1 Introduction

Nowadays, the academic analysis of the production of a group or academic entity is calling attention, specially by quantity and quality valuation of production and evaluation of researcher’s merits (^{KOCHHANN, 2021}). This scenario repeats among Nursing undergraduate and graduate programs, yet with fewer attention to undergraduate courses (^{SILVA et al., 2009}; ^{MENDONÇA et al., 2018}; ^{FERREIRA; TAVARES; KEBIAN, 2018}; ^{GIACCHERO; MIASSO, 2016}).

Studies about Brazilian Nursing teams don’t approach research intentions and networks between researchers work groups. We consider “research intention” as a proposition of a research object that may or may not become a product, such as an article or a book or other academic productions. Those intentions can be evaluated by different means, like organization of research groups (formal or informal), production of research projects and interaction or intention of collaboration between researchers, professors and students (^{SOUSA et al., 2019}). Another important source of information available nowadays is the curriculum vitae of professionals. Internationally, many from these curriculums are grouped in free web platforms like ResearchGate, ORCID and others.

Lattes Platform is a public research curriculum vitae academic database widely used in Brazil. It comprehends researcher’s academic history and interactions with scientific groups or institutions, as well as personal academic history, published articles, and other actions on education, research and extension (^{DIAS et al., 2016}). Lattes Platform is managed by the National Council for Scientific and Technological Development (CNPq) a federative organ and linked to the Department of Sciences, Technology, Innovation and Communication of Brazil. In addition, it is a public data platform.

Therefore, research projects and collaboration networks have great importance when demonstrating professors’ interactions (such as research group leaders) and professionals from a study group for example, or even as evaluation tool of focus and inter-relation of research objects. A network can be defined as connection method visualization between at least two elements based in connection criteria. Thus, networks or social network analysis as a methodology is widely used for interaction analysis and allow diverse possibilities and actions to identify main actors, engaged in knowledge mediation processes and information flux, as well as bonds and criteria built in network development (^{MACHADO; IPIRANGA, 2013}; ^{LANDIM et al., 2010}; ^{ANDRADE; DAVID, 2015}; ^{CAVALCANTE et al., 2018}). As an example of connection criteria between researchers is participating in a research group and sharing a research project. In this case, both criteria are good indicators on research intention. Additionally, these relationships could be easily accessed by, for example, exploring the public data available from CNPq or Lattes Platform.

Based on that, this study evaluated research intentions network among professors of an undergraduate Nursing course, from a Brazilian, public and federal institution, based on CNPq registered research groups and Lattes Platform registered research projects.

2 Material e Methods

2.5 Results

Twenty-five research groups were identified in the CNPq Directory Research Group, of four areas: Botany, Nursing, Medicine and Public Health. Seven institutions were affiliated: Federal University of Catalão, Goiás state (UFCAT; 4.00%, 1 of 25 groups); Federal University of Goiás, Goiás state (UFG; 4.00%, 1 of 25 groups); State University of Campinas, São Paulo state (UNICAMP; 4.00%, 1 of 25 groups); Federal University of São Paulo, São Paulo state (UNIFESP; 4.00%, 1 of 25 groups); Federal University of Triângulo Mineiro, Minas Gerais state (UFTM; 24.00%, 6 of 25 groups); University of São Paulo, São Paulo state (USP; 32.00%, 8 of 25 groups); and UFU (28.00%, 7 of 25 groups) (Table 1). Among the professors that had no registration in the directory, three were active and five were inactive (retired).

Table 1 Research groups between professors in a Brazilian Nursing Undergraduate course in March of 2021 in CNPq database (Brazilian National Council for Scientific and Technological Development - CNPq, in Portuguese: Conselho Nacional de Desenvolvimento Científico e Tecnológico) 

CNPq research group names (translated from Portuguese)	Area	Institution1	Code
Comprehensive Child and Adolescent Health Care - AISCA	Nursing	UFU	G1
Botany-UFU	Botany	UFU	G2
Nursing and women's health	Nursing	UFTM	G3
Spirituality and Cardiology: Interprofessional Approach - UFCAT	Medicine	UFCAT	G4
Spirituality and Cardiology: Interprofessional Approach - UFG	Medicine	UFG	G5
GIPHA - Interdisciplinary Research Group on Hypertension	Nursing	USP	G6
Study group on the rehabilitation of surgical and oncologic patients	Nursing	USP	G7
Study Group on Prevention of Harmful Use of Alcohol and/or Drugs - GRUPAD	Nursing	USP	G8
Study group and research in evidence-based practice and patient safety in the care process	Nursing	UFTM	G9
Study group and research on the use of human resources in nursing	Nursing	USP	G10
Study group on mental health and its interfaces with other chronic conditions - GESMIC	Public Health	UFU	G11
Research group on rehabilitation and quality of life	Nursing	USP	G12
Cardiovascular Health Research Group	Nursing	UFU	G13
Collective Health Research Group	Public Health	UFTM	G14
Advanced Laboratory in Gender and Violence Studies	Nursing	UFU	G15
Laboratory of studies and research in suicide prevention and posvention (LEPS)	Nursing	USP	G16
Emergency League in Nursing- LUREEN	Nursing	UFU	G17
Center for Study and Research in Epidemiology and Quantitative Methods in Health	Public Health	UFTM	G18
Center for Studies of Prevention and Infection Control in Health Services (NEPECISS)	Nursing	USP	G19
Center for studies on psychiatric disorders: care and research - NUDPAS	Nursing	USP	G20
Research Center in Education, Nursing and Health	Nursing	UFU	G21
General Pathological Processes and Infectious and Parasitic Diseases	Medicine	UFTM	G22
Women's and Newborn Health	Nursing	UNICAMP	G23
Safety, Technology and Care (SEGTEC) - Nursing research group on patient safety, pediatric intensive care and intravenous and drug therapy	Nursing	UNIFESP	G24
Adult Living and Health	Nursing	UFTM	G25

¹Institution name abbreviation adopted by Brazilian federal institutions

Source: prepared by the authors (2022)

Fifty-one Nursing professors from UFU were registered in research groups. Registrations in Nursing area had only 12.50% of students (5 in 40 registrations), while other areas have 18.18% (2 in 11 registrations), with no difference between the two groups (Fisher's Exact Test p = 0.635). Registrations from other units outside UFU had 33.33% of students registered (7 in 21 registrations). When only UFU registrations were evaluated, 100% (30 in 30) are classified as researcher.

Evaluating the admission time until 2014, showed that 42.85% of registers are classified as students (6 in 14 registrations). From 2015 on, only 2.70% are students (1 in 3 registrations), with statistical difference between groups (Fisher's Exact Test p = 0.001).Only admission time was able to predict professors role in the group as student (constant = −3.584; p < 0.001; B1 = 3.296; p = 0.004), that is, being admitted until 2014 leads to a 27 times higher chance that the registration was done as a student (Odds Ratio = 27.00, 95%CI: 2.84, 256.51).

These findings support the hypothesis that these registrations can be associated to previous bounds before professor’s admission, since registrations are associated to external institutions, and in some cases the same institution of professors’ graduation. Moreover, low registration prevalence of students from 2015 on can be justified by post-doctoral leave, or even outdated registrations. Association between kind of registration and the degree level was not tested.

From the total 252 research projects, 28.57% (n = 72) were analyzed, or by still being developed (28.17%, n = 71) or were concluded in 2021 (0.40%, n = 1). Six institutions related to the coordinators were identified: Federal University from Rio Grande do Sul (UFRGS; 4.17%, n = 3), University of Triângulo Mineiro (UFTM; 4.17%, n = 3), Oswaldo Cruz Foundation (FIOCRUZ; 1.39%, n = 1), University of Pernambuco (UPE; 1.39%, n = 1), University of São Paulo (USP; 2.78%, n = 2), and the Federal University of Uberlândia (UFU; 86.11%, n = 62). Projects linked to UFU were predominant. It is important to highlight that those affiliations were different between the groups and research projects.

In research groups network from CNPq was noticed different behavior between the networks (Figure 1, Table 2). In the research group network four professors were linked to only one research group, and found that only one research group (G21) is highly linked to professors (12 Professors) (Figure 1A, Table 2). Majority of professors connect only with Nursing area, while Public Health area is the one that connects the most with Professors connected to Nursing (Figure 1B). Two Professors do not connect with Nursing (Figure 1B). Botany connects with one professor and Medicine with two professors (Figure 1B). Majority of professors connect only to UFU, and with that group majority of connections are with professors from UFTM and USP (Figure 1C, Table 2). Four institutions were connected to only one professor (UNICAMP, UFCAT, UFG e UNIFESP), and one of them did not connect to UFU (Figure 1C, Table 2).

Source: prepared by the authors.

Figure 1 Interaction network between Professors from the Nursing undergraduate Course from Federal University of Uberlândia (UFU), obtained from research groups registered in CNPq. A. based on research groups. B: based on research area. C. based on institution affiliation. 

Table 2 Descriptive network metrics observed between professors in a Brazilian Nursing Undergraduate course in March of 2021 based on research group database from CNPq and research projects from Lattes database. 

	Connectors
	Research groups				Research projects
Statistic	Group	Area	Affiliation		All Projects	Project with collaboration	Affiliation
n professors	24	24	24		24	23	24
n groups	25	-	-		-	-	-
n projects	-	-	-		72	30	-
n areas	-	4	-		-	-	-
n affiliation	-	-	7		-	-	6
n links	51	51	51		146	104	146
Minimum aresta professor	1	1	1		1	1	1
Maximum aresta professor	5	4	2		13	8	3
Minimum aresta connector	1	1	1		1	2	1
Maximum aresta connector	12	18	22		11	11	24

Source: prepared by the authors (2022).

The majority of Lattes research projects network are associated to only one professor, 42 projects have no collaboration and 30 have collaboration of the course professors (Figure 2A, Table 2), and one professor was linked to only one project. Another analysis, excluding projects with no collaboration among Nursing professors of UFU, was done to provide a better visualization of connections. Most professors were connected to the network and only a few of them were connected to only one aresta (Figure 2B, Table 2). Most connections of professor with project coordinators affiliation were relative to UFU, while the remaining affiliations presented few arestas (Figure 2B, Table 2).

Source: prepared by the authors.

Figure 2 Professors from Federal University of Uberlândia (UFU) Nursing under graduation Course from interaction network, obtained from Lattes registered research projects and ongoing projects. A. based on all active or that ended in 2021 projects. B. based on previous projects excluding the projects without collaboration with other nursing professor of UFU. C. based on institution coordinator affiliation. 

Two active professors had no connection to any CNPq research group. One retired professor maintained active registration in research groups. Two active professors had no registration to any research project in Lattes, and no retired professor had ongoing projects. Professors with no connection were not plotted in any network.

CNPq research group corpus obtained 98 observations distributed in 66 words. The 15^th most cited were: health (9); care (5); nursing (5); patient (3); prevention (3); safety (3); women (2); spirituality (2); cardiology (2); interprofessional (2); approach (2); rehabilitation (2); drug (2); process (2) and infection (2) (Figure 3A). Lattes research project corpus obtained 655 observations distributed in 351 words. The 17^th most cited were: health (18); patient (16); care (15); university (11); nursing (10); women (9); clinical (8); hospital (8); study (8); unit (7); safety (7); service (7); family (6); use (6); student (6); public (6) and perception (6) (Figure 3B).

Source: prepared by the authors.

Figure 3 Word cloud of the most common 50 words in the evaluated body of the corpus, from the Professors research intention of a Nursing under graduation Course corpus. A. based in the research groups title in CNPq corpus. B: based in ongoing or that ended in 2021 research projects title in Lattes corpus. 

3 Discussion

Analyzing the interactions in network research intentions between Nursing undergraduate professors, it was observed non structured networks, with biased relations (actors with too much or too little connections), isolated actors and neglected links. The identification of determinant factors of these relations was not made. Not including all network connection sources of interaction or the factor of prediction of these connections is similar to other studies and reflects a common limitation to this kind of analysis (^{GUIMARAES et al., 2009}). Some factors have been reported influencing network structures such as the presence of graduation course, previous relation of supervision and internationalization of the group (^{SILVA; MARTINS, 2016}; ^{NOGUEIRA; SILVA, 2017}). Therefore, presuming that the inclusion of other variables not approached in this study may explain more clearly the connections between professors from this university and their determinants is valid and suitable to other investigations.

Networks with few groups or even those who approach isolated actors are common in literature. The evaluation of cooperation structure aspects between researchers in the field of public administration and social management highlighted institutions with no collaboration or just a few (^{ROSSONI; HOCAYEN-DA-SILVA; FERREIRA JUNIOR, 2008}), as found here. It is still necessary to insert data to construct strong or weight networks (connection strength, regarding the number of projects and groups that interact), and better understand the relations and strength of collaborations inside the research institution. Independently, groups and actors were already highlighted, and in some cases assume more than one pattern in different networks, even though other variables and more reliable data (validated, since some inconsistencies were observed between curricula) should be approached to network construction. Similar behavior to the observed here highlights isolated actors and the formation of few collaboration groups (^{SILVA et al., 2006}).

Neglected links are evident in the interaction between professors and affiliations that were distinct between research groups and projects. Possibly, the groups biding is previous than the projects and reflect each professor’s formation history, although it was observed some projects bounded to institutions and with no bound to any research group in the same institution. These findings emphasize that projects or groups are more related to the research intention, since even though some professors are bounded to a research group of an institution, the same professors do not have any research project bounded to that institution, or vice versa. It is important to notice that a research project bounded to a research group may be from authorship of an UFUs professor as a coordinator, and that the coordinator may not register the collaborator institution, by that losing the intention.

It was not possible to bind all the actors of the projects and the research groups, since in most curricula authors were not identified (the authors name linked to the respective Lattes curricula). Previous orientation and research relations are common in networks (^{SILVA; MARTINS, 2016}; ^{NOGUEIRA; SILVA, 2017}). We infer that many group bonds are long standing and related to the Professors formation in Doctorate and Post-Doctorate programs, and that they are not ongoing in some cases. In some cases, this speculation is strengthened by the former registration in the research group and in the condition of student and absence of ongoing research projects with the group or in the theme.

The present study brings to discussion the “quantitative stress”, concerning the high pressure to produce among professors and researchers (^{MEIS et al., 2003}), that are constantly evaluated by bibliographic production metrics. In the Brazilian system these evaluation indicators metrics have the capability of validate and classify research groups or institutions, rebounding classification systems and international rankings in a global scope that includes Nursing (^{SCOCHI et al., 2013}; ^{NOBRE; FREITAS, 2017}). In contrast, the products are measured and classified, but the generation process of the products still remains unknown in most times (^{OECD, 2011}). Thus, many times actors see themselves obligated to collaborate or change their collaboration methods to enhance their productivity, regardless the knowledge of the theme and even other ethical aspects (^{SANTOS; RABELO, 2017}). In this context, it is important to comprehend how the research intention is structured and subsequently evaluate how it actually converts into a research product and the research relations in the future.

The exponent increase of works with collaboration and facility of access to its metadata makes the professors production and research group networks analysis assist the comprehension of interpersonal relations (^{DIAS, T.; MOITA; DIAS, P., 2019}), likewise the ones observed in this study. These networks can be heterogeneous, not dense and with minimum medium paths relatively low, and, by analyzing publication titles, regional expressions are highlight by its frequency in the word cloud (^{DIGIAMPIETRI, 2015}). The word clouds we have here are little informative and do not reflect very much about the studied themes. Alternatively, the projects could be categorized in research areas in future studies.

When analyzing the research theme (or research line), it was verified that education elements of Nursing Administration were linked to other themes, especially those related to management, education, politics and work, which supports that the technique application in more specific themes, lines or areas can present more informative results, even arising more generic words (^{FIGUEREDO; MACEDO, 2020}). Majority of occurrence of non-informative words have been observed in Nursing abstracts on thesis and dissertations, being the five most common words: Nursing, Nurse, data, study, and patient (^{AMORIM et al., 2020}). These finding are similar to the observed here, and emphasizes that the most common words health, care, Nursing, patient and university are little informative about the theme, actuation area or focus of the projects.

Lattes Platform and research groups registered in CNPq database are the main open access tool, at least in Brazil, to researchers’ data and their production in various areas, due to its easy access (^{DIAS, T.; MOITA; DIAS, P., 2019}; ^{BRITO; QUONIAM; MENA-CHALCO, 2016}). The present study used public data in those databases to establish the dynamic relation and evaluate research intentions. We highlight the importance of using other techniques and databases already described in literature to the same purpose, such as predicting link loss technique which evaluates the presence or absence of the researcher in determined group, with the model “Support Vector Machines (SVM)” (^{PEREZ-CERVANTES et al., 2013}). To determine which researcher is more “productive”, “collaborative”, “influent” and that can be an essential actor in the collaboration process is an important aspect, however, that can be a form of duress, pressure and stress to enhance productivity and collaboration. Those techniques must be used to strengthen the group, emphasize points that can be worked and not only as an evaluation metric or to discriminate a professor before another one.

4 Conclusions

Application of the network theory in CNPq research groups and Lattes registered projects in a Nursing undergraduate Course was effective in demonstrating the relations between the unit professors in different perspectives. This methodology was able to highlight fragilities (isolated actors, actors with few links, neglected links) and strengths (well-connected actors, low frequency of actors without interactions) of the relations in the research intentions of the group. These results can provide basis to future research stimuli policies to the Professors investigated.