Study on the multi-dimensional paths of ecological ethics influencing the realisation of the value of herders’ grassland ecological products Authors: Jingjin Li, Juan Yao, Jiran Hu, Ali Zhu, Haoran Wei, Huijuan Du Published: 2026 ABSTRACT The efficient and orderly realisation of the value of grassland ecological products is the key to achieving the strategic goal of transforming ‘green mountains’ into ‘golden mountains’. The behaviour of herders, as the direct promoters of the realisation of the value of grassland ecological products, is crucial to achieving this goal. Therefore, it is of great significance to carry out research on the behaviour of micro-entrepreneurs in order to accelerate the cracking of the problem between the shortage of ecological products and the constraints of natural resources, and to promote balanced development. Based on the eco-ethical perspective, the contradictory factors and research frameworks affecting the behaviours of grassland eco-products value realisation were extracted from five aspects: natural relationship cognition, grassland ecological function cognition, grassland ecological statute awareness, pro-environmental attitudes, and knowledge of grassland ecological husbandry. A mixed method of structural equation modelling (SEM) and fuzzy set qualitative comparative analysis (fsQCA) was used to explore the influencing mechanisms and antecedent configurations of the behaviours of value realization of grassland ecological products in China. The results showed that natural relationship cognition, grassland ecological function cognition, grassland ecological statute awareness, pro-environmental attitude and grassland ecological animal husbandry knowledge all significantly influenced the intention and behaviour of grassland ecological product value realisation. And the intention to realise the value of grassland ecological products played a partial mediating effect between the ecological ethical view and the behaviour of realising the value of grassland ecological products. Secondly, policy support played a positive moderating role in the intention and behaviour of grassland ecological product value realization. Finally, six nested configuration pathways leading to the behaviour of value realisation of grassland ecological products have been identified. Introduction With the depletion of global resources and the increasing deterioration of the ecological environment, issues of ecology and sustainable development have increasingly become a focal point of attention for countries around the world1. To address this issue, the Third Plenary Session of the 20th Central Committee of the Communist Party of China called for the improvement of the institutional framework for ecological civilisation and the refinement of mechanisms for realising the value of ecological products, thereby accelerating the implementation of the principle that ‘lucid waters and lush mountains are invaluable assets’. As China’s ecological civilisation continues to advance, the ‘Two Mountains’ philosophy has become a widely accepted consensus across society. The realisation of the value of ecological products serves as a vital bridge for putting this philosophy into practice; it not only transforms ecological resource advantages into economic advantages2, but also overcomes the ‘resource curse’ theory. Consequently, there is an urgent need to actively explore models for realising the value of ecological products, refine the mechanisms for such realisation, and facilitate the conversion of ecological product value. This not only aligns with the fundamental requirements for thoroughly implementing Xi Jinping’s Thought on Ecological Civilisation for a New Era, but also provides a sustainable development pathway for regions possessing significant ecological advantages yet situated in economic depressions. Secondly, from the perspective of resource distribution, grasslands—as China’s largest terrestrial ecosystem—are endowed with abundant ecological resource stocks and ecological product production capacity, serving as a vital ecological barrier and strategic resource for the nation. Through the effective and rational utilisation of grassland resources, herders provide a diverse range of grassland ecological products, including livestock products and grassland cultural tourism. Furthermore, from a supply and demand perspective, the realisation of the value of ecological products differs from exploitative resource development models. Ecological resources do not diminish abruptly with increased consumption; rather, through conservation, they generate a continuous self-sustaining supply capacity and meet the growing demand for high-quality ecological products. Ecological ethics, with the harmonious coexistence of humanity and nature at its core, serves as a vital set of values and a theoretical foundation for regulating the ecological behaviour of individual actors and underpinning ecological governance in pastoral areas3,4. The nomadic peoples of north-western China have developed indigenous ecological ethical concepts with distinctive regional characteristics, embodying ecological wisdom such as the coexistence of humanity and nature, reverence for nature, and the prudent use of grasslands; these constitute a localised ecological resource distinct from Western theoretical systems5,6. Ecological ethics in pastoral areas primarily comprise two dimensions: ecological customary norms and ecological literacy. The former consists of informal local norms formed through the long-term productive practices of nomadic peoples, covering natural taboos, the balance between grass and livestock, and resource management. These norms can operate independently of external institutional constraints, continuously guiding herders to engage in conservation-oriented grazing practices and maintaining the ecological stability of the grasslands7–9; the latter manifests as herders’ ecological cognition, attitudes and value orientations, serving as a vital vehicle for the practical implementation of ecological ethics. Academic research generally explores these through the dimensions of ecological knowledge, ecological attitudes and ecological behaviour, confirming that they can positively drive the ecological participation of farmers and herders10–12. Existing empirical research further indicates that, as a concrete expression of ecological ethics, ecological cognition serves as the core driving variable for farmers’ and herders’ ecological conservation behaviour. Drawing on theoretical models such as the Theory of Planned Behaviour (TPB) and the Value-Based Normative (VBN) framework, it has been found that elements inherent in ecological ethics—such as individual norms and ecological worldviews—can positively activate herders’ willingness to engage in ecological conservation through multidimensional pathways, and can interact with ecological compensation policies to reinforce the endogenous motivation for ecological behaviour13–15. Ecological ethics is a core endogenous variable shaping herders’ productive and ecological behaviour, and is also key to analysing the micro-level logic underlying the behaviour of value realisation for grassland ecological products. Current academic research on herders’ behaviour tends to focus on external, explicit factors such as policy incentives, economic returns and institutional constraints, interpreting their production and ecological decisions by treating them as economically rational agents16; it generally overlooks the profound shaping influence of endogenous factors such as indigenous ecological ethics and local cultural norms on herders’ behaviour. In fact, the differentiated behaviours of herders—such as pasture utilisation, management and restoration, and production transformation—are the result of the combined influence of both internal and external factors; meanwhile, indigenous ecological ethics rooted in pastoral areas serve as the core intrinsic driving force guiding herders to actively engage in ecological conservation and the rational utilisation of pasture resources17. Relying solely on external policies and economic incentives makes it difficult to explain the behavioural differentiation among herders in pastoral areas and fails to elucidate the underlying logic of micro-level behaviour, thereby highlighting the necessity of research from an ecological ethics perspective. As the primary stewards and users of grassland ecosystems, herders’ varied ecological behaviours and production decisions ultimately determine the quality of supply, sustainability and effectiveness of value conversion of grassland ecosystem services; they constitute the core micro-foundation for the behaviour of value realisation of grassland ecological products. The theory of ecosystem services has laid a solid foundation for the construction of a research framework on ecological products18–20. Domestic scholars, drawing on the practice of ecological civilisation, have defined a localised concept of ecological products, propelling the realisation of their value to the forefront of ecological governance research21. The academic community has now conducted systematic research on the conceptual analysis of ecological products, the construction of value accounting systems, and diverse pathways for value realisation, forming a macro-level research paradigm underpinned by GEP accounting, ecological compensation, and the industrialisation of ecology22–25. Overall, the pathways for realising the value of ecological products identified in existing research can be categorised into four mainstream models, which constitute the core practical framework for the current transformation of ecological product value. The first is the ecological compensation pathway, which is also the mainstream model with the highest level of academic recognition. Relying on government regulation and market-based collaborative compensation mechanisms, it balances the costs of ecological protection with development benefits, thereby achieving the synergistic enhancement of ecological and environmental protection alongside socio-economic benefits26–28. The second is the resource-factor empowerment pathway. Grounded in the perspective of ecological resource indices, this approach promotes the efficient coupling of natural endowments, human-made modifications and human capital to achieve complementary empowerment between ecological resources and market capital, thereby revitalising existing ecological resources and unlocking incremental ecological value29,30. The third is the path of industrialised ecological governance. Relying on a multi-stakeholder collaborative model involving government guidance, scientific research support, corporate operation and the participation of farmers and herders, this approach harnesses green industries such as eco-tourism and specialised livestock farming to generate welfare dividends. It utilises industrial development to monetise ecological products in the market, whilst emphasising the establishment of a community of shared interests and a collaborative governance framework to enhance the long-term sustainability and stability of value realisation31–34. Fourthly, the ecological restoration and value-enhancement pathway targets areas of ecological degradation and environmental damage. Through systematic ecological governance and restoration projects, it restores the structure and functions of grassland ecosystems, continuously enhances the capacity to supply ecological products, and consolidates the resource foundation for realising the value of ecological products, thereby achieving mutual empowerment between ecological restoration and value enhancement35,36. Based on literature on the realisation of the value of ecological products indexed in Web of Science between 2012 and 2026, this paper employs CiteSpace software to conduct bibliometric and visualisation analyses, providing an intuitive representation of research progress and evolving trends in this field. According to the keyword-time-period map (Fig. 1), research themes shifted gradually from early-stage macro-governance, climate change and biodiversity conservation to policy instruments centred on government and carbon markets (2015), before expanding to elements and process mechanisms such as China’s ecological products, ecosystem services, nitrogen use efficiency and soil microbial communities (2020). By around 2025, a trend towards modelling and value quantification emerged, with the emergence of cutting-edge directions such as support-pressure models, entity-value linkages and specific biological species. Overall, this reflects the field’s continuous deepening from basic ecological governance towards market mechanisms, multi-factor coupling and system modelling. Fig. 1A temporal map of key terms relating to the realisation of the value of ecological products. A temporal map of key terms relating to the realisation of the value of ecological products. Existing research has laid a solid theoretical foundation for this study. By systematically reviewing the existing literature and constructing a literature map, this paper has confirmed the actual existence and specific boundaries of these gaps, thereby identifying significant content gaps and research limitations that still exist in the current academic discourse. First, regarding the tendency towards a macro-level research perspective: existing studies generally focus on top-level macro-policy frameworks and meso-level industrial development, whilst there is a relative lack of exploration into the internal logic and underlying ethical drivers of herders’ micro-level behavioural choices. Although existing research has confirmed that there are significant differences in the operational mechanisms of the entities responsible for realising the value of different types of ecological products37,38, they predominantly analyse universal patterns by treating diverse ecological products across the entire region as their research subjects. Targeted, scenario-based studies specifically focused on grassland ecological products are scarce. This has resulted in a research landscape that prioritises the macro over the micro, and the general over the specific37, making it difficult to precisely align with the practical demands of grassland ecological governance and the efficient conversion of ecological product value in ecologically fragile ethnic regions of Northwest China. Second, regarding the limited depth of research: existing studies largely offer a superficial analysis of the dichotomous conflict between herders’ livelihoods and ecological conservation. Explanations regarding the internal processes by which ecological ethics are internalised into herders’ ecological behaviour and form transmission mechanisms remain rather superficial. This, to a certain extent, constrains a deep understanding of the challenges facing the coordinated development of ‘ecology, production and livelihood’ in pastoral areas. Third, regarding the suitability of research paradigms: Academic circles have largely adopted mainstream Western behavioural research paradigms directly, whilst paying insufficient attention to the embedded nature of local ecological wisdom, cultural norms and social structures in pastoral areas. This transplantation of paradigms may lead to a disconnect between theoretical explanations and practical experience in pastoral areas. Fourth, regarding the narrowness of research methods: existing methods predominantly employ linear analytical frameworks, with little consideration given to the complex influence of multi-factor configuration effects on herders’ behaviour; consequently, it is difficult to comprehensively reveal the intrinsic patterns of subject behaviour. Fifth, regarding the absence of systemic mechanisms and implementation pathways: existing research often discusses ecological ethics and herders’ ecological behaviour separately, or analyses individual components in isolation, whilst lacking optimised pathways and implementation methods for realising values that are tailored to the ethnic characteristics of pastoral areas and take into account the effects of multi-factor configurations4,14. Current research into ecological behaviour in rural areas predominantly employs an analytical paradigm that combines a general environmental value orientation with the TPB; yet these studies are generally characterised by superficial variable embedding and homogenised analytical frameworks. They lack theoretical refinement and scenario optimisation tailored to the ecological endowments and indigenous cultural characteristics of grassland ethnic regions. Furthermore, there has long been a disconnect between research on the core of ecological ethics and that on the behaviour of value realisation for grassland ecological products, making it difficult to accurately elucidate the endogenous driving mechanisms and the logic behind the differentiation of herders’ ecological value transformation behaviours. Consequently, this paper focuses on herders in ecologically fragile grassland regions of western China and identifies two core scientific questions: how can a refined ecological ethics system tailored to pastoral contexts be constructed to overcome the limitations of existing universal TPB analytical frameworks and expand the theoretical explanatory scope regarding the behaviour of value realisation for grassland ecological products? Amidst the interplay of diverse contextual factors, how do ecological ethics and external environmental factors interact to form differentiated pathways for the realisation of ecological product value in pastoral areas? Centring on these questions, this paper conducts systematic research at theoretical, empirical and practical levels. Theoretically, the study rejects the generalised value-embedding models common in traditional rural research, refines the multidimensional sub-dimensions of ecological ethics, and achieves a deep integration of indigenous ecological ethics with the TPB. This approach addresses the explanatory shortcomings of Western theoretical paradigms in pastoral contexts, bridges the academic divide between ecological ethics and research on the transformation of ecological value, and enriches the theoretical framework of grassland ecological governance. Empirically, based on micro-level survey data from pastoral areas and employing a mixed-methods approach combining Structural Equation Modelling (SEM) and Fuzzy Set Qualitative Comparative Analysis (fsQCA), this study overcomes the limitations of traditional linear analysis—which can only identify single net effects—to reveal the configuration mechanisms driven by multiple factors and the differentiated pathways to realisation, thereby enhancing the depth and precision of research on ecological behaviour in pastoral areas. In practical terms, by grounding the analysis in the incentive logic of herders’ endogenous ecological ethics, this study provides localised and actionable pathways to guide local stakeholders in actively participating in ecological conservation, revitalising grassland ecological resources, and advancing the value transformation of ecological products. It offers effective guidance for the construction of an ecological civilisation and the high-quality development of the ecological economy in ecologically fragile grassland regions. Theoretical analyses and research hypotheses Ecological ethics extends moral concern from relationships between people and between people and society to the relationship between humans and nature. Its core objective is to establish a development paradigm based on the harmonious coexistence of humans and nature, clarifying the moral obligations and codes of conduct of humanity as the primary agent of responsibility within the ecological community39. The grassland pastoral areas of Xinjiang constitute a typical human-land coupling system; the grasslands serve not only as the means of production upon which herders depend for their survival, but also as the cornerstone for the supply of regional ecological products and the maintenance of ecological functions. The pastoralists’ utilisation and management of the grasslands balance the dual dimensions of livelihood improvement and income generation with the ecological evolution of the grasslands. Essentially, this represents a dynamic equilibrium between ecological responsibility and economic imperatives, which aligns closely with the core research scope of ecological ethics. Consequently, it provides the central theoretical underpinning for this study’s interpretation of the indigenous value logic underlying pastoralists’ ecological behaviour. As a classic behavioural decision-making theory originating from a Western context40,41, the TPB fails to incorporate foundational elements such as local ecological ethics, indigenous culture and institutional characteristics. Consequently, its direct application proves ill-suited to the production methods and ecological cultural characteristics of Xinjiang’s nomadic peoples. It suffers from inherent limitations, including insufficient explanatory power in the local context and poor contextual adaptability, as well as an intrinsic contradiction between the Western theoretical paradigm and the local research context. In response, this study adopts a cross-cultural adaptation approach, acknowledging the contextual limitations of the original TPB theory. Rather than simply replicating its original framework, it refines and adapts the theory by defining its scope of application and embedding it within the localised ecological ethics system of pastoral areas. Significant gaps remain in existing research on the integration of TPB and ecological ethics: most studies focus on single public-benefit acts within ecotourism, failing to construct a hierarchical local ecological ethics system and neglecting the market-oriented transformation of grassland ecological products16; research in pastoral areas often employs generic environmental constructs, failing to undertake localised reconstruction in conjunction with nomadic customs14; conclusions from studies of fishing and farming contexts are difficult to adapt to the land-human relationships and management models of grassland nomads13,42; some studies emphasise policy and technological perception perspectives, failing to sufficiently explore deep-rooted local ecological ethical values15. Overall, existing research suffers from four major shortcomings: limited empirical contexts, overly generalised operationalisation of concepts, narrow research scopes, and superficial theoretical integration; there is a lack of specialised studies on the mechanisms specific to Xinjiang’s fragile grasslands and the Kazakh nomadic communities. Compared with existing work, this study focuses on the unique ecological and cultural characteristics of Xinjiang’s frontier pastoral areas. By integrating Kazakh pastoral practices, it constructs a five-dimensional indigenous ecological ethics system that balances the value realisation of both ecological governance and income generation for herding households, thereby achieving a deep coupling between indigenous ecological ethics and the hierarchical behavioural logic of the TPB. clarifying the mechanisms through which ethical values are translated into ecological behaviour. By achieving breakthroughs in empirical settings, conceptual reconstruction, research scope and theoretical mechanisms, this study effectively addresses the inadequacies of Western theoretical adaptation to local contexts and the numerous shortcomings of existing research. Building on the complementary nature of these two theoretical frameworks, this study constructs an integrated analytical framework comprising ‘indigenous ecological ethics as a foundational element’ and ‘an interpretation of the TPB behavioural logic’, thereby establishing a clear theoretical framework: ecological ethics provides the value foundation for the harmonious coexistence of humans and nature, explaining the intrinsic ethical causes of herders’ ecological behaviour; the TPB theory provides a classic hierarchical logic for the formation of individual behaviour, offering an important theoretical basis for the delineation of the ecological ethics dimension in this study. Drawing on the behavioural analysis logic of the TPB, this paper reconstructs a five-dimensional system of localised ecological ethics in pastoral areas, treating each dimension of ecological ethics as a core antecedent driving herders’ ecological behaviour. It thereby corrects the erroneous approach found in the original text—which substituted the TPB structure with ethical dimensions—and thoroughly avoids theoretical logical conflicts. Relying on the progressive mechanism of ‘cognition—values—norms—emotions—practice’, this study establishes a complete transmission chain comprising ‘tracing the origins of ecological ethical values—psychological transformation of behaviour—generation of behavioural intentions—implementation of practical behaviour’ during the behaviour of value realisation for grassland ecological products (Fig. 2). Fig. 2Logical framework for realising the value of grassland ecological products from the perspective of pastoralists. Logical framework for realising the value of grassland ecological products from the perspective of pastoralists. Ecological ethics embodies a comprehensive logical framework comprising ‘cognitive foundation—value establishment—normative constraint—emotional identification—practical transformation’[5–10]. This study combines the nomadic culture of Xinjiang with the production characteristics of pastoral areas to undertake a localised and operationalised reconstruction of ecological ethics constructs. It systematically clarifies the one-to-one transmission relationships between the five major dimensions of ecological ethics and the TPB attitude, subjective norm, and perceived behavioural control. It clearly elucidates the adaptability advantages and theoretical contributions of local ecological ethics constructs compared to Western general models, highlighting the enhanced explanatory power and academic innovation value of theoretical refinement within the pastoral context. Specifically, this study draws upon the behavioural hierarchy of the TPB—comprising attitude, subjective norm and perceived behavioural control—and combines this with the local characteristics of pastoral areas to delineate a five-dimensional ecological ethics system. Each dimension forms a precise correspondence with the core behavioural logic of the TPB: natural relationship cognition and grassland ecological function cognition constitute the foundation of individual ecological cognition and values, corresponding to and shaping herders’ attitudes towards ecological behaviour; Grassland ecological statute awareness, grounded in local pastoral norms and ethnic consensus, corresponds to the logic of individual subjective norms; pro-environmental attitudes and knowledge of grassland ecological pastoralism, operating at the levels of emotional identification and practical skills, correspond to and underpin an individual’s perceived behavioural control. Each dimension progresses sequentially and fulfils its distinct function; drawing upon classical behavioural formation logic, they achieve the complete transformation of ethical values into the intention for value realisation of grassland ecological products and the behaviour of value realisation. Natural relationship cognition constitutes an individual’s stable understanding of the fundamental relationship between humans and nature, as well as their fundamental stance on this matter. It serves as the logical starting point for all ecological concepts and behaviours, and is characterised by its foundational and universal nature. Humanity and nature form a mutually beneficial and symbiotic organic whole; whilst humans rely on nature to obtain the resources necessary for survival and development, they must respect ecological laws and take into account the environment’s carrying capacity43,44. For herders, a clear understanding of the symbiotic relationship between humans and the land can effectively guide them to proactively adapt to the ecological laws of the grasslands, actively participate in grassland ecological conservation and the conversion of ecological product value, and thereby enhance their intention for value realisation of grassland ecological products and their behaviour related to that realisation. Based on this, the following hypotheses are proposed: H1. Herders’ natural relationship cognition positively influences their intention for value realisation of grassland ecological products. H2. Herders’ natural relationship cognition positively influences their behaviour related to the value realisation of grassland ecological products. Grassland ecological function cognition represents a concrete extension of natural relationship cognition within the grassland context; it constitutes herders’ value judgements regarding core ecological functions such as water conservation, soil and water conservation, and the maintenance of biodiversity. Grassland ecosystems provide a diverse range of ecosystem services and form the cornerstone of the supply of ecological products in pastoral areas45. The deeper herders’ grassland ecological function cognition, the more they recognise the economic and ecological value of grassland ecological products, and the more willing they are to actively engage in ecological production, supply high-quality ecological products, and promote the intention for value realisation of grassland ecological products. Based on this, the following hypotheses are proposed: H3. Herders’ grassland ecological function cognition has a positive influence on their intention for value realisation of grassland ecological products. H4. Herders’ grassland ecological function cognition has a positive influence on their behaviour of value realisation of grassland ecological products. The core tenet of ecological ethics is to regulate human exploitation of nature through moral principles, thereby translating the concept of harmonious coexistence between humanity and nature into concrete behavioural norms46. Over time, the Kazakh herders of Xinjiang have developed a nomadic ethic and folk traditions characterised by reverence for nature and care for the grasslands. Relying on village regulations and ethnic consensus, they have established a stable sense of identification with ecological conservation norms, forming a unique social mechanism of ecological constraint amongst the herders. Based on social identity theory, individuals tend to conform to group norms and behavioural paradigms. The endogenous ecological consensus formed through geographical and kinship interactions in pastoral areas can effectively curb extensive grazing practices and guide the adoption of green production methods47,48. Pro-environmental attitudes belong to the category of affective attitudes; they represent an individual’s emotional concern for the natural environment, value identification, and the degree of acceptance and endorsement of environmental protection behaviours49. They possess universal and affective characteristics, emphasising environmental inclinations at the level of subjective attitudes and focusing on the stance taken towards environmental protection behaviours. Specifically, this manifests as an individual assuming a degree of responsibility for specific behaviours in particular contexts50, thereby leading to proactive environmental protection through their own efforts51,52. Pro-environmental attitudes are grounded in an individual’s value system, reflecting the composition and prioritisation of diverse values within that system53,54. The stronger the sense of responsibility herders feel towards their own living environment, the more likely they are to engage in pro-environmental behaviour55, and consequently, the more willing they are to provide grassland ecological products and to increase the extent of their practical actions. Consequently, herders are influenced by both individual and group effects in their decision-making processes. If herders identify with social pressure, expectations and moral standards originating from themselves, those around them or the group, and if they are inclined to meet these expectations, they will demonstrate corresponding willingness and behaviour56. Based on this, the following hypothesis is proposed: H5. Herders’ grassland ecological statute awareness has a positive influence on their intention for value realisation of grassland ecological products. H6. Herders’ grassland ecological statute awareness has a positive influence on their behaviour of value realisation of grassland ecological products. H7. Herders’ pro-environmental attitudes have a positive influence on their intention for value realisation of grassland ecological products. H8. Herders’ pro-environmental attitudes have a positive influence on their behaviour of value realisation of grassland ecological products. Knowledge of grassland ecological animal husbandry constitutes a practical system of knowledge and skills that enables herders to adapt to the ecological laws of the grasslands and engage in sustainable livestock production. The accumulation of ecological knowledge and production skills can help herders overcome production barriers and optimise their husbandry practices, thereby enhancing the quality and efficiency of ecological product supply whilst protecting the grassland ecosystem57. The more extensive herders’ knowledge base in ecological livestock farming, the stronger their capacity and initiative to engage in ecological production, and the more pronounced their intention for value realisation of grassland ecological products and behaviour regarding value realisation of grassland ecological products. Based on this, the following hypotheses are proposed: H9. Herders’ knowledge of ecological livestock farming on the grasslands has a positive influence on their intention for value realisation of grassland ecological products. H10. Herders’ knowledge of ecological livestock farming on the grasslands has a positive influence on their behaviour of value realisation of grassland ecological products. Behavioural intention serves as a direct precursor to actual behaviour and acts as a core mediating variable between antecedent factors and actual behaviour41. The various dimensions of herders’ ecological ethics drive the implementation of actual ecological production behaviour by shaping a positive intention for value realisation of grassland ecological products. Based on this, the following hypothesis is proposed: H11. Herders’ intention for value realisation of grassland ecological products has a positive influence on the behaviour of value realisation of grassland ecological products. H12. The intention for value realisation of grassland ecological products mediates the relationship between herders’ ecological ethical views and their behaviour. The policy context constitutes a key boundary condition for the transformation of intention into behaviour. Grassland ecological products possess the typical attributes of public goods, and fragile grassland pastoral areas have long been trapped in a vicious cycle of ‘ecological fragility and herder poverty’58. The implementation of policies such as grassland ecological compensation, grazing bans and rest periods, and grass-livestock balance has effectively compelled the transformation of traditional pastoral models, creating conditions for the realisation of the market value of grassland ecological products; accompanying policy publicity can further strengthen the motivation for herders to transform their ecological behaviour59. Based on this, the following hypotheses are proposed: H13. Policy support exerts a positive moderating effect on the process by which herders’ intention for value realisation of grassland ecological products is translated into behaviour. Materials and methods Data sources To scientifically investigate the realisation of the value of grassland ecological products, this paper conducts a questionnaire survey from the perspective of herders to identify the factors influencing their value-realisation behaviours, thereby providing micro-level empirical data to support subsequent efforts to address the challenges of realising the value of grassland ecological products and optimise practical pathways. This study selects the grasslands of the Ili River Valley in Xinjiang as the core survey area. This region is one of the areas with the richest grassland resources in Xinjiang, with a total area of natural grassland in the valley exceeding 20 million hectares and an average grassland coverage of 65%–85%. The grassland ecosystem is intact and diverse in type: the low-mountain areas serve as excellent pastures in spring and autumn; the mid-mountain areas are densely forested with spruce; and the high-mountain areas provide high-quality summer pastures. Combining ecological fragility with ethnic settlement characteristics, it is a typical region for studying the realisation of the value of grassland ecological products. The specific distribution of the survey area and an overview of grassland resources are shown in Table 1; Fig. 3. This survey employed a strategy combining multi-stage stratified random sampling with purposive sampling. First, sample strata were defined according to grassland ecological types and management scale; subsequently, townships, villages and herding households were selected at random, thereby balancing geographical representativeness with group heterogeneity. Finally, respondents were selected using random household sampling to ensure that the sample data effectively reflect the overall characteristics of local herders’ behaviour of value realisation of grassland ecological products. The questionnaire for this study is primarily divided into three modules: individual characteristics of herding households, herders’ ecological ethics, and their willingness and behaviour of value realisation for grassland ecological products. Variable measurement strictly adheres to existing authoritative research findings. The ecological ethics scale has been adapted to local contexts by incorporating grassland regional culture and the cognitive characteristics of herders. For willingness, behaviour, and other mediating and control variables, established academic scales and standard items have been selected. All variables are measured using a 5-point Likert scale, with values ranging from 1 to 5 corresponding to ‘strongly disagree’, ‘disagree’, ‘neutral’, ‘agree’, and ‘strongly agree’, accurately quantifying herders’ level of agreement with each item. Among these, the dependent variable—behaviour of value realisation of grassland ecological products—includes measurement items centred on the types, frequency, and diversity of herders’ actual participation, covering four practical behaviours: livestock sales, participation in cultural tourism, claims for ecological compensation, and grassland transfer; Due to the constraints of field research conditions in pastoral areas, the study was unable to obtain objective empirical data such as administrative ledgers or transaction records. Consequently, a five-point Likert scale was employed to conduct self-reported measurements among herders, with ratings assigned according to participation frequency from low to high, focusing on measuring the status of herders’ actual past behaviours. Furthermore, this study strictly distinguishes between the conceptual meanings of variables in the questionnaire wording: behavioural intention focuses on subjective willingness to participate in the future, whilst actual behaviour focuses on past actual participation. By clarifying the boundaries between the two at the levels of measurement semantics and content, the study effectively avoids the issue of homogenisation between the measurement of intention and behaviour. Following the completion of the questionnaire, two rounds of pilot testing were conducted. In the first round, the appropriateness of the survey questions was discussed with relevant experts and scholars, and ambiguities in the wording were corrected. In the second round, through preliminary research, ineffective items were eliminated, resulting in the finalised questionnaire. This ensured that the questionnaire content was scientifically sound, logically clear, and easy for herding households to understand and complete. The survey was conducted through face-to-face interviews between researchers and professional interpreters. Given that some local herders have limited proficiency in Mandarin, local interpreters who are fluent in the local ethnic languages and familiar with pastoral production were engaged to accompany the team throughout the process, ensuring the accurate transmission of information and unimpeded communication. Data collection took place between 1 June and 1 August 2024, coinciding with the peak grazing season on the grasslands, meaning the survey data more accurately reflects the actual production conditions of herding households. The research team strictly adhered to standardised procedures when conducting the fieldwork, and enlisted the assistance of professionals to facilitate communication during the survey period, thereby ensuring the authenticity and accuracy of the data. Once the questionnaires had been collected, the research team undertook a unified process of data collation and rigorous screening. A total of 356 questionnaires were distributed for this survey. In accordance with established data cleaning protocols, invalid questionnaires—including those with incomplete responses, logical inconsistencies or random entries—were excluded, resulting in a final total of 330 valid questionnaires, representing a response rate of 92.70%. The study clearly defined criteria for identifying invalid samples and comprehensively clarified the entire process—from questionnaire distribution and on-site collection to sample exclusion. This effectively mitigated selection bias, ensuring the scientific rigour and representativeness of the research sample, and providing robust and reliable data support for subsequent empirical analysis. It is hereby confirmed that all research involving human subjects and the associated methods have been conducted in accordance with relevant guidelines and regulations, and that all protocols have been approved by the Institutional Review Board of Xinjiang Agricultural University. Furthermore, it is confirmed that informed consent has been obtained from all research participants. Fig. 3Research Area. Research Area. Table 1Distribution of the survey sample.County (City)Townships/Scenic AreasSample Size (People)Proportion (%)Zhaosu CountyKarasu Town3711.212Hongnahai Town3510.601Chahanwusu Mongolian Township3610.909Akdala Township4814.545Tekesi CountyQiaolaketiere Township6319.091Koktirek Township3610.909Kara Dala Town7522.727Total330100.00 Table 1Distribution of the survey sample.County (City)Townships/Scenic AreasSample Size (People)Proportion (%)Zhaosu CountyKarasu Town3711.212Hongnahai Town3510.601Chahanwusu Mongolian Township3610.909Akdala Township4814.545Tekesi CountyQiaolaketiere Township6319.091Koktirek Township3610.909Kara Dala Town7522.727Total330100.00 Distribution of the survey sample. Sample characterisation The majority of respondents in this survey were male herders, accounting for 74.545%. Most herders on the grasslands belong to traditional nomadic ethnic groups, with men typically undertaking tasks such as herding that involve greater contact with the outside world, whilst women are more responsible for domestic affairs. The respondents were predominantly middle-aged; this is because older workers lack the sufficient energy to engage in activities such as herding, with herders aged 56 and over accounting for only 12.727%. The educational attainment of herders is generally low, with the majority holding a junior secondary school qualification or below, accounting for 70% of the total sample. In terms of political affiliation, the majority of herders are ordinary citizens, accounting for 70%, followed by Party members at 19.091%. Regarding household structure, the majority of herding households have two or fewer working-age members, accounting for 80% (Table 2). Table 2Description of the distribution of sample features.NormCategorisationFrequencyPercentage (%)NormCategorisationFrequencyPercentage (%)SexMale24674.545Social StatusThe Masses23170.000Female8425.455Political Party Member6319.091Age25 Years And Under3510.606Village Cadres175.15226–35 Years7121.515Cooperative Leaders20.60636–45 Years10832.727Teachers92.72746–55 Years7322.121Government Postings82.42456 Years And Over4212.727Participation In The Agricultural Labour Force1 People9729.394Educational AttainmentPrimary And Below6218.7882 People16750.606Junior High School16951.2123 People4513.636High School/Secondary School5817.5764 People164.848University/Tertiary4112.4245 People And Above51.515 Table 2Description of the distribution of sample features.NormCategorisationFrequencyPercentage (%)NormCategorisationFrequencyPercentage (%)SexMale24674.545Social StatusThe Masses23170.000Female8425.455Political Party Member6319.091Age25 Years And Under3510.606Village Cadres175.15226–35 Years7121.515Cooperative Leaders20.60636–45 Years10832.727Teachers92.72746–55 Years7322.121Government Postings82.42456 Years And Over4212.727Participation In The Agricultural Labour Force1 People9729.394Educational AttainmentPrimary And Below6218.7882 People16750.606Junior High School16951.2123 People4513.636High School/Secondary School5817.5764 People164.848University/Tertiary4112.4245 People And Above51.515 Description of the distribution of sample features. Research Methods Given that the herders’ behaviour of value realisation for grassland ecological products is driven by a combination of multiple factors and exhibits characteristics that encompass both linear causal relationships and configurational coupling effects, this study adopts a mixed-methods approach combining Structural Equation Modelling (SEM) with Fuzzy Set Qualitative Comparative Analysis (fsQCA) to overcome the analytical limitations of single research methods. This design is not a simple superimposition of methods, but rather a scientific choice based on the nature of the research questions, the complementary nature of the methods, and academic rigour. The behaviour of value realisation of grassland ecological products by herders is a complex socio-economic phenomenon influenced by the combined effects of multi-dimensional factors, including ecological ethics. Its driving mechanisms encompass both linear direct influences from individual factors and configurational effects arising from the interaction and coupling of multiple factors; a single method is insufficient to fully reveal its underlying patterns. The hybrid SEM-fsQCA approach enables a dual analysis of ‘net effects of independent variables and multi-factor synergistic effects’, precisely aligning with the core objective of exploring the complexity of herders’ behaviour and the multi-dimensional mechanisms influencing it[60]. Although the two methods differ in their analytical logic, improper application can easily lead to inconsistencies. This study establishes a closed-loop, complementary logic of ‘basic validation—in-depth expansion’ through the orderly sequence of ‘first validating linear net effects, then mining configuration pathways’: SEM undertakes the functions of preliminary variable screening and net effect testing, providing fsQCA with a scientific and robust set of conditional variables, thereby avoiding arbitrariness in variable selection during configurational analysis61; fsQCA, in turn, builds upon the SEM analysis results to deeply explore multi-factor joint driving pathways, compensating for SEM’s inability to depict synergistic effects and equivalence mechanisms. By mutually supporting and deeply integrating with one another, these two approaches not only effectively overcome the limitations of individual methods, ensuring the rigour and precision of the research, but also fully address the complexity of real-world issues. This enables a comprehensive analysis of the behaviour of value realisation for grassland ecological products, thereby ensuring the systematic, comprehensive and scientific nature of the research conclusions. This paper first employs SEM to conduct an empirical analysis. As a mainstream method of multivariate statistical analysis, SEM can effectively identify direct and indirect effects between latent variables, thereby reducing measurement error and systematic bias62; This method imposes no strict constraints on data distribution or sample size, and is capable of accurately estimating path coefficients between latent variables and their statistical significance63, thereby clearly identifying the core linear influence pathways in herders’ behaviour regarding the realisation of the value of ecological products, clarifying the direction and intensity of the independent effects of individual variables, and providing a basis for variable selection in subsequent configurational analysis. Regarding the choice of software for SEM analysis, this study employs Covariance-Based Structural Equation Modelling (CB-SEM) rather than Partial Least Squares Structural Equation Modelling (PLS-SEM). The primary rationale for this choice is that the core objectives of this study are theoretical testing and model validation, rather than sample prediction. Existing research indicates that PLS-SEM is more suitable for exploratory analysis, prediction-oriented models, complex formative indicator structures, or small-sample scenarios64–66; whereas CB-SEM is better suited to confirmatory research, with its application premised on large samples and the assumption of multivariate normal distribution67. Furthermore, CB-SEM offers significant advantages in terms of overall model fit assessment, estimation of latent variable relationships, and hypothesis robustness testing, thereby meeting the methodological requirements of this study for rigorous testing of theoretical models68. This approach achieves comprehensive estimation through the structural model (Eq. 1) and measurement models (Eqs. 2 and 3), ensuring the standardisation and rigour of the empirical analysis.1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upsigma\:}={\Gamma\:}{\uptau\:}+\mathrm{B}{\upsigma\:}+{\upzeta\:}$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{X}={{\varnothing}}_{\mathrm{X}}{\uptau\:}+{\upvarphi\:}$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{Y}={{\varnothing}}_{\mathrm{Y}}{\upsigma\:}+{\upomega\:}$$\end{document} Here, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\uptau\:}$$\end{document} is the exogenous latent variable, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upsigma\:}$$\end{document} is the endogenous latent variable, the estimated parameters and regression coefficients are \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\Gamma\:}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{B}$$\end{document}, respectively, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upzeta\:}$$\end{document} is the error term of the model. Exogenous and endogenous latent variables are denoted by X and Y, respectively. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\varnothing}}_{\mathrm{X}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\varnothing}}_{\mathrm{Y}}$$\end{document} are the loading matrices of X and Y with their corresponding latent variables, and the error terms for X and Y are \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upvarphi\:}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\upomega\:}$$\end{document}, respectively. However, traditional SEM methods can only test single causal relationships between predefined variables, focusing on identifying the net effects of independent variables. They struggle to reveal the configuration effects arising from the interaction of multiple variables, nor can they explain the equivalence mechanism whereby ‘different combinations of factors can lead to the same outcome’69; consequently, they fail to adequately account for the complexity and non-linear characteristics of the process by which herders’ behaviours are formed. Consequently, this study draws on previous research to introduce the Fuzzy Set Qualitative Comparative Analysis (fsQCA) method, thereby addressing the limitations of SEM in the analysis of multi-factor interactions60,61. fsQCA is case-oriented and grounded in set theory and configurational thinking. Through counterfactual analysis and Boolean algebraic operations, it organically integrates the depth of qualitative analysis with the objectivity of quantitative analysis, effectively overcoming the reliance of traditional regression analysis on the net effects of individual variables70. This method focuses on the sufficient and necessary conditions for the occurrence of outcomes, and is capable of revealing differentiated driving pathways under the coupled effects of multiple antecedent variables71. In terms of evaluation metrics, fsQCA employs consistency (Eq. 4) and coverage (Eq. 5) as core indicators, further deriving raw coverage (Eq. 6), unique coverage (Eq. 7) and overall solution coverage (Eq. 8). These metrics enable a systematic, multi-dimensional assessment of the explanatory power of configuration pathways, thereby precisely meeting the analytical requirements of this study regarding multi-factor synergistic driving mechanisms.4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{C}_{n}\left({X}_{i}\le\:{Y}_{i}\right)=\frac{\sum\:_{i=1}^{N}min\left({X}_{i},{Y}_{i}\right)}{\sum\:_{i=1}^{N}{X}_{i}}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{C}_{v}\left({X}_{i}\le\:{Y}_{i}\right)=\frac{\sum\:_{i=1}^{N}min\left({X}_{i},{Y}_{i}\right)}{\sum\:_{i=1}^{N}{Y}_{i}}$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:RC=\frac{{\sum\:}_{i=1}^{N}\left|{S}_{i}\right|}{\left|R\right|}$$\end{document}7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:UC=\frac{{\sum\:}_{i=1}^{N}\left|{S}_{i}\setminus\:{\bigcup\:}_{j\ne\:i}{S}_{j}\right|}{\left|R\right|}$$\end{document}8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:SC=\frac{\left|{\bigcup\:}_{i=1}^{N}{S}_{i}\right|}{\left|R\right|}$$\end{document} Here, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{C}_{n}$$\end{document} is consistency and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{C}_{v}$$\end{document} is coverage. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{i}$$\end{document} is the condition combination and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}_{i}$$\end{document} is the result. i is the number of samples and N is the total number of cases. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}$$\end{document} is the set of results covered by the ith condition combination. R is the set of all results. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left|\:\:\right|$$\end{document} is the sum of affiliation. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\setminus\:$$\end{document} is the set difference operation, e.g., A\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\setminus\:$$\end{document}B denotes an element that belongs to A but not to B. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\bigcup\:$$\end{document} is the set concatenation operation, e.g., \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\bigcup\:}_{i=1}^{N}{S}_{i}$$\end{document} denotes the concatenation set of all grouped condition-covered results. Result Analysis of correlations between variables To verify the validity of the data structure among the variables across the five dimensions of ecological ethics and to avoid issues of multicollinearity, this study conducted bivariate correlation analyses on each core variable; the results are shown in Fig. 4. All study variables exhibited significant positive correlations at the 0.01 significance level, consistent with the research hypotheses and model predictions. Specifically, the pairwise correlation coefficients for the five core ecological ethics constructs ranged between 0.372 and 0.443, all significantly below the multicollinearity threshold of 0.6. This indicates that the variables across dimensions are independent of one another, with no serious issues of multicollinearity, and that the estimated path coefficients in the model were not distorted by multicollinearity bias. Furthermore, the correlation coefficients between the five ecological ethics dimensions and behavioural intention, as well as ecological behaviour, were highly similar, confirming that the driving effects of different ecological ethics dimensions on herders’ willingness and behaviour regarding ecological conservation were generally balanced. This result is a true reflection of the homogeneity of ecological cognition within the study sample and does not constitute a statistical anomaly. Furthermore, behavioural intention and ecological behaviour exhibited a significant and strong correlation (r = 0.678, p < 0.01), effectively validating the rationality of the mediating mechanism in this study and providing a reliable data foundation for subsequent path analysis and mediation effect testing. Fig. 4Heatmap of correlations between core variables. Heatmap of correlations between core variables. Reliability and validity tests In this paper, the latent variables were tested for reliability and validity using SPSS 28.0 and AMOS 28.0 software, respectively. The reliability of scale data is generally measured by Cronbach’s α coefficient and combined reliability coefficient (CR). As can be seen in Table 3, the Cronbach’s α values for all latent variables reached 0.8 or more, indicating strong reliability for the overall scale. In addition, the CR values for all groups of dimensions were greater than the commonly used standard value of 0.7, which indicates that the scales designed in this study have good internal consistency and the data have high reliability. In this study, convergent and discriminant validity of the scale were tested separately using validated factor analysis. The standardised factor loadings of each question item ranged from 0.7 to 1, all of which met the requirement of standardised factor loadings greater than 0.6, indicating that each variable was highly representative. In terms of convergent validity, the Average Variance Extracted (AVE) values of all variables were above 0.5 as required by the standard, indicating good convergent validity72. In addition, the discriminant validity was mainly based on the Fornell-Larcker criterion73, and it can be found through Fig. 5 that the correlation coefficients between the variables are less than the square root of the AVE values of the variables on the corresponding diagonal, which indicates that there is a large difference between the measurement items and the discriminant validity of each potential variable is good. Table 3Reliability and validity test results.VariableSubjectEstimateCronbach’s αAVECRNatural Relationship CognitionNRC1Life comes from water, without water there is no life.0.7940.8560.6660.857NRC2Man is small in the face of nature.0.825NRC3Animals are man’s friends.0.828Grassland Ecological Function CognitionGEFC1Grasslands are an important place to raise livestock.0.8550.8970.6880.898GEFC2Grassland improves the soil and facilitates the growth of pasture grasses.0.784GEFC3The distinctive characteristics of the nomadic culture of the steppe, which is conducive to the development of tourism.0.851GEFC4Grasslands can absorb carbon dioxide.0.826Grassland Ecological Statute AwarenessGESA1The village rules and regulations explicitly call for the protection of the environment.0.8070.8900.6680.890GESA2The village code requires us to use animal and plant resources in moderation.0.818GESA3Our village negotiates pasture boundary disputes with surrounding neighbours.0.812GESA4Our village negotiates with its neighbours to deal with the problem of mixed herds of cattle and sheep.0.833Pro-Environmental AttitudePA1I can bury the hole I left when I picked up the mushrooms.0.7720.8990.6420.900PA2I can bury domestic waste in transit.0.814PA3I can pick up plastic bags, bottles and other waste while I’m grazing.0.827PA4I can set up sheep and cattle pens on rocky outcrops and away from water.0.798PA5The number of cattle and sheep that can be authorised after I have complied with the stock reduction in my grazing.0.793Knowledge Of Grassland Ecology And Animal HusbandryKGEAH1I’m able to accurately determine the best time to graze a pasture.0.8560.9050.7050.905KGEAH2I can judge the health of animals by observing their behaviour and appearance.0.794KGEAH3I am familiar with basic disease prevention and treatment methods.0.860KGEAH4I prefer to use natural forage rather than processed feeds.0.847Intentions For Value Realisation Of Grassland Ecological ProductsI1I am willing to provide ecological products.0.7840.8680.6230.869I2I am willing to participate in different forms of emerging ecotourism operations.0.792I3I am willing to reduce the amount of livestock breeding in order to promote the restoration of grassland ecosystems.0.782I4I’m willing to stream the pasture.0.800Behaviour Of Value Realisation Of Grassland Ecological ProductsB1Livestock sold through multiple channels.0.7910.8730.6360.875B2Various ways of participating in tourism operations.0.819B3Access to multiple types of compensation for ecological resettlement.0.773B4Multiple ways to carry out pasture turnout or leasing.0.807Policy SupportPS1Regular publicity and education campaigns on grassland ecological conservation are conducted by the Government or the community.PS2The Role of Government Ecological Compensation Policies and Funds on Grassland Ecological Conservation. Table 3Reliability and validity test results.VariableSubjectEstimateCronbach’s αAVECRNatural Relationship CognitionNRC1Life comes from water, without water there is no life.0.7940.8560.6660.857NRC2Man is small in the face of nature.0.825NRC3Animals are man’s friends.0.828Grassland Ecological Function CognitionGEFC1Grasslands are an important place to raise livestock.0.8550.8970.6880.898GEFC2Grassland improves the soil and facilitates the growth of pasture grasses.0.784GEFC3The distinctive characteristics of the nomadic culture of the steppe, which is conducive to the development of tourism.0.851GEFC4Grasslands can absorb carbon dioxide.0.826Grassland Ecological Statute AwarenessGESA1The village rules and regulations explicitly call for the protection of the environment.0.8070.8900.6680.890GESA2The village code requires us to use animal and plant resources in moderation.0.818GESA3Our village negotiates pasture boundary disputes with surrounding neighbours.0.812GESA4Our village negotiates with its neighbours to deal with the problem of mixed herds of cattle and sheep.0.833Pro-Environmental AttitudePA1I can bury the hole I left when I picked up the mushrooms.0.7720.8990.6420.900PA2I can bury domestic waste in transit.0.814PA3I can pick up plastic bags, bottles and other waste while I’m grazing.0.827PA4I can set up sheep and cattle pens on rocky outcrops and away from water.0.798PA5The number of cattle and sheep that can be authorised after I have complied with the stock reduction in my grazing.0.793Knowledge Of Grassland Ecology And Animal HusbandryKGEAH1I’m able to accurately determine the best time to graze a pasture.0.8560.9050.7050.905KGEAH2I can judge the health of animals by observing their behaviour and appearance.0.794KGEAH3I am familiar with basic disease prevention and treatment methods.0.860KGEAH4I prefer to use natural forage rather than processed feeds.0.847Intentions For Value Realisation Of Grassland Ecological ProductsI1I am willing to provide ecological products.0.7840.8680.6230.869I2I am willing to participate in different forms of emerging ecotourism operations.0.792I3I am willing to reduce the amount of livestock breeding in order to promote the restoration of grassland ecosystems.0.782I4I’m willing to stream the pasture.0.800Behaviour Of Value Realisation Of Grassland Ecological ProductsB1Livestock sold through multiple channels.0.7910.8730.6360.875B2Various ways of participating in tourism operations.0.819B3Access to multiple types of compensation for ecological resettlement.0.773B4Multiple ways to carry out pasture turnout or leasing.0.807Policy SupportPS1Regular publicity and education campaigns on grassland ecological conservation are conducted by the Government or the community.PS2The Role of Government Ecological Compensation Policies and Funds on Grassland Ecological Conservation. Reliability and validity test results. Fig. 5Distinctive validity of variables. Note: The diagonal of the matrix is the square root of AVE, and below the diagonal is the matrix of correlation coefficients. Distinctive validity of variables. Note: The diagonal of the matrix is the square root of AVE, and below the diagonal is the matrix of correlation coefficients. Test for normality Prior to estimating the SEM model, this study conducted normality tests on all observed variables to determine whether the data met the distributional assumptions of the maximum likelihood method. The results in Table 4 show that the absolute values of the skewness coefficients for each observed variable ranged from − 0.590 to -1.166, whilst the absolute values of the kurtosis coefficients ranged from − 0.757 to 1.197; both fall within the acceptable range generally recognised in social science research. Further results from the multivariate normality test indicate that the Mardia multivariate kurtosis coefficient is 66.707, suggesting that the sample data deviate significantly from a multivariate normal distribution. Table 4Results of the multivariate normal distribution.Variableminmaxskewc.r.kurtosisc.r.PA52.0005.000−0.692−5.129−0.444−1.648PA11.0005.000−0.667−4.9450.0750.279PA22.0005.000−0.665−4.9310.1860.688PA32.0005.000−0.605−4.490−0.430−1.596PA41.0005.000−0.606−4.497−0.252−0.935B42.0005.000−0.903−6.6950.2020.748B31.0005.000−1.166−8.6511.1524.273B21.0005.000−1.057−7.8410.7382.738B12.0005.000−1.150−8.5260.7502.780I42.0005.000−0.789−5.8490.4411.635I32.0005.000−0.704−5.2230.4121.526I22.0005.000−0.604−4.4810.1630.605I12.0005.000−0.735−5.4540.2841.053KGEAH11.0005.000−1.088−8.0660.5171.917KGEAH22.0005.000−0.880−6.5280.0950.353KGEAH31.0005.000−0.991−7.3480.4881.811KGEAH42.0005.0000−0.998−7.4030.2400.892GESA11.0005.000−1.015−7.5301.1104.117GESA21.0005.000−0.812−6.0250.3751.390GESA31.0005.000−1.103−8.1831.1974.439GESA41.0005.000−1.024−7.5950.9193.409GEFC11.0005.000−0.742−5.5010.0360.133GEFC22.0005.000−0.651−4.829−0.378−1.401GEFC32.0005.000−0.590−4.373−0.757−2.807GEFC42.0005.000−0.688−5.101−0.436−1.617NRC11.0005.000−0.748−5.5500.2580.956NRC21.0005.000−0.866−6.4260.4251.575NRC31.0005.000−1.013−7.5100.5321.973Multivariate66.707 Table 4Results of the multivariate normal distribution.Variableminmaxskewc.r.kurtosisc.r.PA52.0005.000−0.692−5.129−0.444−1.648PA11.0005.000−0.667−4.9450.0750.279PA22.0005.000−0.665−4.9310.1860.688PA32.0005.000−0.605−4.490−0.430−1.596PA41.0005.000−0.606−4.497−0.252−0.935B42.0005.000−0.903−6.6950.2020.748B31.0005.000−1.166−8.6511.1524.273B21.0005.000−1.057−7.8410.7382.738B12.0005.000−1.150−8.5260.7502.780I42.0005.000−0.789−5.8490.4411.635I32.0005.000−0.704−5.2230.4121.526I22.0005.000−0.604−4.4810.1630.605I12.0005.000−0.735−5.4540.2841.053KGEAH11.0005.000−1.088−8.0660.5171.917KGEAH22.0005.000−0.880−6.5280.0950.353KGEAH31.0005.000−0.991−7.3480.4881.811KGEAH42.0005.0000−0.998−7.4030.2400.892GESA11.0005.000−1.015−7.5301.1104.117GESA21.0005.000−0.812−6.0250.3751.390GESA31.0005.000−1.103−8.1831.1974.439GESA41.0005.000−1.024−7.5950.9193.409GEFC11.0005.000−0.742−5.5010.0360.133GEFC22.0005.000−0.651−4.829−0.378−1.401GEFC32.0005.000−0.590−4.373−0.757−2.807GEFC42.0005.000−0.688−5.101−0.436−1.617NRC11.0005.000−0.748−5.5500.2580.956NRC21.0005.000−0.866−6.4260.4251.575NRC31.0005.000−1.013−7.5100.5321.973Multivariate66.707 Results of the multivariate normal distribution. Model fit test In this study, AMOS 28.0 software was used to first test the fitness of the model. According to the results in Table 5, it can be seen that all indicators meet the criteria, and it can be considered that the model constructed in this study has a good fit. Table 5Model fitness test results.IndexAbsolute Fitness IndicatorsValue-Added Fitness IndicatorsΧ2/dfRMSEAGFIAGFIIFITLICFIFitting Value1.4560.0370.9070.8850.9750.9720.975Reference Value1-3 is excellent< 0.05 is excellent> 0.9 is excellent (> 0.8 is acceptable) Table 5Model fitness test results.IndexAbsolute Fitness IndicatorsValue-Added Fitness IndicatorsΧ2/dfRMSEAGFIAGFIIFITLICFIFitting Value1.4560.0370.9070.8850.9750.9720.975Reference Value1-3 is excellent< 0.05 is excellent> 0.9 is excellent (> 0.8 is acceptable) Model fitness test results. SEM test results In accordance with the R2 threshold criteria proposed by Hair et al. (2021)74, whereby R2 values of 0.25, 0.50 and 0.75 correspond to weak, moderate and strong explanatory power respectively, the R2 values for the endogenous latent variables in this study’s model range from 0.596 to 0.764, all exceeding the threshold for moderate explanatory power. In particular, the R2 value for actual value realisation behaviour reached 0.764, indicating strong explanatory power. Overall, the model variables demonstrate a high degree of explanatory power, and the model performs well. Building on this, a path analysis was conducted on the latent and observed variables in the model; the results of the tests for each research hypothesis are detailed in Table 6; Fig. 6. The results in Table 6 show that natural relationship cognition (β = 0.204, p < 0.001), grassland ecological function cognition (β = 0.205, p < 0.001), grassland ecological statute awareness (0.216, p < 0.001), pro-environmental attitudes (β = 0.241, p < 0.001), and grassland ecological livestock knowledge (β = 0.189, p < 0.001) all positively and significantly affected the intention for value realisation of grassland ecological products, and hypotheses H1, H3, H5, H7 and H9 were validated. It can be seen that the contributions to the intention for value realisation of grassland ecological products were ranked in descending order as pro-environmental attitude, grassland ecological statute awareness, grassland ecological function cognition, natural relationship cognition, and grassland ecological animal husbandry knowledge. In addition, natural relationship cognition (β = 0.184, p < 0.001), grassland ecological function cognition (β = 0.174, p < 0.001), grassland ecological statute awareness (β = 0.187, p < 0.05), pro-environmental attitude (β = 0.172, p < 0.001), and grassland ecological livestock knowledge (β = 0.167, p < 0.001), intention for value realisation of grassland ecological products (β = 0.248, p < 0.05) all positively contributed to the behaviour of value realisation of grassland ecological products, and hypotheses H2, H4, H6, H8, H9, H10, and H11 were fully supported. Among them, intention for value realisation of grassland ecological products is the key factor to enhance the behaviour of value realisation of grassland ecological products. In summary, the ecological ethical view positively influences the herders’ intention for value realisation of grassland ecological products and their behaviour. Table 6Results of hypothesis testing.HypothesesPathPath CoefficientS.E.C.R.P-valueHypothesis Testing ResultsH1NRC→I0.2040.0533.367***set upH3GEFC→I0.2050.053.607***set upH5GESA→I0.2160.0513.682***set upH7PA→I0.2410.0524.205***set upH9KGEAH→I0.1890.0473.515***set upH2NRC→B0.1840.0473.361***set upH4GEFC→B0.1740.0453.405***set upH6GESA→B0.1870.0463.524***set upH8PA→B0.1720.0473.316***set upH10KGEAH→B0.1670.0423.445***set upH11I→B0.2480.0713.44***set upNote: *p < 0.05,**p < 0.01,***p < 0.001, same below. Table 6Results of hypothesis testing.HypothesesPathPath CoefficientS.E.C.R.P-valueHypothesis Testing ResultsH1NRC→I0.2040.0533.367***set upH3GEFC→I0.2050.053.607***set upH5GESA→I0.2160.0513.682***set upH7PA→I0.2410.0524.205***set upH9KGEAH→I0.1890.0473.515***set upH2NRC→B0.1840.0473.361***set upH4GEFC→B0.1740.0453.405***set upH6GESA→B0.1870.0463.524***set upH8PA→B0.1720.0473.316***set upH10KGEAH→B0.1670.0423.445***set upH11I→B0.2480.0713.44***set upNote: *p < 0.05,**p < 0.01,***p < 0.001, same below. Results of hypothesis testing. Note: *p < 0.05,**p < 0.01,***p < 0.001, same below. Fig. 6Path coefficient modelling diagram. Path coefficient modelling diagram. Mediation effect test This study utilised AMOS 28.0 software to test the mediation model using the Bootstrap method. With natural relationship cognition, grassland ecological function cognition, grassland ecological statute awareness, pro-environmental attitudes and knowledge of grassland ecological animal husbandry as independent variables, and behaviour of value realisation of grassland ecological products as the dependent variable, with the intention to realise the value of grassland ecological products as the mediating variable, the analysis was conducted at a 95% confidence level with 2,000 bootstrap repetitions. The test results are shown in Table 7. The 95% confidence intervals for all paths do not include 0, indicating that the mediating effects are statistically significant and are all partial mediations. Specifically, the standardised direct effects of the five pathways were 0.159, 0.152, 0.160, 0.155 and 0.145 respectively, whilst the standardised indirect effects were 0.044, 0.045, 0.046, 0.054 and 0.041 respectively; the proportion of the total effect accounted for by the mediating effects was 21.569%, 22.843%, 22.330%, 27.136% and 21.693%, respectively. To further quantify the actual explanatory power of the mediating effects, this study employed the f2 effect size to assess the contribution of the mediating pathways. The results show that the f2 values for the mediating effects of each path range from 0.131 to 0.144, falling within the small effect size range (0.02 ≤ f2 < 0.15). This indicates that although the mediating variables’ additional explanatory power for the dependent variable is statistically significant, their actual mediating strength is relatively weak. Taking into account the combined results regarding direct and indirect effects and effect sizes, it can be seen that the intention for value realisation of grassland ecological products played a statistically significant partial mediating role across all pathways; however, the influence of the independent variables on the dependent variable remained predominantly via direct pathways, with the mediating variables performing only a limited auxiliary role in the transmission mechanism. Accordingly, research hypothesis H12 has been partially confirmed. Table 7Mediation effect test results.Path RelationshipEffect RelationshipEstimateLowerUpper P Proportion of the effectf2NRC→I→BTotal effect0.2040.1060.3440.001Direct effect0.1590.0520.3010.00277.941%Indirect effect0.0440.0090.1120.00521.569%0.140GEFC→I→BTotal effect0.1970.1150.2910.001Direct effect0.1520.0610.2430.00377.157%Indirect effect0.0450.0160.1040.00222.843%0.136GESA→I→BTotal effect0.2060.1120.3220.001Direct effect0.1600.0690.2580.00177.670%Indirect effect0.0460.0150.1070.00122.330%0.144PA→I→BTotal effect0.1990.0990.3060.001Direct effect0.1550.0470.2900.00477.889%Indirect effect0.0540.0170.1240.00127.136%0.131KGEAH→I→BTotal effect0.1890.1090.2820.001Direct effect0.1450.0680.2470.00076.720%Indirect effect0.0410.0120.0960.00221.693%0.140 Table 7Mediation effect test results.Path RelationshipEffect RelationshipEstimateLowerUpper P Proportion of the effectf2NRC→I→BTotal effect0.2040.1060.3440.001Direct effect0.1590.0520.3010.00277.941%Indirect effect0.0440.0090.1120.00521.569%0.140GEFC→I→BTotal effect0.1970.1150.2910.001Direct effect0.1520.0610.2430.00377.157%Indirect effect0.0450.0160.1040.00222.843%0.136GESA→I→BTotal effect0.2060.1120.3220.001Direct effect0.1600.0690.2580.00177.670%Indirect effect0.0460.0150.1070.00122.330%0.144PA→I→BTotal effect0.1990.0990.3060.001Direct effect0.1550.0470.2900.00477.889%Indirect effect0.0540.0170.1240.00127.136%0.131KGEAH→I→BTotal effect0.1890.1090.2820.001Direct effect0.1450.0680.2470.00076.720%Indirect effect0.0410.0120.0960.00221.693%0.140 Mediation effect test results. Moderating effects test This paper takes intention for value realisation of grassland ecological products as the independent variable, behaviour of value realisation of grassland ecological products as the dependent variable. and consider the moderating effect of policy support from 2 dimensions: policy propaganda and policy effect. The results in Table 8 show that the coefficient of the interaction term of policy support is significant and has the same sign as the main effect, indicating that there is a positive moderating effect of policy support in the influence of herders’ intention for value realisation of grassland ecological products to behaviour, and Hypothesis H13 is verified. It may be due to the fact that, when the policy publicity of grassland ecological protection is more extensive and the policy effect of ecological compensation is better, the more herders’ attention to environmental protection issues can be aroused, thus accelerating the transformation of their intention for value realisation of grassland ecological products to behaviour. Table 8Results of the moderating effects analysis.VariableBBBetatP-valueBetatP-valueI0.70219.171***0.52913.313***PS0.3248.858***0.35710.587***Interaction term0.3248.095***R20.5640.637F211.388***190.576*** Table 8Results of the moderating effects analysis.VariableBBBetatP-valueBetatP-valueI0.70219.171***0.52913.313***PS0.3248.858***0.35710.587***Interaction term0.3248.095***R20.5640.637F211.388***190.576*** Results of the moderating effects analysis. To visually illustrate the mechanism of action of the moderator variable PS, a simple slope moderation plot was constructed, with the results shown in Fig. 7. The x-axis represents the levels of the independent variable I, whilst the y-axis represents the dependent variable B; the blue dotted line represents the group with low levels of the moderator variable PS, and the red solid line represents the group with high levels of PS. As can be seen from the figure, when the moderator PS is at a low level, the positive influence of the independent variable I on the dependent variable B is relatively weak, and the trend line is almost flat; when PS is at a high level, the positive driving effect of the independent variable I on the dependent variable B is significantly enhanced, and the slope of the trend line increases substantially. The fact that the two trend lines intersect indicates that the moderating effect of PS varies markedly depending on the context: when the level of independent variable I is low, the promotional effect of high-level PS on dependent variable B is weaker than that of low-level PS; as the level of independent variable I increases, the positive effect of I on B under high-level PS is significantly amplified, ultimately far exceeding that of the low-level PS group. Combined with the regression results showing a significant positive interaction term, it can be seen that PS exerts a significant positive moderating effect between independent variable I and dependent variable B; that is, the higher the level of PS, the stronger the positive facilitating effect of independent variable I on dependent variable B. Fig. 7Modulation effect diagram. Modulation effect diagram. Configuration analysis results Data calibration Since fsQCA is derived from the concept of set affiliation, each variable needs to be assigned an affiliation on the set (0, 1) first to calibrate the data and make the antecedent and outcome variables set meaningful. Since fsQCA only applies to a single variable and the single variable in this paper contains multiple items, the arithmetic mean of each variable item should be calculated before calibration. Referring to Ragin’s findings, the anchor points for full affiliation, intersection, and full unaffiliation points were set at 95%, 50%, and 5%, respectively, and the samples were calibrated accordingly75. Table 9 presents the calibration of the variables. Table 9Description of fuzzy set anchor thresholds.VariableFull Affiliation Point(0.95)Calibration Intersection(0.50)Full Unaffiliated Point(0.05)NRC5.0004.3332.667GEFC5.0004.2502.500GESA5.0004.2502.250PA4.8004.2002.400KGEAH5.0004.5002.638PS5.0004.5002.550I5.0004.2502.250B5.0004.7502.750 Table 9Description of fuzzy set anchor thresholds.VariableFull Affiliation Point(0.95)Calibration Intersection(0.50)Full Unaffiliated Point(0.05)NRC5.0004.3332.667GEFC5.0004.2502.500GESA5.0004.2502.250PA4.8004.2002.400KGEAH5.0004.5002.638PS5.0004.5002.550I5.0004.2502.250B5.0004.7502.750 Description of fuzzy set anchor thresholds. Univariate necessity analysis The analysis of the necessity of the condition is necessary to verify whether a single antecedent variable is a behaviour of value realisation of grassland ecological products and should precede the dissection of the conditional grouping of the variables76. A conditional variable is necessary if the consistency value exceeds the threshold of 0.9. Table 10 was calculated to obtain that the consistency value of each antecedent variable was below 0.9, indicating that the individual variables were weak in explaining the outcome and were not necessary for the outcome variable77. Therefore, it is necessary to further combine multiple antecedent variables to analyse their formation and action mechanisms. Table 10Necessity analysis results.Conditional VariableConsistencyCoveragenrc0.8560.687~nrc0.5990.560gefc0.8410.668~gefc0.5700.540gesa0.8660.713~gesa0.6480.589p.a.0.8410.665~pa0.5890.561kgeah0.8240.678~kgeah0.6360.579ps0.8020.625~ps0.5580.541i0.8820.767~i0.6930.595 Table 10Necessity analysis results.Conditional VariableConsistencyCoveragenrc0.8560.687~nrc0.5990.560gefc0.8410.668~gefc0.5700.540gesa0.8660.713~gesa0.6480.589p.a.0.8410.665~pa0.5890.561kgeah0.8240.678~kgeah0.6360.579ps0.8020.625~ps0.5580.541i0.8820.767~i0.6930.595 Necessity analysis results. fsQCA results The data were analysed empirically with the help of fsQCA 3.0 software. After conducting the necessity test, this study set the consistency threshold at 0.8, the threshold for the frequency of cases at 1, and the PRI consistency threshold at 0.7. Based on the above settings, the group conditions were obtained as shown in Table 11. The core condition is a condition variable that has a strong causal relationship with the outcome variable, while the edge condition is the opposite. According to the group state representation set by Ragin and Fiss (2008)75, ● and ⊗ are used to represent the existence and non-existence of the core condition, respectively, · and × are used to represent the existence and non-existence of the auxiliary condition, respectively, and blank represents that this condition may or may not exist and is an irrelevant condition. Solution consistency in the fsQCA table indicates the degree to which the set of conditional groupings is a subset of the set of results, and its value is greater than or equal to 0.8; in QCA research, the degree of coverage is a key indicator for assessing the relevance of the empirical evidence, in which the overall coverage is generally required to be above 0.5. According to Table 11, there are 6 combinatorial paths formed, and the solution consistency of each combinatorial path reaches more than 0.964 (> 0.8), and the solution coverage reaches 0.655 (> 0.5). This indicates that the overall interpretation of the 6 groupings is strong. Table 11Configuration analysis results.Conditional VariableBehaviour Of Value Realisation Of Grassland Ecological Products123456NRC●●●●×GEFC·●×●GESA×····PA·×··×·KGEAH×·×·●·PS●●●●●●I●●●●●Raw Coverage0.3320.3520.4100.3540.3590.361Unique Coverage0.0210.0120.0300.0270.0110.048Consistency0.9830.9760.9870.9710.9720.984Solution Consistency0.964Solution Coverage0.655 Table 11Configuration analysis results.Conditional VariableBehaviour Of Value Realisation Of Grassland Ecological Products123456NRC●●●●×GEFC·●×●GESA×····PA·×··×·KGEAH×·×·●·PS●●●●●●I●●●●●Raw Coverage0.3320.3520.4100.3540.3590.361Unique Coverage0.0210.0120.0300.0270.0110.048Consistency0.9830.9760.9870.9710.9720.984Solution Consistency0.964Solution Coverage0.655 Configuration analysis results. The key criteria for configuration paths can be summarised into six models, as analysed in detail below: Model 1: Policy-incentivised and motivation-driven. The core prerequisites are policy support and a willingness to realise personal value, whilst grassland ecological statute awareness and knowledge of ecological animal husbandry are marginally absent; auxiliary prerequisites include grassland ecological function cognition and pro-environmental attitudes, whilst natural relationship cognition is an irrelevant factor in this model. This configuration reveals a concise and robust pathway: provided that policies are effectively implemented and individuals’ intention for value realisation of grassland ecological products is fully activated, herders can still effectively engage in behaviour of value realisation of grassland ecological products, even if they lack grassland ecological statute awareness and have insufficient professional knowledge of ecological pastoralism. The simultaneous absence of grassland ecological statute awareness and knowledge of ecological pastoralism is a typical structural feature of this pathway, indicating that, under the combined influence of external policy incentives and internal subjective willingness, insufficient normative awareness and gaps in professional knowledge do not constitute a substantial obstacle to the realisation of ecological value. This model is termed the ‘Policy Incentive and Will-Driven Model’, aiming to highlight the central and dominant role of external institutional incentives and individual internal willingness within this pathway. Model 2: Rational cognition replaces emotional cognition. The core conditions are the existence of natural relationship cognition, grassland ecological function cognition, and policy support; pro-environmental attitudes appear in a non-supportive form. The auxiliary conditions are the existence of grassland ecological statute awareness and knowledge of ecological animal husbandry; the willingness to realise values is an irrelevant condition for this model. This model exhibits a distinct counterintuitive feature among the six pathways: the willingness to realise value and pro-environmental attitudes—two subjective variables typically regarded as core elements in the generation of behaviour—do not play a role at the core level in this pathway. However, the natural relationship cognition and grassland ecological function cognition together form a dual foundation at the level of rational cognition; under the institutional safeguards provided by policy support, they can still independently drive the generation of behaviour of value realisation of grassland ecological products. This model is named the ‘rational cognition substituting for the emotional type’, aiming to reveal the structural substitution mechanism whereby the completeness of the cognitive dimension replaces emotional attitudes and subjective intentions in specific contexts. Model 3: Cognitive-Institutional-Intentional Synergy Type. The core existence conditions are natural relationship cognition, policy support and the intention to realise value, whilst knowledge of grassland ecological animal husbandry serves as a supplementary missing condition; grassland ecological statute awareness and pro-environmental attitudes are supplementary existence conditions, and grassland ecological function cognition does not play a role in this configuration. This configuration has an original coverage rate of 0.410, the highest value among the six pathways, indicating the broadest scope of real-world applicability and representing the most universal combination of sufficient conditions for facilitating behaviour of value realisation of grassland ecological products. This pathway is constructed upon a core framework formed by elements from the three major dimensions of cognition, institutions and intention, establishing a synergistic operational logic underpinned by cognitive foundations, institutional safeguards and intention-driven action. Knowledge of grassland ecological animal husbandry is in a state of auxiliary absence, indicating that when the three core categories of comprehensive cognition, institutional incentives and subjective willingness are simultaneously present, professional knowledge does not play a significant direct driving role. This model is named the ‘Cognition-Institution-Intentional Synergy Type’, serving to summarise the mechanism of action whereby multiple core conditions interact and complement one another. Model 4: Macro-cognitive compensation for specialised cognitive type. The core conditions for existence include natural relationship cognition, policy support and a willingness to realise value, whilst grassland ecological function cognition is a missing auxiliary condition; grassland ecological statute awareness and pro-environmental attitudes are auxiliary conditions for existence, and knowledge of grassland ecological pastoralism is an irrelevant condition in this configuration. This configuration shares the three core presence conditions of natural relationship cognition, policy support, and willingness to realise value with Pattern 3; the primary distinction between the two lies in the fact that grassland ecological function cognition is in a state of auxiliary absence. This characteristic holds significant theoretical significance: herders do not need to possess specialised knowledge of the ecosystem services provided by grasslands. When individuals form a basic understanding of the relationship between humans and nature, receive policy support, and develop a willingness to realise values, a lack of specialised knowledge regarding grassland ecological function cognition does not constrain the emergence of relevant behaviours. It is thus evident that macro-level natural relationship cognition can functionally compensate for the lack of specialised grassland ecological function cognition. This model is termed the ‘macro-cognition compensating for specialised cognition’ type, reflecting the structural substitution effect of cognition regarding the natural relationship on the formation of grassland ecological function cognition. Model 5: Function-Cognition-Dominated Substitute Attitude Type. The core conditions for this configuration are natural relationship cognition, grassland ecological function cognition, policy support and willingness to realise value; pro-environmental attitudes are a secondary missing condition; knowledge of grassland ecological pastoralism is a secondary present condition; and grassland ecological statute awareness is an irrelevant condition. This configuration shares the same characteristic as Model 2, namely that pro-environmental attitudes are in a state of auxiliary absence. This indicates that a deficiency in emotional-level attitudes does not hinder the occurrence of behaviour of value realisation of grassland ecological products. The distinction between the two lies in the fact that Model 5 possesses a more comprehensive combination of core conditions, forming a complete driving system jointly constituted by a dual cognitive foundation, institutional safeguards and intrinsic willingness. Within this, the natural relationship cognition and grassland ecological function cognition form a dual rational drive from the dimensions of macro-cognition and functional cognition, respectively. Combined with policy incentives and subjective willingness, this constructs a driving framework where the four elements work in concert, effectively replacing the emotional driving role of pro-environmental attitudes. Grassland ecological livestock knowledge serves as a supporting condition providing technical backing, whilst grassland ecological statute awareness exerts no significant influence. This implies that when the cognitive system, institutional environment and subjective willingness are all in an ideal state, normative cognition does not constrain the final behaviour. This model is termed the ‘Grassland Ecological Function Cognition-Dominated Attitude Substitution Model’, aiming to reflect the structural substitution role of dual cognitive elements for pro-environmental attitudes, whilst simultaneously highlighting the synergistic supportive value of policy and willingness. Model 6: Institutional Will Compensating for Cognitive Gaps. The core prerequisite for this configuration is policy support and the willingness to realise value, whilst natural relationship cognition is in a state of auxiliary non-existence; auxiliary prerequisites include grassland ecological statute awareness, pro-environmental attitudes and knowledge of ecological animal husbandry, whilst grassland ecological function cognition is an irrelevant condition for this mode. This configuration has a unique coverage rate of 0.048, ranking first among the six driving pathways, indicating that this pathway possesses a unique explanatory scope that cannot be fully covered or substituted by other configurations. The core logic of this model is that even if herders have shortcomings in their natural relationship cognition, under an institutional safeguard system formed by policy support, with the intention for value realisation of grassland ecological products as the core intrinsic driving force, and relying on the normative constraints of grassland ecological statute awareness, the emotional support of pro-environmental attitudes, and the technical empowerment of ecological livestock knowledge as multi-dimensional auxiliary conditions, it is still possible to effectively facilitate the emergence of behaviour of value realisation of grassland ecological products. Through the synergistic coordination of these diverse supporting factors, this pathway compensates for structural deficiencies in the macro-cognitive dimension, fully demonstrating a compensatory mechanism where institutional frameworks and subjective willingness act as dual drivers to address cognitive shortcomings. This model is termed the ‘Institutional-Willingness Model for Compensating Cognitive Deficiencies’, a designation that encapsulates its operational logic: utilising external institutional incentives and internal subjective willingness as core drivers, whilst relying on the coordinated action of multi-dimensional supporting factors to compensate for cognitive deficiencies. Conclusion Against the backdrop of the ‘Two Mountains’ theory, this paper constructs an analytical framework for the behaviour of herders in terms of the value realisation of grassland ecological products, grounded in the theory of ecological ethics. It not only inherits the forward-looking mindset of Marxism but also highlights the importance of considering ecological construction and environmental protection from the standpoint of dialectical materialism; furthermore, it lays the groundwork for enriching and refining the theory of the ‘community of life’ and the concept of green development. It is precisely because this paper has incorporated the logical and methodological strengths of Marx’s ecological ethics that it possesses rich philosophical and methodological implications. Furthermore, by employing a mixed-methods approach combining SEM and fsQCA to analyse the data, the paper clarifies the mechanisms through which various variables influence the behaviour of value realisation for grassland ecological products, as well as the pathways of their multifaceted interactions, leading to the following conclusions: Firstly, factors related to ecological ethics exert a significant positive influence on herders’ intention for value realisation of grassland ecological products and behaviour. Natural relationship cognition, grassland ecological function cognition, grassland ecological statute awareness, pro-environmental attitudes, and knowledge of grassland ecological animal husbandry are all key predictor variables influencing herders’ intention for value realisation of grassland ecological products and behaviour. These factors can effectively encourage herders to engage in practices related to the value realisation of grassland ecological products, and constitute an important foundation for the formation of herders’ ecological behaviour. Secondly, the willingness to realise values plays a mediating role, whilst policy support exerts a positive moderating effect. The willingness to realise value acts as a partial mediator between ecological ethics and herders’ behaviour of value realisation of grassland ecological products; ecological ethics can influence herders’ practical behaviour directly, whilst also indirectly facilitating the implementation of such behaviour by activating individuals’ subjective intentions to realise value. At the same time, policy support can positively moderate the relationship between the intention to realise value and actual behaviour, effectively enhancing the efficiency of transforming herders’ subjective intentions into practical ecological actions, and improving the effectiveness of converting the value of grassland ecological products. Thirdly, the herders’ behaviour of value realisation for grassland ecological products follows multiple, differentiated nested pathways, and the underlying causal mechanisms exhibit distinct structural characteristics. Through fsQCA configuration analysis, this study identified six distinct nested pathways that can effectively lead to the behaviour of value realisation of grassland ecological products. It demonstrated that there is no single optimal driving model for herders’ behaviour of value realisation of grassland ecological products; different combinations of multiple conditions can all achieve the same behavioural outcome, fully reflecting the core characteristics of causal complexity and model equivalence within the research context. A systematic comparison of the six configuration pathways further reveals the hierarchical patterns of multi-factor drivers: at the core condition level, policy support is consistently present across all models, forming the institutional foundation for behaviour generation, whilst the willingness to realise value plays a central driving role in most pathways; together, these constitute the core driving force behind behavioural generation; At the level of factor substitution, cognitive factors and emotional attitudes exhibit significant functional complementarity and equivalent substitution effects. When the emotional factor of pro-environmental attitudes is absent, natural relationship cognition and grassland ecological function cognition can compensate for the motivational gap either individually or in combination, confirming the interchangeable logic of rational cognition and emotional attitudes within behavioural drive mechanisms; At the level of conditional elasticity, the intensity of the roles played by ecological compliance awareness and ecological livestock knowledge exhibits marked situational heterogeneity, fluctuating dynamically according to the configuration of internal factors within a given scenario. When the combination of core driving factors is complete, these two conditions may be weakened or participate in the driving process only in a supplementary capacity, highlighting the non-linear interactive characteristics among the various antecedent conditions. In summary, the herders’ behaviour of value realisation of grassland ecological products is not a linear outcome determined by a single factor, but rather a complex configuration resulting from the differentiated combination and synergistic interaction of multi-dimensional factors such as institutional incentives, subjective intentions, cognitive structures, normative awareness and knowledge reserves. The findings of this study clarify the linear influences, boundary effects and synergistic driving mechanisms of these multiple factors, thereby providing a solid theoretical foundation and practical guidance for the differentiated and targeted advancement of efforts to realise the value of grassland ecological products. Discussion Firstly, by contextualising and integrating local pastoral ecological ethics with the TPB, this study has constructed a framework for analysing herders’ behaviour of value realisation for grassland ecological products, tailored to the pastoral context of frontier regions, thereby making a significant theoretical contribution. This study breaks away from the homogenised paradigm of existing research that directly replicates the Western TPB model. It clearly defines the boundaries of TPB’s applicability within the cross-cultural context of pastoral areas, acknowledging the universal theoretical value of its hierarchical analysis of behaviour. At the same time, it points out that indigenous theories, due to a lack of deep foundational dimensions such as local ecological culture, nomadic ethics and local conventions, struggles to explain the coupled behaviour of nomadic groups regarding ecological conservation and livelihood transformation. By retaining the core behavioural logic of TPB and embedding it within a localised ecological ethics system, the study has successfully carried out a contextualised theoretical adjustment, effectively resolving the inherent conflict between local research arguments and Western theoretical frameworks, and providing a standardised cross-cultural adaptation paradigm for research on ecological behaviour in pastoral areas. Furthermore, this study utilised the hierarchical logic of TPB to reconstruct a five-dimensional ecological ethics construct specific to pastoral areas—comprising ‘cognition, values, norms, emotions and practice’—thereby overcoming the limitations of current research in agro-pastoral regions, which generally relies on generic environmental scales and suffers from insufficient conceptual localisation. This approach achieved the operationalisation and contextual deconstruction of nomadic ecological ethics, effectively enhancing the explanatory validity of behavioural theory regarding the eco-economic behaviour of nomadic groups in frontier grasslands. Furthermore, this study extends the application of TPB theory from traditional, single-dimensional public-interest ecological behaviour to the composite behavioural context of ‘ecological conservation plus livelihood enhancement’—the behaviour of value realisation for grassland ecological products. It reveals the hierarchical transmission mechanism through which indigenous ethical values drive herders’ psychological transformation and ecological-economic practices, thereby enriching the theoretical framework of human-land behavioural coupling in arid pastoral regions. Overall, this study has achieved a deep integration and complementary refinement of indigenous nomadic ethical logic and classical Western behavioural theories, providing a new localised theoretical paradigm for research into the micro-mechanisms of ecological governance and the value transformation of ecological products in frontier pastoral areas inhabited by ethnic minorities. Secondly, regarding the impact of ecological ethics on the realisation of the value of grassland ecological products, this paper broadens the perspective based on existing research. Much of the existing literature draws on household economics theory, emphasising the influence of livelihood capital, resource endowments and market conditions on the realisation of the value of ecological products, and posits that farmers or herders make rational choices based on cost-benefit analyses78. In contrast, this paper introduces ecological ethics and decomposes it into five dimensions: natural relationship cognition, grassland ecological function cognition, grassland ecological statute awareness, pro-environmental attitudes, and knowledge of grassland ecological pastoralism. It seeks to reveal the normative constraints and ethical drivers within the value realisation process. However, it should be noted that international literature has not reached a fully consistent conclusion on this matter. Some studies suggest that ethical beliefs and environmental attitudes can significantly enhance individuals’ ecological conservation behaviour and participation in green production79; however, other studies have found that environmental values do not necessarily translate consistently into actual behaviour. Particularly in contexts characterised by significant income pressures, pronounced market volatility and a lack of alternative livelihoods, individuals may still prioritise short-term economic gains, thereby giving rise to an ‘attitude–behaviour disconnect’80. In other words, whilst ecological ethics may constitute an important explanatory variable, the strength of its influence is often jointly affected by institutional safeguards, resource constraints and livelihood risks. Looking further, as the primary agents deriving benefits from the natural environment, herders—in the process of converting ecological value into economic value—are prone to inducing problems such as overexploitation of resources, ecosystem degradation and ‘ecological deficits’ if they pursue the maximisation of economic benefits to an excessive degree81. The overconsumption of resources and ecological destruction not only undermine the sustainable supply of ecological products, but also divert humanity from its original aspiration of achieving a better life. Consequently, herders must bear the risks and costs that their economic activities impose on the ecological environment, ensuring that the benefits they derive from the natural environment are commensurate with the ecological costs they should bear82. Ecological ethics advocate respect, protection and benevolence towards nature, confining human activities to the bounds permitted by ecological ethics and the laws of nature; this is of great significance for maintaining the balance between humans and the natural system65. However, this paper must also guard against a ‘one-dimensional idealisation’ of ecological ethics. This is because in certain regions, even where herders possess a strong sense of ecological ethics, their behaviour may still be constrained by practical limitations if mechanisms for trading ecological products are inadequate, infrastructure is weak, or policy implementation falls short83. In other words, ecological ethics are not a sufficient condition for the realisation of values, but rather an intrinsic driving force that is likely to be activated only within specific institutional and market contexts. Consequently, this paper argues that herders should align themselves with the region’s unique cultural customs and cultivate sound ecological ethics; however, they must also foster synergy with market mechanisms, institutional safeguards and local governance capabilities in order to truly resolve the tension between ethnic tourism development and natural environment governance. Thirdly, the SEM results indicate that ecological ethics play a significant role in promoting herders’ intention for value realisation of grassland ecological products and behaviour related to this activity. Among these factors, pro-environmental attitudes have the greatest influence on the intention for value realisation, followed by grassland ecological statute awareness; this is consistent with the fundamental premise of the TRA that ‘intentions are primarily determined by attitudes and subjective norms’84. However, it should be noted that there are differing views in international research: some studies have found that, compared to attitudinal factors, subjective norms, policy signals and even perceived benefits play a more dominant role in explaining group behaviour in resource-dependent regions85; other studies have pointed out that, in livelihood environments highly dependent on natural resources, actors do not always act in accordance with stable attitudes, but may be more influenced by household income needs, community peer pressure and market expectations86. It is thus evident that the finding in this paper—that pro-environmental attitudes exert the strongest influence—may be related to the cultural context of the case study sites, the characteristics of the sample, and the methods used to measure the variables. For instance, if the sample exhibits a strong sense of local identity and attachment to the grasslands, the attitude variable may be more likely to show statistical significance; whereas in regions with higher levels of commercialisation and greater mobility, the explanatory power of attitudes may not necessarily be predominant. Fourthly, the TPB suggests that intention often acts as a mediator between antecedents and behaviour43. This study confirms that herders’ intention to realise the value of ecological products exerts a partial mediating effect between ecological ethics and practical behaviour, and that ecological ethics can directly influence behaviour. This conclusion effectively explains why the attitude–behaviour gap, which is prevalent in academic circles, is not significant among pastoralist communities: ecological ethics have been deeply internalised as pastoralists’ value standards and identity, and can thus influence behaviour through the mediation of willingness, whilst also directly guiding behavioural decisions through intrinsic value constraints. These dual pathways effectively bridge the gap between attitudes and behaviour. Fifthly, the study found that policy support plays a significant positive moderating role in the translation of intentions into actions. Ecological compensation, financial support and policy advocacy that align with herders’ interests can both strengthen the behavioural guidance provided by intentions and reduce implementation costs, thereby enhancing the feasibility of action59,87. By incorporating relevant policy dimensions within the pastoral context, this study has further expanded the scope of research on the intention–behaviour conversion mechanism. Existing international research on this topic reveals marked divergences: some studies acknowledge the positive role of subsidy incentives88, whilst others argue that regulatory policies are more effective89; there are also views advocating the combined use of policy instruments90. In light of the findings of this study, it can be critically concluded that the positive moderating effect of policy support is more applicable in pastoral contexts characterised by stable policy implementation, sufficient public awareness, and strong government credibility, thereby clarifying the scope of application of this study’s conclusions. In summary, driven by both endogenous ecological ethics and external policy regulation, herders in pastoral areas have achieved a unity between ecological attitudes and practical behaviour. Local governments need to continuously refine ecological support policies, strengthen publicity and guidance, foster herders’ sense of ecological responsibility, and adopt a multi-pronged approach to facilitate the successful realisation of the value of grassland ecological products82,91. Sixthly, existing research largely relies on linear models to analyse the independent net effects of variables, making it difficult to uncover the complex mechanisms driven by the synergy of multiple factors. This study integrates SEM linear effects with fsQCA configuration analysis to identify six distinct behavioural driver pathways, confirming that herders’ behaviour in realising ecological value is the result of the interplay of multiple conditions, with different paths leading to the same outcome. The two research perspectives complement one another; the conclusions not only corroborate each other but also offer distinct interpretations. These six pathways fully reflect asymmetric causal relationships, such as compensation for missing factors, cognitive complementarity and substitution, and multidimensional synergistic empowerment. They effectively explain the causal logic that is difficult to elucidate through single-linear analysis, achieving an integrated interpretation of linear transmission patterns and multiple concurrent causal relationships, thereby highlighting the methodological advantages of an integrated research paradigm. Due to limitations in the sample and calibration standards, the resulting configuration pathways are applicable only to the research context of this study and do not possess universal applicability60,61. Overall, by adopting an ecological ethical perspective, this paper enriches the interpretative framework for research on the behaviour of value realisation for grassland ecological products, whilst also supplementing existing research at two levels: the translation of intentions for value realisation and complex causal pathways. It should be emphasised, however, that the conclusions of this paper do not imply that ecological ethics, intentions or policy support exert an influence of equal intensity and direction in all contexts. On the contrary, the variations observed in the international literature demonstrate that the behaviour of value realisation for grassland ecological products is a complex process embedded within cultural, institutional, market and resource constraints. Only by taking a holistic view that integrates ethical identity, economic incentives and institutional safeguards can we gain a more comprehensive understanding of the underlying logic behind herders’ behaviour, and provide theoretical support for the realisation of ecological product value that is both more explanatory and practically relevant. Limitations and future research This study has several key limitations that require addressing in future research. Firstly, the sample is not sufficiently representative, as the study primarily covers specific regions or types of herder groups, failing to fully reflect differences across various grassland types, economic conditions and cultural backgrounds; future research should expand the sample scope to enhance the external validity of the findings. Secondly, due to the inherent limitations of cross-sectional data—where all variables are measured at a single point in time—it is difficult to determine the temporal sequence and causal relationships between variables. Thirdly, based on the TPB framework, which requires a time lag between behavioural intention and actual behaviour, the conversion cycle for the behaviour of value realisation of grassland ecological products is particularly pronounced; this study utilised single-time-point cross-sectional data to examine the relationship between intention and behaviour, which fails to satisfy the temporal assumptions of the TPB, making it difficult to avoid bias in causal inference. Therefore, future research is advised to adopt a two-stage lagged survey design: measuring behavioural intention and antecedent variables at T1, and actual behaviour at T2, to enhance the rigour of causal inference. Fourthly, future research should further explore the interactions between these factors, as well as how effective interventions can be used to shape and enhance herders’ ecological ethics, thereby more broadly promoting the behaviour of value realisation for grassland ecological products and making a greater contribution to achieving sustainable development goals. Fifth, given the multitude and complexity of factors influencing behaviour, this study has not covered all relevant factors; subsequent research may continue to develop a multi-factor theoretical model of the behaviour of value realisation for grassland ecological products. Sixth, in terms of research methodology, future studies may adopt a mixed-methods approach by combining other quantitative research methods, with a view to yielding more research findings of practical value.