An Integrated Mixed-Methods Framework for Studying Expert Geographical Reasoning in Climate Change Education: A Spain–France Comparative Design

Abstract

The teaching of climate change in geography is frequently reduced to a curricular content, whereas it is a systemic, multi-scalar and axiologically contested phenomenon whose understanding calls for expert geographical reasoning. This article is a conceptual and methodological contribution: it sets out the epistemological foundations and the research design of a comparative study of climate change teaching in Spanish and French geography education, and analyses no new empirical dataset. The framework rests on four analytical continua—organisation of knowledge, causal structure, treatment of scale and the fact–value relation—derived from a single operative definition of expert geographical reasoning and formulated so that one grid may interrogate curricular texts, focus groups, interviews and a questionnaire within a multiphase exploratory sequential design of qualitative priority. Its first phase has been conducted and published: a focus-group study that calibrated the precursor categories of the present continua and showed that value-laden and epistemically regulated registers coexist within expert discourse. Those findings are used here as prior published evidence; the three remaining phases are presented as a proposed design. It is argued that the Spain–France comparison operates as a device of epistemological denaturalisation, and that coherence derives not from accumulating instruments but from subordinating them to one analytical lens, whose risks are also discussed.

1. Introduction, Statement of the Problem and Aim of the Study

Over the past two decades, European education systems have moved from models centred on the transmission of content towards approaches oriented to the development of competences that enable learners to understand, interpret and act upon complex problems [1]. This shift entails accepting that learning geography does not consist solely in acquiring conceptual knowledge about territory, but in developing specific forms of reasoning capable of analysing multi-causal phenomena, taking informed decisions and participating critically in society, which requires articulating geographical thinking with the spatial competences proper to the discipline. Within this context, climate change constitutes one of the most demanding educational scenarios, since it requires the integration of scientific knowledge, geographical thinking, critical literacy, spatial competences and the capacity to operate in contexts characterised by uncertainty, controversy and socially situated decision-making [2,3,4]. Climate change education has become consolidated as one of the central challenges of contemporary geography education [5,6]. Reducing that challenge to the incorporation of a content item—ensuring that students know the causes, consequences and measures relating to global warming—nevertheless amounts to an insufficient reading of the problem. Climate change is not one more curricular object: it is a systemic phenomenon, with cross-sectoral effects at different scales. This makes it a socially controversial problem [7], whose understanding demands modes of reasoning that the traditional teaching of geography, descriptive and encyclopaedic [8], rarely mobilises.

To this must be added a sociological context in which post-truth, disinformation and the erosion of scientific consensus are especially visible around extreme episodes such as heavy rains, cut-off low events, heatwaves or wildfires, which, given their repercussions, are charged with emotions and visceral feelings close to the local and to identity, and which generate the spread of hoaxes and fake news [9]. This context, together with the magnitude of the phenomenon, makes climate literacy a central concern [10]. From this perspective, climate literacy must be understood as a transversal competence and as a complex cognitive process, and not as the simple acquisition of information about climate. Understanding climate change requires interpreting data, locating processes, recognising spatial patterns, relating scales, distinguishing evidence from opinion and assessing the social and territorial implications of human decisions.

Geography education, from the holistic dimension of geographical knowledge that analyses the interaction between physical, human, spatial, temporal and environmental phenomena at various scales, thus occupies a privileged position, since it makes it possible to articulate spatial thinking, systemic reasoning, territorial understanding and critical citizenship around a phenomenon that is simultaneously physical, social, economic, political and ethical. What counts as legitimate and powerful knowledge? How is evidence distinguished from opinion, and what role do values play in the interpretation of socio-environmental phenomena? To what extent should the cognitive processes proper to geographical reasoning be subordinated to merely conceptual knowledge? From this perspective, the didactic problem branches out in several directions. On the one hand, understanding climate change requires systemic thinking capable of integrating multiple causality, feedbacks, accumulations and/or delays—processes that even highly educated populations tend to simplify through linear schemas [11,12]—and a multi-scalar sensitivity that articulates the global and the local [13] within such multi-systemic schemas (the notion of “glocality”).

On the other hand, and through the prevailing sociological context, the character that climate change takes on as a socially live issue—a question socialement vive in the French didactics tradition: a question that is simultaneously live in social debate, live in the scientific field and live in the knowledge that students and teachers already bring to the classroom, and which therefore cannot be taught as settled curricular content [14,15]—places teachers before the tension between the transmission of scientific evidence and education in values: between educating “about” climate and educating “for” climate [16,17]. That tension is resolved neither through activism nor through passive neutrality; rather, it refers to the capacity to distinguish facts, interpretations and value judgements without dissolving the boundaries between science and normativity [18]. Climate change thus condenses the features of powerful geography—relational, multi-scalar and with conceptual depth [19,20]—and, at the same time, the most persistent deficits of its teaching: fragmentation, reductionism and moralisation [21]. At this triple crossroads (didactic, sociological and disciplinary), teachers operate as mediators between science and ethical values, the prescribed curriculum and classroom practice, the local territory and globalised rhythms and forms of knowledge, politics—in a broad sense, but also education policy—as well as ideology and social controversies [22,23]. To all this, the ultimate purpose should not be forgotten: students as the centre of the teaching and learning process [24].

Teachers’ professional knowledge—in the tradition of pedagogical content knowledge [25,26]—is not reduced to what they know about climate change, but comprises the way in which they organise, hierarchise and communicate that knowledge in real educational situations. Inquiring into how teachers’ reason, and into how normative and curricular frameworks enable or constrain that reasoning, therefore constitutes the route of access to the problem [27]. Much of the research on conceptual change has nonetheless assumed that teachers hold consolidated expert conceptions, and has concentrated on students’ alternative conceptions [28,29,30], neglecting the way in which teachers themselves construct—or fail to construct—expert geographical reasoning about climate [21,31].

The study, whose design underpins much of this article, addresses that gap concerning the construction of expert geographical reasoning by teachers, on the basis of a transnational study within the European Education Area. The choice of Spain and France responds to diverse factors, not only to methodological convenience—proximity, accessibility or availability of data—but also to an epistemological decision. Both systems face the challenge of teaching climate change within a secondary school subject that in each country combines geography and history, and in both, the subject is taught predominantly by teachers whose initial training is historical rather than geographical. They approach that shared challenge, however, from divergent disciplinary and curricular traditions. Current Spanish regulation articulates the “climate emergency” through an ecosocial, competence-based and axiologically explicit approach [24], whereas the French tradition inscribes it in the notion of global change and in a school geography structured by multi-scalarity and by case study, within a transversal education for sustainable development that has itself been criticised for its normative drift [32,33]. Confronting two ways of taking for granted how climate change should be taught renders visible the epistemological presuppositions that each system naturalises and that, examined separately, would remain unexamined; difference is used here as an analytical instrument.

Three differences between the two systems bear directly on the object of this study and make the comparison analytically productive rather than merely binational. The first concerns the governance of the curriculum: the French programmes are defined nationally, by cycle and by subject, and their interpretation is regulated through accompanying national resources and through the inspectorate and the académies, whereas the Spanish framework distributes the definition of the curriculum across the state, the autonomous communities and the school, so that the prescribed text is a layered object in Spain and a comparatively unified one in France. The second concerns the emphasis placed on the relation between disciplinary knowledge and competence, which is a difference in tradition and not an opposition between the two terms: the Spanish framework of 2022 organises the subject around specific competences and basic knowledge and states its ecosocial purpose explicitly, while the French tradition retains the disciplinary architecture of histoire-géographie and channels competence through capacities and methods subordinated to disciplinary content. The third concerns the model of school geography itself: the French case study installs multi-scalar reasoning as a method of the discipline, whereas the Spanish formulation approaches multi-scalarity through the competence framework rather than through a prescribed method.

These characterisations are drawn from the normative texts and from the literature cited, and what they licence is an expectation and not a finding. It may be expected that a comparatively unified and methodologically prescriptive framework will yield a more homogeneous treatment of scale and a more reticent treatment of values than a layered and axiologically explicit one, and, therefore, that the contrast between the two systems will not be reducible to the presence or absence of climate change in the curriculum. Establishing whether that expectation holds is a task of the curricular phase and not a premise of the design; Section 3.3 states these differences and the analytical consequence that follows from each.

This problem raises a methodological difficulty that constitutes the specific object of the present article. Any study of an educational phenomenon is conditioned from the outset by the paradigm from which it departs [34], and the phenomenon at issue here is simultaneously conceptual, curricular, discursive and axiological [14,15]. Addressing it by adding instruments to one another would reproduce, at the level of the research, the very fragmentation that is criticised at the level of the classroom: four sources would yield four parallel analyses rather than one account. The difficulty is therefore not one of coverage but of coherence, and it consists of ensuring that every source interrogates the same object. The position argued here is that such coherence is pursued through a single system of categories derived from expert geographical reasoning—organisation of knowledge, causal structure, treatment of scale and the fact–value relation [21]—applied transversally to materials of differing nature as one and the same interpretive grid.

Consequently, the aim of this article is to ground, epistemologically and methodologically, the design of a comparative study of climate change teaching in Spanish and French geography education, and to show how a system of categories of expert geographical reasoning organises the convergence of heterogeneous sources within a mixed-methods design of qualitative priority. The study that the design serves is governed by a single overarching question: how do the Spanish and French education systems, and the teachers who mediate them, construct the treatment of climate change in school geography, and to what extent does that construction correspond to expert geographical reasoning? Because that question bears on objects of different constitution—a prescriptive text, a collectively produced professional discourse, an individual professional discourse and a distribution across a population—it is decomposed into four subquestions, each addressed to one source, and each answered through the same four analytical continua:

SQ1. Which forms of reasoning about climate change do the curricular and normative texts of each system prescribe, at the transnational, national and regional levels?

SQ2. Which features distinguish expert from non-expert professional reasoning about the teaching of climate change when that reasoning is constructed collectively?

SQ3. How does the individual teacher articulate scientific knowledge, curricular prescription, territory and controversy, and how do institutional conditions modulate that articulation in each system?

SQ4. How are the patterns identified qualitatively distributed among secondary geography teachers in the two systems, and which of them discriminate between the systems?

The second of these subquestions has already been addressed in the completed phase reported in [21]; the remaining three correspond to the phases presented in Section 3 as a proposed design. A further question, of a methodological order, is the specific object of the present article: how can a research design preserve epistemic coherence when it interrogates an object of this heterogeneity from sources of this diversity? Section 2 sets out the integrated theoretical framework from which the continua are derived, Section 3 develops the methodological architecture and states the correspondence between each subquestion, its source, its method, the continua it mobilises and its intended output, and Section 4 and Section 5 discuss the contribution of the design, the risks inherent in it and its limitations.

2. Theoretical Framework: An Integrated Apparatus for a Complex Object

Grounding a coherent methodological design requires that, beforehand, the categories with which all the elements that make up climate education in formal contexts are to be interrogated should not derive from the superimposition of unconnected theories, but from their assembly within one and the same process of inquiry. Climate change, with all its teaching dimensions, calls upon at least four thematic fields in the specialised literature: research on conceptual change and alternative conceptions in geography [29,31], systems thinking [35,36], spatial and geographical thinking [37,38] and, finally, the topics proper to human and physical geography applied to subject-specific didactics, such as studies on risk as a social construction or on first- and second-order metaconcepts [39,40] proper to geography. The thesis that structures this Section 2 is that these four bodies of knowledge must converge on a single definition of what it means to think geographically about climate in an expert manner, and that this convergence is what enables an integrated system of categories.

This article does not claim to present a systematic review. What it offers is a problem-driven theoretical synthesis, and the criterion of selection can be stated. Each of the four bodies of knowledge names one dimension that recurs in the literature on climate change education as a condition of an expert treatment of the phenomenon—how knowledge is organised, how causality is represented, how scale is handled and how facts are related to values—and each supplies criteria that can be observed in discourse rather than only postulated. Dimensions such as temporal reasoning, the treatment of uncertainty, futures thinking, environmental justice and action competence are not held to be irrelevant. Uncertainty is addressed within the continuum concerning the fact–value relation, through the notion of epistemic regulation, and environmental justice within the same continuum through the social construction of risk; futures thinking and action competence concern the outcomes of instruction in students rather than the structure of teachers’ reasoning, and are therefore left outside a framework whose object is the latter. The framework is accordingly heuristic and provisional rather than exhaustive, and it remains open to revision by the emergent codes that the analysis of each source may produce.

2.1. Geographical Thinking and Powerful Geographical Knowledge as the Overarching Framework of the Study

The anchoring point is the notion of geographical thinking and its contemporary formulation as powerful geographical knowledge [19,20,41]. Thinking geographically does not consist in locating facts on a map, but in explaining how and why phenomena occur where they occur, and what social, environmental and territorial implications they entail [39]. Uhlenwinkel et al. [13] offered an operative metaphor by distinguishing between the “vocabulary” of geography—concepts, data and descriptions—and its “grammar”—the theories and competences that make it possible to organise that vocabulary in order to interpret the world—proposed as metaphors by Jackson [39]. Possessing vocabulary alone amounts to accumulating inert concepts; it is grammar that structures, relates and interprets. Teaching geography in order to think like a geographer therefore implies mobilising competences of spatial thinking, geographical understanding and spatial citizenship [37], which, far from being exclusive to the discipline, constitute the condition for addressing climate change without reducing it to an inventory of causes and effects.

This framework performs a decisive function in the design: it establishes that the remaining theoretical dimensions—the systemic, the conceptual and the axiological, as explained below—are not external contributions that are “applied” to geography, but internal components of expert geographical reasoning itself. It is this integration that justifies the possibility that one and the same system of categories may run through sources as different as a curriculum (prescribed or enacted) and an in-depth interview.

This connection makes it possible to link the notion of powerful geography with the literature on GeoCapabilities [42] and with current competence-based approaches. Thinking geographically implies not only knowing disciplinary concepts, but having intellectual tools with which to expand subjects’ capacities for understanding and action in the face of the problems of the world. In the case of climate change, those capacities take shape in the possibility of interpreting territorial interdependencies, recognising socio-spatial inequalities, analysing risks, assessing sustainability scenarios and participating in an informed way in public debates. In this way, geographical competences act as a bridge between disciplinary knowledge, expert reasoning and spatial citizenship.

2.2. Alternative Conceptions and Expert Conceptions from Geographical Perspectives

Research on conceptual change [28,30,43] has shown that learning complex scientific phenomena does not consist in adding information to a “blank mind”, as though it were a matter of pages waiting to be written on, but in restructuring frequently resistant prior conceptions. Applied to climate change, this literature has documented a stable repertoire of alternative conceptions—the confusion between weather and climate, the assimilation of the greenhouse effect to the ozone hole, the failure to understand accumulation processes—that persists even after instruction [29,44]. In turn, these alternative conceptions intervene not only in the construction of the theoretical concept, but in the way it is approached. As described in Section 2.3 and Section 2.4, scientific literacy from epistemically regulated perspectives, as well as systemic and complex modes of thinking, also form part of expert thinking.

In this sense, the displacement introduced by this work consists of shifting the focus: it is not centred on the learner’s error, but on the structure of expert reasoning. In turn, it should be noted, as a defining feature of the article, that what distinguishes an expert conception is not an error-free repertoire but the architecture of the explanation: propositions that are ranked, mutually constraining and answerable to scientific criteria, an architecture built over time through disciplinary study, professional practice and reflection upon it [43,45]. These constructions are not, however, fixed in many of their elements, but depend on the idiosyncrasy of each of the branches of scientific knowledge. Addressing expert knowledge about climate change from physics or from chemistry, for example, will entail a variation in the ways of thinking about it with respect to other sciences such as geography, especially given its condition as a discipline that assumes postulates from both the natural and the social sciences. This redefinition is what makes it possible to treat the distance between expert and non-expert reasoning as an analytical continuum and not as a judgement of competence: what distinguishes the expert is not knowing more, or knowing in a specific way, but how that knowledge is organised in a particular manner. In this sense, there is no single typology of the expert, nor is it a matter of placing people into dichotomous groups; rather, it must be understood as a system of tensions depending on the object (at times discursive) of study [21]. Such a distinction, originally formulated for student learning [46], is thus projected onto teachers and onto the curricular text itself, both of which may be interrogated with the same rubric.

2.3. Systems Thinking: The “Grammar” of Geographical Complexity

If the structures of expert thinking supply part of the “expert/non-expert” axis, systems thinking supplies the content of what expert reasoning must articulate. With its genesis in General System Theory [47] and incorporated into geography through the quantitative revolution and the concept of the geosystem [48], systems thinking designates the form of reasoning required when a phenomenon cannot be explained by isolating its components: interdependence, feedback and the displacement of effects in time and space become the object of the explanation rather than complications of it [35,49]. Its relevance for the teaching of climate is twofold. On the cognitive plane, the Structure–Behaviour–Function model (SBF) [11] shows that the learner’s difficulty does not lie in being unaware of the components of a system (structure), but in articulating its non-visible mechanisms (behaviour) and its purposes (function). As indicated in the previous subsection, the expert does not accumulate elements, but explains causally and organises the system by functions.

On the empirical plane, the study by Sterman and Sweeney [12] on the mental models of climate demonstrated that even qualified graduates fail when reasoning about accumulations and flows within the climate system, assuming that it is possible to stabilise the atmospheric concentration of greenhouse gases while emissions exceed their rate of absorption. This finding is central to the design: it proves that the failure to understand climate is not due to a deficit of information, but to a mode of thinking alien to the systemic, and that the “systemic/linear” opposition is not rhetorical but an observable feature of reasoning [12]. From a geographical perspective, systems thinking is furthermore articulated with multi-scalarity—the capacity to change scale and to understand the interdependencies between spatial levels [39]—which generates the “multi-scalar/single-scale” axis.

2.4. The Concept of Risk, Citizenship and Values, and the Epistemic Regulation of Values

The fourth theoretical body, taking geography from paradigms that combine the physical and the human, addresses the dimension that distinguishes climate change from any other physical phenomenon: its condition as a socially live issue, traversed by values, ideologies and controversies [50]. The conception of risk as a social construct [51,52] holds that risk does not depend solely on the occurrence of a hazardous phenomenon, but on the exposure and the social and territorial vulnerability that determine the magnitude of its impacts. Teaching climate from geography therefore implies treating it as a spatially unequal phenomenon, linked to human decisions and to territorial, political and lifestyle-related explanatory frameworks, which connects with the concerns for environmental justice present both in critical geography and in French-language didactics [32,33].

This dimension introduces the final axis of the system of categories: fact–value relation. What Weber [53] formulated as value neutrality, and what the sociology of knowledge subsequently qualified [18], is read here as a requirement of regulation rather than of absence: the expert end of this continuum does not describe discourse without values, but discourse in which the value component is identified as such and held apart from the empirical claim and from the interpretation that connects them [16]. For that reason, the continuum is designated here by the terms “epistemically regulated” and “undifferentiated, moralising or prescriptive” reasoning, rather than by the terms “non-axiological” and “axiological” used in the previously published study [21], with which they are equivalent in content. It should be specified—and this is a nuance that the design explicitly assumes—that the axiological and the non-axiological do not constitute the two ends of this continuum and do not constitute a mutually exclusive dichotomy: the evidence [50,54] suggests that both registers coexist in tension even within expert reasoning, so that the category does not seek to classify discourses as “pure”, but to describe the dominant direction and the epistemic regulation that a subject or a text exercises over their own judgements.

Furthermore, the “non-axiological” does not mean neutrality in the analysis of a phenomenon or the absence of values [55], a perspective that this account rejects and that dissolves when attention is paid to the scientific thinking proper to the social sciences. Unlike the ideal of detachment inherited from the natural sciences, its object—territory, risk, society, climate as a social problem—is constituted by meanings, interests and collective decisions, so that objectivity cannot be founded on the suppression of values, but on their regulation [16]. For teaching, the consequence is a role defined by the management of cognitive conflict rather than by the promotion of a conclusion: the distinction between teaching about climate and teaching for climate is not thereby dissolved, but made available to students as an object of deliberation. An education oriented towards citizenship, human rights or risk reduction does indeed renounce neutrality, but it can be sustained without sliding into moralistic discourse as long as it preserves the epistemic regulation of the discipline: what Kelly [55] termed committed impartiality.

2.5. Convergence in a Single System of Categories

The four axes explained are integrated into an operative definition of expert geographical thinking about climate and, with it, into a system of deductive categories that constitutes the true integrator of the design (

Table 1. The four analytical continua of expert geographical reasoning and their theoretical provenance. Own elaboration. Note: the ends of each continuum are analytical positions and not categories for classifying individuals; the fourth continuum was designated “non-axiological/axiological” in the previously published study [21], and the terminology adopted here is equivalent in content; the four continua are analytically distinguishable but are not assumed to be empirically independent, and the quantitative phase is designed to examine whether linear causal reasoning, single-scale treatment and the conflation of facts and values co-occur.

Analytical Continuum	Expert End of the Continuum	Non-Expert End of the Continuum	Theoretical Anchoring
Knowledge organisation	Hierarchical and coherent explanation	Isolated propositions, alternative conceptions	Chi et al. [45]; Driver et al. [43]; Posner et al. [28]
Causal structure	Systemic (multiple causality, feedback)	Linear (simple cause-and-effect, simplification)	Hmelo-Silver et al. [11]; Sterman & Sweeney [12]
Treatment of scale (and of phenomena in relation to scale)	Multi-scalar (global–local articulation)	Single-scale (a single scale)	Jackson [39]; Uhlenwinkel et al. [13]
Fact–value relation	Epistemically regulated (facts, interpretations and value judgements distinguished, and their status made explicit)	Undifferentiated (moralisation, opinion and prescription conflated with empirical claims)	Dearden [16]; Jasanoff [18]; Weber [53]

). The “expert/non-expert” axis derives from conceptual change; the “systemic/linear” and “multi-scalar/single-scale” axes, from systemic and spatial thinking; the fact–value axis, from studies on risk, values and citizenship, as well as from the very definition of the social sciences. A highly relevant point, which the following section develops and which has already been made explicit in the preceding sections, is that these categories are not conceived for a single source, but as a common interpretive network, capable of being applied to the different elements and sub-elements that make up the education system from geographical perspectives (curriculum and syllabuses, educational contexts, teachers’ discourses and conceptions, students, mass media and social networks, and so on). This provides a common analytical basis for the epistemic coherence discussed above. It does not guarantee it: what a shared grid supplies is traceability between sources, and the design states below both how that coherence is to be pursued and how the risks it carries are to be controlled.

3. Methodological Framework: Establishing an Integrated Research Design

The design grounded in this article follows from what has been set out above: if climate change is at once a conceptual, curricular, discursive and axiological object, its study cannot be entrusted to a single instrument nor to the mere accumulation of several. The answer is not methodological in an instrumental sense, but “meta-methodological”. That is, it consists of subordinating the choice of each technique to one and the same analytical approach—the system of categories in Table 1—so that the different sources converge upon the same object instead of generating parallel analyses, or analyses that are merely equivalent in an interpretive sense. The following subsections set out this methodological “architecture” and, above all, justify the reasons for each decision taken, since these are a consequence of the research question.

3.1. Epistemic Coherence: Mixed Methods with a Qualitative Emphasis

The nature of the object of study requires a mixed-methods design, but one with a greater weight of the qualitative. The question—how climate change is taught and to what extent that treatment aligns with expert geographical reasoning—is primarily interpretive: it bears on meanings, conceptions and forms of reasoning, which an experimental or causal-explanatory design is not equipped to reconstruct [56]. Hence, the qualitative predominance. The quantitative component is not incorporated in order to “validate” the qualitative in a neo-positivist sense, but to examine the extent to which constructs previously identified in a qualitative manner can be observed in broader samples, and to make two education systems comparable at a scale that interviews and focus groups do not reach.

The design is, more precisely, a multiphase exploratory sequential mixed-methods design with qualitative priority. The order of the phases is not a calendar but a dependency: the continua are calibrated on professional discourse grounded in professional discourse before being applied to curricular texts and to individual reasoning, and the quantitative instrument is derived from the codes produced in the qualitative phases, so that it could not have been constructed first. The design does not therefore claim that the phases are chronologically indifferent; what it claims is that their coherence does not derive from their ordering. Mixing occurs at three identifiable points: at the construction of the instrument, where qualitative codes are transformed into indicators and items; at the sampling of the quantitative phase, which is stratified by the profiles that the qualitative phases make visible; and at the interpretation, where the results of each source are read against the others in order to produce meta-inferences [57,58]. Agreement between sources is treated as convergence, complementarity as the normal case, and disagreement as a finding in its own right rather than as error.

This design must be distinguished from triangulation understood as the simple convergence of instruments upon the same datum. What provides coherence here is not that three techniques “point” to the same fact, but that all of them operationalise the same categories. Triangulation is therefore epistemic before it is instrumental. Curricular analysis, expert focus groups, teacher discourse and the questionnaire are readings of the same system of categories over different materials. The qualitative components of the design are governed by the criteria of credibility, transferability, dependability and confirmability proper to interpretive research [59]. These criteria do not replace validity and reliability: the quantitative component requires its own, and a mixed-methods design is answerable to both sets rather than to either alone.

Applying the same continua to sources of different constitution does not make those sources equivalent. A curricular provision, a turn in a group discussion, an individual account of one’s own practice and a response to a closed item differ in linguistic form, in illocutionary status and in what each can be taken to evidence; a prescription is not a belief, and a declaration about practice is not that practice. The continua therefore operates at the level of the analytical construct, while the observable indicators are specific to each source. Positions on a continuum are consequently comparable across sources as positions, and never as equivalent empirical facts. A source-specific coding protocol, currently in use in the phase under way and available from the corresponding author on request, states for each continuum what counts as evidence in each of the four materials, together with the units of analysis, the inclusion and exclusion rules, the procedure for locating a segment along a continuum, the treatment of ambiguous and contradictory cases, the incorporation of emergent codes and the correspondence between qualitative codes and questionnaire items. The correspondence between the continua and the observable indicators of each source are set out at the end of Section 3.4.

Moreover, the system of categories as an integrating element entails a mode of operation that is not exhausted by the deductive character of the “expert/non-expert” instrument. Its deductive character is what makes it possible to apply one and the same grid to diverse sources and data. But the design does not close upon the deductive. Following thematic analysis in a variant that combines prior categories with emergent codes [60,61], the established system for analysing expertise remains open to the data introducing nuances, tensions or new categories.

3.2. The Completed Phase: Deductive Construction and Focus-Group Calibration of the Analytical Continua Precursor Categories

The first phase of the design has already been carried out, and its findings have been published elsewhere [21]. What follows therefore reports no new analysis: it restates, as previously published evidence, the empirical basis on which the framework rests, and specifies what the present article adds to it. In that study, two focus groups were convened—one composed of university teaching staff and secondary teachers whose academic training and professional trajectory lay in geography, geography education and climatology, and one composed of pre-service secondary geography teachers—and their discourse was analysed thematically, combining categories derived deductively from the literature with inductively emergent codes. Expertise was assigned on grounds external to the reasoning subsequently observed—disciplinary degree, doctoral training and professional role—and not on the systemic or multi-scalar quality of the discourse itself, so that the categories were not validated against the attribute they were intended to characterise. Participants, recruitment, procedure, coding and the participant-information, informed-consent and data-protection arrangements are reported in full in that publication and are not reproduced here. Coding are reported in full in that publication and are not reproduced here; the ethical framework under which it was conducted is stated in the Institutional Review Board Statement.

The instrument was not incidental to the phase. A grid intended to capture how expertise organises reasoning cannot be calibrated against an attribute presumed to pre-exist in individuals and to be retrievable through a closed instrument; it requires a setting in which expertise is exercised. Group interaction supplies that setting, since it obliges participants to justify, contest and qualify their claims before peers, so that the organisation of their reasoning becomes observable instead of merely declared [62].

Three results of that study bear directly on the present design [21]. First, the theoretical distinctions proved observable in professional discourse and discriminated between the two profiles, which licensed their use as an analytical grid rather than as a purely conceptual construct. Second, the contrast between profiles lay in the organisation of reasoning—density of interconnection, recourse to change in scale, explicit regulation of the epistemic status of claims—rather than in the volume of declarative knowledge, which supports defining the first continuum by structure and not by content. Third, and with the greatest consequence for the framework, value-laden and epistemically regulated registers did not sort neatly onto the two profiles: both were present, in tension, within the discourse of the more expert group. That result rules out treating the fact–value relation as a dichotomy and requires its formulation as a continuum, which is the form adopted in Table 1.

The present article does not re-examine those data; it reformulates the apparatus that produced them. The published study operated with three conceptual axes—geographical thinking, systems thinking and the epistemic regulation of values—operationalised for a single source and a single analytical purpose, namely the comparison of two professional profiles. What is proposed here is a system of four continua, which separates the organisation of knowledge from the causal structure and from the treatment of scale, and which is designed from the outset for application to four sources of different epistemic and linguistic constitution—a prescriptive text, a collective discourse, an individual discourse and a response to a closed item—across two education systems. Table 1 is accordingly not reproduced from the earlier publication but reformulates its axes for that purpose. The contribution of this article is therefore the transformation of a discourse-analytic category system into an integrator for a multi-source comparative design, together with the translation rules that make that transfer explicit.

3.3. Multi-Level Analysis of Curricula, Official and Enacted, in Relation to Geography Education

The second phase applies the grid to the statutory and curricular texts of both systems through a content analysis organised at three levels, which reproduces the structure of curricular governance itself [63]. The macro level examines the transnational reference frameworks: the Sustainable Development Goals, the UNESCO guidelines on education for sustainable development, the Lucerne Declaration on Geographical Education for Sustainable Development of the International Geographical Union, and the European framework of sustainability competences. The meso level examines the national regulations of each system: in the Spanish case, the royal decrees establishing the minimum requirements for compulsory secondary education and for the baccalaureate [24,64]; in the French case, the official histoire-géographie programmes for collège and lycée and the circulars on education for sustainable development. The micro level addresses the regional concretion of the Spanish curriculum and its implementation by the French académies, together with the teaching materials, projects and departmental decisions through which the prescribed text becomes an enacted one. The expression enacted curriculum is used for this last level in place of hidden curriculum, since materials, projects and departmental decisions are neither concealed nor implicit and do not constitute a hidden curriculum in the strict sense.

The corpus is defined by three inclusion criteria. A document is included if (i) it is in force during the period of analysis, (ii) if it is prescriptive or officially recommended for the levels at which geography is taught in secondary education, and (iii) if it addresses climate, global change or education for sustainable development either as content or as a transversal purpose. Documents addressed exclusively to primary education, to vocational training or to non-formal contexts are excluded, as are commercial textbooks, which are the object of a separate line of work. Every text is dated and its version recorded, and the analysis distinguishes texts in force from texts already published but not yet applicable and from texts still at the draft stage. That distinction is not a formality in the French case, where the secondary corpus is at present in three different states: the programmes of cycle 4, which covers the 5e, 4e and 3e years, remain those established by the arrêté of 9 November 2015 as modified by the arrêté of 17 July 2018; the programmes of cycle 3, which includes the first year of collège, were replaced by the arrêté of 22 April 2026 (NOR MENE2608631A, Bulletin officiel no. 22 of 28 May 2026), which nevertheless does not apply to that year until the 2027–2028 school year; and the revision of the history and geography programmes of cycle 4, commissioned from the Conseil supérieur des programmes in March 2024 and published as a draft in July 2025, has not been adopted by arrêté. At the lycée, the programmes in force are those of the arrêté of 17 January 2019 for the seconde générale et technologique and for both première streams, and those of the arrêté of 19 July 2019 for both terminale streams. The Spanish corpus requires the same treatment rather than a presumption of stability: the state decrees [24,64] are concretised by the Aragonese orders ECD/1172/2022 of 2 August (Boletín Oficial de Aragón of 11 August 2022), amended by order ECD/867/2024 of 25 July (BOA of 7 August 2024), and ECD/1173/2022 of 3 August, amended by order ECD/886/2024 of 25 July (BOA of 8 August 2024). Each text is therefore cited with its status, and the comparison is conducted between corpora whose stability is itself an object of analysis rather than an assumption.

Regional and académie-level documents are not analysed exhaustively. In Spain the analysis takes as a case of curricular concretion the autonomous community in which the subsequent empirical phases are conducted; in France it takes the académie de Bordeaux, whose territory includes the French side of the Pyrenean border and the institution at which the French phase of the research is conducted. The unit of coding is the propositional segment—an assertion, an objective, an assessment criterion or a descriptor that can be assigned a position on one of the four continua—rather than the paragraph or the document, so that a single provision may be coded on more than one continuum. The comparative procedure proceeds in two movements: each system is first characterised in its own terms, so that its categories are described as it formulates them; only then, are the two characterisations confronted, so that the comparison bears on the position of each system along each continuum rather than on the presence or absence of a lexical item.

The analysis therefore does not consist in an inventory of the presence of climate change. It is the application of the same grid to texts of divergent constitution that converts the comparison into a device of denaturalisation rather than a double description: divergences—such as the ecosocial and axiologically explicit approach of the Spanish regulations against the multi-scalar framing of changement global in the French tradition—become legible as epistemological decisions and not as differences in content.

Table 2. Institutional and epistemological differences between the Spanish and French systems and their relevance to the analytical design. Own elaboration. Note: the characterisations are drawn from the normative texts and from the literature cited in Section 1; they state expectations to be examined in this phase, not findings.

Dimension	Spain	France	Relevance to the Design
Curriculum governance	State framework of minimum requirements, developed by the autonomous communities and concretised by the school	National programmes by cycle and subject, with accompanying resources and regulation through the inspectorate and the académies	Determines the number of levels the curricular analysis must cover and the interpretive scope left to teachers
Knowledge and competence	Specific competences and basic knowledge; ecosocial purpose stated explicitly	Disciplinary architecture of histoire-géographie; capacities and methods subordinated to disciplinary content	Conditions where an expert treatment is expected to be made explicit in the prescribed text
Model of school geography	Multi-scalarity approached through the competence framework	Multi-scalarity installed as method through the case study	Bears directly on the continuum concerning the treatment of scale
Framing of climate change	Climate emergency within an ecosocial and axiologically explicit approach	Changement global within a transversal education for sustainable development	Bears directly on the continuum concerning the fact–value relation
Norm governing the teacher position	No general norm of reserve; citizenship education with explicit commitments	Laïcité and devoir de réserve as institutional norms of neutrality	Frames how the stances of Table 3 and stances of Table 4 can be expressed in each system

summarises the institutional and epistemological differences between the two systems that bear on the design and states, for each, the analytical consequence that follows from it.

3.4. The Operationalisation of Geographical Competences in the Design

Before examining teachers’ reasoning in the interviews (Section 3.5), the design requires specifying what, in operational terms, the geographical competences presupposed by the grid of categories consist of. Spatial thinking, geographical understanding and spatial citizenship are not interchangeable synonyms, but progressive levels of one and the same geographical meta-competence: the first operates on the location, distribution and interaction of phenomena in space; the second articulates those elements into causal and multi-scalar explanations; the third translates that understanding into a disposition to act upon the territory [13,37,65]. This progressive architecture—from the spatial gaze to territorial action—has recently been systematised in a model sequenced into four blocks for secondary education [66], whose logic is pertinent for this design because it makes it possible to situate expert geographical reasoning on a continuum of progressive acquisition and not in a static dichotomy. Previous work had furthermore warned that the way in which school curricula formulate these competences varies appreciably between apparently close education systems [67], which reinforces the pertinence of a comparative curricular analysis such as the one undertaken in Section 3.3.

From the perspective of GeoCapabilities, the acquisition of geographical competences cannot be understood apart from the disciplinary knowledge that sustains them, given that the capacity to act upon the territory depends on the quality of the geographical reasoning that the student develops [38]. Within the design, this framework of geographical competences operates as an additional validation criterion for the “multi-scalar/single-scale” axis: where teacher discourse or the curricular text articulates spatial thinking, geographical understanding and spatial citizenship in an integrated manner, a treatment closer to the expert pole of the grid may be expected (Table 1). The operationalisation of geographical competences requires distinguishing several levels. In the first place, spatial thinking makes it possible to identify locations, distributions, patterns, distances, connections and spatial relations. In the second place, geographical understanding makes it possible to explain those patterns by means of concepts such as scale, place, territory, interaction, risk, vulnerability, inequality and change. In the third place, spatial citizenship introduces the capacity to assess alternatives, to argue with evidence and data through perspectives proper to the sciences, and to participate in decisions that affect the territory. Taken together, these levels make it possible to translate expert geographical reasoning into observable indicators for curricular analysis, the interviews and the questionnaire. Table 3 condenses that operationalisation: for each continuum and each source, it states the observable indicators at both ends of the continuum, the inclusion and exclusion criteria, and the kind of evidence that the source can provide.

Table 3. Observable indicators of the four continua in each source, with the inclusion and exclusion criteria and the kind of evidence that each source can provide. Own elaboration. Note: positions are comparable across sources as positions and never as equivalent empirical facts; where a segment makes no claim bearing on a continuum, non-activation is recorded rather than the assignment omitted.

Source and Unit of Analysis	Continuum	Observable Indicators (Expert End/Non-Expert End)	Inclusion and Exclusion	Kind of Evidence
Curricular and normative texts. Unit: the normative segment with complete sense within its structural unit (specific competence, item of basic knowledge, assessment criterion, objective, article, theme)	E1 Organisation of knowledge	Conceptual hierarchy; anchoring in the body of knowledge of the subject; dedicated curricular space; problematised formulation/purposive, attitudinal or organisational statement; isolated mention	In: prescriptive or officially recommended texts in force for the secondary levels at which geography is taught that address climate, global change or education for sustainable development. Out: primary and vocational texts, commercial textbooks, procedural preambles, repeated boilerplate	A prescription: what the system requires or recommends, and not what teachers believe or do
	E2 Causal structure	Concurrence of factors of different nature; society–environment interaction; feedback; temporal depth; explicit non-linearity/simple chain; isolated consequence; unidirectional anthropisation
	E3 Treatment of scale	Articulation of levels; change of scale prescribed as a skill; territorial inequality of risk; regional effects of a global process/statement at a single level
	E4 Fact–value relation	Distinction of epistemic statuses; object presented for analysis; value thematised; judgement tied to an explicit criterion (intermediate position)/attitudinal prescription; militant framing; action as outcome
Focus-group discourse [21]. Unit: the intervention, delimited by change of speaker and subdivided where it develops two distinct claims	E1 Organisation of knowledge	Hierarchisation of claims; theoretical relevance; recognition of prior conceptions; method invoked as warrant/theoretical confusion; oversimplification; complexity named as an obstacle	In: substantive contributions on the teaching of climate change. Out: organisational talk, moderator questions, unattributable fragments	An interactional claim, produced before peers able to contest it
	E2 Causal structure	Interdependence between factors; indirect or deferred effect; reinforcing loop/linear description; attribution to a single agent
	E3 Treatment of scale	Change of scale doing analytical work; glocal articulation; local scale as entry point/one level throughout the passage
	E4 Fact–value relation	Evidence-based argumentation; use of official sources; theoretical prudence; metacognitive regulation; value stated with its criterion (intermediate position)/moralisation of action; ethnocentric framing; single solution pathway
Interview discourse. Unit: the thematic passage, a stretch of talk sustaining one line of reasoning across one or more turns	E1 Organisation of knowledge	The discourse indicators, and the justification of the sequence of content by conceptual dependency/sequence justified by the availability of materials	In: reasoning about content, its didactic treatment, territory and controversy. Out: biographical framing without analytical content; talk about the interview itself; passages the informant withdraws	A declaration about the informant’s own reasoning and practice, and not that practice
	E2 Causal structure	The discourse indicators, applied to the explanation of the phenomenon and of its teaching
	E3 Treatment of scale	The discourse indicators, and the articulation of the proximate territory with wider processes
	E4 Fact–value relation	The discourse indicators, coded separately for the settled scientific core and for the socio-political dimension (Section 3.5)
Questionnaire responses. Unit: the item response	E1 Organisation of knowledge	Item family on the structure of explanations: subordination of propositions/accumulation of unrelated facts	In: complete responses to items retained after expert judgement, cognitive interviewing and piloting. Out: items failing the pilot criteria; non-differentiated response patterns	A response to a closed stimulus, interpretable only through the code from which the item derives
	E2 Causal structure	Item family on accumulation and flow and on indirect and deferred effects; reversed items required
	E3 Treatment of scale	Item family on the relation between levels: relating levels/recognising a single level
	E4 Fact–value relation	Item family on the epistemic status of assertions and on settled as against legitimately controversial questions; reversed items required

Table 3. Observable indicators of the four continua in each source, with the inclusion and exclusion criteria and the kind of evidence that each source can provide. Own elaboration. Note: positions are comparable across sources as positions and never as equivalent empirical facts; where a segment makes no claim bearing on a continuum, non-activation is recorded rather than the assignment omitted.

Source and Unit of Analysis	Continuum	Observable Indicators (Expert End/Non-Expert End)	Inclusion and Exclusion	Kind of Evidence
Curricular and normative texts. Unit: the normative segment with complete sense within its structural unit (specific competence, item of basic knowledge, assessment criterion, objective, article, theme)	E1 Organisation of knowledge	Conceptual hierarchy; anchoring in the body of knowledge of the subject; dedicated curricular space; problematised formulation/purposive, attitudinal or organisational statement; isolated mention	In: prescriptive or officially recommended texts in force for the secondary levels at which geography is taught that address climate, global change or education for sustainable development. Out: primary and vocational texts, commercial textbooks, procedural preambles, repeated boilerplate	A prescription: what the system requires or recommends, and not what teachers believe or do
	E2 Causal structure	Concurrence of factors of different nature; society–environment interaction; feedback; temporal depth; explicit non-linearity/simple chain; isolated consequence; unidirectional anthropisation
	E3 Treatment of scale	Articulation of levels; change of scale prescribed as a skill; territorial inequality of risk; regional effects of a global process/statement at a single level
	E4 Fact–value relation	Distinction of epistemic statuses; object presented for analysis; value thematised; judgement tied to an explicit criterion (intermediate position)/attitudinal prescription; militant framing; action as outcome
Focus-group discourse [21]. Unit: the intervention, delimited by change of speaker and subdivided where it develops two distinct claims	E1 Organisation of knowledge	Hierarchisation of claims; theoretical relevance; recognition of prior conceptions; method invoked as warrant/theoretical confusion; oversimplification; complexity named as an obstacle	In: substantive contributions on the teaching of climate change. Out: organisational talk, moderator questions, unattributable fragments	An interactional claim, produced before peers able to contest it
	E2 Causal structure	Interdependence between factors; indirect or deferred effect; reinforcing loop/linear description; attribution to a single agent
	E3 Treatment of scale	Change of scale doing analytical work; glocal articulation; local scale as entry point/one level throughout the passage
	E4 Fact–value relation	Evidence-based argumentation; use of official sources; theoretical prudence; metacognitive regulation; value stated with its criterion (intermediate position)/moralisation of action; ethnocentric framing; single solution pathway
Interview discourse. Unit: the thematic passage, a stretch of talk sustaining one line of reasoning across one or more turns	E1 Organisation of knowledge	The discourse indicators, and the justification of the sequence of content by conceptual dependency/sequence justified by the availability of materials	In: reasoning about content, its didactic treatment, territory and controversy. Out: biographical framing without analytical content; talk about the interview itself; passages the informant withdraws	A declaration about the informant’s own reasoning and practice, and not that practice
	E2 Causal structure	The discourse indicators, applied to the explanation of the phenomenon and of its teaching
	E3 Treatment of scale	The discourse indicators, and the articulation of the proximate territory with wider processes
	E4 Fact–value relation	The discourse indicators, coded separately for the settled scientific core and for the socio-political dimension (Section 3.5)
Questionnaire responses. Unit: the item response	E1 Organisation of knowledge	Item family on the structure of explanations: subordination of propositions/accumulation of unrelated facts	In: complete responses to items retained after expert judgement, cognitive interviewing and piloting. Out: items failing the pilot criteria; non-differentiated response patterns	A response to a closed stimulus, interpretable only through the code from which the item derives
	E2 Causal structure	Item family on accumulation and flow and on indirect and deferred effects; reversed items required
	E3 Treatment of scale	Item family on the relation between levels: relating levels/recognising a single level
	E4 Fact–value relation	Item family on the epistemic status of assertions and on settled as against legitimately controversial questions; reversed items required

3.5. Conducting In-Depth Interviews to Address Conceptions and Tensions Surrounding Expert Thinking

The third phase of the design shifts the focus to the subjects. Where the focus group captures expert knowledge as a collective construction, in-depth interviews—stratified by level of expertise and by education system—make it possible to examine how each individual teacher articulates geographical reasoning about climate and how they mediate between science, the prescribed curriculum, the territory and students. Teachers stand here as the figure that lends unity to the didactic problem: their professional knowledge [25,26] is the point at which scientific content, curricular transposition and social controversies converge. The in-depth interview is a good instrument for this phase because it provides access to conceptions, to possible alternative conceptions and to the epistemic-axiological regulation of discourse, which a closed instrument would reach with greater difficulty since, as stated in Section 2, reality is not dichotomous in terms of “expert–non-expert”. In sum, the same categories that ordered the curricular analysis now order the analysis of teacher discourse, which makes it possible to relate what the regulations prescribe to what teachers think and declare that they do.

The phase is designed as follows. The target population is secondary school teachers who teach geography within the subject that combines geography and history in each system. The planned sample comprises approximately twenty teachers, distributed evenly between the two systems and stratified by degree of expertise, which is defined by criteria external to the reasoning to be analysed: disciplinary degree, specific training in geography or in climate-related fields, years of practice and institutional role. These criteria are not assumed to converge: disciplinary background, professional experience, pedagogical knowledge and institutional role may generate different forms of expertise, and the stratification is designed to keep them distinguishable rather than to collapse them into a single scale. Recruitment proceeds through educational administrations, teacher-training institutions and professional networks; selection is purposive and not probabilistic, and the sample is designed for analytical contrast between profiles and between systems rather than for representativeness. The interview guide is structured in twelve blocks and incorporates a spontaneous elicitation protocol: climate change is not named by the interviewer until the informant introduces it, and the moment of its first unprompted appearance is itself recorded as data. Data collection continues until thematic sufficiency is reached, assessed through the criterion of information power [68] rather than through a predetermined number of interviews. Sessions are recorded with informed consent, transcribed verbatim, pseudonymised at the point of transcription and analysed in the original language; only the segments reported in publications are translated, and each translation is checked against the source by a second reader. The unit of analysis is the thematic passage—a stretch of talk sustaining one line of reasoning, which may span several conversational turns—rather than the isolated utterance.

The qualitative phase addresses the teaching of climate change as a spatially and socially live issue (controversial topics). Scientific knowledge, territory, risk and citizenship are interwoven here, and their didactic treatment is reducible neither to the transmission of evidence nor to the prescription of attitudes [14,69]. On this premise, the way in which teachers position themselves in relation to the evaluative dimension of the problem does not constitute noise that the interview should neutralise, but a central datum: the teacher’s position regarding neutrality is itself a specification of the fact–value axis that structures the system of categories and that is applied in a homologous manner to the curriculum, the interviews, the focus groups and the questionnaire.

In order to operationalise that position, the interview guide and its subsequent coding incorporate Kelly’s [55] typology of the teacher’s role in the face of controversial issues, which distinguished four stances (see Table 3 and Table 4). It must be specified that the authors’ position regarding committed impartiality as the most justifiable worldview is the thesis advocated by Kelly [55] and not an isolated judgement of the author. The use of this typology in geography education nonetheless requires a distinction that constitutes one of the contributions of the design to the fact–value axis. The scientific core of the phenomenon—the anthropogenic origin of warming, in this case—is not a legitimately open question, and treating it as such by means of false balance introduces an epistemic distortion rather than an exercise in plurality [70,71].

The socio-political and distributive dimensions of the problem—such as mitigation policies or environmental justice—are, by contrast, legitimately controversial. This separation corresponds to the distinction between settled questions and open questions proposed by Hess and McAvoy [72], and makes it possible to sustain that one and the same teacher may be neutral with respect to the scientific core and, at the same time, exercise committed impartiality on the evaluative plane. In this sense, renouncing neutrality about what science establishes does not amount to abandoning rigour, but to sustaining it. Such a reading is consistent with the fact that the axiological and non-axiological registers and value-laden and epistemically regulated registers are not mutually exclusive, but coexist in tension in expert discourse [21].

The coding-categorisation process therefore does not assign a fixed label to each participant, but crosses two planes—the Kelly stance that the discourse reveals and the object upon which that stance is exercised (scientific core as against socio-political dimension)—in relation to the informant’s degree of expertise. Table 4 synthesises this crossing and guides both the drafting of the items and the subsequent thematic analysis.

Table 4. Teachers’ stances on the climate issue on the basis of Kelly [55], and the analytical risk associated with each stance according to the object of controversy. Own elaboration. Note: the expressions in the two right-hand columns name risks identified in the literature and are not coding categories; each is assigned only where the corresponding stance has been identified on both of the planes described in Section 3.5, and no stance is assigned on the basis of a single utterance.

Teacher Stance [55]	Teacher Conduct	In the Face of the Scientific Core (Anthropogenic Consensus)	In the Face of the Socio-Political and Axiological Dimension
Exclusive neutrality	Avoids the issue	Risk that the problem disappears from the classroom	Evades the civic and ethical implications of the phenomenon
Exclusive partiality	Imposes a position without admitting alternatives	Risk of indoctrination if aligned with denial; legitimate transmission of knowledge if confined to established evidence	Risk of moralisation or indoctrination: undifferentiated register
Neutral impartiality	Presents opposing positions and withholds their own	Risk of false balance: presenting as open what is settled	Plurality is possible, although it may evade the teacher’s mediating function
Committed impartiality	Makes a reasoned position explicit and presents the alternatives equitably	Inappropriate as “balance”: knowledge is taught as knowledge, not as opinion in dispute	Stance consistent with the objectivity/neutrality distinction: a grounded position and openness to dissent

Table 4. Teachers’ stances on the climate issue on the basis of Kelly [55], and the analytical risk associated with each stance according to the object of controversy. Own elaboration. Note: the expressions in the two right-hand columns name risks identified in the literature and are not coding categories; each is assigned only where the corresponding stance has been identified on both of the planes described in Section 3.5, and no stance is assigned on the basis of a single utterance.

Teacher Stance [55]	Teacher Conduct	In the Face of the Scientific Core (Anthropogenic Consensus)	In the Face of the Socio-Political and Axiological Dimension
Exclusive neutrality	Avoids the issue	Risk that the problem disappears from the classroom	Evades the civic and ethical implications of the phenomenon
Exclusive partiality	Imposes a position without admitting alternatives	Risk of indoctrination if aligned with denial; legitimate transmission of knowledge if confined to established evidence	Risk of moralisation or indoctrination: undifferentiated register
Neutral impartiality	Presents opposing positions and withholds their own	Risk of false balance: presenting as open what is settled	Plurality is possible, although it may evade the teacher’s mediating function
Committed impartiality	Makes a reasoned position explicit and presents the alternatives equitably	Inappropriate as “balance”: knowledge is taught as knowledge, not as opinion in dispute	Stance consistent with the objectivity/neutrality distinction: a grounded position and openness to dissent

This treatment attains its full meaning within the comparative device (between contexts and between sources of information). For example, the French tradition inscribes teaching within an intense institutional norm of neutrality—associated with laïcité and with the devoir de réserve—which French-language didactics of socially live issues has itself problematised [14], whereas the current Spanish configuration (LOMLOE) operates on axiological coordinates. It may be expected that the “state-regulatory” plane will modulate the stances recorded among the participants. This expectation is stated as a theoretical proposition to be examined, and not as an anticipated finding. Verifying this constitutes one of the functions of the comparison, which operates here as a resource for denaturalising what each system takes for granted regarding the place of teachers in the face of controversy. Recent work on practising social science teachers reinforce the pertinence of inquiring into these stances together with disciplinary training and declared classroom practices [73].

3.6. The Quantitative Questionnaire

The fourth phase operationalises the continua in a quantitative instrument. Its purpose is to examine the extent to which the patterns of reasoning identified qualitatively can be observed in broader samples of teachers, and to establish comparisons between the two systems at a scale that interviews and focus groups do not reach. The target population is secondary geography teachers in Spain and France; extension to further countries is a possible continuation of the research programme and not part of the present design.

The epistemic coherence of the whole depends critically on this phase, since the questionnaire is not constructed from external constructs, but as a translation of the same categories—systems thinking, multi-scalarity, epistemic regulation of values—that ran through the preceding phases. This qualitative-to-quantitative derivation is what distinguishes the instrument from an opinion survey: each item refers to a dimension that is theoretically grounded and empirically observed in the previous phases. It should be underlined—in coherence with the theoretical–methodological character of this article—that the questionnaire is presented here as a justified design and a derived operationalisation, not as an already validated instrument: its psychometric validation (content validity through expert judgement, analysis of the factor structure and reliability) constitutes ongoing work and is one of the anticipated contributions of the study.

The construction of the instrument follows a sequence that the design states in advance. Items are generated from the qualitative codes, each item remaining traceable to the continuum and to the code from which it derives. Content validity is assessed by expert judgement [74], and the resulting version is subjected to cognitive interviewing with a small group of teachers in each system in order to detect divergences of interpretation. A pilot administration precedes the analysis of the factor structure and of internal consistency. Linguistic adaptation does not rely on translation and back-translation alone: conceptual and functional equivalence is examined item by item, and measurement invariance between the Spanish and the French versions is tested before any comparison of scores is attempted [75]. Until that sequence is complete, the instrument is a proposal and not a validated measure, and any inference about the distribution of the constructs in the population remains conditional upon the sampling strategy and upon the demonstration of measurement equivalence. The unit of analysis in this phase is the individual item response. Sampling is purposive and stratified by education system and by the expertise profiles that the qualitative phases make visible (Section 3.1); the analyses envisaged are the examination of the factor structure and of internal consistency, the test of measurement invariance between the two versions and, where invariance holds, the comparison of distributions between the two systems. The analyses envisaged are specified as follows. The assumptions of the comparative tests are checked beforehand: normality of the distributions through the Shapiro–Wilk test and homogeneity of variances through Levene’s test; parametric tests are applied where those assumptions hold and their non-parametric alternatives where they do not, and effect sizes are reported where appropriate. Because the items are ordered categorically, the factor structure is first explored through exploratory factor analysis on a calibration subsample and then tested through confirmatory factor analysis on an independent validation subsample, in both cases on polychoric correlations. The minimum sample size is fixed before data collection by Monte Carlo simulation for the model derived from the qualitative phases, with a floor of approximately 200 respondents per system in the validation subsample, so that the model can be estimated in each country separately before it is estimated jointly [76]. Estimation uses weighted least squares with mean- and variance-adjusted test statistics (WLSMV), preferred to maximum likelihood for ordinal indicators [77]. Internal consistency is reported for each factor through ordinal omega rather than through Cronbach’s alpha [78]. Measurement invariance between the Spanish and the French versions is tested hierarchically at the configural, threshold, loading and intercept levels, under the identification conditions established for ordered categorical outcomes [75,79]; latent means are compared between the two systems only where at least partial scalar invariance is supported, and where only lower levels hold, the comparison is restricted to the pattern of relations between constructs. Administration is likewise conditional on ethical approval and informed consent, and responses are collected anonymously.

4. Discussion

As has been stressed throughout the manuscript, the design as a whole is not a chronological sequence of independent phases, but the projection of one and the same lens onto successive sources (Figure 1). The focus group calibrates the categories; the focus-group phase provided the empirical basis for the categories; the curricular analysis applies them to the prescribed framework; the interviews, to the subject who mediates it; and the questionnaire, to the population scale. Each technique answers a question that the previous ones could not answer, and all share the same integrator. The ultimate justification for each decision does not reside in its abstract methodological value, but in its necessity with respect to the research question: the Spain–France comparison requires multi-level curricular analysis; inquiring into expert reasoning requires the focus group and the interviews; and the aim of examining how widely the identified patterns are distributed, in order to compare the two systems, calls for the quantitative questionnaire. It is this derivation—from question to method, and not the reverse—that confers on the design its epistemic coherence.

The scientific community in social sciences education has consolidated in recent years a robust body of research on spatial and temporal thinking and on school geographical and historical knowledge, citizenship education and the use of sources in the classroom [80]. That body of work has demonstrated, with empirical solidity, which competences and which contents are desirable. The contribution of the design grounded here is not situated on that plane; it does not propose a new inventory of what should be taught about climate, but is situated on a meta-methodological plane that has received less attention: the problem of how to study, in an epistemically coherent manner, an object that is simultaneously conceptual, curricular, discursive and axiological within the field of knowledge. The literature on climate education has tended to address these dimensions separately—curricular analyses on the one hand, studies on students’ conceptions on another, and didactic proposals on a third—reproducing at the level of research the same fragmentation that is criticised at the level of the classroom. Against this, the specific contribution of this work is to show that a single system of categories derived from expert geographical thinking can run transversally through heterogeneous sources and articulate their analysis without loss of coherence while pursuing coherence. This contribution engages with the notion of powerful geography [20,41] by taking it a step beyond its usual use. If that notion is normally employed to characterise the knowledge that students should acquire, here it becomes an analytical criterion for interrogating both teachers and the curricular text: what defines an expert treatment of climate—systemic reasoning, multi-scalar articulation, epistemic regulation of values—becomes measurable and comparable, operationalisable and potentially comparable. The displacement of the focus of study towards the structure of expert reasoning [31,45] furthermore opens up a line seldom travelled in social sciences education: that of conceptual change, studied not in students, but in those who mediate their learning.

Two clarifications situate that contribution. The first concerns its relation to existing frameworks. Powerful geographical knowledge, GeoCapabilities, climate literacy and education for sustainable development each specify what should be learned, and the design proposed here does not replace any of them; what it does is to convert a selection of the dimensions they name into an analytical structure that can be applied, with source-specific indicators, to heterogeneous evidence about how that learning is prescribed and mediated. The second concerns sustainability education. The four continua correspond to competences that the literature on education for sustainable development treats as central: systems thinking is the object of the causal continuum; anticipatory and normative competence presuppose, respectively, multi-scalar articulation and an epistemically regulated relation between facts and values; critical thinking and climate literacy presuppose the organisation of knowledge; and citizenship and climate-action education presuppose the distinction, within the fact–value continuum, between the settled scientific core and the legitimately controversial socio-political dimension. The correspondence is not an equivalence, and it is not claimed that the framework measures those competences; what is claimed is that the didactic conditions under which they can be taught become legible through the continua, which situates a meta-methodological contribution to geography education within the broader field of sustainability education and within the concerns of target 4.7 of the Sustainable Development Goals.

4.1. Internal Coherence of the Design

The main strength of the design resides in its internal coherence, understood as the traceability that links the research question, the theoretical framework, the categories and each instrumental decision. The categories are not imported from unconnected traditions, but are derived from the theoretical assembly set out in Section 2. Moreover, the instruments do not overlap, but operationalise those same categories over different materials; and the sequence of phases does not obey a timetable, but a logic in which each technique answers a question that the previous ones could not resolve. This articulation is intended to promote comparability while preserving contextual and theoretical sensitivity [56] by keeping the grid open to emergent codes that capture what is specific to each system. The resulting triangulation is, strictly speaking, epistemic before it is instrumental, since its value derives not from accumulating sources, but from subjecting them to one and the same lens.

4.2. Limitations

The design presents limitations that is appropriate to make explicit. The first is inherent to its qualitative predominance and to the situated character of expert knowledge: no claim of distributional equivalence is made, since what the design compares are configurations of reasoning; the transferability of the qualitative findings to other contexts must accordingly be formulated with caution [56,59]. The second affects the curricular analysis: the examination of state regulations does not exhaust the curriculum as actually implemented, given that regional specification in Spain and mise en œuvre by the académies in France introduce a variability that the level of analysis at intermediate and local scales does not fully capture. To all this must be added the regulatory instability of the French system during the study period, which requires precise dating of which texts are in force and which remain at the draft stage. The third concerns the questionnaire: its condition as a derived instrument demands rigorous psychometric validation—expert judgement, analysis of the factor structure and reliability—whose soundness conditions the quantitative phase and which is acknowledged as ongoing work. Finally, transversal to any interpretive design, the reading of the research team intervenes in every coding decision; the design answers this not with a claim to neutrality but with documented decision rules, recorded analytical memos and independent double coding of a subsample.

A second order of limitation concerns the common lens itself, and it is not dissolved by the internal consistency of the design. Applying one grid to heterogeneous materials may impose constructs that the material does not warrant; it may orient the analyst towards what the grid predicts and so reinforce confirmation bias; it may suppress meanings specific to one source or to one national context; it may privilege the normative conception of expertise held by the researchers themselves; and it may produce an appearance of coherence precisely where the underlying constructs are not equivalent. The provision for emergent codes does not by itself answer these risks. The design therefore commits to four procedures. Coding is double: a subsample of each corpus is coded independently and disagreements are resolved by explicit rule rather than by discussion alone. Discrepant cases are recorded as such, and the analysis reports the cases that the continua fail to accommodate instead of assimilating them. Rival interpretations of a segment are documented in analytical memos and reported where they remain undecided. And the framework is treated as revisable: evidence that contradicts it is evidence about the framework and not an anomaly to be removed, and the conditions under which a continuum would be reformulated are stated before the analysis begins.

5. Conclusions

The grounding set out here sustains a thesis of general scope for research in the didactics of geography that goes beyond the didactics of climate change. When the object of study is complex and heterogeneous, the coherence of a design is not achieved by choosing each technique well in isolation, but by subordinating all of them to one and the same analytical lens derived from theory. In the case of the present study, the view proposed by the authors is a system of categories of expert geographical thinking that makes it possible to interrogate, by one and the same criterion, the prescribed curriculum, teachers’ reasoning and, once the instrument has been validated, the distribution of the identified patterns in broader samples.

Furthermore, the comparison between Spain and France operates as an epistemological device that denaturalises the presuppositions that each system takes for granted. And teachers, as mediators between science, the curriculum, the territory and social controversies, are the figures that confer unity on the didactic problem. The framework accordingly locates the didactic problem of climate change, neither in the delivery of settled content nor in the promotion of prescribed attitudes, but in the quality of the reasoning that teachers are able to construct and to make available to their students. The future lines opened up by the design include deepening the study of expert reasoning by means of the Structure–Behaviour–Function model [11], the validation of the quantitative instrument, the translation of the findings into proposals for the improvement of initial and continuing teacher education, and of teachers’ expert development. The design is therefore offered as an instrument for studying, and not yet as evidence about, the conditions under which the teaching of climate change can articulate rigorous knowledge with educational responsibility.

Three claims of different status should be distinguished in closing. What the present article establishes is theoretical and methodological: that a single system of analytical continua, derived from an operative definition of expert geographical reasoning, can be formulated so as to interrogate sources of different epistemic constitution without dissolving the differences between them. What the previously published evidence suggests is narrower: that these continua are observable in professional discourse, that they discriminate between profiles defined on external criteria, and that the relation between facts and values behaves as a gradient rather than as a dichotomy. What remains to be tested is everything else: whether the grid retains its discriminating power when applied to curricular texts and to individual teacher reasoning, and whether it can be transformed into a measurement instrument whose equivalence across the two systems can be demonstrated. The applications that the design anticipates—contributions to initial and continuing teacher education and to the improvement of climate change teaching—are objectives of the broader research programme and not results of the present article.