Introduction

As reflected in the title of this work, its fundamental purpose is to validate a diagnosis of needs through a complex process of methodological triangulation. From this perspective, we find the three elements that compose the gears of this methodological experience of an environmental nature, namely:

Let us remember in this regard that the diagnosis of general needs can be considered as “a planned action aimed at obtaining information about the problem to be treated or the situation that we wish to modify with an educational or assistance intervention” (ÁLVAREZ ROJO, 2002, p. 54). For Álvarez Rojo (2002, p. 57-58), the main reasons for recommending needs assessment when a future social intervention is to be carried out are to:

However, and as the second element of this work, we highlight the term Environmental Education Centre (EEC)². Before trying to define an EEC, we show some key ideas of another term in which it is intimately incarnated, that is, with the term “environment”.

The first noteworthy aspect is that it would be a term of difficult terminological precision due to its complexity, lack of definition and interrelation with diverse and different factors (^{BACA; BOSKER, 2000}; ^{PEREIRO, 2001}; ^{CABANILLAS, 1996}). The second aspect is that despite this, we can say that the term “environment” is related to the concepts of ecosystem, habitat, natural resources and ecology, among others. From here, we can define environment as a concrete scenario formed by many elements in which man, both individually and organized into a social group of any scale and level of complexity, develops its multitude of activities (^{REBORATTI, 2000}). Another interesting definition is that indicated by the RAE (Real Academia Española [Royal Spanish Academy, RSA]) as a term referring to the circumstances surrounding people or things. This meaning coincides with one of the meanings of the word medio (medium), which would lead us to affirm that the expression “environment” is redundant. However, the terms medio and environment are used interchangeably to refer to the same concept (BUSTAMENTE, 1995).

Turning back to the EEC, we can affirm that the manual’s definitions configure them as establishments in which comprehensive programmes and processes of environmental sensitization are implemented based on knowledge and awareness of the defence of natural values, etc. Therefore, we can affirm that EEC becomes

[…] spaces for reflection, inquiry, research and citizen participation in environmental awareness processes in order to facilitate knowledge, sensitivity and the development of pro-environmental actions in the effort to minimize the environmental problems of the context where they are inserted. (^{GRACIA et al., 2013}, p.139).

The research that concerns us has as a frame of reference an Environmental Education Centre located in a rural municipality of the province of Granada (Spain), surrounded by a rich natural environment that has numerous facilities, resources and educational programmes that establish guidelines for sustainable intervention and environmental awareness.

The third element in the list is triangulation. It is not strange —indeed, it is typical—to find research reports, articles, papers, and communications of a scientific nature in which it is unashamedly stated to have implemented processes of methodological triangulation in its different typologies. Without going deeply into fundamental questions that could challenge these claims, we agree with ^{Yeasmin and Ferdousour (2012)}, ^{Torrance (2012)} and ^{Van Drie and Dekker (2013)} that many of these works do not comply with the logic of triangulation despite the fact that they fly its flag. What usually appears as a work of triangulation becomes what ^{Bericat (1998)} calls the strategy or pathway of complementation.

In this sense, complementation is not the same as triangulation because the final product of this type of multi-method design is usually a report with two very different parts, each of which exposes the results achieved by the application of the corresponding research pathways implemented (methods, information-gathering techniques, informants...). Although a degree of integration exists, it is minimal and its legitimacy is based on verifying the degree to which each angle is capable of revealing different and interesting areas of social reality. Faced with this position, however, in the logic of triangulation, the degree of integration increases in relation to complementation, and legitimacy is based on capturing similar facets of social reality from opposite poles, obtaining a final product in which the comparison is sought and is quite convergent and quite divergent. In our case, the pretence of multiple triangulation implemented is the convergent validation in the prioritization of a needs analysis in an environmental facility in the province of Granada (Spain).

Methodological triangulation as a validation strategy

For ^{Erazo (2011)}, triangulation in qualitative research is one of the basic requirements of credibility of the obtained findings. However, we have to go back a long time to the seminal works of ^{Campbell and Fiske (1959)}, ^{Denzin (1978)} and ^{Jick (1979)} to find triangulation as a plausible strategy for validating results. In essence, triangulation consists of documenting information and contrasting it according to different points of view, which is why we talk about the triangulation of techniques, agents, times, and even methodologies. As a contrast technique, it reveals the contradiction, inconsistency or convergence between data obtained by different sources, as proven by, among others, the works of ^{Bjurulf et al. (2013)}, ^{Brown et al. (2015)}, ^{Flick (2005)}, ^{López et al. (2013)}, ^{Teixeira et al. (2012)}, ^{Van Drie and Dekker (2013)}. In addition, with ^{Barusch et al. (2011)}, ^{Bekhet and Zauszniewski (2012)}, ^{Beattie et al. (2005)}, ^{Hind (2017)}, ^{Newman et al. (2013)}, ^{Oelinik (2015)}, ^{Okoe and Boateng (2016)}, ^{Schraeli et al. (2017)}, ^{Stutz (2016)} and ^{Torrance (2012)}, we can affirm that triangulation strategies allow the exercise of quality control over qualitative research processes, offering confidence tests and guarantees that the results and findings proposed there meet minimum requirements of credibility, rigor, veracity and robustness.

Triangulation consists of a procedure by which the researcher takes a series of precautions throughout the planning, design, development and diffusion of the research so that their work does not suffer from serious scientific biases. These strategies do not exclusively affect the techniques and instruments but also affect the agents that collect this information, the different moments in which the research is developed and organized, the methods, the different contexts, the sources of documentation, etc. Triangulation has been defined as a way to improve the results that a researcher obtains after applying a specific technique to his or her work (GORISSEN et al., 2012).

Therefore, triangulation is a strategy for the quality control of qualitative research, which gives us guarantees and confidence regarding the methodological goodness of our findings. The idea of quality in qualitative research is generally associated with the idea of credibility in the work of researchers (^{RODRÍGUEZ et al., 1999}; ^{SANDÍN, 2000}; ^{CRESWELL, 2005}; ^{RODRÍGUEZ-SABIOTE; LORENZO-QUILES, 2005}; ^{MORAL, 2006}; ^{MORSE, 2006}; ^{TÓJAR, 2006}; BROD, TESLER; CHRISTENSEN, 2009; ^{DENZIN, 2010}; OLABUÉNAGA, 2012). Some people, however, consider triangulation as a strategy that is too incomplete and weak compared to what they call crystallization (^{RICHARDSON, 1997}). While the triangle is erected in a rigid figure, fixed in two dimensions […] “the crystal combines symmetry and substance with an infinite variety of shapes, substances, transmutations, multidimensionalities, and angles of approach” (p. 92).

One might ask then if perhaps what this author advocates with an impeccable literary style is not similar to the complex processes of multiple triangulation that combine, contrast, compare, etc., different points and infinite levels of triangulation that comprise a vast and complex structure that can resemble the polyhedral contours that make up the crystallization.

In this context of questioning, it is important that the assessment of quality in design and development can and should be carried out either by the researcher himself or by other agents involved. In both cases, sufficient evidence must be given that the procedure followed in the investigation has been correct, that all relevant cases have been selected without ignoring those that could represent negative evidence in relation to our interests, that the data collection has been adequate and without pitfalls, and that the data analysis procedures have been followed without forcing the results in any tendentious direction. All these strategies constitute fundamental action guidelines for quality research (^{RODRÍGUEZ et al., 1999}); in all of them, the triangulation processes of information play a key role. “An exclusive trust in a method may polarize or distort the researchers view of the particular slice of reality that he or she is investigating” (^{COHEN; MANION, 2002}, p. 332).

The proposal implemented

To prioritize the analysis of the needs of environmental facilities through multiple analytical triangulation strategies, we propose a procedural model structured in various stages. In its development, we will start with a basic approach by ^{Gracia et al. (2013)}, whose fieldwork was developed during the spring of 2011 in an environmental facility of the province of Granada (Spain) and that consisted of the deployment of several techniques for collecting information: internal facility documents, taking photographs and conducting interviews with various agents. These agents (relatives of children with visual needs who visited the facility, environmental education teachers of the facility, and special education teachers not belonging to the facility) were selected by theoretical sampling strategies related to such environmental facilities. In this work, the authors proposed a process of multiple triangulation for the analysis and prioritization of the environmental facility needs of the province of Granada (Spain) that were addressed in a basic manner and in which conclusions were presented. Although this approach is legitimate, we have considered that it could be improved with the methodological triangulation model in phases proposed by ^{Rodríguez-Sabiote and Gutiérrez-Pérez (2005)} and ^{Rodríguez-Sabiote et al. (2006)} and that we will now detail.

The development of some type of methodological triangulation

Initially, it is obvious that it is necessary to implement some type of methodological triangulation in its different typologies. It is true that the logic of the model contemplated favours qualitative methodologies and techniques, given that it is based on presences vs. absences of the needs and less on the quantitative dimension of those needs. In our example, ^{Gracia et al. (2013)} proposed a nested multiple triangulation with two points of triangulation (T1 and T2) and various levels of triangulation:

Triangulation point 1 (T1): Techniques: Internal documents (Centre project), other documents, photographs and interviews (4 levels of triangulation).

Triangulation point 2 (T2): Agents: Relatives of children with visual needs, special education professionals, and environmental education professionals (3 levels of triangulation).

Source: Own elaboration.

Figure 1 Structure of the multiple triangulation developed 

Results

The prioritization of needs from the matrix of presences vs. absences

Another alternative for the prioritization of needs is an analysis of the map of the matrix of the presence vs. the absence of needs. Taking its binary arrangement as 0 = absence and 1 = presence, an ideal method to address its few execution restrictions is called cluster analysis. In our case, we have implemented the hierarchical cluster analysis using the Ward method and taking the quadratic Euclidean distances as dissimilarity measures. The main results after its implementation, the tree diagram (dendogram) and the membership table are shown below. However, before assigning each need to a type of priority cluster, we must calculate its coverage index averages at the ratio and percentage level based on the cluster table initially provided by the inferred model. Once the initial structure of belonging for each need to its corresponding cluster is known, the averages are calculated at the ratio and percentage level with the objective of assigning each need with the same classification used in the percentiles; in other words, we identify the lowest averages with low priority and vice versa.

Table 4 Clusters, according to priority and complementary statistics 

Cluster types, according to priority	Original membership cluster	Number of needs in the cluster	Needs that make up the cluster	Averages of coverage index	Average% coverage index
cluster 1: low priority	3	6	4,5,7,10,11,12	.2167	21.67%
cluster 2: moderate priority	1	7	1,8,9,14,17,19,	.2814	28.14%
cluster 3: high priority	2	8	2,3,6,13,15,16,18,21	.4763	47.63%

Source: Own Elaboration.

Source: Own elaboration.

Graphic 1 Dendrogram obtained after the application of the cluster analysis implemented 

Conclusions

From the results obtained, we can infer two types of conclusions. On the one hand, we make what we can call methodological findings, and on the other hand, we make findings that are strictly related to the object of diagnosis, in our case, the needs of the environmental facility related to the unavoidable task of implementing an intervention project for visitors with sensory impairment. In any case, both are clearly related and it is impossible to understand one without the other.

Of the first, that is, those of a strictly methodological nature, we must bear in mind that the basic objective of this work has been the diagnosis of needs from an actionable research perspective to prioritize the demands presented when beginning the multiple triangulation processes. However, the premise of the approach has been multiple methodological-analytical triangulation as convergence. We have attempted to bring together—from different prisms— the degree of concordance or agreement established in terms of its importance given the needs highlighted in the diagnosis prior to the intervention project.

In this sense, as seen from the agreement coefficient finally obtained (κ = .603), it can be said that an agreement has been established consisting of how different agents or participants and different information collection techniques have highlighted the needs that, in its opinion, were more important in an environmental facility. Consequently, this legitimates the hierarchy of needs established as useful for the intervention project put in place. We must not forget that in the qualitative methodology, one of the most powerful strategies to validate our conclusions is triangulation in its different typologies.

Until now, the problem has been that even though triangulation processes are used and there have been many research reports, only a few of those reports have presented totally combined results. That is, it is typical to conclude that there are watertight findings from the different levels and points of triangulation that are commented on and discussed separately. Thus, in our example, starting from three levels with two triangulation points (3Nx2P), we would obtain results A₁/A₂ + B₁/B₂ + C₁/C₂ and, therefore, conclusions A₁/A₂ + B₁/B₂ + C₁/C₂. From this perspective, our proposal lies in obtaining combined results: A_1/2 B_1/2 C_1/2 and consequently, conclusions: A_1/2B_1/2C_1/2 intimately confronted in search of a possible or probable consensus.

Last but not least is to determine the degree to which the results obtained by the different views considered are similar, provided that the initial premise is both triangulation and convergence. In this sense, the triangulation model implemented will have greater validation power the greater the agreement reached.

Thus, even with all the limitations of this study, we can confirm that methodological triangulation is a useful tool to validate and legitimize results based on qualitative quality criteria such as confirmability, dependence...

Second, we find what we have called findings or conclusions related to the object and research objectives of the study. In this sense, we can affirm that there is a justified consensus that establishes three degrees of importance in which the 21 prioritized needs in this diagnosis are grouped in the following dimensions: there is a group of high priority needs that involve the importance of carrying out intervention projects related to gardens and their adaptation to spaces, resources and the timing of actions. We could call this high-priority group the Adaptation of Socio-environmental Actions for the Sensory Disabled.

However, there is a second level or group of needs of moderate priority related to the importance of developing actions on ecological teaching gardens to promote training, environmental awareness and knowledge of organic farming from processes adapted to the visually disabled group to encourage their integration. We could call this group the Training and Awareness for the Sensory Disabled through Socio-environmental Actions.

Finally, we find a group of low-priority needs that derive from the planning, execution and adaptation of the spaces and the intervention for the development of these socio-environmental initiatives. We could call this group of needs Design and Planning of Socio-Environmental Actions for the Sensory Disabled.

We can conclude, therefore, that to make feasible and develop relevant socio-environmental interventions in environmental education facilities, a good needs diagnosis is required that can be supported at a methodological level in processes involving the methodological-analytical triangulation of data that validate the demands derived from said process.