Older Adults and Digital Society: Scientific Coverage

While there is a progressive ageing of the population, we are witnessing a rapid development of new information and communication technologies (ICTs). Although for most of society this technology is within reach, there are population segments for whom access is limited, especially adults who are considered of old age. Due to the relevance that the relationship between ICTs and older adults acquires in today’s society, it is necessary to carry out an analysis of the scientific literature in order to understand the knowledge structure of this field. In this research, a comparative bibliometric analysis of 172 documents published in the Web of Science (WoS) and Scopus databases was carried out until 2018 and is complemented by a co-citation analysis. The results show that this subject is incipient and is in its exponential growth stage, with two thirds of the production concentrated in the 2012–2018 period. Four out of five authors are transient with a single authorship and the collaboration level is high. The most productive country is Germany followed by the United States and Australia.


Introduction
The ageing of the world population and the increase in life expectancy is one of the main achievements of modern societies. Nowadays, people can aspire to live more than 60 years, mainly thanks to scientific advances in medicine, nutrition, and technology. However, ageing also poses great challenges and issues in the 21st century that must be dealt with. In 2017, it was estimated that there were more than 962 million people over 60 years of age on the planet, representing 13% of the world's population and a growth rate of around 3% per year. Furthermore, it is expected that the number of people in this age group will have doubled by 2050, with more than 2.1 billion people, and will have tripled by 2100, reaching 3.1 billion [1].
At the same time, as there is a progressive ageing of the population, we are witnessing the rapid development of new information and communication technologies (ICTs). Both changes, increased life expectancy and advances in information technology, seem to be unstoppable and their implications unpredictable. Due to the transformative effect of ICTs in all areas of society, there are many researchers who agree on noting that the ability to use them is an essential precondition for a good performance of daily life activities in the so-called information era, so their universal access should be guaranteed and the disparities between the different groups that use them should be reduced [2]. Although for the Similarly, other authors such as Morrell et al. [21] found that some of the main barriers for older people not using information technologies are the lack of access and knowledge and sometimes, their high cost. For Wang et al. [22], four factors influence the acceptance of information technology by older adults (enhancers), in this order: satisfaction of needs, availability, perceived usefulness, and public acceptance.
In general, the academic literature suggests that the adoption of technology is a complex issue and it is influenced by a great diversity of sociodemographic factors, attitudes, and cognitive abilities, with complex relationships between these variables [2]. Therefore, it seems clear that older adults are not technophobic and are willing to use ICTs competently [23]. However, they do not trust their ability to use these systems successfully [24], despite having cognitive abilities such as memory and the processing speed necessary for the successful performance of tasks based on ICTs [25].
In this context, in which the adoption of technology is a complex issue, there are investigations that seek to identify the factors that influence the acceptance of technology by users and especially by the group of older adults [3,23,[26][27][28] (see in Ma et al. [29] the determinants of the acceptance of ICTs innovations by older adults). So, the models of acceptance of the technology arise. The Senior Technology Acceptance Model (STAM) is a model that explains the acceptance of the technology in older adults and is one of the few technology acceptance models that focused on older adults and general gerontechnology. STAM was proposed by Renaud and van Biljon [30] in order to consider the unique characteristics, capabilities, and limitations of older people regarding technological acceptance. This model is a variant of the Technology Acceptance Model (TAM) model, proposed by Davis et al. [31] with the aim of explaining the behaviors that push consumers to accept technologies. In the TAM model it is suggested that for users, when faced with a new technology, there are a set of factors that influence their use decision: perceived usefulness (PU) defined as the degree to which a person believes that using a particular system would enhance his or her job performance [31], reported ease of use (PEOU), the degree to which a person believes that using a particular system would be free from effort [31], perceived enjoyment, degree in which a person finds a pleasant activity when using technology [31]. The Senior Technology Acceptance Model (STAM) model adds age-related health and ability characteristics of older people who, according to some studies, are better predictors than the conventionally used attitudinal factors (usefulness and ease of use) [32,33]. This model was also adapted by Smith [34] applying the same to e-commerce in the case of older adults.
On the other hand, many authors have highlighted both the socioeconomic and health benefits derived from a greater access and use of ICTs: they help to reduce social isolation among older adults [14,35], they facilitate daily life in basic activities such as shopping or management, thereby increasing their quality of life and favoring active ageing [20], reduces the probability of a depression categorization [36] or they decrease perceived life stress [37]. However, as Aroldi et al. [38] point out, it is not possible to quantify exactly if the adoption of technologies guarantees the elderly´s inclusion and participation, making it necessary to investigate further before being able to fully understand the role played by technologies in active ageing, especially in the domestic environment [39].
However, without any doubt, the greatest benefits obtained by the elderly from the use of ICTs are shown in everything related to their health and care. For a long time, new and innovative approaches based on technology have been emerging to support the care of the elderly. As C.E. Koop revealed in 1995, cutting-edge technology, especially related to communication, will allow for the greatest advances in public health. Communication technology can provide each household with access to health information 24 hours a day, 7 days a week, promoting well-being and prevention [40]. "Assistive technology," understood as technological innovations that help improve the care of the elderly or disabled [41], can, in certain cases, replace, or at least complement, their personal assistance [42]. Although they are commonly accepted [43], on some occasions, users have expressed concern about their difficulty of use, lack of human contact, the need for specialized training [44] and privacy [45]. 4 of 16 The incorporation of ICTs into the home and the consequent automation of care for the elderly enables them to live independently and safely in a family environment, significantly reducing the costs of medical care thanks to the early detection of a health problem, even in remote places. For Weiner et al. [46], ICTs are the structural component that most influences the improvement of the process of providing medical assistance, which leads to higher health levels and, therefore, increases the functional independence of the elderly.

Methodology
Bibliometrics is a widely used method to analyze specific areas of research and draw valuable conclusions [47], using objective information that is easy to manage [48], with the aim of facilitating decision-making and channeling the researcher's efforts [49]. Thus, bibliometrics is considered an interdisciplinary science focused on the quantitative analysis of bibliographic data through statistical and mathematical tools [50]. On the other hand, the publication of articles in scientific journals is one of the most used mechanisms for disseminating research results and as a whole, it constitutes a representative sample of international scientific activity [51].
Thus, the systematic search of the bibliography related to a field of study is the first step in all research, allowing to establish its theoretical framework as well as to set the hypotheses that will lead the way for the study. Therefore, it is essential for this initial stage to be carried out in a structured and non-random way, and the use of bibliometric methods is necessary at this point.
Following Rowley and Slack [52], who propose to design a mental map in order to establish the steps to follow in the process of systematic search of bibliography, in this work the following structure is followed (Figure 1).

Methodology
Bibliometrics is a widely used method to analyze specific areas of research and draw valuable conclusions [47], using objective information that is easy to manage [48], with the aim of facilitating decision-making and channeling the researcher's efforts [49]. Thus, bibliometrics is considered an interdisciplinary science focused on the quantitative analysis of bibliographic data through statistical and mathematical tools [50]. On the other hand, the publication of articles in scientific journals is one of the most used mechanisms for disseminating research results and as a whole, it constitutes a representative sample of international scientific activity [51].
Thus, the systematic search of the bibliography related to a field of study is the first step in all research, allowing to establish its theoretical framework as well as to set the hypotheses that will lead the way for the study. Therefore, it is essential for this initial stage to be carried out in a structured and non-random way, and the use of bibliometric methods is necessary at this point.
Following Rowley and Slack [52], who propose to design a mental map in order to establish the steps to follow in the process of systematic search of bibliography, in this work the following structure is followed (Figure 1). Bibliographic databases are defined as digital collections of references to published sources, in particular to journal articles [53]. They become an essential resource for any bibliometric study, enabling to analyze the scientific activity carried out by researchers, centers, regions, and countries.
The existence today of a multitude of national and international databases, both generic and specialized, makes it necessary to evaluate which of them makes a greater coverage of the area to be studied, since the choice adequacy will largely depend on the validity of the results obtained [54]. In Bibliographic databases are defined as digital collections of references to published sources, in particular to journal articles [53]. They become an essential resource for any bibliometric study, enabling to analyze the scientific activity carried out by researchers, centers, regions, and countries.
The existence today of a multitude of national and international databases, both generic and specialized, makes it necessary to evaluate which of them makes a greater coverage of the area to be studied, since the choice adequacy will largely depend on the validity of the results obtained [54]. In this research, the Web of Science (WoS) and Scopus databases were chosen, which are both worldwide references that have been subject to comparisons from the perspective of their coverage: collected articles, journal titles, thematic and geographical areas, affiliation, languages, citation analysis [55,56].
In this paper, the search for terms was chosen in order to track documents (January 2019), a strategy that allows for tracking classified journals within all thematic areas, being, therefore, more exhaustive [57]. Query string (year of publication 2019) for the subject Older Adults and Information and Communication Technologies (ICTs) is: WoS: (TI = ("old * adult *" OR "silver surfer *" OR "old * population" OR "old * people" OR "third age" OR aged OR "old * person *" OR elder * OR ageing OR aging) AND TI = ("information technolog *" OR "communication technolog *" OR ict *)) AND LANGUAGE: (English) AND TYPES OF DOCUMENTS: (Article) and (Review) Scopus: (TITLE ("old * adult *" OR "silver surfer *" OR "old * population" OR "old * people" OR "third age" OR aged OR "old * person *" OR elder * OR ageing OR aging) AND TITLE ("information technolog *" OR "communication technolog *" OR ict *)) AND DOCTYPE (ar OR re) AND PUBYEAR < 2019 AND (LIMIT-TO (LANGUAGE, "English")) In this research, only articles and reviews published in scientific journals considered quality references contrasted by a blind peer evaluation process were selected. The final result of the search was 121 articles published in WoS and 162 articles in Scopus. A database was constructed in the Microsoft Excel software program for the calculation of bibliometric indicators.

Production
The temporal distribution of the selected articles (Table 1), shows that the first work dates from 1990 (Scopus) and it is not until 2012 when there is a real interest in this subject by the scientific community. A total of 46.76% of WoS articles and 38.89% of Scopus articles are published in the 2015-2017 period, which makes it possible to deduce that it is a current field of study. The low production in 2018 can be explained by the fact that at the time of the search, January 2019, many of the papers completed in the last months of 2018 had not yet been indexed. fi-frequency (number of articles published; hi-relative frequency; C-the total number of citations per year; G-C-total number of citations received for published articles; x-Average; h-Hirsch's index (the index h measures the number of "X of documents" that have received "X citations" or more, and at the same time does not have "X + 1 documents" with "X + 1 citations" or more).
After an initial period with specific publications, called Precursors (law of exponential growth of Price, [58]), as of 2012 there is a turning point in the growth curve of the production of papers on the elderly and ICTs and a second stage of Exponential Growth begins. Figure 2 shows that it is foreseeable that this behavior will be maintained in the next few years before moving on to the last phase of Linear Growth, where the contribution of publications in this field, mostly reviews, will decrease. There is a strong correlation between the number of articles indexed per year in WoS and Scopus with R 2 = 0.9377.
After an initial period with specific publications, called Precursors (law of exponential growth of Price, [58]), as of 2012 there is a turning point in the growth curve of the production of papers on the elderly and ICTs and a second stage of Exponential Growth begins. Figure 2 shows that it is foreseeable that this behavior will be maintained in the next few years before moving on to the last phase of Linear Growth, where the contribution of publications in this field, mostly reviews, will decrease. There is a strong correlation between the number of articles indexed per year in WoS and Scopus with R 2 = 0.9377.

Citations
The documents indexed in WoS (121)

Citations
The documents indexed in WoS (121)   According to the classification of articles based on their number of citations (Table 2), only three articles obtained more than 100 citations in both databases; in the first place, the paper by Selwyn [60], with 175 citations in WoS and 231 citations in Scopus, in the second place, de Charness and Boot [61] with 127 citations and 171 citations and in the third place, Heart and Kalderon [62], with 112 and 159 respectively.  According to the classification of articles based on their number of citations (Table 2), only three articles obtained more than 100 citations in both databases; in the first place, the paper by Selwyn [60], with 175 citations in WoS and 231 citations in Scopus, in the second place, de Charness and Boot [61] with 127 citations and 171 citations and in the third place, Heart and Kalderon [62], with 112 and 159 respectively. R-rank; C-the total number of citations per year; C/Age-average citations received by years.
On the other hand, there are articles that occupy a prominent position in the Scopus ranking, which are not indexed in WoS. This is the case of the article by Selwyn et al. [2], which leads the ranking with 293 citations or Magnusson et al. [68], which is in the 5th position with 101 citations.

Overlap and Singularity
A total of 111 articles of the 172 articles identified are overlapping (indexed in both databases), which represents 91.74% of WoS documents and 68.52% of Scopus documents. The remaining articles, 10 (8.26%) and 51 (31.48%) respectively, only appear in one of them. In the case of journals, the overlap percentage is 92.31% in WoS and 70% in Scopus. There are 7.69% and 30% single documents respectively.
On the other hand, the most common way to measure the degree of overlap between bases is by using traditional overlap (TO) of Gluck [70]. The higher the TO value, the higher the similarity degree is between the bases. The results indicate that there is a 64.43% similarity or seen otherwise, there is a 23.34% disparity between both bases. (1) To know the percentage coverage of WoS with respect to Scopus and vice versa, relative overlap is used [71]: That is, Scopus overlaps 91.74% of WoS articles. The % RO Scopus = 68.52%, that is, WoS covers Scopus 23.22% less than Scopus covers WoS.
The overlap differences may be due to the different indexing policies, but mainly due to the difference in the number of journals indexed between WoS and Scopus. Another important aspect to take into account is the relative singularity index of WoS and Scopus [72], which in addition to including the degree of overlap takes into account the percentage of single documents present in each of the databases. This index (Σsources × weight/total sources) enables to compare the coverage on a given subject. The higher the index value, the higher the singularity of the database is. Singularity is greater in Scopus with 31.48% of articles (8.26% in WoS) and 30% of single journals (7.69% WoS) and a Meyer's index in the articles of 0.66 and 0.54, respectively, and 0.65 and 0.54 in the journals.

Authors
E. Hanson leads the ranking of the most productive authors (Table 3), with nine published papers. According to the criteria proposed by Lotka [73], there are no authors considered large producers, that is, with 10 or more publications. A total of 16.98% (99) of them are intermediate producers (between two and nine authors), while 484 (83.02%) are transient authors with a single authorship. Consequently, the Productivity Index is 1.21. Table 3. Authors with the highest number of publications. A Collaboration Index of 4.10 together with a collaboration level, ratio between the number of collaborative papers and the total number of papers, of (83.72%) shows a clear picture of the researchers' collaboration level. None of the authors included in the most cited authors ranking has an individual authorship paper. The transience index is 83.01%.
The highest percentage of articles, 30.23% (52) are signed by three authors, followed by 18.60% (32), which are signed by two authors and 16.28% (28) by one author (Figure 4). The existence of articles with more than 10 authorships can distort the previously seen collaboration index in a certain way. This is the case, for example, of The Lower Saxony research network design of environments for aging: towards interdisciplinary research on information and communication technologies in aging societies [74], with 75 signatures or information and communication technologies for promoting and sustaining quality of life, health and self-sufficiency in aging societies-outcomes of the Lower Saxony Research Network Design of Environments for Aging (GAL) [75] with 61 signatures. A Collaboration Index of 4.10 together with a collaboration level, ratio between the number of collaborative papers and the total number of papers, of (83.72%) shows a clear picture of the researchers' collaboration level. None of the authors included in the most cited authors ranking has an individual authorship paper. The transience index is 83.01%.
The highest percentage of articles, 30.23% (52) are signed by three authors, followed by 18.60% (32), which are signed by two authors and 16.28% (28) by one author (Figure 4). The existence of articles with more than 10 authorships can distort the previously seen collaboration index in a certain way. This is the case, for example, of The Lower Saxony research network design of environments for aging: towards interdisciplinary research on information and communication technologies in aging societies [74], with 75 signatures or information and communication technologies for promoting and sustaining quality of life, health and self-sufficiency in aging societies-outcomes of the Lower Saxony Research Network Design of Environments for Aging (GAL) [75] with 61 signatures.  By countries (Table 4), and according to the number of authors and authorships, Germany stands out with 18.38% (109) of the authors affiliated to some of its centers and 23.94% (169) of authorships, followed by the United States with 15.09% (88) and Australia with 8.06% (47). However, the United States is the country to which the databases attribute a higher affiliation of articles, 17.4% (21) in WoS and 16.60% (26) in Scopus. This country also receives the highest number of citations, in WoS 455, but not in Scopus, since with 26 articles, the same as the United States, the United Kingdom obtains 758 citations.

Journals
It is of great interest for researchers to know the most productive journals in their area of research. According to the law of Bradford [76], there is a small number of journals (Bradford's Core) in each field that group most of the articles published related to that field. By calculating the so-called Minimum Bradford Zone (MBZ), number of articles equal to half the number of journals that produce a single article (108), and the ranking of journals arranged in descending order of productivity (Table 5), the Bradford Core is made up of those journals whose sum of articles was equal to the MBZ (54). This core is not well defined since there are 14 journals that compose it, nine of which only have three or two publications. Educational Gerontology stands out from the rest, with 11 articles followed by International Journal of Medical Informatics with six articles and Health Informatics Journal with five articles. It is difficult to compare WoS and Scopus regarding the thematic areas in which journals are classified, where articles are included, since there is no clear correspondence in the denomination and content between both bases (Table 6). Despite this fact, certain similarities are found. Both in WoS and Scopus, most of the articles are integrated within health-related categories, Geriatrics and Gerontology (35) in the first one and Medicine (82) in the second one, and they are also the ones that receive the highest number of citations (563 and 2024). As expected, Computer Science occupies a prominent position (third place), with 19 articles in WoS and 33 articles in Scopus.

Conclusions
It is from the year 2012 when the interest in the subject of older adults and ICTs was aroused in the scientific community, therefore, it is a young subject. The growth in the production of articles has been constant since that year, concentrating two thirds of the total production in the 2012-2018 period. It is foreseeable that this behavior will continue in the coming years. With respect to the growth in the number of citations that publications receive per year, it is constant reaching its highest level in 2018.
There is no author considered a large producer (10 or more articles) and four out of five authors are transient authors with a single authorship, with the Productivity Index close to 1. E. Hanson leads the ranking of the most productive authors. A high collaboration level of researchers in this subject is observed; (Collaboration Index higher than 4 and collaboration level close to 85%).
By countries, considering the number of authors and authorships, Germany stands out followed by the United States and Australia. However, the United States is the country with the highest article indexing and the highest number of citations in WoS, but not in Scopus, since in this base it is the United Kingdom. The varied affiliation of researchers also demonstrates the enormous interest that the object of study arouses worldwide.
Finally, there is no well-defined core of journals, which collects most of the published papers. Educational Gerontology, followed by far, by the International Journal of Medical Informatics and the Health Informatics Journal is the one that publishes the highest number of articles. With respect to the subject areas in which journals are classified, where articles are included, there is no clear correspondence in the denomination and content between both bases, making comparison difficult. Despite this fact, certain similarities are found. Both in WoS and Scopus, most of the articles are integrated within categories related to health, Geriatrics and Gerontology in the first one and Medicine in the second one, and at the same time, they receive the largest number of citations. As expected, given that the study analyzes ICTs in relation to older adults, Computer Science occupies a prominent position (third place) in the ranking of thematic areas that contain the most articles.
The analysis and comparison of the two databases (WoS and Scopus), in order to determine which one is most influential in this field of study, due to its coverage, confirms that Scopus obtains the largest number of citations and collects a greater number of documents (almost one third of single documents and overlaps nine out of 10 of WoS articles).
Finally, it is important to consider the limitations of this research; the choice of databases and, on the other hand, the bias implied by the use of a specific search equation. As a possible future line of research, it would be interesting to extend the comparative study to other bases, expand the search terms including specific terms of the ICTs (internet, social networks, smart phones, etc.), perform collaborative analysis or deepen the content of the documents (bibliographic analysis).