An Interactive Tool for the Factor Analysis of Environmental Social Representations

Abstract

Wind turbines are a focal point of contemporary environmental debates, symbolizing progress in renewable energy but also generating significant social tension. This study introduces FactoShinySR, a web-based application designed to facilitate correspondence factor analysis (Corr.F.A.) for studying social representations, making complex analyses accessible to researchers without advanced programming skills. Using data from 323 participants who completed free association tasks and Likert-scale questionnaires, the tool was applied to examine social representations, attitudes toward wind turbines, and trust in authorities. Corr.F.A. revealed two primary dimensions: ecological versus critical attitudes and systemic optimism versus concrete opposition. Positive representations highlighted environmental benefits, while negative perceptions focused on local impacts, such as noise, visual disruption, and ecological harm. Trust in authorities emerged as a critical factor shaping attitudes. FactoShinySR proved instrumental in visualizing the complex socio-representational structures surrounding wind turbines, offering a platform for analyzing environmental perceptions. By bridging methodological complexity and practical application, this tool enables researchers and practitioners to better understand and address public social representations.

1. Introduction

Environmental issues generate debates that polarize opinions depending on cultural and political contexts. Themes such as climate change, the transition to renewable energy, deforestation, land use planning, environmental health, and the use of technology often divide societies and animate daily discussions. These social interactions reveal not only a diversity of opinions and attitudes but also the cognitive frameworks and shared values of individuals and social groups.

The theory of social representations, a central concept developed by Moscovici [1,2], makes it possible to explore and understand the “common sense” that structures collective perceptions. It provides a conceptual framework for analyzing how groups collectively construct meanings and interpret their reality, ultimately influencing their behaviors and interactions with their social and physical environment [1,3]. This theoretical model highlights the impact of sociocultural factors on both individual and collective perceptions and is an essential tool for understanding the dynamics of controversy and consensus surrounding environmental issues.

Studying these representations helps identify beliefs and values that either hinder or promote environmental actions. For instance, previous studies have explored social perceptions of natural resource management [4], protected areas [5], and forest plantations and rural lands [6]. These studies illustrate how social representations shape attitudes toward complex environmental objects, which are often interpreted through shared frameworks within specific social groups.

However, a major challenge lies in making these analyses accessible and usable for decision-makers, researchers, and practitioners who lack expertise in statistical methods. To address this, a web application [7] based on R [8] was developed. This application facilitates correspondence factor analysis tailored to social representation data, thereby making it easier to study environmental perceptions without requiring advanced programming skills. It represents an innovative tool for bridging complex analytical methods with practical issues, such as those related to renewable energy.

2. Theoretical Background

2.1. The Theory of Social Representations and Its Relevance in Environmental Research

The theory of social representations, which was first introduced by Moscovici [1,9], provides a robust conceptual framework for analyzing how individuals and social groups collectively construct an understanding of the world around them, as well as the complex phenomena they encounter [10]. Social representations are not static constructs. They are defined as systems of opinions, knowledge, and beliefs developed and shared by a group about a specific social object, event, or situation. They evolve dynamically through social interactions and are continually shaped by cultural, historical, and contextual factors [9,11,12]. These systems not only provide a framework for understanding reality but also contribute to the formation of a common identity [13].

Social representations are therefore not merely isolated reflections or individual beliefs; they are social constructs that emerge through interactions and communication within the group [9,14]. According to Abric [15], they are “what people think of knowing and are persuaded to know about objects, about situations, about given groups” (p. 11), highlighting their shared and often implicit nature. This shared nature is critical for promoting collective action or understanding resistance to change, especially in contexts where environmental challenges demand coordinated efforts. In this sense, social representations make the world more intelligible by simplifying complex concepts and embedding them within frames of reference that are comprehensible to all [16]. This theory emphasizes the interplay between communication processes, social relationships, and the contextual relationship between the group and the object of representation [17].

In environmental studies, the theory of social representations provides essential insights into how groups perceive and interpret environmental challenges such as climate change, pollution, or biodiversity loss. These challenges are often abstract and difficult to grasp, yet they are interpreted through socially shared frameworks that influence individual and collective actions [10,18]. Social representations simplify these complex issues, embedding them within culturally and contextually grounded narratives [19]. Furthermore, they act as mediators between scientific knowledge and public understanding, bridging the gap by translating technical information into culturally meaningful concepts [16]. One of the key applications of social representation theory in this context is the study of perceptions of environmental risks [20], an area where these representations play a central role in explaining both support for and resistance to environmental policies [21].

For instance, the perception of climate change varies significantly among populations. While some view it as a manageable, gradual phenomenon requiring rational mitigation and adaptation strategies [22], others see it as an unpredictable, catastrophic force beyond human control [20]. These representations are shaped not only by scientific knowledge but also by emotional, social, and cultural factors [23,24]. Misconceptions, such as the belief that technology alone can resolve environmental issues without lifestyle changes, highlight the barriers that social representations can impose on pro-environmental behaviors [25,26]. Such beliefs illustrate the tension between technocratic optimism and behavioral inertia, a dynamic that requires targeted interventions to address [18].

Social representations also play a key role in shaping pro-environmental attitudes and behaviors [27,28]. They influence how citizens perceive actions such as recycling, reducing energy consumption, or adopting eco-friendly transportation methods [29,30,31]. For instance, the study by Callaghan et al. [29] on water recycling in Australia highlights the influence of these representations on the acceptance of this practice. The findings reveal a tension between the perceived environmental benefits of recycled water, associated with sustainability and cost savings, and concerns about its quality, particularly its purity. The opposition between “purity” and “contamination” is central to these social representations, generating resistance to the use of recycled water for direct applications, such as drinking. The perception of recycled water as “dirty” or “contaminated” thus presents a significant barrier to its acceptance, despite lacking a scientific basis. Such results raise the question of individual responsibility for the environment [32,33]. The way in which individuals perceive their own ecological impact, as well as the actions they consider desirable or undesirable, is largely influenced by the social norms and beliefs shared within their group [34,35]. These representations shape the very concept of personal responsibility, affecting not only individuals’ willingness to adopt sustainable behaviors but also the way they justify or minimize their actions according to their perception of the role of other actors (governments, corporations, etc.) in the environmental crisis [36].

Another major area of application of social representation theory pertains to the analysis of the impact of communication messages on the formation of social representations of environmental issues [14,37,38]. Indeed, awareness and informational campaigns, while explicitly aiming to influence attitudes and promote more responsible behaviors, can elicit varied effects depending on the dominant social representations within targeted groups [39,40]. Previous studies on climate change communication have highlighted the importance for communicators to segment target groups and consider their interpretative frames [41,42]. As Moser [43] pointed out, “different audiences require distinct frames, goals, messages, and messengers.” p.39. In their research, Maibach et al. [42] found that communicating climate change through the lens of its implications for human health can provide many Americans with an engaging and relevant frame of reference. Social representations therefore play a critical role in modulating how messages are perceived and internalized by the public. Thus, the success of a campaign relies not only on the clarity and relevance of the message but also on its ability to resonate with pre-existing thought patterns and navigate the sometimes contradictory dynamics of understanding and adopting environmental issues. A thorough analysis of these processes provides deeper insights into how environmental messages can reinforce, transform, or, paradoxically, sustain social representations that perpetuate ambivalent or resistant attitudes toward necessary changes.

2.2. Correspondence Analysis: A Method for Revealing the Underlying Structures of Perceptions

Correspondence analysis (CA) is a robust multivariate statistical method that facilitates the exploration of relationships between categorical variables. Widely employed across various disciplines, CA has become a powerful tool for uncovering the latent structures in complex datasets, particularly those related to social perceptions and attitudes. Its ability to analyze and visualize multidimensional data in a simplified, interpretable manner makes it especially valuable for studying social representations [44,45].

At its core, CA is based on singular value decomposition (SVD), a mathematical technique that reduces the dimensionality of a dataset while retaining its most relevant information [46]. This process is akin to principal component analysis (PCA) [47] but is specifically designed for categorical rather than continuous data. By analyzing contingency tables, CA identifies the key dimensions underlying the relationships between variables and represents these dimensions graphically in a perceptual map. This map positions categories as points in a reduced space, with the distances between points reflecting the degree of similarity or dissimilarity between the categories [10,48]. The intuitive visual output of CA allows researchers to detect associations, oppositions, and clusters among categories, facilitating interpretation and analysis. These maps not only summarize the data but also reveal the structure of relationships, enabling a deeper understanding of the patterns within complex datasets [46].

One of the unique strengths of CA lies in its suitability for analyzing social representations, which are inherently multidimensional and influenced by various sociocultural, psychological, and contextual factors. According to Vergès [49], CA enables the “positioning of subjects and items along bipolar dimensions, thus facilitating interpretation in terms of ideological opposition” (p. 559). This feature is particularly useful in the environmental domain, where perceptions of issues such as climate change, biodiversity, and pollution vary widely across demographic, cultural, and socioeconomic groups [50,51,52]. CA’s graphical approach is particularly valuable in environmental research, where the analysis of perceptions often involves disentangling complex, interrelated factors. Social representations of environmental issues, such as natural resource management or renewable energy adoption, are influenced by psychological, social, cultural, economic, and geographical dimensions [50,51]. By visualizing these multidimensional relationships, CA provides a nuanced view of how individuals and groups construct meaning around ecological challenges. For instance, Lo Monaco et al. [48] demonstrated how CA could reveal patterns in representations of biodiversity, highlighting how certain groups associate biodiversity with scientific conservation efforts, while others link it to cultural or esthetic values. Similarly, Wagner et al. [10] showed how CA helps uncover cultural divergences in the framing of pollution, illustrating the role of shared cultural narratives in shaping public attitudes.

Correspondence analysis represents a powerful tool for uncovering the underlying structures of perceptions and social representations. Its ability to visualize multidimensional relationships makes it indispensable for understanding the dynamics of environmental attitudes and behaviors. By mapping the similarities and differences in social representations, CA not only enhances theoretical understanding but also provides actionable insights for addressing ecological challenges. Whether applied to perceptions of climate change, biodiversity, or pollution, CA enables researchers to explore the social, cognitive, and cultural mechanisms that shape environmental representations, fostering a deeper and more nuanced appreciation of ecological issues. By mapping these complex perceptions, CA offers an original approach to studying how individuals and social groups construct and interpret environmental challenges, shedding light on the social, cognitive, and cultural mechanisms underlying these representations. It not only allows for the identification of divergences and convergences in perceptions but also helps to understand the processes of meaning construction that shape representations of ecological issues. Thus, through its graphical approach, CA enables a more nuanced exploration of the social dynamics associated with environmental concerns, providing a deeper understanding of ecological challenges.

3. FactoShinySR Development and Features

The FactoShiny library is an extension to FactoMineR, a reference tool in R [8] for multivariate data analysis. Developed by Husson, Mongé, and Vaissié [53], FactoShiny enables the creation of interactive web interfaces via the Shiny [54] framework, thus facilitating access to complex analysis methods such as correspondence factor analysis and principal component analysis. The aim of FactoShiny is to facilitate the use of these methods by offering users an intuitive interface where in-depth analyses can be carried out without coding thanks to interactive options and adjustable parameters.

FactoShinySR [7] has been specifically designed to meet the needs of social representation analysis, a field in which correspondence factor analysis is often used to explore the representational structures of the social groups studied. This application adapts and extends FactoShiny’s functionalities to include options for integrating multiple variables, managing evoked words, and performing correspondence analysis while taking into account the specificities of social representation data. FactoShinySR incorporates interactive solutions for visualizing significant contributions to the factorial axes and customizing the graphical display, meeting the criteria of clarity and reproducibility that are essential in scientific publications.

The FactoShinySR application has been built using the R programming language (from version 4), which is widely recognized for its advanced capabilities in statistical analysis and data manipulation. The user interface is therefore based on the Shiny framework, which enables interactive web applications to be created directly from R. Shiny integrated HTML, CSS and JavaScript components, offering advanced flexibility and customization of the graphical display and user interactions. The graphical display is based on the ggplot2 library, which has become a benchmark for the creation of customizable graphics in R. Using ggplot2 [55], customization options can be set up, such as bolding and italicizing modalities according to their significant contribution to the factorial axes. Data manipulation and preparation are managed using dplyr [56], a library that is part of the tidyverse. One of the special features of FactoShiny and FactoShinySR is their ability to perform correspondence analyses on stacked contingency tables. This method, adapted from the recommendations of Greenacre [57], makes it possible to visualize the contributions of the modalities of different variables without creating biases linked to the structure of the data.

4. Using FactoShinySR to Analyze Environmental Representations

4.1. Data Import and Manipulation

To use the FactoShinySR application for analyzing social representations, users must first download the application. A link to the application is provided in the Supplementary Materials. The application is an R script that can be run using software such as RStudio (version 2024) or directly through the R console. Once the application is running, users can proceed with importing data and conducting their analyses. Using FactoShinySR to analyze social representations begins with data import and preparation. Users can import data files in .csv format directly into the application. Once the file has been loaded, the interactive interface provides a preview of the data (see Figure 1), ensuring that the structure and quality of the data can be checked before analysis. This step is crucial to ensure that the variables are correctly identified and formatted for the correspondence factor analysis.

Data manipulation in FactoShinySR is facilitated by options for selecting the variables to be used, specifying active variables (such as evoked words, which are referred to here as ‘observations’) and supplementary variables (for example, socio-demographic variables, which are referred to here as ‘other variables’). This helps to explore the nuances in the perception of the objects studied, as we will see with wind turbines, and to better understand how verbal associations and socio-demographic contexts or other variables influence these representations. Thanks to these integrated tools, FactoShinySR offers a complete platform for preparing and organizing data for in-depth analysis.

4.2. Displaying Results: Options and Customisation

FactoShinySR offers advanced functions for visualizing results, enabling researchers to explore the structures of social representations through interactive, customizable graphs. Once the correspondence factor analysis has been carried out, the application generates a graph that represents the modalities of the variables along the main axes (see Figure 2). This graph highlights the relationships and contributions of the modalities, making it easier to interpret the data visually. Users can interact with the graph to adjust the display parameters, such as selecting the axes to be displayed, zooming in on certain sections, and highlighting significant modalities. The graph can be downloaded in JPG, PnG, and PDF formats.

The application also allows you to customize the appearance of the graphs to bring them into line with academic standards (label size and number of occurrences). Grayed frames refer to the (independent) variables. Independent variables (bold) contribute to the formation of Factor 1. Independent variables (italic) refers to the independent variables which contribute to the formation of Factor 2. Observations (bold) refers to the observations which contribute to the formation of Factor 1. Observations (italic) refers to the observations which contribute to the formation of Factor 2. Observations (bold + italic) refers to the observations which contribute to the formation of both Factors 1 and 2.

An essential aspect of the analysis includes taking into account the contribution threshold, as discussed by Deschamps [58]. This threshold allows modalities to be filtered to represent only those that make a significant contribution to the factorial axes, making interpretation clearer and focused on the most relevant elements of the representation. FactoShinySR incorporates this functionality by allowing users to define and adjust the modality contribution threshold in order to display only those elements that have a substantial impact on the overall structure of the analysis.

In addition to displaying contributions, FactoShinySR offers tools for exploring additional dimensions beyond the first two axes, enabling a more complete analysis of the data when relevant structures are hidden in less visible dimensions. In this way, users can navigate between different factorial planes and obtain complementary views for an in-depth understanding of the data. Finally, the ‘CA summary’ tab (see Figure 3) uses FactoMineR’s summary.ca function to display all the elements, summarizing the analysis (contributions, coordinates, and cosine squared).

5. Exploring Social Representations of Wind Turbines: A Case Study

5.1. Wind Turbines as an Object of Environmental Representations

Wind turbines occupy a central role in contemporary debates about the energy transition and the fight against climate change. As societies grapple with the urgent need for sustainable energy solutions, wind turbines symbolize technological innovation and a commitment to environmentally friendly energy production. However, these symbols of progress also evoke polarized reactions, becoming focal points of social and political controversy [59]. The divergent views on wind turbines reflect broader tensions in public discourse about the balance between environmental benefits and social acceptability.

The contentious nature of wind farms arises from multiple factors that influence how local citizens and communities perceive them. Among the most frequently cited concerns are the visual impact of turbines on landscapes, noise pollution, and their alleged effects on wildlife, particularly birds and bats [60,61]. These issues are not merely technical; they are embedded in social representations that shape individual and group interpretations of wind energy projects. For example, concerns about the visual impact often reflect deeper values related to the preservation of natural or cultural heritage, which are particularly strong in rural and coastal areas where wind farms are often sited [62]. Additionally, fears of reduced biodiversity due to turbine installations reinforce resistance, particularly when local communities place intrinsic value on wildlife conservation.

These concerns frequently manifest in public protests and opposition movements. Local associations and citizen groups, motivated by a sense of place attachment and perceived threats to their quality of life, often lead campaigns to prevent the construction of new wind farms [63]. Such opposition is further fueled by economic arguments, including fears about potential negative impacts on tourism or property values [64]. Notably, these protests often position wind turbines as symbols of external imposition, representing decisions made by distant authorities without adequate local consultation or involvement.

Understanding the social representations of wind turbines allows researchers to examine how individuals and groups construct meanings around these objects, which embody both the promise of a greener energy future and sources of social tension. Social representations are shaped by two primary dimensions: (1) attitudes toward wind turbines themselves, encompassing explicit evaluations of their environmental, esthetic, and social impacts and (2) trust in authorities regarding wind turbines, which reflects confidence in the transparency, fairness, and intentions of decision-makers involved in wind energy implementation [59,65]. These dimensions are shaped by both individual beliefs and broader institutional contexts, creating a complex interplay between local and systemic factors.

Attitudes toward wind turbines often hinge on their perceived trade-offs between environmental and social benefits. For instance, while many individuals recognize wind turbines as symbols of clean energy and climate mitigation, others associate them with the industrialization of natural landscapes and disruptions to local ecosystems [62]. This duality is reflected in research showing that perceptions of visual disamenity can outweigh perceived environmental benefits, especially in regions where tourism and scenic beauty are economically significant [61].

Trust in authorities also plays a crucial role in shaping public acceptance or rejection of wind energy projects. Studies indicate that perceptions of fairness and inclusivity in the planning and decision-making processes strongly influence local support [66]. When communities feel excluded from these processes or perceive that developers prioritize profits over local well-being, resistance is likely to emerge. Conversely, projects that engage local stakeholders meaningfully and transparently often enjoy higher levels of acceptance [65].

The study of social representations of wind turbines thus provides a comprehensive framework for identifying the points of tension and beliefs that influence public attitudes. By mapping these representations, researchers can uncover how different social groups interpret the trade-offs associated with wind energy. For example, Gee and Burkhard [62] demonstrated that the acceptability of wind farms is closely tied to their alignment with locally valued landscapes and ecological priorities. Similarly, Devine-Wright [59] highlighted how place attachment and emotional connections to specific environments shape resistance or support for wind projects.

By integrating these insights, policymakers and developers can design wind energy initiatives that better address the diverse concerns and values of local communities. This involves not only improving the technical aspects of turbine installations but also fostering trust and transparency in the planning processes. The social acceptance of wind turbines, therefore, hinges on their ability to resonate with local values, mitigate perceived harms, and involve communities in shaping their implementation. Thus, the study of social representations of wind turbines should help to identify the points of tension and beliefs that influence the acceptance or rejection of this form of renewable energy.

5.2. Materials and Method

The study was conducted using an online questionnaire via the LimeSurvey platform (version 3.23.6+200929). Data collection took place between 18 and 29 November 2024. The invitation message mentioned that the study was about how one perceives the world, with no specific mention of wind turbines. A total of 323 participants freely took part in this study without compensation (68.60% of women, Mage = 37.30, SD = 13.50, and age range: 18–80). After applying the exclusion criteria (legal minimum age, participants who reported being distracted or disturbed, and participants not a resident in France) the final sample consisted of 312 participants (69% of women, Mage = 37.20, SD = 13.70, and age range: 18–80). After freely consenting to participate in the study, participants provided some socio-demographic characteristics (gender, age, and country of residence) and then completed a free association task that included attitudes towards wind turbines and trust in authorities regarding wind turbines.

Free associations: After providing some socio-demographic characteristics, the participants first completed an evocation questionnaire [14,16]. This method is based on free association and involves a prime word (here “wind turbines”; “éoliennes” in French), to which participants must spontaneously list four words, phrases, or feelings. This method aims to identify the cognitions belonging to the social representation [67,68,69].

Attitudes towards wind turbines: Participants completed a Likert-scale questionnaire comprising 15 items designed to assess their attitudes toward wind turbines across several dimensions. The scale evaluates perceptions of environmental and energy benefits (e.g., “Wind turbines are essential for combating climate change”) as well as their esthetic and landscape impact, capturing whether participants view wind turbines as enhancing or degrading natural and rural environments (e.g., “Wind turbines disfigure rural and coastal landscapes”). It also addresses perceived nuisances and risks, such as noise pollution and harm to wildlife (e.g., “Wind turbines negatively impact local wildlife”), and explores beliefs about the economic and social value of wind turbines, including their contribution to job creation and public resource allocation (e.g., “Wind turbines create local jobs and stimulate the economy”). Finally, the scale measures overall support and preference, including participants’ willingness to accept wind turbines locally and their attitudes compared to other renewable energy options (e.g., “I prefer other renewable energy sources over wind turbines”). Some items are reverse-coded to account for response biases, and higher scores reflect more favorable attitudes toward wind turbines. A principal component analysis (PCA) with Varimax rotation revealed a two-factor structure, explaining the underlying dimensions of attitudes. Factor 1, labeled “Attitude Pro-Wind Turbines”, includes items reflecting positive perceptions of wind turbines, particularly regarding their environmental and energy benefits, with strong loadings (e.g., “Installing more wind turbines is good for our energy independence” = 0.784). Factor 2, labeled “Localized Critical Attitude”, captures more critical perceptions related to local impacts such as noise, landscape disruption, and wildlife disturbance, with key loadings (e.g., “Noise from wind turbines unbearable for local residents” = 0.787). Bartlett’s test of sphericity (χ² = 2227, p < 0.001) and the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy (0.926) confirmed the suitability of the data for factor analysis. Reliability analysis demonstrated high internal consistency (McDonald’s ω = 0.907), supporting the scale’s validity. Based on the median scores of each factor, four groups were created to explore subgroup differences: Attitude_Pro+ (high positive attitudes), Attitude_Pro− (low positive attitudes), Attitude_Crit+ (high critical attitudes), and Attitude_Crit− (low critical attitudes). These groups enable a nuanced understanding of how positive and critical attitudes toward wind turbines coexist and vary across individuals. Each participant was independently categorized on both dimensions (pro-attitude and critical attitude) based on median splits. Consequently, individuals could belong to more than one group, reflecting the coexistence of positive and critical evaluations of wind turbines. This approach is consistent with the theoretical framework of social representations, which acknowledges the presence of ambivalence and complexity in collective and individual attitudes [3].

Trust in authorities regarding wind turbines: Participants also completed a short Likert-scale questionnaire that evaluated their level of trust in authorities concerning wind turbines. Distrust of all authorities is a major phenomenon these days [70]. This scale consists of 7 items that measure perceptions of transparency, decision-making fairness, and the motives of authorities (e.g., “I trust authorities to select appropriate locations for wind farms” and “Authorities use wind energy as a pretext to serve hidden economic interests”). One item with a conspiratorial tone (“Authorities use wind energy as a pretext to serve hidden economic interests”) was included to capture mistrust of authorities (conspiracy theory affects many areas [71]). Higher scores reflect greater trust in authorities regarding wind turbines. A principal component analysis (PCA) with Varimax rotation confirmed a unidimensional structure, with all items loading strongly on a single factor (loadings ranging from 0.627 to 0.796). Bartlett’s test of sphericity (χ² = 814, p < 0.001) and the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy (0.853) supported the factorability of the items. Reliability analysis indicated good internal consistency (Cronbach’s α = 0.849), validating the scale’s use for assessing trust in this context. Based on the median trust score, participants were divided into two groups: Trust+ (above the median, indicating higher trust in authorities) and Trust− (below the median, indicating lower trust or greater mistrust), enabling a comparative analysis of attitudes and perceptions across these groups.

5.3. Results

After cleaning up the corpus obtained during the evocation task (typos, homogenization of certain terms, singular/plural, etc.), a preliminary analysis [72] reveals 1248 evocations for 213 different evocations and 134 hapaxes (terms mentioned only once). The diversity index (number of different evocations/total number of evocations) is 17.07%, revealing a certain lexical richness in the responses. This is consistent with a corpus centered on a socially complex and potentially controversial object, such as wind turbines. The following

Table 1. Most frequent evocations and their cognitive salience in social representations of wind turbines.

Evocations	Frequency	Average Rank of Appearance
winds	229	1.54
energy	109	2.23
ecology	79	2.35
electricity	75	2.81
gigantic	46	2.83
disfigured landscapes	39	3.28
renewable energy	35	1.77
blades	33	2.58
noise	28	2.82
renewable	28	2.50

shows the ten most important words with their frequency and average rank of appearance.

The most frequent evocations, such as ‘winds’ (229 occurrences) and ‘energy’ (109 occurrences), highlight the physical and functional aspects of wind turbines (cf. Table 1). Words like ‘ecology’ (79 occurrences) and ‘renewable energy’ (35 occurrences) convey environmental concerns, while more critical terms such as ‘noise’ (28 occurrences) and ‘disfigured landscapes’ (39 occurrences) reflect perceived negative impacts. The average rank of appearance of the evocations provides insights into the cognitive salience of the words for the participants. Evocations with a low average rank, such as ‘winds’ (1.54) and ‘renewable energy’ (1.77), are likely perceived as intrinsically linked to wind turbines and are activated as a priority. Conversely, terms like ‘disfigured landscapes’ (3.28) appear later, suggesting that they are more secondary or derived from initial representations. Wind turbines are represented through multiple and sometimes contradictory dimensions, reflecting both their symbolic role in the energy transition and the resistance they generate.

We conducted a correspondence factor analysis (Corr.F.A.) [58,73,74,75] on the verbal productions of participants. The correspondence factor analysis (CFA) reveals significant patterns in the social representations of wind turbines, as visualized in the graphical representation (see Figure 4). The Corr.F.A. identified two primary dimensions, which together explain 87.98% of the variance (69.54% for Dim 1 and 18.44% for Dim 2). Each dimension corresponds to a set of opposing associations and perceptions related to wind turbines, providing insight into the socio-representational universe that structures collective views.

As a reminder, grayed frames refer to the (independent) variables. Independent variables (bold) contribute to the formation of Factor 1. Independent variables (italic) refers to the independent variables which contribute to the formation of Factor 2. Observations (bold) refers to the observations which contribute to the formation of Factor 1. Observations (italic) refers to the observations which contribute to the formation of Factor 2. Observations (bold + italic) refers to the observations which contribute to the formation of both Factors 1 and 2. Figure 5 shows the visualization of the different clusters. This visualization was carried out outside FactoShinySR using a simple drawing program. The solid line shows the oppositions on axis 1, and the dotted line shows the oppositions on axis 2.

Axis or Dimension 1, named “Ecological vs. critical attitudes”, explains 69.5% of the variance. It reflects a fundamental opposition between positive representations of wind turbines and negative criticisms. On the left, evocations such as “clean energy”, “sustainable”, and “renewable” refer to ecological dimensions and the potential of renewable energies for a sustainable future (cf. Figure 4 and Figure 5). These associations emphasize their functional role in combating climate change. On the right, words like “visual pollution”, “noise”, and “biodiversity” reflect negative perceptions of the environmental and esthetic impact of wind turbines. This juxtaposition underscores the duality in how wind turbines are symbolically and functionally perceived.

Axis or Dimension 2, named “Opposition abstract vs. concrete”, explains 18.4% of the variance. This axis highlights an opposition between systemic and abstract representations, located at the top of the graph, and concrete or political criticisms, located at the bottom. For example, “ecological transition”, “unstable energy”, and “green electricity” evoke an optimistic, global vision of renewable energies. At the bottom of the chart, a lack of confidence (‘Trust−’) is often associated with critical terms such as “controversies”, “false solution”, and “not in my backyard”, suggesting skepticism about the motivations behind wind energy projects. They point to socio-political debates and local resistance often linked to conflicts of interest or a perception of wind turbines as harmful. References such as “biodiversity”, “birds”, and “impact” indicate a concern for ecological consequences, in particular the effects of wind turbines on flora and fauna.

Some elements participate in both axes, reflecting complex and ambiguous dimensions. For example, the term “cost” or “disaster” reflects both concrete criticisms and perceived global impacts. Finally, the explanatory variables reinforce this structuring. ‘Attitude_Pro+’ and ‘Attitude_Crit+’ are positioned oppositely on Axis 1, reflecting a clear polarization between favorable and critical attitudes (Cumulative Contributions: 16.42% and 20.28%, respectively). Similarly, ‘Trust+’ (Axis 2 Contribution: 32.82%) and ‘Trust−’ (Axis 2 Contribution: 33.54%) mainly influence Axis 2, illustrating the central role of trust in shaping representations of wind turbines.

One methodological limitation concerns the use of median splits to define attitudinal and trust groups. While this ensures balanced subgroup sizes for statistical analysis, it may include participants with moderate or ambivalent positions, potentially diluting attitudinal contrasts. Future research with larger samples could implement stricter thresholds (e.g., ±1 SD) to identify more polarized profiles.

In conclusion, the social representation highlights distinct groups of terms that form the socio-representative universe of wind turbines. These groups reflect the heterogeneity of public perceptions and their underlying dimensions:

• Pro-Environmental Cluster: Terms such as “green electricity”, “sustainable”, “clean energy”, and “future” dominate this cluster. These words align with the positive environmental symbolism of wind turbines and their role in mitigating climate change. Participants who evoke these terms likely perceive wind turbines as an essential component of the energy transition.
• Critical/Resistance Cluster: This cluster includes terms such as “noise”, “visual pollution”, “ugly”, and “not in my backyard”. These terms are closely tied to the experiential and esthetic impacts of wind turbines, reflecting resistance based on local and personal considerations.
• Technocratic and Trust Dynamics: Terms like “paradox”, “industry”, “controversies”, and “false solution” cluster around the theme of institutional trust. These words point to the importance of governance and stakeholder engagement in shaping public acceptance.
• Wildlife and Biodiversity Cluster: Words such as “birds”, “biodiversity”, and “impact” indicate a focus on ecological concerns. This cluster underscores the significance of environmental trade-offs in the broader discourse on renewable energy.

6. Conclusions

This study highlights the complex social representations of wind turbines, revealing both their symbolic role in the energy transition and the local resistance they generate. The results demonstrate that wind turbines occupy a contested space within the socio-representational field, as they are strongly associated with environmental progress and sustainability while simultaneously perceived as sources of local disruption and ecological harm. Using correspondence factor analysis (CFA) via the FactoShinySR application, we identified two main dimensions structuring public perceptions: ecological versus critical attitudes and abstract versus concrete oppositions. This duality reflects broader societal tensions between global environmental goals and local impacts, illustrating the multifaceted nature of attitudes towards renewable energy.

FactoShinySR, the interactive tool presented in this article, offers significant advantages in the study of social representations. Its user-friendly interface democratizes advanced statistical methods such as CFA, enabling researchers with no programming expertise to carry out these analyses. The tool’s visualization functions enable the intuitive interpretation of complex data, promoting a deeper understanding of environmental perceptions and, beyond that, of any object of social representation. By incorporating customizable options for filtering and displaying contributions, FactoShinySR ensures the clarity and reproducibility of scientific research.

Future improvements to FactoShinySR could include expanding its functionality to support more diverse datasets and multilingual capabilities, particularly for cross-cultural research on global environmental issues. Additionally, incorporating real-time collaboration features would allow researchers to share and refine analyses more effectively. Finally, the ability to visualize clusters directly without the need for third-party drawing software would be a significant gain. These enhancements would not only extend the tool’s applicability but also contribute to its adoption in interdisciplinary and applied contexts, bridging the gap between academic research and policy-making in renewable energy and other environmental domains.