Mapping and Bridging Neuroscience-Based Research Domains: Neuromarketing, Neuroeconomics, Neurotourism and Neurorobotics

Highlights

What are the main findings?

• The specified author-keyword network is centred on neuroeconomics and neuromarketing, while neurorobotics occupies a more peripheral technological position.
• The analytical corpus contains a small and methodologically heterogeneous neurotourism core, whereas tourism-related neuroscience extends mainly through consumer neuroscience and neuromarketing.

What are the implications of the main findings?

• The four neuroscience-based domains are connected but should not be treated as conceptually or methodologically interchangeable.
• Distinguishing scientific-domain membership from tourism application clarifies how related domains may support the development of neurotourism.

Abstract

Background/Objectives: Neuroscience-based research has expanded across neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics, but these domains differ in disciplinary orientation, levels of consolidation and methodological traditions. This study examines the scientific and thematic evolution represented within the analytical corpus across these domains, the position of neurotourism within this broader landscape, and differences in methodological approaches, key constructs and reported research needs. Methods: A scoping review combined structural evidence mapping with structured abstract-level data charting of 981 English-language articles indexed in Scopus. Bibliometric procedures examined scientific production, citation patterns, author-keyword co-occurrence and thematic evolution, while rule-based and human-validated classification supported cross-domain comparison. Abstracts were available for 972 publications. Results: Scientific production increased substantially over time, while the specified author-keyword network remained centred on neuroeconomics and neuromarketing, with neurorobotics occupying a more peripheral technological position. Thematic evolution showed continuity in these domains alongside increasing specialisation in neuroimaging, machine learning, memory, embodiment and spiking neural networks. Within the analytical corpus, the neurotourism-related core comprised eight publications and was methodologically heterogeneous, with only one primary empirical study involving human participants and direct psychophysiological measurement. A broader body of tourism and hospitality research was located mainly within neuromarketing and consumer neuroscience, with more limited contributions from neuroeconomics and neurorobotics. Reported research needs differed, ranging from methodological consistency and conceptual refinement to technological robustness and empirical consolidation. Conclusions: The four domains are connected but are not conceptually or methodologically interchangeable. Distinguishing scientific-domain membership from application context clarifies how tourism-related neuroscience extends beyond the neurotourism label and how related domains may strengthen its empirical and interdisciplinary development.

1. Introduction

Neuroscience has progressively extended beyond its traditional biomedical foundations into research on consumer behaviour, economic choice, tourism experiences and human interaction with intelligent and robotic systems. This expansion has given rise to several interdisciplinary domains that share an interest in the neural and cognitive mechanisms underlying behaviour but differ in their disciplinary traditions and levels of consolidation. Neuromarketing and consumer neuroscience integrate neuroscience, psychology and marketing research to investigate consumer responses [1,2,3]. Neuroeconomics focuses more specifically on the neural and cognitive foundations of valuation, reward, risk and economic decision-making [4,5,6]. Neurotourism has emerged through the application of neuroscience and related behavioural disciplines to tourism research [7,8], whereas neurorobotics has developed around neuro-inspired systems, embodied interaction and interfaces connecting neural processes with robotic behaviour [9,10].

These domains have also generated a substantial body of synthesis research. Neuromarketing and consumer neuroscience have been examined through conceptual, systematic, methodological, bibliometric and meta-analytical reviews [2,3,8,9,10,11,12,13,14]. Neuroeconomics has an established review tradition centred on decision-making, reward, risk and economic behaviour [5,6,15,16,17]. Neurotourism has more recently been examined through conceptual, systematic and bibliometric approaches [12,18,19,20,21,22], whereas reviews in neurorobotics tend to address specialised technological, sensorimotor or interface-related problems [9,10,11]. The gap therefore does not lie in an absence of previous reviews. These synthesis efforts have largely developed within separate disciplinary, methodological or application-specific streams, leaving a limited comparative understanding of how these neighbouring neuroscience-based domains have evolved scientifically and thematically relative to one another. This raises the first question addressed in the study: what scientific and thematic patterns of development are represented within the analytical corpus across neuromarketing, neuroeconomics, neurotourism and neurorobotics over time?

The consequences of these disciplinary boundaries are particularly visible in neurotourism. Neuroscience-based tourism research has developed comparatively slowly and with limited interdisciplinarity, leading previous reviews to draw explicitly on neighbouring areas such as affective neuroscience, neuromarketing and neuroeconomics when formulating research agendas for tourism neuroscience [22]. At the same time, the literature relevant to tourism extends substantially beyond publications explicitly adopting the neurotourism label. Cardoso et al. [12] identified 45 studies applying neuroscience and neuromarketing theories or methods to tourism, hospitality and travel, of which only three explicitly addressed neurotourism. Cardoso and Fraga [19] similarly showed that neuroscientific research applied to tourism is distributed across multiple themes, methods and disciplinary communities. Empirical studies using neuroscientific and psychophysiological approaches have examined destination preference, advertising effectiveness, visual attention, memory, hotel selection, pricing, gastronomy and tourist emotions, while being positioned within neuromarketing, consumer neuroscience, tourism marketing or consumer behaviour [23,24,25,26,27,28]. The explicitly identified neurotourism-related literature therefore represents only part of a broader body of neuroscience-based research applied to tourism and hospitality. This distinction raises a second question: how is neurotourism represented as a research domain, and how does neuroscience-based research extend into tourism and hospitality through related fields?

Differences across the four domains are also evident at methodological and conceptual levels. Reviews of neuromarketing and consumer neuroscience identify concerns involving ecological validity, measurement integration, reproducibility and generalisability [13,29,30]. In neuroeconomics, unresolved questions concern decision processes, their conceptualisation and the boundaries between neuroeconomics and adjacent areas of decision neuroscience [4,5,31,32]. Neurorobotics faces a different set of problems involving interface robustness, benchmarking, biological plausibility, usability and ethical implications [11,33,34,35,36]. Neurotourism presents another pattern, characterised by limited empirical consolidation, methodological accessibility problems and continued calls for stronger interdisciplinary development [12,20,21,22]. These reviews provide detailed accounts within individual fields, but they do not establish whether methodological approaches, key constructs and reported research needs converge across domains or remain field-specific. This leads to the third question: how do the principal research domains compare in their methodological approaches and key constructs, and what research gaps and future directions are reported across the four domains?

The literature therefore points to a fragmented synthesis landscape rather than a lack of prior knowledge. Although individual domains, technologies and applications have been examined, comparative analysis across the four research areas remains limited. To address these questions, the study adopts a scoping review design combining structural evidence mapping with structured abstract-level data charting. Domain classification is treated separately from the tourism and hospitality application context, while research gaps and future directions are retained only when explicitly reported in the source abstracts. The study has three objectives: (1) to examine the scientific and thematic patterns of development represented within the analytical corpus across the four domains; (2) to delineate the neurotourism-related research core and examine neuroscience-based applications in tourism and hospitality across related domains; and (3) to compare their methodological approaches and key constructs and identify reported research gaps and future directions. The results indicate increasing scientific production, an author-keyword network centred on interconnected neuroeconomics and neuromarketing cores, a comparatively small and heterogeneous neurotourism-related core, and tourism and hospitality applications distributed mainly across neighbouring neuromarketing and consumer neuroscience research. Further comparison shows that the research needs reported across the four domains differ in nature, allowing research priorities to be defined more specifically within each field and indicating where methods or concepts developed in neighbouring domains may inform neuroscience-based tourism research.

2. Conceptual Foundations and Interdisciplinary Interfaces

2.1. Neuroscience-Based Domains: Boundaries, Development and Methodological Differences

Neuroscience-based research has developed across neighbouring domains that share an interest in the mechanisms underlying behaviour but differ in theoretical orientation, objects of analysis and disciplinary traditions. Neuromarketing and consumer neuroscience apply neural and psychophysiological approaches to attention, emotion, memory, preference and choice in consumer contexts [1,2,3]. Neuroeconomics combines neuroscience, economics and behavioural research to examine valuation, reward, risk, uncertainty and choice [4,5,6]. Although the two fields intersect around valuation and decision processes, consumer neuroscience places these mechanisms within consumer and market contexts, whereas neuroeconomics focuses more directly on economic and value-based choice [3,5,6].

Neurorobotics occupies a different position, connecting neuroscientific principles with embodied technological systems, including brain-inspired robotics, sensorimotor control, body representation, neural interfaces and human–robot interaction [9,10,11]. Its links with neighbouring domains arise through learning, decision-making, agency, embodiment and responses to technology, while its primary object of analysis remains distinct. Neurotourism is more permeable because tourism-related neuroscience extends beyond publications explicitly using that label. Reviews have described neuroscience in tourism as comparatively slow to develop and have called for stronger interdisciplinary engagement [22]. Cardoso et al. [12] identified 45 studies connecting neuroscience and neuromarketing with tourism, hospitality and travel, although only three explicitly addressed neurotourism, while Cardoso and Fraga [19] showed that this research spans different methods, themes and disciplinary communities. Empirical applications include destination marketing and preference [24], travel advertising and visual attention [25], and hotel selection [26]. Al-Nafjan et al. [37] similarly grouped 52 tourism studies across different contexts and neuroscientific or psychophysiological modalities. A tourism or hospitality application of neuroscience should therefore not, by itself, be treated as evidence that neurotourism is the scientific domain of the study.

The four areas also differ in scientific consolidation. Neuromarketing and consumer neuroscience have been examined through conceptual, systematic, methodological and bibliometric reviews [3,7,14], while neuroeconomics has an established tradition of reviews and meta-analytical research on economic decision-making and its neural mechanisms [5,15,16,17]. Neurotourism has more recently become the subject of conceptual, systematic and bibliometric synthesis [12,18,21,22,37], whereas neurorobotics reviews tend to focus on more specialised architectures, embodied processes and interfaces [9,10,11]. The resulting synthesis remains largely organised within disciplinary and methodological boundaries, leaving cross-domain relationships less examined.

Their scientific trajectories also differ. In neuromarketing and consumer neuroscience, early work helped distinguish the academic study of consumer behaviour using neuroscientific approaches from the more managerial use of neuromarketing [38,39]. The field later broadened towards attention, affect, memory, reward, preference and decision-making, alongside a wider range of neuroscientific and physiological measures [39,40,41]. More recent research points towards computational and predictive approaches, including EEG-based deep-learning models for purchase decisions [42], combined EEG and eye tracking for preference prediction [43], and neuroforecasting using neural, behavioural and self-report data with machine-learning methods [44].

Neuroeconomics developed around increasingly detailed accounts of valuation and decision-making, including reward learning, risk and ambiguity, intertemporal choice, and emotional and social influences on choice [5]. The neural common currency hypothesis proposed that qualitatively different rewards can be represented on a shared scale of subjective value [45], with evidence implicating the ventromedial prefrontal/orbitofrontal cortex and a broader valuation network [5,45]. The field thus moved beyond neural correlates of choice towards explaining how value is learned, represented and compared.

Neurorobotics followed a more technological and embodied trajectory. Emerging from the convergence of neuroscience, computational neuroscience, artificial intelligence, embodied cognition and robotics, the field became associated with embodied autonomous neural systems [46,47]. More recent developments include neuromorphic computing and spiking neural networks [48], closed-loop brain–machine interfaces [49], and co-adaptive neural interfaces in which users and decoding systems adapt reciprocally [49,50].

In tourism, research has continued to draw on concepts and methods from neighbouring fields, alongside calls for stronger interdisciplinarity and a broader empirical base [12,19,22]. Since the 2010s, work at the tourism–neuroscience interface has incorporated cognitive and affective neuroscience, neuroeconomics, and consumer neuroscience and neuromarketing [12,22,37]. Attention, emotion, preference, choice and decision-making remain central, while more recent agendas have extended towards memory, sensory perception and tourism experience, using methods including EEG, eye tracking and cardiovascular measures [12,22,37,51]. A broader neurotourism agenda may also encompass the cognitive, affective and behavioural processes of other tourism stakeholders [12]. From this perspective, we propose neurotourism as more than the application of neuromarketing to tourism contexts: it can be understood as an interdisciplinary interface between neuroscience and tourism studies capable of addressing tourism as a complex and multistakeholder phenomenon.

The domains also face different methodological challenges. Neuromarketing and consumer neuroscience continue to address measurement quality, multimethod integration, ecological validity and generalisability [13,29]. Neuroeconomics combines neuroimaging, behavioural economic paradigms and computational approaches while confronting conceptual issues such as uncertainty and the boundaries of decision neuroscience [4,5,31,32]. Neurorobotics faces challenges involving interface robustness, benchmarking, biological plausibility and usability [11,33,35,36], whereas neurotourism still requires stronger interdisciplinary and methodological development [12,19,22]. These differences reinforce the need to compare the domains without treating them as conceptually or methodologically interchangeable.

2.2. Interdisciplinary Interfaces, Adaptation and Neuro-Adaptive Ecosystems

Despite these differences, the four domains share mechanisms, constructs and, in some cases, measurement approaches [3,11,22]. Their interfaces can be organised without implying a linear evolution between fields: from neural and computational processes involved in choice and decision-making [52], through consumer behaviour and decision processes [53], to experience-centred approaches in tourism [22]. In parallel, developments at the neuroscience–AI interface are bringing brain-inspired computation closer to embodied artificial systems capable of perceiving, deciding and acting in changing environments [48,54].

Adaptation and co-adaptation provide a further point of contact. Reciprocal exchanges between neuroscience and artificial intelligence allow computational models to inform neuroscience while neuroscientific principles contribute to artificial learning systems [55]. Neuromorphic computing translates features of biological intelligence into adaptive computational architectures [48], closed-loop brain–machine interfaces incorporate feedback, learning and neural plasticity [49], and co-adaptive interfaces allow users and decoding algorithms to adapt to one another over time [50]. Related work on adaptive intelligence likewise considers artificial agents capable of online learning, generalisation and rapid adaptation to environmental change [56].

Building on this conceptual background, we propose Neuro-Adaptive Ecosystems as a prospective interdisciplinary concept for systems in which brains, artificial intelligence, machines, and physical, social and economic environments interact and adapt reciprocally over time. Its added value lies in bringing reciprocal adaptation among brains, artificial intelligence, machines, and physical, social and economic environments into a single prospective framework, rather than treating these adaptive relationships separately. The concept does not seek to merge neuroeconomics, consumer neuroscience and neuromarketing, neurotourism and neurorobotics into a single field. Instead, it identifies a possible research frontier in which understanding and prediction are complemented by adaptation and co-adaptation across human, technological and environmental systems. This perspective may be particularly relevant to tourism, whose knowledge base has historically drawn on multiple contributory disciplines [57]. Neurotourism can therefore engage with neighbouring neuroscience-based domains while retaining its focus on tourism phenomena. This balance between disciplinary distinction and interdisciplinary connection provides the conceptual basis for the analysis that follows.

3. Materials and Methods

3.1. Research Design

This study adopted a scoping review design informed by established methodological guidance [58,59,60], with reporting guided by the PRISMA Extension for Scoping Reviews (PRISMA-ScR) [61]. The design was selected to support the systematic mapping and comparison of a broad and heterogeneous body of research spanning neighbouring neuroscience-based domains.

The review examined the extent, patterns of scientific development represented within the analytical corpus, and substantive characteristics of research spanning neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics. It combined systematic evidence identification and corpus construction with two complementary analytical components: structural evidence mapping at the corpus level and structured abstract-level data charting. Structural mapping was used to characterise the scientific and thematic organisation of the corpus, whereas data charting supported the comparison of methodological characteristics, constructs, application contexts and source-reported research needs.

3.2. Evidence Identification and Corpus Construction

The evidence search was conducted in Scopus using the TITLE-ABS-KEY field and was completed on 17 July 2026. The search strategy combined three mandatory conceptual blocks connected by the Boolean operator AND. The first captured terminology associated with neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics; the second captured neural, cognitive and affective mechanisms; and the third captured neuroscientific, neurophysiological, psychophysiological and biometric methods. The complete search syntax and filtering specifications are provided in Appendix A.1 (Table A1).

The initial search retrieved 2154 records. Application of the predefined subject-area and document-type restrictions reduced the dataset to 1531 records, and restriction to publications in English produced the analytical corpus of 981 articles. No additional records were excluded after corpus construction. Subsequent examination of ambiguous, overlapping or potentially unsupported research-domain assignments concerned research-domain assignment rather than corpus eligibility. Accordingly, all subsequent analyses were conducted on the 981 articles captured by the predefined search architecture and filtering criteria. Domain-specific counts therefore indicate representation within this analytical corpus and should not be interpreted as estimates of the absolute size of the respective research fields. The predefined three-block search architecture was designed to construct this cross-domain analytical corpus rather than an exhaustive historical bibliography of each field. The complete corpus-construction workflow is reported in Appendix A.2.

3.3. Analytical Procedures

Structural evidence mapping was first conducted to examine the scientific and thematic organisation of the corpus. Bibliometric science-mapping procedures were implemented using Bibliometrix/Biblioshiny (v. 5.5.0) [62] and included annual scientific production, citation analysis, author-keyword co-occurrence and thematic evolution, consistent with established approaches to science mapping and bibliometric analysis [63,64,65]. To assess whether the reported author-keyword network structure depended on the selected network size or co-occurrence threshold, a robustness analysis was conducted. The original specification retained 50 nodes, applied a minimum co-occurrence threshold of 2 and removed isolated nodes. Sensitivity was assessed by varying the maximum number of retained nodes across 40, 50, 60, 75 and 100 and the minimum co-occurrence threshold across 1, 2 and 3, yielding 15 network specifications. Author keywords were retained in their original bibliographic form without lexical harmonisation in all specifications. Rank-order stability of betweenness centrality, closeness centrality and PageRank was assessed using Spearman’s rank correlation coefficient (ρ), calculated across nodes shared between each alternative specification and the 50-node, threshold-2 baseline network. The 2020 thematic-evolution cut-off was selected based on the observed acceleration in scientific production after that year. Sensitivity analysis using alternative cut-offs of 2019 and 2021 confirmed the persistence of the principal thematic continuities involving neuroeconomics, neuromarketing and neurorobotics. Across these alternative specifications, 46.9% and 43.8% of the baseline thematic links were preserved, respectively, with Jaccard similarity values of 0.319 and 0.241, indicating stability in the principal thematic continuities alongside variation in the broader configuration of thematic relationships (Appendix A.3, Table A6 and Table A7). The complete analytical settings, network and thematic-evolution parameters, and robustness specifications are reported in Appendix A.3.

Following corpus-level structural mapping, Python-supported (3.14.7) rule-based processing generated candidate assignments to neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics. Domain assignment was non-mutually exclusive, and ambiguous, overlapping or potentially incidental matches were examined contextually using the title, abstract and author keywords. Multiple domain assignments were retained when supported by the available bibliographic evidence. The Python-assisted processing instructions, classification logic and associated operational rules are reported in Appendix A.4 (Table A8).

Structured abstract-level data charting was then used to organise variables aligned with the research questions, including study type and design, technologies and measures, key constructs, application areas, tourism and hospitality relevance, and explicitly reported research gaps and future research directions [58,59,60]. Abstracts were available for 972 of the 981 publications. For the remaining nine records, applicable coding relied on the available title and author keywords; information that could not be established from these bibliographic fields was recorded as not reported rather than inferred. The data-charting dimensions, operational definitions and decision rules are presented in Appendix A.5 (Table A9).

Scientific-domain classification and tourism or hospitality application were treated as analytically distinct dimensions. The neurotourism-related research core was identified using the predefined neurotourism and tourism-neuroscience terminology incorporated into the search strategy together with contextual bibliographic evidence and was therefore not restricted to publications containing the literal term neurotourism. A tourism or hospitality application did not, by itself, result in classification as neurotourism. Information not supported by the available bibliographic evidence was not inferred. The complete charting framework, operational definitions and decision rules are provided in Appendix A.5 (Table A10).

AI-assisted contextual interpretation was used as a supporting procedure for ambiguous classifications, abstract-level charting and organisation of comparative evidence. ChatGPT, using the GPT-5.6 Thinking model developed by OpenAI, was applied through standardised prompts restricted to the available title, abstract and author keywords. The AI-assisted procedure followed the initial rule-based processing and produced provisional classifications, extractions and syntheses. The provisional classifications and other AI-assisted outputs were jointly reviewed against the bibliographic evidence by three authors, and final classification and interpretation decisions were reached by consensus, with the authors retaining responsibility for the final classifications and interpretations. The model did not make autonomous eligibility, inclusion or exclusion decisions. The complete set of standardised prompts, analytical purposes and human-validation rules is provided in Supplementary Table S1.

For the analysis of research gaps and future directions, only limitations, unresolved issues and future research needs explicitly reported in the available abstracts were retained as source-reported evidence. The absence or relative scarcity of methods, constructs, populations, contexts or applications within the corpus was not, by itself, interpreted as a research gap. Review-level patterns were therefore kept analytically distinct from research needs explicitly identified by the source publications. The complete operational rules are reported in Appendix A.6.

4. Results and Discussion

4.1. Scientific and Thematic Evolution Represented Within the Analytical Corpus

The first research question concerned the scientific and thematic evolution represented within the analytical corpus across the four domains over time. The analytical corpus shows a marked increase in scientific production over the period analysed (Figure 1). Publication activity remained limited during the early years but began to increase more consistently from 2007 onwards, followed by substantially stronger growth after 2020. Annual production reached its highest level in 2024, with 87 publications, and remained similarly high in 2025, with 86 publications. This recent acceleration is consistent with evidence of increasing scientific activity and thematic diversification in neuromarketing and consumer neuroscience [7], while the emergence of neuroscience-based tourism research illustrates how neuroscientific approaches have continued to extend into newer application domains [8,9,10]. Nevertheless, because Figure 1 aggregates the four research domains, the observed growth should not be interpreted as evidence that they have expanded at the same pace or reached comparable levels of consolidation.

Figure 1. Annual scientific production of the final corpus, 2002–2026.

The lower number of publications recorded for 2026 (n = 53) should not be interpreted as a decline in scientific production, as data collection was completed on 17 July 2026 and therefore covered only part of the publication year.

4.1.1. Most-Cited Documents

Within the final analytical corpus of 981 Scopus-indexed articles, the ten most-cited documents within the corpus are concentrated in neuroeconomics and in neuromarketing and consumer neuroscience, particularly around valuation, reward processing, decision-making and consumer response (Table 1). These ten documents correspond to the ten highest values obtained by ranking all 981 records in descending order according to the Scopus Cited by field. The tenth-ranked record had 395 citations; therefore, 395 Scopus citations constituted the empirical inclusion threshold for the top 10 within this analytical corpus. Neuroeconomics accounts for the highest-cited contributions within the analytical corpus, while neuromarketing and consumer neuroscience extend these foundations into branding, product design and multimethod consumer research. This concentration around valuation, reward and decision-making reflects the close conceptual relationship between neuroeconomics and consumer neuroscience described by Plassmann et al. [3], while the prominence of branding and consumer-response research is consistent with the thematic structure of neuromarketing identified by Cardoso et al. [12]. No publications assigned to the neurotourism or neurorobotics domains appear among the ten most-cited records within the analytical corpus. Citation counts are those recorded in the Scopus dataset on 17 July 2026.

Table 1. Ten most-cited documents within the final Scopus analytical corpus.

Rank	Authors	Cites	Research Domain(s)	Main Contribution
1	Plassmann, O’Doherty and Rangel [66]	714	Neuroeconomics; Neuromarketing and consumer neuroscience	Linked medial orbitofrontal activity to willingness to pay, supporting a neural representation of subjective economic value.
2	Hare, O’Doherty, Camerer, Schultz and Rangel [67]	673	Neuroeconomics	Distinguished goal-value representation from reward prediction-error processing.
3	Hodgkinson and Healey [68]	573	Neuroeconomics	Examined reflective and reflexive cognition in strategic management.
4	Padoa-Schioppa [69]	491	Neuroeconomics	Consolidated the good-based model of economic choice and subjective-value representation.
5	Koenigs and Tranel [70]	435	Neuroeconomics	Demonstrated the role of ventromedial prefrontal and emotion-related processes in fairness-based decisions.
6	Reimann et al. [71]	429	Neuromarketing and consumer neuroscience; Neuroeconomics	Linked aesthetic package design with behavioural, psychological and neural responses.
7	Plassmann, Ramsøy and Milosavljevic [3]	425	Neuromarketing and consumer neuroscience; Neuroeconomics	Synthesised attention, memory, emotion, reward, valuation and decision-making within consumer neuroscience.
8	Schmitt [72]	405	Neuromarketing and consumer neuroscience	Organised the cognitive, affective and social processes underlying consumers’ relationships with brands.
9	Khushaba et al. [73]	397	Neuromarketing and consumer neuroscience	Demonstrated multimethod integration of EEG and eye tracking in the assessment of consumer responses.
10	Huettel, Song and McCarthy [74]	395	Neuroeconomics	Showed that probabilistic context modulates neural processing during decisions under uncertainty.

The absence of neurotourism and neurorobotics from the ten most-cited documents should not be interpreted as evidence of lower scientific relevance, since citation counts are affected by publication age and differences in the historical development of the respective domains. In the case of neurotourism, its absence is also compatible with a literature that remains comparatively small and still in a process of consolidation [8,12].

4.1.2. Conceptual Structure: Author-Keyword Co-Occurrence

Figure 2 displays the two dominant conceptual communities represented in the author-keyword co-occurrence visualisation, centred on neuromarketing and neuroeconomics. These communities contain the highest-centrality nodes in the network and are connected through decision-related and neuroimaging-related terms, while retaining distinct thematic profiles. The complete network analysis also identifies smaller peripheral configurations that are not represented in Figure 2 and are examined separately below. Because author keywords were analysed as reported in the Scopus records, the network should therefore be interpreted as a representation of author-reported bibliographic terminology rather than a fully harmonised conceptual vocabulary. As terminology associated with the four research domains formed part of the predefined search architecture, the relative prominence of domain labels is partly conditioned by corpus construction; network centrality and community structure are therefore interpreted as properties of the retrieved corpus rather than as independent estimates of prominence in the wider literature.

Figure 2. Author-keyword co-occurrence network across the research corpus.

The first dominant cluster displayed in Figure 2 is organised around neuroeconomics, which has a betweenness centrality of 299.824, together with a closeness value of 0.014 and PageRank of 0.090. The cluster encompasses decision-making, reward, risk, trust, behavioural economics, decision neuroscience, game theory and reinforcement learning. Within it, reinforcement learning occupies a notable intermediary position (betweenness = 32.397), connecting value-based decision research with computational learning approaches.

Closely connected to this decision-oriented core, the second dominant cluster displayed in Figure 2 is centred on neuromarketing, the most structurally influential node in the entire network, with the highest betweenness centrality (569.359), closeness centrality (0.017) and PageRank (0.160). The cluster integrates consumer neuroscience, neuroscientific and psychophysiological methods, cognitive and affective constructs, and marketing applications. Consumer neuroscience is the second most influential node within this cluster (betweenness = 89.936; PageRank = 0.066), while fMRI also occupies an important intermediary position (betweenness = 55.219; PageRank = 0.048). EEG, neuroscience, electroencephalography, eye tracking, emotion, attention, memory, consumer behaviour, advertising, branding and marketing form a dense conceptual structure around neuromarketing.

The proximity of these two dominant clusters is reflected in their substantial structural connectivity. Cross-cluster links involve decision-related and neuroimaging terms, while both cores share concepts associated with decision-making, valuation, reward, risk, preference and consumer choice. Their thematic organisation nevertheless remains distinct: the neuroeconomics cluster is primarily structured around economic valuation and strategic choice, whereas the neuromarketing cluster is more strongly associated with consumer responses, brands, advertising and marketing stimuli. This distinction reflects the different substantive emphases described in the literature, with neuroeconomics more closely concerned with value-based and economic decision processes and consumer neuroscience integrating valuation and choice with attention, emotion, memory and responses to marketing stimuli [3,6,7].

Although Figure 2 focuses on these two dominant communities, the complete network output contains additional, substantially smaller and more peripheral configurations involving neurorobotics, learning and spiking neural networks. Their network positions and centrality values provide additional information about how these specialised themes relate to the dominant conceptual structure, even though they are not displayed in the figure. Within this peripheral part of the network, neurorobotics has limited betweenness (1.615), low closeness (0.006) and a PageRank of 0.014, indicating weaker integration with the dominant cores. Its principal relationships occur with learning and spiking neural networks, reflecting the technologically specialised and brain-inspired orientation characteristic of neurorobotics research [10,11].

Learning, by contrast, occupies a distinctive intermediary position in this peripheral configuration. Although its PageRank is comparatively low (0.009), its betweenness centrality is relatively high (59.603). Its principal connection with the decision-oriented part of the network occurs through reinforcement learning, while additional relationships with neurorobotics and spiking neural networks connect the decision-oriented core with the smaller technological branch.

At the outer edge of this peripheral configuration, spiking neural networks have a PageRank of 0.012, closeness of 0.006 and zero betweenness. It therefore occupies a specialised peripheral position associated primarily with learning and neurorobotics. This positioning is consistent with the use of spiking neural architectures in brain-inspired and neurorobotic systems [11,36].

The network is dominated by the two conceptual cores centred on neuroeconomics and neuromarketing, while a smaller technological periphery develops around learning, neurorobotics and spiking neural networks. Learning, particularly through its connection with reinforcement learning, provides the clearest link between these parts of the network, whereas direct connectivity between the technological periphery and the neuromarketing cluster remains limited. Sensitivity analysis across alternative network specifications supported the stability of the principal centrality hierarchy and the distinction between the two dominant communities, while betweenness centrality and the smaller peripheral configuration were more parameter-sensitive (Appendix A.3, Table A4 and Table A5).

4.1.3. Thematic Evolution

The thematic evolution map traces the continuity and diversification of identified themes between 2002–2020 and 2021–2026 (Figure 3). In the earlier period, the structure was organised around seven themes, with neuromarketing and neuroeconomics as the most prominent. In the later period, both remain central alongside more specialised themes involving neuroimaging, machine learning, memory, cognitive neuroscience and computational approaches.

Figure 3. Thematic evolution. Note: Arrows indicate thematic links from the earlier to the later time period; colors distinguish the themes.

Among the identified themes, neuroeconomics shows the strongest thematic continuity. Its transition to the same theme from 2021–2026 has the largest number of occurrences in the map (193) and a high Weighted Inclusion Index of 0.79. This continuity is supported by the substantial retention of conceptual content from the earlier period, including fMRI, decision-making, reward, risk, behavioural economics, decision neuroscience, amygdala and loss aversion.

At the same time, neuroeconomics also contributes to several more specialised later themes, including neuroimaging (occurrences = 6; Weighted Inclusion Index = 0.36), memory (5; 0.42), trust (8; 0.50), ultimatum game (6; 0.43), computational modelling (2; 0.50) and robot (3; 0.44). By contrast, its transition to neuromarketing involves a comparatively large number of occurrences (52) but a very low Weighted Inclusion Index of 0.02, suggesting a relatively frequent thematic linkage but limited preservation of the earlier theme’s conceptual composition.

A similarly persistent pattern is visible for neuromarketing. Its self-transition comprises 141 occurrences and a Weighted Inclusion Index of 0.59, indicating continuity around consumer neuroscience, EEG, attention, advertising, emotions, eye tracking, branding, galvanic skin response, consumer behaviour and artificial intelligence. However, neuromarketing also diversifies into several adjacent areas. Its strongest proportional transitions are towards perceived value (Weighted Inclusion Index = 1.00), functional magnetic resonance imaging (0.75), neuroimaging (0.64), brain (0.43) and cognitive neuroscience (0.40).

Some of these transitions, however, are based on few occurrences, particularly perceived value, which is based on only two occurrences. The transition from neuromarketing to neuroeconomics includes 35 occurrences but a low Weighted Inclusion Index of 0.08, indicating a relatively frequent connection between the themes but limited preservation of the earlier neuromarketing theme’s internal composition. This combination of continuity and diversification also corresponds with the broadening of neuromarketing from core consumer processes such as attention, emotion, memory and valuation towards a wider range of neuroscientific methods, technologies and application areas [3,7].

Beyond these two dominant domains, several earlier methodological or conceptual themes evolve into more specialised later configurations. The earlier electroencephalography (EEG) theme develops mainly towards machine learning and neuromarketing. Its strongest transition is to machine learning, with a Weighted Inclusion Index of 0.43 and four occurrences, supported by terms such as deep learning and consumer preferences. The transition to neuromarketing is weaker (Weighted Inclusion Index = 0.29; occurrences = 3) and is supported by EEG signal and virtual reality. EEG-related themes are increasingly linked to computational and applied consumer research contexts.

A related but more fragmented pattern characterises the earlier event-related potential theme. Its most coherent transition is towards brain–computer interface research, with both the Weighted Inclusion Index and Inclusion Index equal to 0.33 and four occurrences. Weaker transitions connect cognitive neuroscience, machine learning and neuromarketing. The transition to neuromarketing has the highest number of occurrences among these links (9) but a low Weighted Inclusion Index of 0.11, indicating broader connection without strong thematic continuity as an independent configuration.

The earlier learning theme also evolves in more specific directions. Its strongest transition is towards embodiment, with both the Weighted Inclusion Index and Inclusion Index equal to 0.50. It also contributes to memory (occurrences = 8; Weighted Inclusion Index = 0.25) and robot research (3; 0.44), suggesting that this theme develops simultaneously in cognitive and technological directions.

Within this more technological branch, neurorobotics exhibits marked continuity and specialisation. Its self-transition has a Weighted Inclusion Index of 0.90, 30 occurrences and a Stability Index of 0.09. It also develops spiking neural networks, with a Weighted Inclusion Index of 0.50, seven occurrences and a Stability Index of 0.08, indicating increasing connections to biologically inspired computational architectures. A related but smaller specialised line is visible in the transition from robotics to electromyography, which is based on only two occurrences but has the highest Stability Index reported among the transitions examined (0.14). This specialised trajectory is compatible with the longstanding brain-inspired orientation of neurorobotics and its subsequent development around neural, embodied and interface-based technological systems [10,11].

4.2. Neurotourism Within the Broader Neuroscience-Based Research Landscape

4.2.1. The Explicit Neurotourism Research Core

The second research question concerned how neurotourism is represented as a research domain and how neuroscience-based research extends into tourism and hospitality through related fields. The analysis first delineated the explicit neurotourism research core before examining this broader cross-domain extension.

Within the analytical corpus, the explicitly identified neurotourism core is limited and methodologically heterogeneous. The eight publications forming this core were identified through the predefined neurotourism and tourism-neuroscience expressions used in the domain-classification procedure. They comprise publications explicitly adopting the neurotourism label, broader neuroscience-based tourism research, review-oriented contributions and one neuro-inspired computational study. Table 2 summarises the characteristics of this explicit neurotourism-related core.

Table 2. Publications representing the explicit neurotourism-related research core.

Authors	Study Design	Primary Empirical Study	Relationship to Neurotourism
Güner and Işık [20]	Systematic review	No	Neurotourism is explicitly indicated in the author keywords.
Li et al. [22]	Conceptual and narrative review	No	Explicitly centred on neuroscience in tourism.
Calderón-Fajardo et al. [75]	Experimental study	Yes	Explicitly identified as neurotourism.
Bağçı [76]	Documentary thematic analysis	No	Explicitly identified as neurotourism.
Jiang and Li [77]	Computational modelling study	No	Neuro-inspired computational approach applied to rural tourism planning; included within the predefined neurotourism-related terminology but conceptually distinct from human neuroscience-based neurotourism research.
Ndaguba and van Zyl [18]	Bibliometric review	No	Directly centred on neurotourism discourse in tourism and hospitality.
Arslan et al. [21]	Bibliometric review	No	Directly centred on the emergence of neurotourism.
Cardoso and Fraga [19]	Bibliometric and content analysis	No	Examines neuroscientific methods in tourism and destination research.

Of the eight publications identified, only one was classified as a primary empirical neurotourism study involving human participants and direct psychophysiological measurement. Calderón-Fajardo et al. [75] conducted an experimental study using eye tracking and galvanic skin response, whereas the remaining seven contributions relied on review-based, bibliometric, conceptual, documentary or computational approaches.

Most of these non-primary studies focused on mapping, synthesising or conceptually framing the emerging field. Güner and Işık [20] reviewed food image databases relevant to food services and tourism, while Li et al. [22] provided a conceptual and narrative synthesis of neuroscience research in tourism. This broader effort to define and organise the field is also evident in the bibliometric studies of Ndaguba and van Zyl [18], Arslan et al. [21], and Cardoso and Fraga [19], who combined bibliometric and content analysis to examine neuroscientific methods in smart tourism and destination research. Bağçı [76] extended this non-experimental strand through a neuroscience-informed thematic analysis of tourism promotional materials. This predominance of review, bibliometric, conceptual and documentary contributions over primary human experimentation is consistent with previous assessments of tourism neuroscience as a still-developing research area with a limited empirical base and a continuing need for stronger interdisciplinary research [8,12].

A different trajectory is represented by Jiang and Li [77], who developed and tested a neuro-inspired xLSTM model for rural tourism planning. Although quantitatively evaluated, the study did not examine tourists’ neural, physiological or behavioural responses and therefore represents a neuro-inspired computational application within the neurotourism-related corpus rather than a primary human neuroscience-based neurotourism study.

4.2.2. Neuroscience-Based Tourism and Hospitality Research Beyond the Neurotourism Label

The analysis identified a substantially broader body of tourism and hospitality research outside the explicit neurotourism core. These studies were located mainly within neuromarketing and consumer neuroscience, with more limited contributions from neuroeconomics and neurorobotics. To clarify this cross-domain extension, the studies were organised according to their research-domain classification and principal tourism or hospitality application theme (Table 3).

Table 3. Tourism and hospitality applications beyond the explicit neurotourism label.

Research Domain	Application Theme	Authors	Specific Context
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Bastiaansen et al. [24]	Destination marketing
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	De-Frutos-Arranz and Blasco López [78]	Tourism advertising
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Li et al. [8]	Destination advertising
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Stepchenkova et al. [79]	Tourist-attraction videos
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Jóźwiak [80]	Sustainable tourism advertising
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Lyu et al. [81]	Destination advertising and prior commercial exposure
Neuromarketing/Consumer neuroscience	Destination marketing, advertising and branding	Balzan et al. [82]	City and destination branding
Neuromarketing/Consumer neuroscience	Tourism imagery, heritage and destination perception	Michael et al. [83]	Tourism images and destination preference
Neuromarketing/Consumer neuroscience	Tourism imagery, heritage and destination perception	Garzón Paredes and Royo Vela [84]	Heritage and destination image
			Heritage, architectural styles and destination image
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	Muñoz-Leiva et al. [25]	Travel websites and advertising effectiveness
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	López [85]	Visual attractiveness of hotel websites
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	Gunawan et al. [86]	Online travel agency logo design and purchase intention
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	Kim and Kim [87]	Customer preferences on travel websites
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	Wang et al. [88]	Design, arousal and impulse purchase in mobile travel applications
Neuromarketing/Consumer neuroscience	Digital travel interfaces and booking	Modi and Kumar [89]	Eye-gaze behaviour in online travel booking
Neuromarketing/Consumer neuroscience	Hospitality and hotel-related consumer responses	Hsu and Chen [27]	Subliminal advertising and hotel selection
Neuromarketing/Consumer neuroscience	Hospitality and hotel-related consumer responses	Isa et al. [90]	Consumer decision-making in green hotels
Neuromarketing/Consumer neuroscience	Hospitality and hotel-related consumer responses	Yücel [91]	Neuroaesthetic responses to hotel architecture and visual imagery
Neuromarketing/Consumer neuroscience	Gastronomy, food and restaurant experiences	Mengual-Recuerda et al. [28]	Emotional responses in haute-cuisine dining
Neuromarketing/Consumer neuroscience	Gastronomy, food and restaurant experiences	Savelli et al. [23]	Typical local foods and food-tourism attractiveness
Neuromarketing/Consumer neuroscience	Gastronomy, food and restaurant experiences	Berčík et al. [92]	Aroma marketing, restaurants and food choice
Neuromarketing/Consumer neuroscience	Gastronomy, food and restaurant experiences	Barrera Duque et al. [93]	Gastronomic and restaurant experience
Neuromarketing/Consumer neuroscience	Gastronomy, food and restaurant experiences	Baydeniz et al. [94]	Food, emotions, hospitality and gastrotourism
Neuromarketing/Consumer neuroscience	Sustainability and ecotourism	Isa and Alsharif [95]	Eco-labels, eco-brands, pricing and ecotourism choices
Neuromarketing/Consumer neuroscience	Heritage, museums and cultural tourism	Aytekin et al. [96]	Virtual museum experience
Neuromarketing/Consumer neuroscience	Heritage, museums and cultural tourism	Cheng et al. [97]	Cultural heritage landscape preferences using ERP
Neuromarketing/Consumer neuroscience	Tourism and hospitality neuroscience	Erkayiran [98]	Current and future use of neuroscience tools in tourism and hospitality
Neuroeconomics	Food-related decision-making	Peng-Li et al. [99]	Ambient noise, food craving and cognitive self-regulation
Neuroeconomics	Tourism pricing and consumption	Wei et al. [100]	Tourist-attraction type, price perception and neural mechanisms in tourism consumption
Neuroeconomics	Destination choice	Mas et al. [101]	Tourism choice and psychological game theory
Neurorobotics	Human–robot service interaction	Pozharliev et al. [102]	Physiological emotional responses to human–robot service interactions in hospitality

Most of the identified applications are associated with neuromarketing and consumer neuroscience and span destination marketing and branding, tourism imagery and destination perception, digital travel interfaces and booking, hospitality and hotel responses, gastronomy, sustainability, heritage and cultural tourism. Neuroeconomic applications are considerably fewer and are concentrated on food-related decision-making, tourism pricing and consumption, and destination choice. Neurorobotics is represented by only one tourism and hospitality application in the corpus, involving physiological responses to human–robot service interaction. This dispersion across neighbouring domains supports the view that neuroscience-based tourism research cannot be adequately delimited by the neurotourism label alone, since relevant empirical work is embedded within broader neuromarketing, consumer neuroscience and related behavioural research [8,12,19]. It also reinforces the interdisciplinary character of the field, in which tourism applications draw on conceptual and methodological traditions that have developed largely outside tourism itself [8,12].

4.3. Comparative Methodological and Substantive Characteristics

The third research question concerned how the principal research domains compare in their methodological approaches and key constructs.

The comparative analysis focused on the three principal research domains represented in the corpus: neuromarketing and consumer neuroscience, neuroeconomics, and neurorobotics. The explicit neurotourism literature was not included in this domain-level comparison because of its limited number of publications and heterogeneous methodological composition, as reported in Section 4.2.1. Across the three principal domains, the abstract-level charting identifies differences in the range of methodological approaches, constructs and application contexts represented (Table 4). This comparison is descriptive and does not establish the relative frequency or predominance of individual characteristics within each domain.

Table 4. Comparative methodological and substantive characteristics of the research domains.

Research Domain	Papers	Study Types Identified	Technologies and Measures Identified	Constructs Identified	Application Areas Identified
Neuromarketing and consumer neuroscience	567	Experimental and quasi-experimental studies; observational studies; physiological measures combined with questionnaires or behavioural tasks; reviews; methodological studies	EEG/ERP; eye tracking; pupillometry; fMRI; EDA/skin conductance; heart rate/HRV; facial EMG; fNIRS; virtual reality; behavioural measures; self-report; machine learning	Attention; visual perception; emotion; arousal; valence; memory; preference; engagement; cognitive load; trust; purchase intention; willingness to pay; brand recognition; decision-making	Advertising; branding; packaging; product design; retail; e-commerce; pricing; food and beverages; sustainability communication; social media; digital interfaces; tourism and hospitality applications
Neuroeconomics	346	Controlled behavioural and neuroimaging experiments; economic games; choice and valuation tasks; computational modelling; observational studies; reviews; theoretical contributions	fMRI; EEG/ERP; behavioural economic tasks; eye tracking; EDA; heart rate; PET; MEG; computational modelling; reinforcement-learning models; psychometric scales	Reward; risk; uncertainty; loss aversion; value computation; intertemporal choice; social preferences; fairness; trust; cooperation; pricing; willingness to pay; financial and consumer decision-making	Finance; investment; gambling; auctions; pricing; product and food choice; public policy; health decisions; social and moral decision-making; tourism pricing and destination choice
Neurorobotics	90	Robotic-system development and validation; experimental testing; human–robot interaction; motor-control and rehabilitation studies; simulation; reviews; conceptual research	Neuro-inspired control systems; artificial neural networks; deep learning; reinforcement learning; robotic platforms; BCI; EEG; EMG; motion capture; computer vision; tactile/proprioceptive sensors; virtual reality; physiological and behavioural measures	Motor learning; sensorimotor control; adaptation; autonomy; agency; embodiment; trust; acceptance; emotional response; cognitive workload; intention recognition; human–robot cooperation	Autonomous, assistive and rehabilitation robotics; prosthetics; navigation; social and collaborative robots; service robots; embodied AI; hospitality service applications

The comparison reveals three distinct but partially overlapping research profiles. Neuromarketing and consumer neuroscience display a comparatively broad range of methodological approaches and application contexts in the abstract-level charting, combining neural, psychophysiological, behavioural and self-report measures across consumer, digital and service contexts. The constructs identified include attention, affective response, preference, memory, trust and decision-related outcomes. This combination of multiple measurement approaches with cognitive, affective and decision-related constructs is consistent with the broad conceptualisation of consumer neuroscience developed by Plassmann et al. [3] and with subsequent methodological work emphasising the integration of neurophysiological measures in consumer research [13]. This methodological breadth differs from the decision-oriented profile identified in neuroeconomics, which encompasses valuation and choice alongside related cognitive and behavioural processes.

Neuroeconomics combines neuroimaging, behavioural economic tasks and computational approaches, with identified constructs including reward, risk, uncertainty, loss aversion, fairness, trust and willingness to pay. Its identified application areas include financial, consumer and social decision-making, as well as tourism-related pricing and destination choice. The identification of valuation, reward, risk and choice is consistent with the decision-centred foundations of neuroeconomics identified in earlier and more recent accounts of the field [5,6]. Moving further towards the technological end of the spectrum, neurorobotics presents a distinct methodological profile involving neuro-inspired and embodied technological systems.

The methodological approaches identified in neurorobotics include neuro-inspired control, robotic systems, brain–computer interfaces and sensorimotor interaction, with constructs such as agency, embodiment, autonomy, adaptation and human–robot cooperation. Its identified applications include autonomous, assistive, rehabilitation and social robotics. Hospitality-related service applications were also identified. This technologically and embodied orientation is consistent with the development of neurorobotics around brain-inspired robotic systems, sensorimotor processes and neural interfaces [10,11].

4.4. Research Gaps and Future Research Directions

The third research question also examined the research gaps and future research directions reported across the four research domains. To maintain a clear distinction between source-reported evidence and review-level interpretation, this analysis was restricted to limitations, unresolved issues, underdeveloped areas and future research needs explicitly identified in the abstracts. Patterns inferred solely from the relative frequency or absence of methods, constructs, populations or applications were not classified as research gaps at this stage.

The four domains differ substantially in corpus size and should therefore not be interpreted as quantitatively equivalent. Neuromarketing and consumer neuroscience, neuroeconomics and neurorobotics provide considerably larger evidence bases, whereas the explicitly identified neurotourism core comprises only eight publications. Neurotourism was nevertheless retained in this analysis because several of these publications explicitly identify limitations concerning the development, interdisciplinarity, methodological accessibility and geographical representation of the field. Table 5 summarises the principal gaps and directions supported by the abstract-level evidence.

Table 5. Research gaps and future research directions explicitly identified in the abstracts across the four research domains.

Research Domain	Domain Corpus (n)	Research Gaps and Limitations Reported in the Abstracts	Future Research Directions Supported by the Abstracts	Illustrative Studies
Neuromarketing and consumer neuroscience	567	Limitations in the prediction and measurement of consumer responses outside controlled settings; methodological inconsistency and reproducibility challenges in neurophysiological research; limitations associated with individual variability, measurement quality and generalisability; insufficient methodological and data diversity in parts of the literature; continuing ethical concerns surrounding neural and neurophysiological data; limited cross-cultural representation in some application areas.	Improve external and ecological validity; integrate complementary neural, physiological, behavioural and self-report measures where appropriate; strengthen methodological harmonisation and reproducibility; increase cross-cultural and longitudinal validation; improve transparency, data diversity and ethical safeguards; extend neurophysiological research to naturalistic consumer, retail and service contexts	Lei et al. [30]; Pamungkas et al. [103]; Rancati et al. [29]; Baydeniz et al. [94]
Neuroeconomics	346	Persistent theoretical, conceptual and methodological challenges in decision neuroscience; unresolved conceptualisation and operationalisation of constructs such as uncertainty; incomplete understanding of neural mechanisms underlying socially and contextually embedded decisions; methodological limitations affecting the interpretation and prediction of economic behaviour; porous conceptual boundaries between neuroeconomics and adjacent areas of decision, affective and social neuroscience.	Refine the conceptual and operational definition of economic decision constructs; strengthen integration between neuroscience, behavioural economics and computational approaches; further investigate social and contextual influences on decision processes; improve the measurement and interpretation of neural predictors; develop research capable of connecting neural mechanisms with increasingly complex forms of economic and social decision-making.	Huettel [4]; Serra [32]; Dennison et al. [5]; Müller et al. [31]
Neuro robotics	90	Limitations in the robustness and spatial or temporal resolution of neurorobotic interfaces; lack of appropriate benchmarks for some robotic applications; limited biological plausibility of brain and musculoskeletal models; high cognitive demands associated with some human–machine interfaces; challenges related to system generalisation, usability and embodied modelling; unresolved ethical and social implications of neurorobotic technologies.	Develop more robust and usable neural interfaces; improve benchmarking and validation across neurorobotic systems; increase the biological plausibility of neural and embodied models; develop multimodal and more intuitive human–machine interfaces; improve reproducibility and scalability of neurorobotic approaches; integrate ethical and social considerations earlier in technological development.	Gardner et al. [35]; Aicardi et al. [34]; Steffen et al. [36]; Tolu et al. [33]; Yang et al. [11]
Neurotourism and tourism neuroscience	8	Slow development of neuroscience research within tourism; limited numbers of well-executed and experimental studies; weak interdisciplinarity; high costs and expertise requirements associated with neuroscientific methods; continued underexploration of neuroscience applications in tourism; fragmentation of the emerging field; geographical concentration of research networks, particularly in the Global North; need for more standardised, reproducible and comparable methodological resources in specific areas.	Strengthen interdisciplinary integration between tourism and neuroscience; increase empirically grounded and experimental research; facilitate methodological accessibility and expertise; consolidate dispersed neuroscience-based tourism knowledge; broaden geographical participation, including stronger representation of the Global South; promote standardisation and reproducibility; expand the empirical foundations required for the consolidation of neurotourism as a research field.	Li et al. [22]; Bağçı [76]; Cardoso and Fraga [19]; Arslan et al. [21]; Güner and Işık [20]

The abstract-level evidence indicates that the four domains are characterised by different types of unresolved research needs rather than by a single common methodological limitation. In neuromarketing and consumer neuroscience, the reported challenges concern both measurement and the translation of neurophysiological evidence into more generalisable accounts of consumer behaviour. Lei et al. [30], for example, identify limitations and future opportunities associated with the prediction of consumer behaviour from neural indicators. More recent work reinforces concerns surrounding methodological consistency, reproducibility and real-world deployment of neurophysiological approaches, including variability across individuals, dataset diversity and ethical safeguards [103]. Rancati et al. [29] similarly position methodological rigour, multimodal measurement and the extension of neurophysiological research to retail and service encounters as priorities for the next stage of neuromarketing research. In food and emotion research, Baydeniz et al. [94], for example, report that approximately three quarters of the reviewed studies relied on Western populations and consequently call for cross-cultural validation, longitudinal investigation and naturalistic research designs.

The pattern is different in neuroeconomics, where the abstracts point more strongly towards unresolved theoretical, conceptual and methodological questions surrounding the study of decision processes. Huettel [4] identifies fundamental challenges associated with established approaches to decision neuroscience, while Serra [32] emphasises both unresolved questions and the porous boundaries separating neuroeconomics from affective, social and broader decision neuroscience. More recent evidence indicates that these challenges have not been fully resolved. Dennison et al. [5] identify continuing theoretical, conceptual, methodological and empirical issues within the field, while Müller et al. [31] highlight difficulties in the conceptualisation of uncertainty and the consistent differentiation of related phenomena. These studies indicate that future development in neuroeconomics depends not only on additional neuroscientific measurement but also on greater conceptual precision and stronger integration between neural mechanisms, behavioural economic models and the social and contextual dimensions of decision-making.

In neurorobotics, the research needs reported in the abstracts are more strongly associated with the technological and embodied development of the field. Gardner et al. [35] identify excessive cognitive load as an unresolved challenge in human–machine interfaces for neurorobotic augmentation and address this through multimodal intention detection. Steffen et al. [36] point to the lack of benchmarks specifically designed for robotic applications of spiking neural networks, whereas Tolu et al. [33] identify limited biological plausibility in existing brain and musculoskeletal models as an obstacle to understanding the mechanisms underlying motor and cognitive deficits. Yang et al. [11] similarly report limitations in the spatial and temporal resolution and robustness of existing neurorobotic interfaces and call for improvements in precision, wearability, ergonomics and real-time interaction. Beyond these technological challenges, Aicardi et al. [34] demonstrate that the emerging field also raises unresolved ethical and social questions. The research frontier identified within the neurorobotics abstracts is therefore primarily concerned with improving robustness, biological plausibility, usability and responsible development of neurorobotic systems.

The neurotourism and tourism-neuroscience evidence reveals a fourth and distinct profile. Despite the small number of publications explicitly associated with this domain, several abstracts identify limitations in the consolidation of neuroscience-based tourism research.

Li et al. [22] describe the growth of neuroscience studies within tourism as relatively slow and identify both a limited number of well-executed studies and weak interdisciplinarity. Bağçı [76] similarly highlights the limited number of experimental studies and the high costs associated with neuroscience tools as constraints on research development. Cardoso and Fraga [19] identify a geographically concentrated collaboration structure dominated by actors and organisations from the Global North and consequently point to the need for stronger inclusion of the Global South. More recent bibliometric evidence characterises neuroscience applications in tourism as still relatively underexplored and dispersed across disciplinary and journal boundaries [21]. At the methodological level, Güner and Işık [20] further emphasise the value of standardised resources for producing more reproducible and comparable findings in neuroscience-based food service and tourism research.

5. Conclusions

The first research question concerned the scientific and thematic patterns identified across the four domains over time. Scientific production increased substantially, while the author-keyword network was centred on neuroeconomics and neuromarketing, with neurorobotics occupying a more peripheral technological position. Thematic evolution showed continuity in these domains alongside increasing specialisation in neuroimaging, machine learning, memory, embodiment and spiking neural networks.

This differentiated structure is particularly important for understanding neurotourism. Within the analytical corpus, its explicitly identified research core is small and methodologically heterogeneous, whereas a much broader body of neuroscience-based tourism and hospitality research is located mainly within neuromarketing and consumer neuroscience, with more limited contributions from neuroeconomics and neurorobotics.

The second research question therefore shows that research relevant to tourism extends well beyond the neurotourism label and remains distributed across neighbouring domains.

The third research question further reveals that these domains are neither methodologically nor substantively equivalent. Neuromarketing and consumer neuroscience combine diverse neural, psychophysiological, behavioural and self-report approaches; neuroeconomics is more closely organised around valuation, reward, risk, uncertainty and choice; and neurorobotics presents a technologically specialised profile centred on neuro-inspired systems, neural interfaces, sensorimotor processes and human–robot interaction. Their reported research needs also differ. Neuromarketing and consumer neuroscience emphasise methodological consistency, generalisability, reproducibility and more naturalistic research; neuroeconomics presents continuing conceptual and methodological challenges; neurorobotics is more strongly concerned with robustness, benchmarking, biological plausibility, usability and ethical and social implications; and neurotourism faces more foundational issues of empirical consolidation, interdisciplinarity, methodological accessibility, geographical representation and reproducibility.

The theoretical contribution of the study lies in showing that these neighbouring neuroscience-based domains are connected without being conceptually or methodologically interchangeable. This distinction is particularly relevant to neurotourism because a substantial part of the knowledge applicable to tourism is situated outside its explicitly identified research core. For academic research, the findings provide a clearer basis for locating tourism-related neuroscience within this wider landscape and for identifying the domains, constructs and methodological traditions with which it already intersects. Beyond these established intersections, future research could also explore reciprocal adaptation across human, technological and environmental systems, as proposed by the Neuro-Adaptive Ecosystems concept.

The practical implications follow from the tourism and hospitality applications identified in the corpus. Neuroscience-based research has been applied to destination marketing and branding, tourism imagery and destination perception, digital travel interfaces and booking, hospitality and hotel responses, gastronomy, sustainability, cultural and heritage tourism, pricing and destination choice. For tourism and hospitality organisations, the findings clarify where these approaches have already been investigated and from which disciplinary traditions they originate. The evidence remains uneven, however, and should be interpreted as a map of existing applications rather than as support for the adoption of methods or technologies.

Several limitations should be considered. The corpus was restricted to English-language articles indexed in Scopus and retrieved using a three-block search strategy, which may have excluded relevant studies that did not explicitly report all three required components in TITLE-ABS-KEY. It should therefore be interpreted as a structured rather than exhaustive representation of the relevant literature. Domain classification relied on titles, abstracts and author keywords and was not mutually exclusive, requiring contextual judgement in ambiguous cases. Abstracts were unavailable for nine publications, and the analysis of research gaps and future directions was restricted to issues explicitly reported in the available abstracts. Domain-specific counts therefore describe representation within this corpus rather than the absolute size of the respective fields. Future research can extend the evidence base through additional databases, languages, document types and, where appropriate, full-text analysis. For neurotourism, the more immediate challenge is to strengthen its empirical foundations while clarifying its relationship with the wider body of neuroscience-based tourism and hospitality research distributed across neighbouring domains.

Appendix A.1. Methodological Transparency and Reproducibility

Table A1. Search strategy and filtering specifications.

Methodological Element	Operational Specification
Database	Scopus
Search field	TITLE-ABS-KEY
Search date	17 July 2026
Search architecture	Three mandatory conceptual blocks connected by the Boolean operator AND: (1) terminology associated with neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics; (2) neural, cognitive and affective mechanisms; and (3) neuroscientific, neurophysiological, psychophysiological and biometric methods.
Full Scopus search syntax	TITLE-ABS-KEY((neuromarketing OR “consumer neuroscience” OR “consumer neuro-science” OR “neuroscience of marketing” OR “neuroscience in marketing” OR “neural consumer research” OR “neurophysiological consumer research” OR neuroeconomic* OR “neuro-economic*” OR “neuroscience of economic decision*” OR “neural basis of economic decision*” OR “neural mechanism* of economic decision*” OR “neural correlate* of economic decision*” OR neurotouris* OR “neuro-touris*” OR “tourism neuroscience” OR “neuroscience in tourism” OR “neuroscience of tourism” OR “neuroscience-based tourism” OR “neurophysiological tourism research” OR “psychophysiological tourism research” OR neurorobot* OR “neuro-robot*” OR “neural robotic*” OR “neuroscience of robotics” OR “neuroscience-inspired robotic*” OR “neuro-inspired robotic*” OR “brain-inspired robotic*”) AND (attention OR perception OR memory OR learning OR emotion* OR affect* OR arousal OR valence OR reward OR motivation OR preference* OR choice OR “decision making” OR “decision-making” OR risk OR trust OR empathy OR agency OR “sense of agency” OR autonomy OR control OR embodiment OR “body ownership” OR engagement OR immersion OR “cognitive load” OR “mental workload” OR “cognitive mechanism*” OR “neural mechanism*” OR “neural response*” OR “neural correlate*” OR “brain activ*” OR “brain response*”) AND (neuroscien* OR neurophysiolog* OR psychophysiolog* OR neuroimaging OR fMRI OR “functional magnetic resonance imaging” OR EEG OR electroencephalograph* OR ERP OR ERPs OR “event-related potential*” OR MEG OR magnetoencephalograph* OR fNIRS OR “functional near-infrared spectroscop*” OR PET OR “positron emission tomography” OR “eye tracking” OR eye-tracking OR eyetracking OR pupillometr* OR “electrodermal activit*” OR “galvanic skin response*” OR “skin conductance” OR “heart rate variability” OR electromyogra* OR “facial electromyograph*” OR “brain-computer interface*” OR “neural signal*” OR “neural recording*”))
Subject areas retained	Neuroscience; Business, Management and Accounting; Psychology; Computer Science; Social Sciences; Economics, Econometrics and Finance; Engineering
Document type	Article
Language	English
Initial retrieval	2154 records
After subject-area and document-type restrictions	1531 records
Analytical corpus after language restriction	981 articles
Post-corpus exclusion	None. Subsequent checking concerned research-domain classification rather than corpus eligibility.

Note: The asterisk (*) is a truncation wildcard used to retrieve terms with different endings.

Table A1. Search strategy and filtering specifications.

Methodological Element	Operational Specification
Database	Scopus
Search field	TITLE-ABS-KEY
Search date	17 July 2026
Search architecture	Three mandatory conceptual blocks connected by the Boolean operator AND: (1) terminology associated with neuromarketing and consumer neuroscience, neuroeconomics, neurotourism and neurorobotics; (2) neural, cognitive and affective mechanisms; and (3) neuroscientific, neurophysiological, psychophysiological and biometric methods.
Full Scopus search syntax	TITLE-ABS-KEY((neuromarketing OR “consumer neuroscience” OR “consumer neuro-science” OR “neuroscience of marketing” OR “neuroscience in marketing” OR “neural consumer research” OR “neurophysiological consumer research” OR neuroeconomic* OR “neuro-economic*” OR “neuroscience of economic decision*” OR “neural basis of economic decision*” OR “neural mechanism* of economic decision*” OR “neural correlate* of economic decision*” OR neurotouris* OR “neuro-touris*” OR “tourism neuroscience” OR “neuroscience in tourism” OR “neuroscience of tourism” OR “neuroscience-based tourism” OR “neurophysiological tourism research” OR “psychophysiological tourism research” OR neurorobot* OR “neuro-robot*” OR “neural robotic*” OR “neuroscience of robotics” OR “neuroscience-inspired robotic*” OR “neuro-inspired robotic*” OR “brain-inspired robotic*”) AND (attention OR perception OR memory OR learning OR emotion* OR affect* OR arousal OR valence OR reward OR motivation OR preference* OR choice OR “decision making” OR “decision-making” OR risk OR trust OR empathy OR agency OR “sense of agency” OR autonomy OR control OR embodiment OR “body ownership” OR engagement OR immersion OR “cognitive load” OR “mental workload” OR “cognitive mechanism*” OR “neural mechanism*” OR “neural response*” OR “neural correlate*” OR “brain activ*” OR “brain response*”) AND (neuroscien* OR neurophysiolog* OR psychophysiolog* OR neuroimaging OR fMRI OR “functional magnetic resonance imaging” OR EEG OR electroencephalograph* OR ERP OR ERPs OR “event-related potential*” OR MEG OR magnetoencephalograph* OR fNIRS OR “functional near-infrared spectroscop*” OR PET OR “positron emission tomography” OR “eye tracking” OR eye-tracking OR eyetracking OR pupillometr* OR “electrodermal activit*” OR “galvanic skin response*” OR “skin conductance” OR “heart rate variability” OR electromyogra* OR “facial electromyograph*” OR “brain-computer interface*” OR “neural signal*” OR “neural recording*”))
Subject areas retained	Neuroscience; Business, Management and Accounting; Psychology; Computer Science; Social Sciences; Economics, Econometrics and Finance; Engineering
Document type	Article
Language	English
Initial retrieval	2154 records
After subject-area and document-type restrictions	1531 records
Analytical corpus after language restriction	981 articles
Post-corpus exclusion	None. Subsequent checking concerned research-domain classification rather than corpus eligibility.

Note: The asterisk (*) is a truncation wildcard used to retrieve terms with different endings.

Appendix A.2. PRISMA-ScR Corpus-Construction Workflow

The identification, filtering and construction of the analytical corpus were documented in accordance with the PRISMA Extension for Scoping Reviews (PRISMA-ScR). PRISMA-ScR was used as the reporting framework for documenting the progression from the initial evidence pool to the final corpus.

Table A2. PRISMA-ScR corpus-construction and filtering workflow.

PRISMA-ScR Reporting Stage	Procedure Applied	Records
Identification	Records retrieved through the Scopus TITLE-ABS-KEY search	2154
Database-level filtering	Restriction to the seven selected subject areas and document type Article	1531
Language restriction	Restriction to publications in English	981
Final corpus	Articles retained for structural evidence mapping and substantive analysis	981
Subsequent classification checks	Ambiguous, overlapping and potentially unsupported domain assignments were examined; these checks concerned domain classification and did not result in removal of publications from the corpus	981

Table A2. PRISMA-ScR corpus-construction and filtering workflow.

PRISMA-ScR Reporting Stage	Procedure Applied	Records
Identification	Records retrieved through the Scopus TITLE-ABS-KEY search	2154
Database-level filtering	Restriction to the seven selected subject areas and document type Article	1531
Language restriction	Restriction to publications in English	981
Final corpus	Articles retained for structural evidence mapping and substantive analysis	981
Subsequent classification checks	Ambiguous, overlapping and potentially unsupported domain assignments were examined; these checks concerned domain classification and did not result in removal of publications from the corpus	981

Accordingly, the 981 publications constituted analytical corpus, and subsequent analytical checks concerned classification within that corpus rather than further eligibility assessment.

Appendix A.3. Structural Evidence-Mapping Specifications

Structural evidence mapping was performed using Bibliometrix/Biblioshiny to characterise the scientific and thematic organisation of the final corpus. Technical specifications are reported below to allow the analytical procedures described in the main text to be assessed independently from their substantive interpretation.

Table A3. Structural evidence-mapping parameters and analytical specifications.

Analytical Component	Operational Specification
Software	Bibliometrix/Biblioshiny
Corpus analysed	Analytical corpus of 981 articles
Annual scientific production	Annual distribution of publications across the corpus
Citation analysis	Scopus citation analysis and identification of the documents with the highest citation counts within the corpus
Co-occurrence field	Authors’ keywords
Keyword pre-processing	Authors’ keywords retained in their original form; no lexical harmonisation was applied
Network normalisation	Association strength
Community detection algorithm	Walktrap
Maximum number of nodes	50
Minimum co-occurrence threshold	2
Isolated nodes	Removed
Repulsion	0.1
Node colour by year	Disabled
Labels	Label-overlap avoidance and text adjustment applied for visual legibility
Centrality measures	Betweenness centrality, closeness centrality and PageRank
Inter-cluster relationships	Interpreted from visible network connections and centrality patterns; aggregated inter-cluster edge weights were not calculated
Thematic-evolution field	Authors’ keywords
Period 1	2002–2020
Period 2	2021–2026
Period definition	The cut-off at 2020 was defined manually on the basis of the publication trajectory observed in the corpus
Thematic-evolution measures	Weighted Inclusion Index, Inclusion Index, occurrences and Stability Index

Table A3. Structural evidence-mapping parameters and analytical specifications.

Analytical Component	Operational Specification
Software	Bibliometrix/Biblioshiny
Corpus analysed	Analytical corpus of 981 articles
Annual scientific production	Annual distribution of publications across the corpus
Citation analysis	Scopus citation analysis and identification of the documents with the highest citation counts within the corpus
Co-occurrence field	Authors’ keywords
Keyword pre-processing	Authors’ keywords retained in their original form; no lexical harmonisation was applied
Network normalisation	Association strength
Community detection algorithm	Walktrap
Maximum number of nodes	50
Minimum co-occurrence threshold	2
Isolated nodes	Removed
Repulsion	0.1
Node colour by year	Disabled
Labels	Label-overlap avoidance and text adjustment applied for visual legibility
Centrality measures	Betweenness centrality, closeness centrality and PageRank
Inter-cluster relationships	Interpreted from visible network connections and centrality patterns; aggregated inter-cluster edge weights were not calculated
Thematic-evolution field	Authors’ keywords
Period 1	2002–2020
Period 2	2021–2026
Period definition	The cut-off at 2020 was defined manually on the basis of the publication trajectory observed in the corpus
Thematic-evolution measures	Weighted Inclusion Index, Inclusion Index, occurrences and Stability Index

Conceptually related but lexically distinct author-keyword variants were therefore retained as separate bibliographic expressions. Centrality and community structure consequently refer to author-reported bibliographic expressions rather than to harmonised concepts.

To assess the sensitivity of the author-keyword network to alternative analytical specifications, the 50-node, threshold-2 network was used as the baseline. Networks combining 40, 50, 60, 75 and 100 nodes with minimum co-occurrence thresholds of 1, 2 and 3 yielded 15 specifications, with isolated nodes removed consistently. Centrality stability was evaluated using Spearman’s rank correlation coefficient (ρ) for betweenness, closeness and PageRank across nodes shared with the baseline. Community stability was examined through the persistence of principal baseline community relationships across the same specifications. Results are reported in Table A4 and Table A5.

Table A4. Sensitivity of author-keyword centrality rankings to alternative network specifications.

Network Specification	Betweenness ρ	Closeness ρ	PageRank ρ
40 nodes/threshold 1	0.749	0.931	0.959
40 nodes/threshold 2	0.971	0.997	0.992
40 nodes/threshold 3	0.755	0.914	0.955
50 nodes/threshold 1	0.839	0.906	0.960
50 nodes/threshold 2 (baseline)	1.000	1.000	1.000
50 nodes/threshold 3	0.765	0.875	0.947
60 nodes/threshold 1	0.831	0.903	0.942
60 nodes/threshold 2	0.987	0.996	0.994
60 nodes/threshold 3	0.766	0.875	0.952
75 nodes/threshold 1	0.836	0.906	0.941
75 nodes/threshold 2	0.986	0.993	0.987
75 nodes/threshold 3	0.760	0.874	0.951
100 nodes/threshold 1	0.823	0.915	0.945
100 nodes/threshold 2	0.970	0.989	0.990
100 nodes/threshold 3	0.767	0.867	0.949

Table A4. Sensitivity of author-keyword centrality rankings to alternative network specifications.

Network Specification	Betweenness ρ	Closeness ρ	PageRank ρ
40 nodes/threshold 1	0.749	0.931	0.959
40 nodes/threshold 2	0.971	0.997	0.992
40 nodes/threshold 3	0.755	0.914	0.955
50 nodes/threshold 1	0.839	0.906	0.960
50 nodes/threshold 2 (baseline)	1.000	1.000	1.000
50 nodes/threshold 3	0.765	0.875	0.947
60 nodes/threshold 1	0.831	0.903	0.942
60 nodes/threshold 2	0.987	0.996	0.994
60 nodes/threshold 3	0.766	0.875	0.952
75 nodes/threshold 1	0.836	0.906	0.941
75 nodes/threshold 2	0.986	0.993	0.987
75 nodes/threshold 3	0.760	0.874	0.951
100 nodes/threshold 1	0.823	0.915	0.945
100 nodes/threshold 2	0.970	0.989	0.990
100 nodes/threshold 3	0.767	0.867	0.949

Table A5. Stability of principal author-keyword community relationships across alternative network specifications.

Baseline Community Relationship Examined	Specifications Preserving the Relationship	Stability
Neuromarketing and consumer neuroscience assigned to the same community	15/15	100%
Neuromarketing and neuroeconomics assigned to distinct communities	15/15	100%
Neuroeconomics, reward and risk assigned to the same community	13/15	86.7%
Neurorobotics, learning and spiking neural networks assigned to the same community	7/15	46.7%

Note: Stability refers to the persistence of the selected baseline community relationships rather than to identity of the complete network partition.

Table A5. Stability of principal author-keyword community relationships across alternative network specifications.

Baseline Community Relationship Examined	Specifications Preserving the Relationship	Stability
Neuromarketing and consumer neuroscience assigned to the same community	15/15	100%
Neuromarketing and neuroeconomics assigned to distinct communities	15/15	100%
Neuroeconomics, reward and risk assigned to the same community	13/15	86.7%
Neurorobotics, learning and spiking neural networks assigned to the same community	7/15	46.7%

Note: Stability refers to the persistence of the selected baseline community relationships rather than to identity of the complete network partition.

Across all 15 specifications, neuromarketing, neuroeconomics and consumer neuroscience remained the three highest-ranked nodes by PageRank. The dominant community distinction was also stable, whereas the smaller technological configuration showed greater parameter sensitivity. Overall, the analysis supports the stability of the principal network hierarchy and dominant community structure, while indicating greater sensitivity for betweenness centrality and peripheral community membership.

To assess the sensitivity of thematic evolution to the manually selected 2020 cut-off, the analysis was repeated using alternative cut-offs of 2019 and 2021. All three specifications retained the original Scopus Author Keywords without lexical harmonisation and the same Bibliometrix settings. The results are presented in Table A6.

Table A6. Sensitivity of thematic-evolution results to alternative temporal cut-offs.

Indicator	2019	2020 (Baseline)	2021
Thematic links identified	30	32	40
Neuroeconomics → Neuroeconomics: occurrences	182	193	202
Weighted Inclusion Index	0.60	0.79	0.97
Neuromarketing → Neuromarketing: occurrences	104	141	172
Weighted Inclusion Index	0.60	0.59	0.72
Neurorobotics → Neurorobotics: occurrences	21	30	31
Weighted Inclusion Index	0.70	0.90	1.00

Note: The arrow (→) indicates the thematic link from the earlier to the later time period.

Table A6. Sensitivity of thematic-evolution results to alternative temporal cut-offs.

Indicator	2019	2020 (Baseline)	2021
Thematic links identified	30	32	40
Neuroeconomics → Neuroeconomics: occurrences	182	193	202
Weighted Inclusion Index	0.60	0.79	0.97
Neuromarketing → Neuromarketing: occurrences	104	141	172
Weighted Inclusion Index	0.60	0.59	0.72
Neurorobotics → Neurorobotics: occurrences	21	30	31
Weighted Inclusion Index	0.70	0.90	1.00

Note: The arrow (→) indicates the thematic link from the earlier to the later time period.

Table A7. Stability of thematic links relative to the 2020 baseline.

Stability Indicator	2019 vs. 2020	2021 vs. 2020
Shared thematic links	15	14
Baseline links preserved	46.9%	43.8%
Jaccard similarity	0.319	0.241

Note: Stability was assessed through the persistence of identical source–destination thematic labels across alternative temporal partitions. Nine thematic links were identified across all three specifications. Occurrence counts and Weighted Inclusion Index values describe the thematic relationships within each temporal partition and are not direct measures of sensitivity between cut-offs.

Table A7. Stability of thematic links relative to the 2020 baseline.

Stability Indicator	2019 vs. 2020	2021 vs. 2020
Shared thematic links	15	14
Baseline links preserved	46.9%	43.8%
Jaccard similarity	0.319	0.241

Note: Stability was assessed through the persistence of identical source–destination thematic labels across alternative temporal partitions. Nine thematic links were identified across all three specifications. Occurrence counts and Weighted Inclusion Index values describe the thematic relationships within each temporal partition and are not direct measures of sensitivity between cut-offs.

Appendix A.4. Python-Assisted Domain Classification

Python-supported rule-based processing was used to assist bibliographic processing and initial assignment of publications to the four research domains. The computational procedure generated candidate classifications that were subsequently examined against the available bibliographic evidence where classification was ambiguous, overlapping or potentially incidental.

Table A8. Operational rules for Python-assisted research-domain classification.

Processing Element	Operational Rule
Input	Bibliographic records from the final 981-article corpus
Bibliographic fields used	Title, abstract and author keywords
Initial computational processing	Bibliographic parsing and rule-based text matching
Matching approach	Domain-specific terminology and regular-expression rules derived from the search strategy
Candidate research domains	Neuromarketing and consumer neuroscience; neuroeconomics; neurotourism; neurorobotics
Domain exclusivity	Non-mutually exclusive; more than one domain assignment was permitted when supported by the bibliographic evidence
Direct domain match	Domain-specific terminology generated a candidate assignment to the corresponding research domain
Ambiguous match	Examined contextually using the title, abstract and author keywords
Overlapping match	Multiple assignments could be retained when the available evidence substantively supported more than one research domain
Potentially incidental match	A lexical occurrence alone was not considered sufficient when the term was used incidentally, metaphorically, technologically or outside the substantive meaning of the research domain
Contextual verification of candidate assignments	Applied to candidate domain assignments rather than eligibility for the final review corpus
Final classification rule	Domain assignment was based on rule-supported detection combined, where required, with contextual validation of the available bibliographic evidence

Table A8. Operational rules for Python-assisted research-domain classification.

Processing Element	Operational Rule
Input	Bibliographic records from the final 981-article corpus
Bibliographic fields used	Title, abstract and author keywords
Initial computational processing	Bibliographic parsing and rule-based text matching
Matching approach	Domain-specific terminology and regular-expression rules derived from the search strategy
Candidate research domains	Neuromarketing and consumer neuroscience; neuroeconomics; neurotourism; neurorobotics
Domain exclusivity	Non-mutually exclusive; more than one domain assignment was permitted when supported by the bibliographic evidence
Direct domain match	Domain-specific terminology generated a candidate assignment to the corresponding research domain
Ambiguous match	Examined contextually using the title, abstract and author keywords
Overlapping match	Multiple assignments could be retained when the available evidence substantively supported more than one research domain
Potentially incidental match	A lexical occurrence alone was not considered sufficient when the term was used incidentally, metaphorically, technologically or outside the substantive meaning of the research domain
Contextual verification of candidate assignments	Applied to candidate domain assignments rather than eligibility for the final review corpus
Final classification rule	Domain assignment was based on rule-supported detection combined, where required, with contextual validation of the available bibliographic evidence

The Python-assisted processing followed the operational instruction:

Identify candidate research-domain assignments using domain-specific terminology derived from the search strategy. Permit multiple assignments when substantively supported. Examine ambiguous, overlapping or potentially incidental matches using the title, abstract and author keywords. Revise a candidate assignment only when the available bibliographic evidence supports the decision. Rejection of a candidate domain assignment must not result in exclusion of the publication from the final review corpus.

For the neurotourism domain, rule-based terminology was not restricted to the literal term neurotourism. It included the predefined expressions incorporated into the search strategy, namely neurotouris*, neuro-touris*, tourism neuroscience, neuroscience in tourism, neuroscience of tourism, neuroscience-based tourism, neurophysiological tourism research and psychophysiological tourism research.

The Python-supported stage therefore provided rule-based candidate classification, whereas contextual examination of ambiguous or potentially unsupported assignments was subsequently supported by the AI-assisted procedure reported in Supplementary Table S1. The same AI-assisted protocol also supported abstract-level charting and organisation of comparative evidence, as specified in Supplementary Table S1. Final classifications remained subject to author validation. Because domain assignment was non-mutually exclusive, domain counts are not additive. They represent the distribution of assignments within the analytical corpus and should not be interpreted as estimates of the absolute size of the respective research fields. Following contextual validation, 952 publications were assigned to one research domain, 28 to two domains, and one to three domains. No publication remained without a validated research-domain assignment. The final validated domain assignments comprised 567 publications in neuromarketing and consumer neuroscience, 346 in neuroeconomics, 90 in neurorobotics, and 8 in neurotourism.

Appendix A.5. Structured Abstract-Level Data Charting Framework

Structured data charting was conducted at the abstract level using variables aligned with the research questions. The principal documentary unit consisted of the title, abstract and author keywords.

Abstracts were available for 972 of the 981 publications. For the remaining nine publications, the available title and author keywords were used for applicable classification and coding. Information that could not be established from the available bibliographic evidence was not inferred.

Table A9. Structured abstract-level data-charting dimensions and decision rules.

Charting dimension	Operational Definition	Decision Rule
Study type and design	Type of study or research design described in the available bibliographic evidence	Code only when supported by the title, abstract or author keywords
Technologies and measures	Neuroscientific, neurophysiological, psychophysiological, biometric, behavioural or related measurement approaches explicitly reported	Do not infer a method from the research topic, domain or construct
Key constructs	Cognitive, affective, neural, physiological or behavioural constructs explicitly investigated	Record only constructs supported by the available bibliographic evidence
Application areas	Empirical, methodological or substantive contexts in which the research was applied	Multiple application areas could be retained when explicitly supported
Tourism and hospitality relevance	Explicit relationship with tourism, hospitality, travel, destinations, gastronomy, heritage, museums or related service contexts	Code independently from scientific-domain membership
Neurotourism-related research core	Publications satisfying the predefined neurotourism and tourism-neuroscience terminology together with contextual evidence	Identification was not restricted to literal occurrence of the term neurotourism
Research gaps	Limitation, unresolved issue or research need explicitly reported in the available abstract	Do not infer from corpus frequency or absence
Future research directions	Explicit recommendation, proposal or need for future research reported in the abstract	Code only when directly supported by the source
Information unavailable	Information relevant to a charting dimension but not stated in the available bibliographic evidence	Record as not reported in the abstract rather than infer absence

Table A9. Structured abstract-level data-charting dimensions and decision rules.

Charting dimension	Operational Definition	Decision Rule
Study type and design	Type of study or research design described in the available bibliographic evidence	Code only when supported by the title, abstract or author keywords
Technologies and measures	Neuroscientific, neurophysiological, psychophysiological, biometric, behavioural or related measurement approaches explicitly reported	Do not infer a method from the research topic, domain or construct
Key constructs	Cognitive, affective, neural, physiological or behavioural constructs explicitly investigated	Record only constructs supported by the available bibliographic evidence
Application areas	Empirical, methodological or substantive contexts in which the research was applied	Multiple application areas could be retained when explicitly supported
Tourism and hospitality relevance	Explicit relationship with tourism, hospitality, travel, destinations, gastronomy, heritage, museums or related service contexts	Code independently from scientific-domain membership
Neurotourism-related research core	Publications satisfying the predefined neurotourism and tourism-neuroscience terminology together with contextual evidence	Identification was not restricted to literal occurrence of the term neurotourism
Research gaps	Limitation, unresolved issue or research need explicitly reported in the available abstract	Do not infer from corpus frequency or absence
Future research directions	Explicit recommendation, proposal or need for future research reported in the abstract	Code only when directly supported by the source
Information unavailable	Information relevant to a charting dimension but not stated in the available bibliographic evidence	Record as not reported in the abstract rather than infer absence

Scientific-domain classification and tourism/hospitality application were treated as analytically distinct dimensions. A publication could therefore remain classified within neuromarketing and consumer neuroscience, neuroeconomics or neurorobotics while simultaneously being coded as having a tourism or hospitality application. The presence of a tourism-related application did not, by itself, result in reclassification of the publication as neurotourism. This distinction supported the separate identification of the neurotourism-related research core and neuroscience-based tourism and hospitality applications located within neighbouring research domains.

A non-inference rule was applied throughout the structured abstract-level charting. When a methodological, substantive or contextual characteristic could not be established from the title, abstract or author keywords, it was treated as not reported in the abstract. Lack of reporting was not interpreted as evidence that the corresponding method, construct, population, procedure or characteristic was absent from the original study. This distinction reflects the documentary scope of the analysis and prevents conclusions about the full study from being derived from information that was not available in the bibliographic record.

Appendix A.6. Source-Reported Research Gaps and Future-Direction Rules

Research gaps and future research directions were charted using a restrictive evidence rule designed to distinguish statements explicitly reported by the source publications from patterns identified subsequently at review level.

Table A10. Operational criteria for source-reported research gaps and future research directions.

Evidence Category	Operational Decision Rule
Explicit limitation	Retain when the abstract explicitly identifies a limitation affecting current knowledge, evidence, theory, methodology or application
Unresolved issue	Retain when the abstract explicitly identifies an unresolved theoretical, methodological, empirical or applied issue
Research need	Retain when the abstract explicitly states that further or additional research is required
Future research direction	Retain when the abstract explicitly proposes, recommends or calls for subsequent research
Low-frequency method or construct	Do not classify as a source-reported gap solely because it appears infrequently in the corpus
Absent or uncommon population, context or application	Do not classify as a source-reported gap solely on the basis of absence or scarcity within the corpus
Review-level pattern	Keep analytically distinct from source-reported research gaps
No explicit statement available	Record as not reported in the abstract

Table A10. Operational criteria for source-reported research gaps and future research directions.

Evidence Category	Operational Decision Rule
Explicit limitation	Retain when the abstract explicitly identifies a limitation affecting current knowledge, evidence, theory, methodology or application
Unresolved issue	Retain when the abstract explicitly identifies an unresolved theoretical, methodological, empirical or applied issue
Research need	Retain when the abstract explicitly states that further or additional research is required
Future research direction	Retain when the abstract explicitly proposes, recommends or calls for subsequent research
Low-frequency method or construct	Do not classify as a source-reported gap solely because it appears infrequently in the corpus
Absent or uncommon population, context or application	Do not classify as a source-reported gap solely on the basis of absence or scarcity within the corpus
Review-level pattern	Keep analytically distinct from source-reported research gaps
No explicit statement available	Record as not reported in the abstract

The governing rule was: a research gap or future research direction was coded as source-reported evidence only when explicitly supported by the abstract. The absence or relative scarcity of a method, construct, population, context or application within the corpus was not, by itself, interpreted as a source-reported research gap. This procedure maintained a distinction between what the reviewed publications explicitly identified as requiring further investigation and patterns that may emerge from comparison of the corpus as a whole.