Positive Organizational Psychology: A Bibliometric Review and Science Mapping Analysis

Positive organizational psychology (POP) is a research area that focuses on the positive aspects of optimal functioning at work. Although consolidated and with a large volume of publications, no bibliometric analysis has been performed that allows knowing its high-level structure, developments, and distribution of knowledge since its origins. The objective is to analyze the 7181 articles published in POP on the Web of Science Core Collection (WoSCC). A retrospective bibliometric analysis and science mapping were performed. The title, authors, institutions, countries, scientific categories, journals, keywords, year, and citations were extracted from WoSCC. Impact factor, quartile, and country were collected from Journal Citation Reports (JCR) 2019. Authors were classified according to the proposal of Crane, and Bradford’s law was calculated. The results show that it is an area with more than 100 years of experience, divided into three stages of different productivity and visibility, highlighting a decrease in its visibility in recent years. With a multidisciplinary and international interest, psychology and business and economics stand out, especially in countries such as the United States, the United Kingdom, and the Netherlands. Four popular study topics emerged: well-being at work, positive leadership, work engagement, and psychological capital.


Introduction
Martin Seligman, in his speech as president of the American Association of Psychology in 1998, stated that positive psychology (PP) is the scientific study of optimal human functioning [1]. Years later, Linley, Joseph, Harrington, and Wood [2], in their work of 2006, analyzed different definitions given by referents of PP, integrating them into the following: "The scientific study of the possibilities of optimal human development, at the metapsychological level, aims to theoretically reorient and restructure the imbalance existing in psychological research and practices, granting greater importance to the study of the positive aspects of human life experiences, integrating them with those that cause suffering and pain; at a level of pragmatic analysis, it addresses the means, process, and mechanisms that make it possible to achieve a higher quality of life and personal realization" [2] (p. 8).
In short, PP goes from focusing on what is wrong with people to what is right, that is, it applies the scientific method to study the positive experiences, strengths, and psychological resources of people, groups, and institutions to achieve optimal development and functioning. As a result, the main research topics addressed by PP are oriented towards subjective well-being (happiness); positive moods and emotions; sensory, intellectual, and aesthetic pleasures; strengths and virtues; healthy practices such as optimism, gratitude, meditation, physical exercise, or artistic expression; interests, skills, and achievements; positive personal relationships; positive institutions (educational, labor, political); or transcendence and the meaning of life, among others.

Design of the Study
An observational, descriptive retrospective study was conducted using bibliometric methodology and science mapping. As it does not include human beings, it did not need the approval of any institutional review board.

Data Collection
On 7 February 2021, a thorough search was conducted in the WoSCC of the ISI Web of Science (Thomson Reuters, Philadelphia, PA, USA). The following terms and retrieval strategy were used: (positiv* AND organization*) OR (positive* AND organization* AND psychol*) OR (positive* AND organization* AND scholar*) OR (positive* AND psychol* AND work*) OR (positive* AND workplace) OR (positive* AND organization* AND behave*) OR (positive* AND human* AND resour*) OR (positive* AND leader*) OR (psychol* AND capital*) OR (organizat* AND virtuous*) OR ("organizat* resil*") OR ("resil* organizat*") OR (happy* AND work*) OR (happi* AND work*) OR (well-bein* AND work*) OR (wellbeing* AND work*) OR ("flow* at work*") OR (cope* AND work* AND stress*) OR (coping* AND work* AND stress*) OR (work* AND engage*) OR (organizat* AND ethic*). The searches were conducted in the title field. These terms were chosen from among those obtained as a result of the review of Donaldson and Ko [19] on the status of the POP area between 2001 and 2009. A total of 10403 works were obtained ( Figure 1). We used the following inclusion and exclusion criteria: (1) there were no restrictions on language or availability; (2) works published in 2021 were not included; (3) according to the document type, only the article, article early-access, review, and review early-access were selected. Finally, 7181 articles were included in the bibliometric analysis.

Design of the Study
An observational, descriptive retrospective study was conducted using bibliometric methodology and science mapping. As it does not include human beings, it did not need the approval of any institutional review board.

Data Collection
On 7 February 2021, a thorough search was conducted in the WoSCC of the ISI Web of Science (Thomson Reuters, Philadelphia, PA, USA). The following terms and retrieval strategy were used: (positiv* AND organization*) OR (positive* AND organization* AND psychol*) OR (positive* AND organization* AND scholar*) OR (positive* AND psychol* AND work*) OR (positive* AND workplace) OR (positive* AND organization* AND behave*) OR (positive* AND human* AND resour*) OR (positive* AND leader*) OR (psychol* AND capital*) OR (organizat* AND virtuous*) OR ("organizat* resil*") OR ("resil* organizat*") OR (happy* AND work*) OR (happi* AND work*) OR (well-bein* AND work*) OR (wellbeing* AND work*) OR ("flow* at work*") OR (cope* AND work* AND stress*) OR (coping* AND work* AND stress*) OR (work* AND engage*) OR (organizat* AND ethic*). The searches were conducted in the title field. These terms were chosen from among those obtained as a result of the review of Donaldson and Ko [19] on the status of the POP area between 2001 and 2009. A total of 10403 works were obtained ( Figure 1). We used the following inclusion and exclusion criteria: (1) there were no restrictions on language or availability; (2) works published in 2021 were not included; (3) according to the document type, only the article, article early-access, review, and review early-access were selected. Finally, 7181 articles were included in the bibliometric analysis.

Data Analysis
WoSCC was used to obtain data on the years of publication of the works, the citations received, the scientific categories in which the journal that published the work is classified, the authors and their affiliation (institution or organization and country), the journals, and the keywords provided by the authors and the KeyWords Plus automatically generated from the titles of the articles. From the Journal Citation Report (JCR) edition 2019, the data related to the journals were obtained, such as the impact factor, the thematic category of each journal, the quartile of each journal in its category, and the country of the publisher. We followed the proposal of Crane [20] for the classification of authors according to their

Data Analysis
WoSCC was used to obtain data on the years of publication of the works, the citations received, the scientific categories in which the journal that published the work is classified, the authors and their affiliation (institution or organization and country), the journals, and the keywords provided by the authors and the KeyWords Plus automatically generated from the titles of the articles. From the Journal Citation Report (JCR) edition 2019, the data related to the journals were obtained, such as the impact factor, the thematic category of each journal, the quartile of each journal in its category, and the country of the publisher. We followed the proposal of Crane [20] for the classification of authors according to their volume of publications, in aspirants (1-4 papers published in POP), moderate producers (4-10 papers), and high producers (more than 10 papers). Moreover, the law of Bradford [21,22] was calculated to determine the objective weight of each journal, as it distributes the journals into three productivity zones with a similar number of articles and an increasing number of journals.
Microsoft Excel 2018 was used to conduct the descriptive study of the years of publication, citations received, scientific categories, authors and signatures, affiliations, countries, and journals. This software also allowed calculating the growth prediction of the publications until 2024, as well as creating the bubble plot of the temporal distribution of the works and citations received according to the scientific category. The tool MapInSeconds (mapinseconds.com) (accessed on 3 March 2021) of Darkhorse Analytics Inc. was used for the representation of the global geographical distribution of the articles. VOSviewer 1.6.15 [23] was used to extract information, obtain the clusters, create network maps on cooperation by country in the publication of articles, establish the collaborative network between authors, and perform the keyword co-occurrence analysis. Prior to the completion of each network, the data were reviewed and standardized manually so that each country, author, or keyword had a single written form. In the network interpretation, each node represents a country, author, or keyword. The size of each node indicates its frequency, so that the larger it is, the higher the occurrence. The line that joins the nodes represents the existence of a connection between countries, authors, or keywords. The thicker these lines are, the greater the relationship between the nodes. Finally, the colors of the nodes represent the clusters of countries, authors, or keywords, such that those of the same color belong to the same cluster.

Temporal Distribution of Published Articles and Cited Papers
The 7181 articles obtained in the search for POP were published over 116 years and have been cited 174,158 times. Of these, 41 can be considered citation classics or highly cited papers, as they received 400 or more citations, and most of them (80.50%) were published over a period of 10 years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011).
Three time periods are observed ( Figure 2). The first one, the incubation period , begins in 1904 with the work of Betty Eicke and Anne S. Bussell "The shortcomings of the teaching methods of the present training-schools for nurses from the standpoint of the graduate nurse engaged in private work" [24]. During these 90 years, 369 works (M = 4.10 papers/year) were published irregularly with 18925 citations (M = 51.29 citations/work). The most cited article of William A. Khan, "Psychological conditions of personal engagement and disengagement at work" [25] (2800 citations), belongs to this period. volume of publications, in aspirants (1-4 papers published in POP), moderate producers (4-10 papers), and high producers (more than 10 papers). Moreover, the law of Bradford [21,22] was calculated to determine the objective weight of each journal, as it distributes the journals into three productivity zones with a similar number of articles and an increasing number of journals. Microsoft Excel 2018 was used to conduct the descriptive study of the years of publication, citations received, scientific categories, authors and signatures, affiliations, countries, and journals. This software also allowed calculating the growth prediction of the publications until 2024, as well as creating the bubble plot of the temporal distribution of the works and citations received according to the scientific category. The tool MapInSeconds (mapinseconds.com) (accessed on 3 March 2021) of Darkhorse Analytics Inc. was used for the representation of the global geographical distribution of the articles.
VOSviewer 1.6.15 [23] was used to extract information, obtain the clusters, create network maps on cooperation by country in the publication of articles, establish the collaborative network between authors, and perform the keyword co-occurrence analysis. Prior to the completion of each network, the data were reviewed and standardized manually so that each country, author, or keyword had a single written form. In the network interpretation, each node represents a country, author, or keyword. The size of each node indicates its frequency, so that the larger it is, the higher the occurrence. The line that joins the nodes represents the existence of a connection between countries, authors, or keywords. The thicker these lines are, the greater the relationship between the nodes. Finally, the colors of the nodes represent the clusters of countries, authors, or keywords, such that those of the same color belong to the same cluster.

Temporal Distribution of Published Articles and Cited Papers
The 7181 articles obtained in the search for POP were published over 116 years and have been cited 174,158 times. Of these, 41 can be considered citation classics or highly cited papers, as they received 400 or more citations, and most of them (80.50%) were published over a period of 10 years (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011).
Three time periods are observed ( Figure 2). The first one, the incubation period , begins in 1904 with the work of Betty Eicke and Anne S. Bussell "The shortcomings of the teaching methods of the present training-schools for nurses from the standpoint of the graduate nurse engaged in private work" [24]. During these 90 years, 369 works (M = 4.10 papers/year) were published irregularly with 18925 citations (M = 51.29 citations/work). The most cited article of William A. Khan, "Psychological conditions of personal engagement and disengagement at work" [25] (2800 citations), belongs to this period.   The second period, the initiation (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007), reveals the first increase in the number of works (n = 932) published on a regular and incremental basis (M = 71.69 works/year), which received a total of 66635 citations (M = 71.50 citations/work). As of 2008, a third time period of exponential growth (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020) of 13 years was also initiated, with 5880 articles (M = 452.31 works/year). When focusing on the prediction for the next four years, the exponential increase observed in the last stage is confirmed, revealing its maximum real value in 2020 with 1136 works. However, although 88,598 citations were received in the latter period, the highest value of the three, the average number of citations per work was 15.07, much lower than the previous ones.

Scientific Categories
The classification of each article into a scientific category was based on the classification carried out by the WoSCC. If a work belonged to multiple categories, it was multi-assigned, so the final count would be higher than that of the articles in the initial search.
POP articles were classified into 132 different thematic categories. Figure 3 shows the temporal evolution of the works and citations in categories with 50 or more works. "Business and Economics" is the area where the largest number of works were published (n = 2179), with a broad time-span since 1962 and with sustained growth. It is also the area that received the highest number of citations (n = 82,145; 37.70 citations/work), showing a marked decline since 2006. Secondly, the "Psychology" setting presents 1995 articles and a development similar to that of the previous area. It has a broader time-span (since 1959), a steady increase in the number of works and citations (n = 80,175 citations; 40.19 citations/work), and a considerable reduction in its volume of citations as of 2006. 62% of the citation classics or highly cited papers belong to these two categories.

Scientific Categories
The classification of each article into a scientific category was based on the classification carried out by the WoSCC. If a work belonged to multiple categories, it was multiassigned, so the final count would be higher than that of the articles in the initial search.
POP articles were classified into 132 different thematic categories. Figure 3 shows the temporal evolution of the works and citations in categories with 50 or more works. "Business and Economics" is the area where the largest number of works were published (n = 2179), with a broad time-span since 1962 and with sustained growth. It is also the area that received the highest number of citations (n = 82,145; 37.70 citations/work), showing a marked decline since 2006. Secondly, the "Psychology" setting presents 1995 articles and a development similar to that of the previous area. It has a broader time-span (since 1959), a steady increase in the number of works and citations (n = 80,175 citations; 40.19 citations/work), and a considerable reduction in its volume of citations as of 2006. 62% of the citation classics or highly cited papers belong to these two categories. The results show that the POP area is very interdisciplinary and with variations over time. During the incubation period, the works were published in 70 scientific categories, highlighting "Psychology" (94), "Business and Economics" (86), "Public, Environmental, and Occupational Health" (39), "Social Sciences-Other Topics" (38), and "General and Internal Medicine" (34). In the initiation stage, there were 86, also with "Psychology" (239), "Business and Economics" (236), "Social Sciences-Other Topics" (135), and "Public, Environmental and Occupational Health" (72) at the top. In the exponential growth stage, the number of categories increases to 122, again with the highest volume of publications in "Business and Economics" (1857), "Psychology" (1662), "Public, Environmental, and Occupational Health" (700), and "Social Sciences-Other Topics" (637). The results show that the POP area is very interdisciplinary and with variations over time. During the incubation period, the works were published in 70 scientific categories, highlighting "Psychology" (94), "Business and Economics" (86), "Public, Environmental, and Occupational Health" (39), "Social Sciences-Other Topics" (38), and "General and Internal Medicine" (34). In the initiation stage, there were 86, also with "Psychology" (239), "Business and Economics" (236), "Social Sciences-Other Topics" (135), and "Public, Environmental and Occupational Health" (72) at the top. In the exponential growth stage, the number of categories increases to 122, again with the highest volume of publications in "Business and Economics" (1857), "Psychology" (1662), "Public, Environmental, and Occupational Health" (700), and "Social Sciences-Other Topics" (637).

Geographical Distribution of Countries and Institutions and Cooperation Network
The works on POP involved 123 countries on the five continents ( Figure 4) although only 60 countries contributed more than 10 works. The 10 countries that made the highest number of contributions were the United States (1895)

Geographical Distribution of Countries and Institutions and Cooperation Network
The works on POP involved 123 countries on the five continents ( Figure 4) although only 60 countries contributed more than 10 works. The 10 countries that made the highest number of contributions were the United States (1895)   The analysis of international participation over time (Table 1) shows that in the first 90 years (incubation period), the works were signed by authors from 16 countries, with the United States being the most productive and most cited. In the initiation period, the The analysis of international participation over time (Table 1) shows that in the first 90 years (incubation period), the works were signed by authors from 16 countries, with the United States being the most productive and most cited. In the initiation period, the number of countries increased to 46, and the United States also led the productivity and number of citations. Finally, during the period of exponential growth, up to 122 countries were included. The United States also contributed the most articles and citations. The top five of international participation included the United Kingdom in all three time periods, Canada in the initial two, and the Netherlands in the last two.     A total of 1000 institutions signed these works. Fifty-four percent did so with five or fewer participations. The 10 most productive were Utrecht University in the Netherlands (127), Erasmus University in the Netherlands (122), North-West University in South Af-

Collaborative Co-Author Network Analysis
These publications involved 17,831 different authors, an average of 2.48 authors per article. Following the proposal of Crane [20], 17,438 authors are aspirants because they have participated in less than four works, 393 are moderate producers, as they have signed between 4 and 10 papers, and 38 are high producers, with more than 10 works.
When analyzing the moderate and high producers (Figure 6), 70 clusters or collaboration groups were obtained. The one that groups the largest number of documents and collaborations is also the largest, with 28 authors collaborating with 185 authors, signing 308 documents that received 33,440 citations. Arnold B. Bakker (93), Evangelia Demerouti (35), Sabine Sonnentag (21), Toon W. Taris (19), and Despoina Xanthopoulou (13) in the Netherlands and Greece stand out in this group as high producers.

Collaborative Co-Author Network Analysis
These publications involved 17,831 different authors, an average of 2.48 authors pe article. Following the proposal of Crane [20], 17,438 authors are aspirants because they have participated in less than four works, 393 are moderate producers, as they have signed between 4 and 10 papers, and 38 are high producers, with more than 10 works.

Most Relevant Journals in POP
The works were published in 2320 different journals. When applying the law of Bar ford (1934), a distribution in three areas was obtained. The first or core area includes 64 journals that published between 236 and 16 papers, the second area includes 490 journal that published between 15 and 3 articles, and the third area includes 766 journals tha published 1 or 2 papers.
The analysis of the 64 journals of the core area ( Table 2 presents the data from the 20 most productive journals in the core area) shows that 29 were published in the United Kingdom; 20 in the United States; 4 in the Netherlands; 2 in South Africa and Germany  (14), Jessica de Bloom (13), and Asko Tolvanen (11) are at the forefront of the productivity of this Finnish group.

Most Relevant Journals in POP
The works were published in 2320 different journals. When applying the law of Barford (1934), a distribution in three areas was obtained. The first or core area includes 64 journals that published between 236 and 16 papers, the second area includes 490 journals that published between 15 and 3 articles, and the third area includes 766 journals that published 1 or 2 papers.
The analysis of the 64 journals of the core area ( Table 2 presents the data from the 20 most productive journals in the core area) shows that 29 were published in the United Kingdom; 20 in the United States; 4 in the Netherlands; 2 in South Africa and Germany; and 1 in Japan, Romania, and Spain. Most are related to the areas of "Psychology," "Business and Economics," and "Public, Environmental and Occupational Health" (65.52%). The impact factor ranges from 0.453 to 6.941 (mean 2.34 ± 1.29). Most (68.42%) are in the first and second quartiles of their category.  Table 3 shows the articles that have received the highest number of citations in the years corresponding to the initiation and exponential growth periods (1995-2020). Together they received 8.56% of the total citations. They are published in 20 different journals, mostly (90%) located in quartiles 1 and 2 of their category, belonging to the categories of "Psychology" (44.74%) and "Business and Economics" (42.11%). Their authors are mostly from the United States (38.99%), the Netherlands (16.67%), and Spain (8.33%). Concerning their study design, most are cross-sectional, which use a multivariate methodology for verifying the fit of causal models (15). The reviews, especially theoretical reviews and discussion of topics (7), are also noteworthy. To a lesser extent, three works of instrument creation or validation and one longitudinal work are included.

Keyword Co-Occurrence Analysis
An analysis of the 16,555 author keywords and KeyWords Plus included in the 7181 articles was carried out. Of them, 81.90% only appear once or twice. When performing a co-word cluster analysis with terms that have a frequency of 40 or higher, four clusters were obtained that are represented in Figure 7 from left to right. The keywords were reviewed and modified so that each meaning was represented by a single written form (for example, well-being instead of wellbeing).

Keyword Co-Occurrence Analysis
An analysis of the 16,555 author keywords and KeyWords Plus included in the 718 articles was carried out. Of them, 81.90% only appear once or twice. When performing co-word cluster analysis with terms that have a frequency of 40 or higher, four cluster were obtained that are represented in Figure 7 from left to right. The keywords were re viewed and modified so that each meaning was represented by a single written form (fo example, well-being instead of wellbeing).   (62), individual-differences (58), construct (57), and core self-evaluations (55).

Discussion
The objective of this work was to carry out a comprehensive analysis of the evolution of research publications in the field of POP. A bibliometric methodology and science mapping were used in order to analyze publication and citation trends, the scientific categories, the top players in terms of journals and institutions, the collaborative networks between authors and countries, and the central research topics. Moreover, in this section, we discuss research gaps and some promising scientific directions for future research in the POP field, which can be shown as follows: • The field of POPs has evolved in three main stages over time and can be considered interdisciplinary, mainly around two predominant areas of research. The 7181 articles analyzed were published over a broad 116-year period, in which three temporal periods can be distinguished: one more extensive period, between 1904 and 1994, with a very irregular frequency of publication; a second period of initiation between 1995 and 2007, with a progressive increase; and a third period of exponential growth, from 2008 to the present moment. The same trend has been observed in the study by Sott et al. [12] in their review of 100 years of scientific evolution in the field of work and organizational psychology. As of this third stage, interest in a proactive and positive approach to the study of organizations is consolidated, which, as indicated by Luthans [7], focuses more on strengths, that is, what is right in organizations, teams, leaders, and employees rather than on what is wrong. This interest, which arises from the current PP that began in the 2000s, quickly spread to professionals and academics in the field of management and organization, as well as in industrial and organizational psychology.
This fact is confirmed by the results of this study on the analysis of scientific areas and journals. Although it is an area of multidisciplinary study with variations throughout its century of existence, the two areas that are present since its inception and that have published together 43.54% of the works on POP are those of "Psychology" and "Business and Economics." This is also confirmed in the results of the journals because, according to Bradford [21], 13 of the 20 that make up the core zone belong to these two areas. This is consistent with the definition of POP [6], as it covers both organizational and work aspects and psychological aspects.
Scholars and experts can explore new categories from other perspectives, such as mathematics or computer sciences, to acquire more meaningful conclusions. Although this category does not appear in Figure 3, mathematics and computer sciences may be a potential research opportunity in the future. For example, the topic of people analytics is future-oriented. In recent years, there has been talk of the future of work, with stories of all jobs being replaced by robots and automated processes. Automation is here and it will impact the job landscape, but the key point is that with automation comes data, and so for us to ignore this new source of data is foolhardy. Digital technologies can provide new sources of people insight and new ways of collaborating across an organization and help with resolving challenges and issues that had not been considered before [51]. The emergence of new technologies is important for researchers and scholars to improve the precision of prediction of organizational behavior.

•
The field of POP has grown and continues to grow exponentially, but for a few years, this growth has not been accompanied by the number of citations these papers have received. Concerning the analysis of citations carried out, it should be noted that their amount has been increasing in the first two periods of publication. This is an objective measure of performance, value, recognition, influence, and impact of an investigation and its researchers [52]. However, the exponential increase in the third period was not matched by an increase in the number of citations received, but rather the contrary. Between 2012 and 2020, this number has been declining dramatically, going from 34.26 citations per work in 2012 to 1.19 in 2020.
As indicated by Garfield in 1977 [53], the reasons for citing a publication are varied. These would include providing a background reading, identifying a methodology, or identifying an original publication describing concepts or discussing related ideas, among others. In any case, we consider it necessary to deepen the study of the evolution of the aforementioned works in the POP area as of 2012 to be able to explain why a decrease in the number of citations and an increase in the number of works are produced.
Firstly, there is evidence for and against the greater likelihood of the oldest works being cited [54]. In addition, it also appears that those published in the last 15 years have less chance of becoming citation classics or highly cited papers in their area [55]. In the case of POP, these two aspects are confirmed. Most of the citation classics, those that have generated high interest in the scientific community, were published between 2001 and 2011. This fact is confirmed because, to this date, the most cited articles each year (except for four of them) received more than 400 citations, highlighting "The measurement of work engagement with a short questionnaire-A cross-national study" [36] with 2386 citations and the work of Luthans and collaborators "Positive psychological capital: Measurement and relationship with performance and satisfaction" [37] with 1220 citations. Therefore, we could conjecture that this last stage could be the frontier of knowledge in POP. As Cole indicated [56,57], in the scientific progress of an area, a distinction can be made between core and frontier knowledge. The core includes the fundamental theories of a field, whereas the more ephemeral frontier knowledge is produced in recent years, including descriptive analyses with little temporal transcendence, and it does not function as a basis for further knowledge development. It is important that academics and experts who research in the field of POP make their work visible to the scientific community.

•
Concerning international representation, although very extensive, with the participation of 1000 institutions from 123 countries on five continents, the analysis of the data shows that in each of the three time periods analyzed, the United States, the United Kingdom, and the Netherlands have contributed the most publications and received the most citations. They are also the ones with the greatest international collaboration, along with Australia, China, and Spain.
The United States is also the country that shows a greater collaborative network with countries such as China, Canada, the United Kingdom, Australia, and South Korea. Its high economic capacity and its institutions with a high level of research cause many researchers from other countries to be interested in establishing research links [58]. It has shown clear leadership in POP since its inception, with Fred Luthans from the University of Nebraska and James B. Avey from Central Washington University standing out among all the authors.
The Netherlands also stands out, with the teams led by Arnold B. Bakker from Erasmus University Rotterdam and Wilmar Schaufeli from Utrecht University. Although the United Kingdom is noteworthy for its productivity, it does not present any referent team in this field. Finally, although China has recently entered the scene with high scientific productivity and a wide collaborative network, it has not been accompanied by high visibility, with very low levels of citation. Therefore, as Zhu and Liu [59] indicate, China may need a change in its research evaluation standards and should focus on improving the quality of research. It would be interesting to the scientific community to increase international collaboration between countries and research groups to consolidate POP.

•
The results of the keywords co-occurrence analysis showed a distribution in four thematic areas, in addition to highlighting the most popular topics in POP. This is in line with the findings of Soot et al. [12] in defining them as new themes unrelated to earlier periods in the development of work and organizational psychology. The terms of the first cluster refer to the research topic of well-being at work, which can be understood as the quality of life at work, which is considered the greatest determinant of productivity at the individual, organizational, and social level [60]. The importance of well-being in the workplace can be seen from the positive outcomes (e.g., health, mental health, social support, happiness, life satisfaction, coping) of well-being and the negative impact (e.g., stress, depression, occupational stress, conflict, family conflict, negative affectivity) when it is not fulfilled [61]. Studies on instruments for evaluation and intervention in this area are also noteworthy.
The second cluster, related to positive leadership, includes terms such as transformational leadership, ethical leadership, and authentic leadership. These studies show the characteristics of positive leadership and its potential benefits for the leader, the employees, and the organization [19], such as positive organizational behaviors, management, improved performance, job satisfaction, or organizational commitment, among others.
Thirdly, the study of work engagement is mainly related to the job demands-resources model [62] and the concept of burnout [63], with terms such as burnout, resources, job resources, demands, job demands, and demands-resources model. Lastly is the research on psychological capital. Following the definition of Luthans, Avey, Avolio, Norman, and Combs [64], the keywords grouped in this cluster refer to the positive psychological state related to self-efficacy when assuming and making the necessary efforts to succeed in tasks of some difficulty, optimism about current and future success, hope of success and perseverance in the objectives, and resilience to achieve success in the face of problems and adversity. The emergence of new research topics, the combination of different approaches, and the determination of how to use different methodologies to discover more meaningful results in the prediction of POP can be viewed as an interesting direction for the future.

Conclusions
During the past few years, there have been no reviews about POP and no in-depth scientometrics analysis. This work analyzes the evolution of research in POP as of its origins, using a bibliometric methodology and science mapping.
According to the above bibliometrics analysis, the main conclusions of this study can be summarized as follows: (1) this area of research began in 1904 and developed around three time periods, consolidating this positive approach in the study of organizations as of 2008; (2) its volume of publications, as well as its visibility, was increasing until 2011, at which point a significant decrease in the number of citations received began; (3) although it has had a multidisciplinary approach since its inception, it has been of particular interest for Psychology and Business and Economics; (4) it has generated research interest in 123 countries on five continents, with the United States, the United Kingdom, and the Netherlands standing out particularly, both for their productivity and for their international collaborative networks; (5) over the years, four research topics have been highlighted as the most popular: well-being at work, positive leadership, work engagement, and psychological capital.
This study presents some limitations. Only the WoSCC was used for the search, so the number and type of works, as well as the number of citations received, depend on it and would have been different if the search had been completed with searches in other databases such as Scopus or Google Scholar. Although the WoSCC does not include the entirety of peer-reviewed publications, it does allow access to extensive and multidisciplinary information from more than 250 sciences, with more than 21,100 peer-reviewed, highquality scholarly journals, books, and conference proceedings published worldwide since 1800. It is also possible that some important articles with great influence in POP are not included in this work due to the search profile used, which, despite being very extensive, may have left out terms of great relevance to this area of knowledge. We also recognize the limitation of not taking into account the self-citations, or whether the citations received by the works are positive or negative, so the level of agreement or criticism of the scientific community towards the cited articles cannot be determined [65].