Concerns regarding methodological rigor and validity of the findings in different research works have existed for some time now in the Social Sciences. According to Ollaik and Ziller (2012), the concept of validity derives from quantitative methods, and aims to achieve a better understanding of the extent to which a given measure correctly represents the proposal of a study. In this way, it is possible to confirm how the differences between quantitative research (objective) and qualitative research (subjective and interpretative) lead researchers to adopt only one of them to implement their scientific work (TEIXEIRA; NASCIMENTO; CARRIERI, 2012). This might help explain the preferences towards quantitative research, as well as the existence of a significant amount of researchers who are very cautious in keeping their works highly scientific (BUELENS et al., 2008).

In spite of the intention of validating a single aspect, of remaining distance from some epistemological paradigms (e.g. Post-modernism), particularly because they are closer to Positivism, diverse qualitative approaches in research intend to create only works that provide more credibility. Kvale (1995), for instance, suggests some validity approaches in qualitative research, despite criticizing the so called “Holy Trinity" – meaning the three concepts: reliability, validity, and generalization. In this matter, one of the most popular ways of seeking confirmation in qualitative research is through the strategy of triangulation, which among other aspects, advocates the use of multiple methods. Such research strategy has also been described as convergent methodology, multi-method/multi-traced (CAMPBELL; FISKE, 1959) and convergent validation (WEBB et al., 2012).

The research interest in triangulation has not been recent. Ever since the idea was first conceived in the 1950s, with some studies in the field of psychology by Campbell & Fiskie (1959) with their "multiple operationism", followed by its popularization in the 1980s (BAZELEY, 2002; 2009) triangulation has remained a matter of debate and discussion, which has provided opportunities for further advances and contributions. The history of scientific research is full of attempts to combine or mix diverse collection, analysis and interpretation methods within the same study. Triangulation is a widely debated subject, in diverse areas, as in the case of human sciences (e.g., MORAN-ELLIS, et al., 2006; KONECKI, 2008; HUSSEIN, 2009; among others), health sciences (e.g., GUION, 2012; among others), and more specifically, applied social sciences (e.g., YAUCH; STEUDEL, 2003; IKEDA, 2009; YIN, 2015; TEIXEIRA; NASCIMENTO; CARRIERI, 2012; OLLAIK; ZILLER, 2012; ZAPPELLINI; FEUERSCHÜTTE, 2015; GIBSON, 2017).

There is no agreement among the different authors regarding the methodological perspective of triangulation to be used. Sometimes groups of researchers even criticize the strategy (e.g., BAZELEY, 2002; BRYMAN, 2007), while others advocate it (e.g., DENZIN, 1978; JICK, 1979; FLICK, 2005a; 2005b; PATTON, 2014). Moreover, some are constantly looking for ways to enhance it by trying to find alternatives for validation (CAMPBELL; FISKIE, 1959; GOLAFSHANI, 2003; SHENTON; 2004).

Diverse specialists have noted the convenience of combining methods, given the weaknesses found in projects oriented by a single methodological route. However, the main previous promoters of triangulation (SMITH, 1975; DENZIN, 1978; FLICK, 1992; WEBB et al., 2012; among others) do not explain how it ought to be executed. Most graduate programs prepare students so that they can make use of a given methodological approach or method, but such programs rarely prepare them for the possibility of combining methods (multi-method). For this reason, it is relevant to undertake deeper studies regarding the methodological approach, as a means to provide enough detail to accurately explain the ways in which the convergent data are collected, examined and interpreted. Thus, the main objective of this paper is to discuss quality in qualitative research, through a systematic analysis of the typologies of the triangulation strategy, highlighting the following quality dimensions in research: reliability, credibility, transferability, and confirmability.

In order to reach the objective proposed in this essay, an exhaustive theoretical research has been carried out, by means of which it a deep look into seminal works has been possible. Such works have introduced discussions ranging from mixed methods of research in the 1970s (e.g. SMITH, 1975; DENZIN, 1978; JICK, 1979) to papers recently published in specialized journals, such as the Journal of Mixed Methods Research (e.g. FLICK et al., 2012; FIELDING, 2012; KERN, 2016) and Organizational Research Methods (e.g. TURNER; CARDINAL; BURTON, 2017; GIBSON, 2017).

This paper is divided into five other sections, in addition to this introduction. The second section briefly provides readers with literature on triangulation and its role in Social Sciences. The third section introduces the main objectives and types of triangulation. A Table 1 is presented to sum up the state of the art on purpose of triangulation and their types. The fourth section brings an important contribution, as it systematizes the role of triangulation in the quality of qualitative research works. A Table 2 is presented explaining the central aspects for the enhancement of credibility in qualitative research. Also, dimensions like transferability, confirmability and reliability are examined. In the fifth section, the limitations of triangulation are addressed, and attention is paid to avoid deviations. The last section provides our final considerations and the contributions made to the field of study.

The origins of the concept of triangulation are not linked to human and social sciences, but to military science. “Due to navigation and topography, triangulation is frequently understood as a method used for establishing a position” (COX; HASSARD, 2005, p. 109). Currently, with the new satellite technologies available, triangulation is used by the military for discovering the exact location of a cell phone, a radio transmitter or any other communication equipment used by the enemy. For this purpose, the direction and intensity of the waves being transmitted by two devices are measured, and a third receiver or transmitter is located. Thus, basic principles in geometry guarantee greater precision achieved from the incidence of multiple points of view.

In Human and Social sciences, the term ‘triangulation’ is used in a less literal manner, as well as in a more ambiguous one. When a researcher looks at an object from a given perspective, he or she needs to think from at least another couple of viewpoints as perspectives capable of adjusting the right ‘distance’ and ‘angulation’ of concepts, thus, relocating him or herself definitively. Accordingly, organizational researchers have the opportunity of improving the precision of their assessments, if they proceed with diverse methodologies, collecting data in different ways, analyzing such data with different methods, or even including the participation of other researchers who study the same phenomenon (FLICK, 1992; 2005b).

In 1959, the term ‘triangulation’ was used in psychology by Campbell and Fiske, who intended to empirically test the results obtained by using different quantitative techniques (CAMPBELL; FISKE, 1959). They introduced the concept as a synonym for convergent validity when they presented their multi-method matrix. A few years later, in 1966 Webb, Campbell, Schwartz, and Sechrest used Campbell and Fiske’s idea, and broadened its concept, suggesting that data collection from different sources, as well as their analysis, would enhance the validity of the results (WEBB et al., 2012). Those authors, and also Jick (1979), redefined Campbell and Fiske’s idea of 1959 as the combination of qualitative and quantitative methods, stating that such approaches would better be seen as complementing rather than competing with each other.

In 1978, Denzin, with the help of this concept, claimed that a hypothesis tested with different methods should be considered more valid than another hypothesis, tested only with one method (DENZIN, 1978). The author defined triangulation as a combination of study methods for the same phenomenon. Thus, the convergence or the agreement between two methods is likely to support the belief according to which results are valid, and not only a methodological artifact. Maxwell (2012) states that triangulation reduces the risk of having the conclusions of a study impaired due to the shortcomings and limitations of a single method and, for this very reason, ends up producing more credible conclusions. In the 2000s, Denzin and Lincoln stated that in Social Sciences, triangulation does not consist of a single tool or strategy for validation; yet, it is an alternative to it. The use of combined methodological perspectives, of diverse empirical materials and the participation of several researchers in a single study need to be seen as a strategy for more rigorousness, scale, complexity, variety and depth in any research work (DENZIN; LINCOLN, 2011). So, triangulation means being able to look at the same phenomenon, or research topic, through more than one source of data. Information coming from different angles may be used to confirm, develop or illuminate the research problem. This strategy limits personal and methodological biases, and increases the possibility of reproducing the findings.

From these concepts, it is possible to infer that triangulation, or the use of multiple methods (convergent methodology, convergent validation or other synonyms) is a strategy that can be used by researchers in diverse areas. This concept does not only embody one of the ways of combining qualitative methods (FLICK, 2005a; 2005b), or articulating quantitative and qualitative methods (FIELDING; SCHREIER, 2001; FLICK, 2005c). It is also a concept that breaks with the methodological hegemony of those advocates of mono-method approaches (TASHAKKORI; TEDDLIE, 1998). The following sections include descriptions of some of the possibilities and limitations of the proposed strategy.

Triangulation can combine methods and collection sources of qualitative and quantitative data (interviews, questionnaires, observation and field notes, documents, among others) as well as different methods for data analysis (content analysis, discourse analysis, descriptive and/or inferential methods and statistics, among others). Its objective is to make a contribution, not only by examining the phenomenon from multiple perspectives, but also by widening the understanding of the matter among researchers, making new and deeper dimensions possible. Triangulation makes a contribution to the creation of inventive methods; it consists of new ways of grasping a problem so as to balance with conventional data collection methods.

The use of multiple methods can even help researchers discover misleading dimensions in a phenomenon. Different points of view can bring elements to that, which do not fit a theory or model. In this way, old theories are modified and new ones are developed. The proposed strategy can also lead to the synthesis or integration of theories. A skillful researcher will use the qualitative data to enrich and illuminate the results of quantitative methods, and vice-versa (JOHNSON; ONWUEGBUZIE; TURNER, 2007; GREENE, 2008; OTTOBONI, 2009). To sum up, Vergara (2015) states that triangulation can be seen from two different perspectives: as a strategy that contributes to construct validity, and as an alternative option to obtain new knowledge, through new viewpoints.

In terms of typologies, in the end of the 1970s, Denzin (1978) identified four types of triangulation: (i) data triangulation; (ii) theoretical triangulation; (iii) researcher triangulation; and (iv) methodological triangulation. Data triangulation means collecting data in different periods of time and from different sources, so as to obtain a richer and more detailed description of the phenomena. By differentiating the subtypes of data triangulation, Denzin (1978) suggested studying the phenomenon in different moments (so as to explore time differences), places (for comparative research), and with different individuals.

Theoretical triangulation refers to the possibility of exploring multiple theories, as a means of interpreting the same data group. According to Guion (2002), this type of triangulation aims at involving researchers with different theoretical approaches and from diverse areas of knowledge, into the analysis of the same phenomenon. In this way, the author offers a complementary view to that of Denzin´s (1978). It is likely to be a doubtful theory, in case researchers’ choices end up being misled.

Researcher triangulation consists of involving diverse researchers in the study of the same matter, while assuming that different researchers could contribute with varied perspectives, considerations and analyses. This is similar to that which Lincoln and Guba (1985) called member checking. The use of diverse researchers in the same study allows to obtain multiple observations in a given field, at the same time enabling the participants to debate their points of view, which helps reduce biases. In this case, the intention is to compare the influence of several researchers on both problems and results. Unlike the previous case, only the theory of a single field of knowledge is taken into account.

Methodological triangulation refers to the use of multiple methods as a way of obtaining more complete and detailed data about the phenomenon. This kind of triangulation is the most applied one. It involves the combination of methods, such as interview and observation, in order to better understand a given reality. In practice, the objective is to avoid the biases of a mono-method strategy. The articulation between interview and observation can be found in leadership studies, in which the efficiency of a leader can be studied by means of both interviews with leaders and, concomitantly, the study of their behavior. Complementarily, it is also possible to assess performance through document collection and analysis. The premise is that if multiple and independent measures lead to the same conclusions, then they are able to provide a more accurate picture of the leadership phenomenon.

As a way of better presenting several objectives of the triangulation and their respective types, we organized Table 1, which summarizes the state of the art on purpose of triangulation and their types.

Table 1 Table 1 The state of the art on purpose of triangulation and their types

Purpose of Triangulation	Suggested Typology	Authors
The authors described a triangulation as a convergent methodology performed by using different techniques and methods.	Multi-method and multi-trace	CAMPBELL; FISKIE (1959)
Triangulation as a combination of study methods for the same phenomenon. Thus, the convergence or the agreement between two methods is likely to support the belief according to which results are valid, and not only a methodological artifact.	Data triangulation, theoretical triangulation; researcher triangulation and methodological triangulation.	DENZIN (1978)
Increasing the quality of research in social sciences from the combination of quantitative and qualitative methods of analysis.	The same proposals by Denzin (1978).	JICK (1979)
The author argues triangulation as a way to increase validity of evaluation and research finding. Additionally, it warns that all resulting from triangulation should be discussed: inconsistency, contradiction, and convergence.	Data triangulation and methodological triangulation.	MATHISON (1988)
Increasing the quality of research in social sciences from the combination of quantitative and qualitative methods of analysis.	Simultaneous triangulation and sequential triangulation.	Morse (1991)
The triangulation is ‘re-presented’ as ‘metaphorization’—in terms of process and movement between researcher-subject positions. Rethinking the lines and angles of enquiry in triangulation, the author suggests a shift from the ‘triangulation of distance’ tradition to a more reflexive consideration of ‘researcher stance’. This movement is represented across three perspectives: the researcher as a follower of nomothetic lines; the researcher as the taker of an ideographic overview; and the researcher as the finder of a particular angle. The triangulation can provide a more complete and comprehensive picture of the phenomenon to be investigated.	They did not present typology.	COX; HASSARD (2005)
The authors presented a final discussion on the limitations, advantages and potential applications of metatriangulation to position it in the field of organizational theory.	Metatriangulation theory-bulding strategy.	LEWIS; GRIMES (2005)
The authors set out different rationales for using more than one method, then they develop a definition of integration of methods as a specific kind of relationship among methods.	Triangulation as a way for (i) integrated methods; (ii) separate methods, integrated analysis; (iii) separate methods, separate analysis, theoretical integration.	MORAN-ELLIS, J.; ALEXANDER, V. D.; CRONIN, A.; DICKINSON, M.; FIELDING, J.; SLENEY, J.; THOMAS, H. (2006)
The author performs triangulation of methods and data in order to gain a complex image of reality, active reconstruction of reality, and double or multiple comparison.	The author used triangulation of data, methodological triangulation and, in part, theoretical triangulation.	KONECKI (2008)
The author dealt with the dichotomy of paradigms between quantitative and qualitative. She also pointed to the paradigm of mixed methods of research. It is concluded that the advance of the science of the Administration depends on the presence of multiple paradigms.	Multiparadigmatic perspective.	OTTOBONI (2009)
The author used "the within-and between types" of triangulation to combine two or more methodological approaches, theoretical perspectives, data sources, investigators and analysis methods to study the same phenomenon.	The use of triangulation for confirmatory and for completeness purposes.	HUSSEIN, A (2009)
Increase the validity of the results from data collection in different sources and analysis by different methods. Quantitative and qualitative methods should be seen as complementary and not as rivals.	Multi-method and multi-trace, convergent validation.	WEBB; CAMPBELL; SCHWARTZ; SECHREST (2012)
Triangulation (2.0) is defended by the author as a potential form of social change.	A mono-paradigmatic approach, based on the use of critical interpretive methodologies.	DENZIN (2012)
Enhancement of illustration, convergent validation and development of analytical density or "richness".	Data triangulation.	FIELDING (2012)
The authors indicated that the combination of methods has been adopted mainly in the purpose of convergent validation without discussions between different worldviews and different scientific knowledge views.	Multi-paradigmatic perspective.	TEIXEIRA; NASCIMENTO; CARRIERI (2012)
Triangulation is a way to check and establish validity	Data, investigator, theory, methodological and environmental triangulation.	GUION (2012)
Triangulation as a means to assess three different conceptions of validity: (i) related to the research formulation phase (prior validity); (ii) related to the research development phase (internal validity), and; (iii) related to the research results phase (external validity).	Data, investigator, theory, methodological, environmental, peer valuation triangulation, focus groups, protocol, transparency, coherence and communicability.	OLLAIK; ZILLER (2012)
The authors introduce and theorize different aggregation strategies commonly used in triangulation, such as weighted and simple averages or ‘the winner takes it all’ strategy.	Data triangulation.	LEUFFEN D.; SHIKANO S.; WALTER S. (2013)
The authors indicated the existence of several types of triangulation, although most follow the classification of Denzin (2005), which distinguished the triangulation of data, theory, researchers and methodology.	They presented the different types of triangulation: data, method, methodological, and evidences of triangulation.	ZAPPELLINI; FEUERSCHÜTTE (2015)
The author relies on the proposal of Leuffen, Shikano, and Walter (2013) to discuss the triangulation of data sources.	Data triangulation.	KERN (2016)
To increase validity and to provide generality, accuracy and simplicity in the construction of theory. The authors propose a roadmap to assist researchers in designing triangulation-based mixed methods research.	Convergent Triangulation, Holistic Triangulation, and Convergent and Holistic Triangulation.	TURNER; CARDINAL; BURTON (2017)
The author proposes that triangulation is a means to increase the value of research.	Multi-paradigmatic perspective, multi-method and proposal of innovative and less conventional alternatives, such as Content Analysis combined with Quantitative Survey Data.	GIBSON (2017)

Source: Produced by the authors.

Within methodological triangulation, Denzin (1978) identified two subtypes of methodologies: intra-method or ‘inside the method’ triangulation and inter-method or ‘between methods’ triangulation. Intra-method triangulation uses diverse techniques of a given method to collect and interpret data. In the case of quantitative methods, such as opinion surveys, this can take the form of multiple scales or indicators focused on the same concept. Inter-method triangulation will be addressed in more detail in the following section. It is important to note that besides the types of triangulation proposed by Denzin (1978), Guion (2002) emphasizes environmental triangulation. In this type of triangulation, data collection techniques are used under different environmental circumstances, such as the time of day, day of the week, season of the year or temperature, among other elements and variables that might influence the phenomenon under examination. In other words, the use of mixed methods in research requires a lot more than just combining quantitative and qualitative approaches, because there are situations in which the epistemological paradigms require much more than mere combinations. We need to understand deeply the investigated context, under a time-space perspective (DENZIN, 2012). Working with mixed methods in research sometimes requires that data be synthesized, because they are equivalent and proportionate. In other cases, contradictions must be highlighted, especially when resulting from different epistemologies, and must be resolved conceptually (FIELDING, 2012). The between method triangulation (or mixed methods research) will be better dealt with in the next session.

Triangulation refers to the use of multiple methods (or mixed methods research), techniques and data sources, as it tries to overcome the flaws of an investigation or method. For some researchers, this strategy leads to a more consistent and objective picture of reality. According to Gaskell and Bauer (2010), the strategy of triangulation is an institutionalization method, of theoretical perspectives and methods, aiming to reduce the inconsistencies and contradictions of research. It contributes to validity and reliability, by providing a more trustworthy picture of the phenomenon, through convergence (PATTON, 2014). Thus, Jick (1979) states that triangulation is a research strategy of convergent validation, both of multiple methods and of multi-approach to data.

There are three important reasons for employing mixed methods research: (i) illustration; (ii) convergent validation; and (iii) analytic density (or “richness”). The (i) illustration reflects how to demonstrate the reality of research. Cold statistical data may be better interpreted or understood by interviews, giving more life to numbers; (ii) convergent validation occurs when different data methods point to similar results, giving greater strength to the collection made and analysis; and (iii) analytic density materializes by the greater breadth and depth given the particular object of study, through the use of mixed methods research (FIELDING et al., 2012). The mixed method research is appropriate, especially for the investigation of complex phenomena either because of the complexity of the field of study or the problem to be solved, or because of the need for multiple levels of perception (FLICK et al., 2012).

An interesting practical guidance for triangulation says that if the sources are systematically biased in the same direction, the “winner takes it all” strategy is the most accurate one. In short, the choice for triangulation strategies depends on the number of sources, their reliability and on one’s level of independence "(LEUFFEN et al., 2013, p. 49). Kern (2016) presented practical guidance on how to apply the strategy of triangulation using three different types of sources for convergence and divergence of light sources.

Similarly to quantitative theories, and with the intention of contributing to rigor, scale, variety and depth in research, Guba (1981) proposed four criteria to be considered by qualitative researchers to produce reliable studies: (i) credibility, corresponding to internal validity in quantitative research; (ii) the possibility of transferability, in spite of external validity or generalization; (iii) confirmability in spite of objectivity; and (iv) reliability, also approached in positivist research. These elements are detailed in the next sections.

Triangulation has some shortcomings. First of all, replication is difficult to achieve in qualitative methods, particularly, due to the difficulty in replicating them, mainly because their basic concepts are tied to the researcher’s point of view. Moreover, caring for the non-use and consequent legitimization of a dominant method or any preferred method by the researcher is important. The same goes for data collection and analysis techniques. Obviously, a particular method or technique may, in fact, be stronger or more suitable for a given phenomenon under examination, but whichever the method or technique chosen, it has to be made explicit and justified. Otherwise, the objective of triangulation is very likely to be subverted. As a matter of fact, data that had previously derived from different methods being used to produce a unitary vision that must pass as “truth” seems very simplistic. Once reality is regarded as multifaceted, ‘truth’ starts functioning merely as a limit and a guideline (ALMEIDA; PINTO, 1986).

According to Becker (1996), even if some methods supposedly represent ‘truth’ more than others, ‘truth’ itself is a misleading view. Based on the premise that different paradigms underlie qualitative and quantitative methods, which at the same time represent different ways of conceiving reality, it is possible to believe, in line with Fielding and Fielding (1986), that researchers can incorrectly interpret the converging and diverging points among the collected data, with very diverse methods.

Another limitation lies in the variety of existing perspectives on validity, according to the following paradigm: whereas a positivist model requires rigorousness in the process of the method application, a constructivist model, for instance, is no less rigorous; yet, it concedes that it is the interpretation processes that make the finding of the truth possible and not the methods (LINCOLN; GUBA, 2012). Fielding and Schreier (2001) observe that the option of multiple sources of information may be misleading as a way of preventing mistakes from happening, since if there are actually any mistakes in any of the methods, the duplication of those mistakes will follow suit. Consistent with this view, hardly ever results coming from different methods will in fact confirm one another.

Cox and Hassard (2005) point to the fact that triangulation is not restricted to validity; it also allows obtaining a more thorough and holistic picture of the phenomenon under examination. On this matter, Kelle (2001) emphasizes that taking the notion of complementarity of methods into account, convergence or divergence of results alone should not be taken as a path. In other words, for the author confirming or invalidating results obtained from applying different methods does not match the use of triangulations intentions; rather, triangulation has to do with producing a better picture of the phenomenon under examination, a more complete one, compared to the case in which a single method is used. Therefore, triangulation requires a good deal of creativity and ingenuity in the data collection process, as well as accurate interpretations of it all.

After the aforementioned discussions signaled here, triangulation can be said to be an alternative option for validating processes (DENZIN; LINCOLN, 2011). Diverse actions such as applying a group of methodological perspectives together, diversifying empirical materials and making the participation of various researchers in a single study possible need to be seen as an opportunity to increase rigorousness, diversity and depth in research in the field of Social Sciences. This paper aimed at systematically discussing quality in qualitative research through the use of triangulation. It is possible to see this contribution evidenced in the discussion developed in the fourth topic, particularly, in Table 2, which presents some guidelines for the improvement of credibility in qualitative research.

Studies using a single method, approach, or research technique seem to be more susceptible to mistakes when compared to other studies that use multiple ones. Triangulation or other multi-method strategies provides a powerful alternative paradigm that makes it possible to reach more informative, complete, useful and balanced research results (MATHISON, 1988). In the area of Management, a multidisciplinary field involving the presence of multiple paradigms, there are different contexts as compared to the field of Natural Sciences, in which one paradigm replaces the other. As a means of enabling meaningful advances in the fields of Management Sciences, making studies more relevant and keeping them as scientifically rigorous as possible is necessary. Partly, it depends on the abilities of the researchers to find their own epistemological path(s). Thus, triangulation emerges as an alternative option for encountering such paths. However, the development of mixed research methods has been put off so far, since researchers still bear concerns surrounding the (im)practicability of integrating qualitative and quantitative research methods (BRYMAN, 2007).

Apparently, one of the major concerns of qualitative researchers is the difficulty of dealing with existing prejudices surrounding quantitative methods or techniques, and vice-versa. Furthermore, one thing that goes particularly against qualitative research is the difficulty experienced by researchers when achieving validity and generalization standards that quantitative research does allow researchers to reach. In this sense, promoters of qualitative research question the objectivity of positivist studies and their impracticability in the Social Sciences. The multi-method approach has become a strategy that enables researchers to visit both paradigms.

Additionally, this paper leads to some advances in the discussions about the use of triangulation and its relative achievements in terms of quality of what is produced, particularly in terms of rigour, scale, complexity, variety and depth in research in the Social Sciences. For this purpose, it has been necessary to discuss and systematize the concepts of credibility (as corresponding to internal validity), transferability (as referring to generalization or external validity), confirmability (as parallel to objectivity), and reliability. We hope that with this contribution, in addition to the systematization proposed in Table 2, we could help to bridge the gap revealed by Turner, Cardinal and Burton (2017) pointing out that, in spite of the wealth of studies on triangulation, its application is still rare in Organizational Studies.

Indeed, any study requires setting its philosophical foundation. Moreover, all researchers need to be aware of the worldview that implicitly or paradigmatically guides their studies (CRESWELL; CLARK, 2010). This means that even though researchers normally live surrounded by a high degree of uncertainty, it is necessary for them to find their own epistemological and ontic route, as a way of guiding their own conclusions and methodological decisions. Such is the path leading altogether theory and practice, by widening existing knowledge and propelling new findings (MORIN, 2010).

Finally, it is important to consider that scientific knowledge, particularly in the case of Applied Social Sciences, does not aim at simply decoding reality and creating a final understanding about the way in which facts arise; rather, it aims at comprehending reality as the result of a social construction and emplacement. The empirical verification in the process of research makes it possible to unravel the relevance of rational conceptions and developed theories by observing parts of the world as they are experienced (BACHELARD, 1996).