17.1%), or to a standard format such as Bibtex (11.4%), RIS (5.7%) and XML (2.9%).

Regarding value-added services that enable greater dissemination of the repository and its contents, only 18.8% of the repositories had a profile on social networking sites, primarily on Facebook and Twitter and sometimes on YouTube, Instagram or Slideshare. In addition, the content of most repositories was listed on aggregators (97.9%) and indexed in Google Scholar (95.8%): however, this information was only publicised as an incentive by 61.7% and 43.5% of the repositories, respectively.

Conclusions

The results obtained in this study provide an overview of the frequency and characteristics of value-added services offered by public university repositories in Spain. The most outstanding conclusions in this regard can be summarized in the following points:

• In general, the results were not disappointing, since 70% of the repositories offered more than half of the value-added services analysed. However, we found wide variability between repositories, whereby some offered the vast majority of the value-added services analysed whereas others only provided a few of them. This finding suggests that some repositories are committed to giving every facility to users while others are merely content to be document deposits that do not cater to users.
• There were a number of value-added services offered by almost all repositories, many of which corresponded to the dimension of “Information on the repository”. These services could be considered basic and were the first to be provided by repositories and the easiest to implement.
• The least frequent services corresponded to information on records and instructions for use and dissemination of repository content. These were the most recently included services in the repositories and are related to statistics, social networking sites, author or researcher identifier systems, citations and alternative metrics. These services are available on other platforms such as academic social networking sites, and are very attractive to and in high demand by researchers. Consequently, we believe that it is important to include them, together with any other service offered by other platforms that users find helpful.

Institutional repositories are a crucial element in the development of open access, as well as being an excellent means for researchers to disseminate their work and monitor use of their publications. Consequently, repositories should not be designed as mere deposits of documents but should also provide a range of services and features to fulfil their function correctly and ensure use by researchers. They therefore need to be dynamic, constantly updated platforms in order to meet changing user demands and preferences and be perceived by users as useful and attractive tools. The under-utilisation of institutional repositories mentioned in the literature would be reduced if they were to provide quality value-added services, because many of those that we analysed are directly related to the provision of information (the dimension of 'Information on the repository') and add value to the repositories (the other two dimensions). Repositories could address users' lack of knowledge by providing sufficient information on all these questions, and the inclusion of value-added services related to citations, downloads, views, author profiles, indexing in Google Scholar and social networking sites could render the repositories more useful and attractive.

Limitations and future research

The main limitation of this study is that data were obtained via external observation of the repositories and we only considered a series of value-added services selected in accordance with a literature review. In order to shed further light on the importance of value-added services in repositories, it would be interesting to complement this study by conducting future research that extends the results reported here. For example, researchers or those responsible for the repositories could be asked directly about their perceptions of each of the value-added services. Another more ambitious goal for future research would be to determine whether the inclusion of these value-added services (or others) leads to a greater use of the repository in terms of queries and deposits alike. We believe that further research of this nature is required, but including other emerging value-added services such as those offered by some of the repositories analysed here (information on query devices and search terms, assessment of records, labelling of records, APIs, applications for mobile applications, download statistics in CSV files, etc.), in order to determine trends in value-added services in repositories.

Based on the data collected in this research, we have been unable to identify what factors determine whether a repository offers a greater or lesser number of value-added services. We have seen that there is no direct relationship between this number and the size of the universities, the budget of the library, or the staff that work in it. However, we consider that perhaps, the number of people dedicated exclusively to management and maintenance of the repository could be a factor that does influence the number of value-added services offered. Thus, we have seen how two of the repositories with the most value-added services, Alicante and Carlos III Universities, had a team of people dedicated exclusively to their management, while, for example, the University of Leon repository, with very few value-added services, does not have a specific staff, but rather some librarians who mainly perform other functions are in charge of its management. In this research we have not had access to data of full-time staff dedicated to the repository of all universities and therefore we have not been able to establish a relationship between both variables, but it could be considered as a line of future research should accurate information be available/when accurate information is available.

Acknowledgements

The authors appreciate the copy-editor, and the anonymous reviewers for their valuable comments and suggestions.

About the authors

How to cite this paper

Appendices

Appendix I: Universities and repositories included in the study

Appendix II: Data collection checklist