Dating Violence: A Bibliometric Review of the Literature in Web of Science and Scopus

This study has the general purpose of improving the understanding and description of the field of violence in young couple relationships by means of a bibliometric analysis. A descriptive and transversal-retrospective methodology is used, the objective of which is to describe in a quantitative way the information obtained from the production of 842 references registered in the Scopus and Web of Science databases. The results show that during 2017 and 2018, the majority of publications were concentrated, highlighting that the United States is the country with the highest amount of scientific production on violence in intimate relationships. It is important to highlight that more and more countries are investigating this subject, highlighting an increase in production from 2015 onwards. The violence that occurs in the relationships of young couples is a global social and health problem that requires research to be able to deepen its knowledge and in the prevention of this social scourge.


Introduction
Dating violence is understood as the threat or use of sexual, verbal or physical abuse by one partner against the other (Leen et al. 2013;Rodríguez-Castro and Alonso-Ruido 2015;Rubio-Garay et al. 2017), which means that to properly identify situations of violence, physical and/or sexual aggression must be assessed alongside other forms of violence such as verbal aggression and/or threats, affecting couples who have either been together for a long time or are starting a relationship (Ontiveros et al. 2020).
Violence in young couples' relationships is a very important and fairly widespread public health and social problem. Numerous studies (Ruel et al. 2017;Navarro-Pérez et al. 2020;Dodaj et al. 2020;Taquette et al. 2020;Rodríguez-Domínguez et al. 2020) reveal a high prevalence of this type of violence among young people in different countries. The levels transcend social class, ethnicity, educational level or sexual orientation (Ortiz and Morales 2011). Similarly, in recent years, several systematic reviews have been conducted on the prevalence of violence experienced by young couples (Jennings et al. 2017;Rubio-Garay et al. 2017;Garthe et al. 2017;Yanez-Peñúñuri et al. 2019;Taquette and Monteiro 2019;Bundock et al. 2020), highlighting the existence of health problems such as depression or anxiety, as well as immediate negative effects on well-being and physical, mental and reproductive health, affecting the quality of life and the development of the individual (Sugg 2015;Meza 2018;Duval et al. 2020). Therefore, and according to the WHO (2017), this type of violence is a health problem in different parts of the world, which is why action programmes and policies have been implemented, mainly in the school context, a scenario of socialisation and construction of identity in adolescents (Bundock et al. 2020). But the problem still needs to be addressed in programming for the prevention of violence that often occurs in mixed-sex formats in a critical understanding of the differences and similarities of sex in the perpetration of violence and, finally, the effectiveness of programming efforts aimed at reducing it could be refined and improved (Dardis et al. 2015).
Several studies (De Puy et al. 2015;Parker et al. 2017;Cho and Huang 2017;Marganski and Melander 2018;Park and Kim 2019;Storer et al. 2019;Tussey and Tyler 2019) argue that more multivariate and longitudinal research is needed, with a lack of gender-specific analytical approaches and focus on contextual factors. In addition, more longitudinal research is needed to cover multiple experiences of violent victimization in various social contexts. There is also a need for a greater critical understanding of gender differences and similarities in the perpetration of domestic violence by young couples, which would help refine and improve knowledge about domestic violence (Dardis et al. 2015;Wincentak et al. 2017;Sianko et al. 2019).
Given this scenario, it is important to determine the research trend, and review the literature related to the concept of "dating violence", taking into account the publications on the Web of Science and Scopus platforms. It is also pertinent to know the management of the programs and consider responding to this social problem that arises and is characterized by the great negative effect that it can have on the health of the person. A bibliometric review of the literature was performed to explore the dating violence with special emphasis on methodological factors and gaps in the literature. Consequently, this bibliometric review aims to analyse existing research on violence in young couples' relationships, helping to understand the needs for its prevention. Bibliometric analysis has been an effective tool to quantitatively analysing academic publications and to assess research trends in different research fields (Wu et al. 2020).

Materials and Methods
The general objective of this article is to analyse the current existing research in the Web of Science and Scopus databases on violence in young couples' relationships. The following are proposed as specific objectives: This research followed a methodological strategy of bibliometric study, of a descriptive type and of a transversal-retrospective nature, whose objective is to describe in a quantitative way the information obtained (Fuentes and Ortega 2019;Suleiman-Martos et al. 2020). The development of these studies allows us to know the main contributions to the state of the art, as well as to emphasize the need to visualize the results from other research as a starting point to make new scientific contributions (Cabrera-Ramos 2020;Abad Robles et al. 2020). To meet the research objective, firstly a descriptor was defined for the search in the corresponding databases, "dating violence", the most commonly used concept to define violence in young couple relationships, from the Thesaurus ERIC. The choice of this descriptor is mainly focuses on the possibility of bringing together the subject matter of the study, bearing in mind that violence in young couples is the central focus of our research.
Representative and reliable sources are used in the various bibliometric studies carried out in the different areas of research (Fuentes Cabrera et al. 2019;Eckhardt and Massa 2019;Cebrino and Cruz 2020;Badenes-Sastre and Expósito 2021). For this reason, the Web of Science and Scopus databases were taken into consideration. Both databases group together the scientific production with the greatest impact, facilitating access to cited documents, basing their data on the calculation of the h factor to measure an author's journey (Granda-Orive et al. 2011;Granda-Orive et al. 2013).
To carry out the descriptive analysis of the quantitative data, the SPSS version 25 programme was used and the analysis and interpretation of co-occurrences of keywords through the VOSviewer programme. The analysis variables (Table 1)  Impact indicators: type of document, country of publication, language, the affiliation of authors, journals, references with greater impact, authors with greater production and bibliometric map.

Variable Inclusion Criteria
Year of publication From the year 2015 to the year 2019 (the year 2020 has been eliminated for not being finalized) Type of document All is considered Affiliation All is considered Language All is considered Publications with greater impact More than 50 citations Research internationalization More than 10 publications Authorship of publication More than 10 publications Reference journals More than 10 publications Co-occurrence of content Title, abstract and keywords In the following table, we can see the variables analysed and the criteria are taken into account for their analysis.
From the year 2015 to the year 2019 (the year 2020 has been eliminated for not being finalized) The statistical variable studied was scientific production following a longitudinal statistical sample between 2015 and March 2020. The data collection took place in July 2020. The data collection took place in July 2020. The final sample of references was 842 (n = 842), of which 495 belong to Web of Science and 347 to Scopus. The initial search resulted in 2234 articles, with 35 removed due to duplication. Two reviewers independently screened the articles by abstract and title to determine eligibility for further examination. One thousand one hundred and twenty were retained for further analysis. The reviewers then conducted an in-depth review of the articles resulting in 842 articles meeting inclusion criteria. Refer to Figure 1 for a flowchart of the study selection process. In the following flow diagram, we can see the process followed until obtaining the final sample. Soc. Sci. 2021, 10, x FOR PEER REVIEW 4 of 16

-Diachronic productivity
Taking into account the diachronic productivity and using the variable of the year of publication, in both databases, it can be observed that there has been a linear growth in the research topic under study, however, in the In the Web of Science database, it is noted that the highest scientific production occurred during the year 2017 (n = 103; 20.1%) and in Scopus in 2018 (87; n = 25.07%). The WoS database collects the greatest scientific production during the number of years studied.

-Diachronic productivity
Taking into account the diachronic productivity and using the variable of the year of publication, in both databases, it can be observed that there has been a linear growth in the research topic under study, however, in the In the Web of Science database, it is noted that the highest scientific production occurred during the year 2017 (n = 103; 20.1%) and in Scopus in 2018 (87; n = 25.07%). The WoS database collects the greatest scientific production during the number of years studied.  When verifying Price's law of the exponential growth of scientific information, it is fulfilled since there is a phase of linear growth, in which the growth rate remains constant or independent of the size of the system, highlighting that the research is in phase 3-linear growth (Price 1986). The amount of production and its percentage percentages remain very stable, around 85-103 publications and 17%-20% in WoS and 54-87 publications and 15%-25% in Scopus.

-
Personal productivity Lotka's law or the law of the productivity of scientific authors indicates that only a small proportion of authors is responsible for the majority of scientific works, so that as the amount of work on a given increases subject, the number of authors decreases (Urbizagastegui 2011). In the following graph ( Figure 4)   When verifying Price's law of the exponential growth of scientific information, it is fulfilled since there is a phase of linear growth, in which the growth rate remains constant or independent of the size of the system, highlighting that the research is in phase 3-linear growth (Price 1986). The amount of production and its percentage percentages remain very stable, around 85-103 publications and 17%-20% in WoS and 54-87 publications and 15%-25% in Scopus.

-
Personal productivity Lotka's law or the law of the productivity of scientific authors indicates that only a small proportion of authors is responsible for the majority of scientific works, so that as the amount of work on a given increases subject, the number of authors decreases (Urbizagastegui 2011). In the following graph ( Figure 4)  When verifying Price's law of the exponential growth of scientific information, it is fulfilled since there is a phase of linear growth, in which the growth rate remains constant or independent of the size of the system, highlighting that the research is in phase 3-linear growth (Price 1986). The amount of production and its percentage percentages remain very stable, around 85-103 publications and 17%-20% in WoS and 54-87 publications and 15%-25% in Scopus.
-Personal productivity Lotka's law or the law of the productivity of scientific authors indicates that only a small proportion of authors is responsible for the majority of scientific works, so that as the amount of work on a given increases subject, the number of authors decreases Soc. Sci. 2021, 10, 445 6 of 16 (Urbizagastegui 2011). In the following graph ( Figure 4) it can be seen that the correlation between the number of authors and the number of scientific production in both databases is positive. Soc. Sci. 2021, 10, x FOR PEER REVIEW 6 of 16 between the number of authors and the number of scientific production in both databases is positive. The Pearson correlation coefficient in the Web of Science database is r = −0.948 and in Scopus r = −0.969, so both variables are correlated in the opposite direction, with a linear correlation with a perfect negative slope, assuming an absolute determination between both variables.

Indicators of Dispersion
Bradford's law or dispersion law describes that most articles about a specialized subject could be published by a few specialized journals, together with certain general or dispersal journals, therefore there is a relationship between the published articles and the journals of a given area, stating that a reduced number of journals belong to the nucleus, and this concentrates a similar number of articles than a large number of journals grouped in different areas of greater dispersion (Alvarado 2016). There are a reduced number of journals, which make up the core, which concentrates a similar number of articles as a large number of journals, grouped in areas of greater dispersion (zone 1, 2 and 3). In total we have 17 journals and 342 references distributed in four areas with an average of 86 in each area, so we observe that the nucleus with only 2 journals has 83 references, a number very similar to the other nuclei, which have a greater number of journals ( Figure 5). The Pearson correlation coefficient in the Web of Science database is r = −0.948 and in Scopus r = −0.969, so both variables are correlated in the opposite direction, with a linear correlation with a perfect negative slope, assuming an absolute determination between both variables.

Indicators of Dispersion
Bradford's law or dispersion law describes that most articles about a specialized subject could be published by a few specialized journals, together with certain general or dispersal journals, therefore there is a relationship between the published articles and the journals of a given area, stating that a reduced number of journals belong to the nucleus, and this concentrates a similar number of articles than a large number of journals grouped in different areas of greater dispersion (Alvarado 2016). There are a reduced number of journals, which make up the core, which concentrates a similar number of articles as a large number of journals, grouped in areas of greater dispersion (zone 1, 2 and 3). In total we have 17 journals and 342 references distributed in four areas with an average of 86 in each area, so we observe that the nucleus with only 2 journals has 83 references, a number very similar to the other nuclei, which have a greater number of journals ( Figure 5).
Both variables are correlated in the opposite direction, with a linear correlation of perfect negative slope between both variables (number of articles and number of journals). The Pearson correlation coefficient in the Web of Science database is r = −0.747 and in Scopus r = −0.991 (Figure 6).

-Type of document
Considering this variable corresponding to the type of document found (Figure 7), it is verified that in both databases the highest percentage of references are scientific articles (

-Type of document
Considering this variable corresponding to the type of document found (Figure 7), it is verified that in both databases the highest percentage of references are scientific articles (n = 276/79.5% in Scopus and n = 92.34/92.34% in WoS). On the other hand, in Scopus we

-Type of document
Considering this variable corresponding to the type of document found (Figure 7), it is verified that in both databases the highest percentage of references are scientific articles (n = 276/79.5% in Scopus and n = 92.34/92.34% in WoS). On the other hand, in Scopus we found book chapters with only 7.4% (n = 26), in addition to reviews (n = 16; 4.6%) and books (n = 2; 0.57%). However, in the Web of Science database, we highlight the review (n = 25; 5.04%), the books (n = 7; 1.41%) and conference proceedings (n = 6; 1.21%).
found book chapters with only 7.4% (n = 26), in addition to reviews (n = 16; 4.6%) and books (n = 2; 0.57%). However, in the Web of Science database, we highlight the review (n = 25; 5.04%), the books (n = 7; 1.41%) and conference proceedings (n = 6; 1.21%). -Affiliation The following variable seeks to analyze the institution to which the authors of the documents found are affiliated. Table 2 shows the affiliations of the authors that contain more than 10 publications. The number of references distributed among the 8 institutions with the highest number of affiliations are 5.76%, 5.47% and 4.89% from the Ohio University, The University of Texas at Arlington and the Centers for Disease Control and Prevention respectively in the Scopus database, the number being higher in the WoS database and coinciding in the Centers for Disease Control and Prevention with a similar percentage (6.25%) ( Table 3).  -Affiliation The following variable seeks to analyze the institution to which the authors of the documents found are affiliated. Table 2 shows the affiliations of the authors that contain more than 10 publications. The number of references distributed among the 8 institutions with the highest number of affiliations are 5.76%, 5.47% and 4.89% from the Ohio University, The University of Texas at Arlington and the Centers for Disease Control and Prevention respectively in the Scopus database, the number being higher in the WoS database and coinciding in the Centers for Disease Control and Prevention with a similar percentage (6.25%) ( Table 3). The language variable (Figure 8) in the scientific production on violence in intimate relationships, the highest percentages are in English in both databases (n = 330; 95.1% Scopus and n = 413; 83.26% Wos), followed by Spanish (n = 16; 4.61% in Scopus) and Korean in WoS (n = 44; 8.87%).  The language variable (Figure 8) in the scientific production on violence in intimate relationships, the highest percentages are in English in both databases (n = 330; 95.1% Scopus and n = 413; 83.26% Wos), followed by Spanish (n = 16; 4.61% in Scopus) and Korean in WoS (n = 44; 8.87%).  (Table 4) with the greatest impact, we found that the article "Teen dating violence: A meta-analytic review of prevalence rates", from the journal Psychology of Violence is the most cited with a total of 102 citations. Followed by "Physical dating violence, sexual violence, and unwanted pursuit victimization: A comparison of incidence rates among sexual-minority and heterosexual college students" from the Journal of Interpersonal Violence with 83 citations in the Scopus database. In contrast, we check in the WoS database, the first most cited article "Teen Dating Violence  (Table 4) with the greatest impact, we found that the article "Teen dating violence: A meta-analytic review of prevalence rates", from the journal Psychology of Violence is the most cited with a total of 102 citations. Followed by "Physical dating violence, sexual violence, and unwanted pursuit victimization: A comparison of incidence rates among sexual-minority and heterosexual college students" from the Journal of Interpersonal Violence with 83 citations in the Scopus database. In contrast, we check in the WoS database, the first most cited article "Teen Dating Violence (Physical and Sexual) Among US High School Students Findings From the 2013 National Youth Risk Behavior Survey" has 132 citations and the second coincides with the Scopus database, but in this case with 78 citations. -

Internationalization of research
Regarding the countries with the greatest scientific production of violence in young couples, the United States stands out, with 70.31% (n = 244) in Scopus and 61.09% (n = 303) in WoS with a large number regarding to the rest of the countries, as can be seen in Table 5 which represents the countries that have from 10 published documents on the subject. Four were extracted containing a total of 321 of the 347 analyzed for the Scopus database where Spain is in the second position (n = 42; 12.10%) followed by Canada (n = 24; 6.91%) and England with only 3.17% (n = 11), as in the Web of Science database where Spain is the second country (n = 61; 12.30%) followed by Canada (n = 25; 5.04%) and Mexico (n = 17; 3.43%). Of the total number of authors with the greatest scientific production (Tables 6 and 7), those with more than 10 publications on the subject were selected, Shorey, RC (5.76%) being the one with the most publications, followed by Stuart, GL with 15 publications (4.32%) and Temple, JR with 3.45% of the total, in the Scopus database. In contrast, we found a greater number of authors in WoS, although it is highlighted that the aforementioned authors coincide with the Scopus database, highlighting that Temple, JR has a greater number (0.1%). -Reference journals Table 8 shows the matching reference journals in both databases; it can be seen that the journal with the most publications is Journal of interpersonal violence (9.7% respectively), however, in the Scopus database, the journal Violence and Victims is the journal with the lowest reference (2.6%) and in the Web of Science database the journal Adolescent Dating Violence Theory Research and Prevention (2.2%). -Content co-occurrence The content co-occurrence analysis applied to the title, abstract and keywords of the analyzed scientific production shows that there are relationships between them. As can be seen in the figure, there are five thematic clusters (green, blue, red, purple and yellow) as we can see in Figure 9. The size of the concepts is proportional to the frequency of appearance and the number of existing connections with other concepts. In the green cluster, the largest concept is "teen", this group is related to descriptors that refer to factors surrounding adolescents. The red group is led by the term "woman" and "dating violence", which allows describing that violence in intimate relationships is directed mainly towards women. On the other hand, the purple cluster is headed by the descriptor "program" and includes descriptors related to its effectiveness and evaluation. The blue cluster encompasses research terms in the field, highlighting the words "literature" and "review". Finally, the yellow cluster highlights "intimate partner violence". This analysis allowed us to know that the research on violence in the relationships of young couples focuses on identifying the factors that influence this type of violence. Soc. Sci. 2021, 10, x FOR PEER REVIEW 12 of 16 allowed us to know that the research on violence in the relationships of young couples focuses on identifying the factors that influence this type of violence.

Discussion
Traditionally, violence in young couples' relationships has been studied and identified in formal contexts such as marriage and committed or monogamous relationships, but it was only recently that researchers suggested that this type of violence is a significant problem in young, single adults (Duval et al. 2020). The scientific review undertaken consolidates the idea that further work is needed on issues such as violence prevention programming, a critical understanding of gender differences and similarities in violence perpetration that could refine and improve the effectiveness of programming efforts to reduce violence (Dardis et al. 2015). This coincides with the purple cluster, which is headed by the descriptor "program" including descriptors related to its effectiveness and evaluation. Several studies (Dardis et al. 2015;Marganski and Melander 2018;Wincentak et al. 2017;Spencer et al. 2020) conclude that further research is needed on the subject matter for a more critical understanding of gender differences and similarities, and the results show exponential growth in output, which is evidence of an increase since 2015. These studies show that violence in relationships among young people is conceived within the framework of psychological and physical aggression to resolve conflicts that are essential to the relationship itself, increasing the risk of committing subsequent aggressions (Rubio et al. 2015). Numerous authors such as Rubio et al.  Duval et al. (2020) show a greater psychological and sexual victimization in women, as well as the perpetration of psychological aggressions of women towards men and sexual aggressions of men towards women. These results coincide with the red cluster, which is headed by the term "woman", highlighting that violence in intimate relationships is directed mainly towards women.
It is necessary to offer a process of preventive intervention, which allows for a more effective and context-adjusted orientation according to the duration of the relationship,

Discussion
Traditionally, violence in young couples' relationships has been studied and identified in formal contexts such as marriage and committed or monogamous relationships, but it was only recently that researchers suggested that this type of violence is a significant problem in young, single adults (Duval et al. 2020). The scientific review undertaken consolidates the idea that further work is needed on issues such as violence prevention programming, a critical understanding of gender differences and similarities in violence perpetration that could refine and improve the effectiveness of programming efforts to reduce violence (Dardis et al. 2015). This coincides with the purple cluster, which is headed by the descriptor "program" including descriptors related to its effectiveness and evaluation. Several studies (Dardis et al. 2015;Marganski and Melander 2018;Wincentak et al. 2017;Spencer et al. 2020) conclude that further research is needed on the subject matter for a more critical understanding of gender differences and similarities, and the results show exponential growth in output, which is evidence of an increase since 2015. These studies show that violence in relationships among young people is conceived within the framework of psychological and physical aggression to resolve conflicts that are essential to the relationship itself, increasing the risk of committing subsequent aggressions (Rubio-Garay et al. 2015). Numerous authors such as Rubio-Garay et al.  Duval et al. (2020) show a greater psychological and sexual victimization in women, as well as the perpetration of psychological aggressions of women towards men and sexual aggressions of men towards women. These results coincide with the red cluster, which is headed by the term "woman", highlighting that violence in intimate relationships is directed mainly towards women.
It is necessary to offer a process of preventive intervention, which allows for a more effective and context-adjusted orientation according to the duration of the relationship, degree of commitment, parental influence, among others (Ravi et al. 2020), as well as prevention programmes focused on the development of optimal emotional regulation and impulse control skills with emphasis on relationships (Ontiveros et al. 2020).

Conclusions
This article aims to present a description of the area of research on violence in young couples' relationships, taking into account indicators of production (diachronic and personal productivity), dispersion (correlation between authors and articles) and impact (type of document, country of publication, language, author affiliation, publications, references with greater impact, authors with greater production and bibliometric map) in the Web of Science and Scopus database. Mainly, it is highlighted that the years 2017 (Web of Science) and 2018 (Scopus) led the research of violence in couple relationships, with the United States being the country with the highest amount of scientific production. The correlation between the number of authors and the number of scientific productions in both databases is negative; the authors with the highest number of publications coincide in both databases. The most frequent type of document is scientific articles published in journals with specific themes of violence. There is no doubt that more and more countries are researching this subject, reflecting an increase in production since 2015. Since the violence in young relationships is a global health problem, more research is being done to understand and prevent this type of violence.
This research focused on analyzing the scientific production published during the specified period of time, but several limitations stand out; bibliometric review studies, for which information may be lost due to not using the correct descriptors or due to the lack of feasibility of covering all the databases that currently exist. However, due to the novelty and specificity of the term, it was decided to introduce only the concept "dating violence" as an element of the search engine, the study method does not allow to judge the quality of the analyzed research, and the option of using only the Scopus and WoS database may have limited sample size and diversity.
As a future line of research, it is proposed to systematically analyze the sample in this article, including new descriptors taking into account the main objectives, methodological design, variables under study, analyzed sample and information collection instruments, to give a greater scope to the objectives proposed.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest. Note 1 They are conceptually detailed in the results section.