The Impact of Tourism Experience in Museum Agglomeration Areas on City Image Promotion

Abstract

Based on the stimulus–organism–response (S–O–R) framework, this study explored the psychological spillover mechanism through which tourism experiences in Museum Agglomeration Areas (MAAs) enhance city image and influence behavioral intentions. Structural equation modeling (SEM) based on survey data yielded several key findings. First, information visibility, content visibility, and the quality of amenities and the operational environment played critical roles in shaping tourists’ internal states, including perceived experiential value, affective response, immersion, and satisfaction. In addition, the social atmosphere emerged as an important factor in enriching these evaluations. Second, accessibility and connectivity were identified as factors that reduce friction along the visitor journey, thereby enhancing experiential continuity and immersion. Third, experiential value and immersion were found to be the primary mediators among the internal-state variables, transmitting the effects of environmental stimuli to city-level perceptions and behavioral intentions, such as revisit and recommendation intentions. These findings suggest that the competitiveness of MAAs lies not merely in spatial agglomeration itself but also in their ability to provide engaging and meaningful content, maintain safe and enjoyable operational environments, and develop integrated circulation and information systems. By conceptualizing MAAs as sites of district-scale tourism experiences, this study extends the application of the S–O–R framework to a multi-site urban cultural context and clarifies how differentiated internal states mediate the spillover from district experience to city-level perceptions and behavioral intentions. Rather than proposing a fundamentally new theoretical framework, the study offers a context-specific refinement of the organism layer and provides empirically grounded implications for design and operational strategies in culturally clustered urban districts.

1. Introduction

1.1. Background and Purpose

In the post-industrial era, the paradigm of urban competitiveness has shifted from a focus on industrial infrastructure and tangible assets to an emphasis on intangible resources, such as cultural identity, creativity, international networks, and the influence of cultural industries [1,2,3]. This transformation has encouraged cities to sharpen their strategic focus on gaining competitive advantages through soft power rather than traditional hard power. As a result, cultural attractiveness and symbolic capital have become key drivers of urban development [4,5,6,7]. Amid this broader cultural turn, museums have evolved beyond their traditional functions of collection, preservation, and exhibition to assume the role of urban anchor facilities that reconstruct a city’s sense of place and embody its local cultural identity [8,9,10].

Global cities are increasingly moving away from singular, isolated landmarks toward the development of Museum Agglomeration Areas (MAAs), which strategically cluster museums, galleries, and cultural venues within defined districts [11,12,13,14]. The establishment and development of MAAs enhance urban cultural competitiveness and shape local identity, contributing more than merely a concentration of physical structures [15,16,17]. Individual museums and galleries function not only as independent cultural institutions but also as mediators that amplify cultural symbolism and facilitate the diffusion of a city’s image through mutual linkages and complementarity [18,19,20]. Furthermore, agglomeration areas create contextualized experiences for visitors through inter-institutional interaction and synergy, generating complex cultural value that exceeds the sum of experiences offered by individual facilities [21,22,23].

From the perspectives of urban marketing and place branding, a city’s image is a crucial intangible asset that extends beyond a simple impression and becomes a core component of urban competitiveness [24,25,26]. A city’s image is a multidimensional construct shaped by the integration of cognitive evaluations of its physical and functional attributes with the affective responses individuals hold toward it [27,28,29,30]. Furthermore, from the perspective of the experience economy, visitors perceive a place not merely as a geographical location but also as an object of consumption defined by the unique experiences it offers [31,32]. In this context, MAAs have the potential to function as brand communicators that foster positive perceptions of a city by encapsulating its cultural identity and encouraging visitor engagement.

However, most existing research on MAAs has primarily focused on spatial configuration or the identification of agglomeration economies [33,34] and has tended to be biased toward studies centered on museum service quality and visitor satisfaction [35,36,37,38,39]. As a result, the psychological mechanisms through which the holistic experiences provided by the spatial environment spill over into city image perception remain insufficiently explored. In other words, the differential effects and causal pathways through which experiential factors within MAAs, such as aesthetic atmosphere, intellectual exploration, and social interaction, influence city brand associations have yet to be clearly identified.

Accordingly, this study seeks to empirically examine the structural relationships and underlying mechanisms through which tourism experiences in Museum Agglomeration Areas (MAAs) shape city image. Rather than simply applying an existing S–O–R model to another tourism setting, the study uses the MAA context to theorize a different analytical unit of experience: a district-scale visit chain formed across clustered cultural facilities, public spaces, and urban services. In this perspective, the relevant stimulus is not the service environment of a single museum alone but the combined environmental conditions encountered across institutions and the spaces between them; the organism is not treated as a single undifferentiated reaction but as a layered sequence of evaluative, emotional, attentional, and retrospective states; and the response extends beyond facility-level satisfaction to city-level image formation and behavioral intention.

The empirical focus is Seoul, which was selected as the metropolitan context of the study because it is both a major international tourism destination and one of the most culturally concentrated urban environments in the Republic of Korea. In addition, Seoul contains a dense concentration of museums, galleries, heritage sites, and other cultural facilities that collectively support museum agglomeration at the urban scale. Within Seoul, Jongno-gu exemplifies this condition particularly well, as it combines palaces, museums, traditional streetscapes, cultural facilities, and tourism-oriented commercial areas within a spatially integrated urban core [40]. As the historic center of Seoul and a district long recognized as a focal area of culture and administration, Jongno-gu provides an especially suitable setting for examining museum agglomeration areas as cluster-based urban tourism environments rather than as isolated institutions. In this way, the study treats MAAs not simply as an abstract conceptual construct but as a concrete metropolitan setting in which museums, heritage sites, public spaces, and tourism-oriented commercial functions are spatially integrated to produce district-scale tourism experiences. More specifically, the study contributes in three ways. First, it reconceptualizes MAAs as district-scale tourism environments rather than as mere agglomerations of cultural facilities. Second, it redefines accessibility and connectivity as forms of visit-chain friction rather than as simple measures of transport convenience. Third, it differentiates the organism layer into experiential value, affective response, immersion, and satisfaction in order to clarify how district-level cultural experiences are translated into city-level perceptions and behavioral intentions through distinct psychological pathways. These contributions should be understood as a contextual extension rather than a fundamental reformulation of the S–O–R framework. More specifically, the study uses the MAA setting to refine how the organism layer may be interpreted when the tourism experience unfolds across multiple connected venues and public spaces at the district scale.

1.2. Literature Review

Research on MAAs can be broadly categorized into three primary areas: discussions of spatial structure and planning, examinations of agglomeration effects and performance outcomes, and investigations of service quality and visitor satisfaction in cultural facilities. From an urban planning perspective, studies have explored how physical conditions, such as the layout of cultural facilities, pedestrian networks, and information systems, affect user convenience and movement behavior [11,41,42,43]. In the context of urban economic revitalization, agglomeration efficiency has been explained through performance variables associated with agglomeration economies, including increased visitor numbers, longer lengths of stay, higher levels of consumption, and spillover effects on surrounding commercial areas [44,45,46,47]. Meanwhile, research on cultural facility management and services has primarily focused on users’ evaluations of service quality, exhibition satisfaction, and intentions to revisit specific museums or galleries.

Although these strands of research effectively conceptualize MAAs either as a form of “spatial agglomeration” or as a “mechanism for generating performance outcomes,” they fall short of theorizing and empirically examining the distinctive nature of MAAs as a “district-scale tourism experience.” A visit to an MAA involves more than the experience of a single facility; rather, it constitutes a contextualized experience shaped by continuous interactions, including movement between facilities, time spent on the street, the atmosphere and symbolism encountered along circulation routes, information searching, and itinerary planning [48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64]. However, many existing discussions oversimplify these experiences by focusing solely on satisfaction with individual facilities or by reducing them to the economic outcomes of agglomeration [65,66,67,68,69,70,71,72,73,74,75,76,77]. As a result, they provide only a limited explanation of the experience-based pathways through which specific environmental conditions are translated into city-level perceptions through particular forms of experience.

In the present study, these organism variables are treated as conceptually adjacent but functionally non-equivalent internal states within the visit process. Experiential value refers to a visitor’s evaluative judgment of whether the itinerary is worthwhile in terms of learning, enjoyment, escape, and time–money trade-off. Affective response captures the immediate emotional tone elicited during the visit, such as excitement, comfort, or inspiration. Immersion refers to sustained attentional absorption and the perceived continuity of the experience as visitors move across institutions and public spaces. Satisfaction, by contrast, represents a more retrospective summary appraisal formed after the experience is compared with prior expectations. Accordingly, these constructs are not introduced as fully isolated psychological domains but as related stages that capture different functional moments in the transformation of district-level experience into city-level response. This distinction is especially important in MAA contexts, where visitors do not consume a single bounded service encounter but instead accumulate evaluation, emotion, attention, and post hoc appraisal across a multi-stage urban itinerary.

Conversely, research on city image has consistently identified it as a multidimensional concept combining cognitive evaluations of urban attributes with affective impressions. However, many studies have tended to focus on overall urban experiences or general destination experiences when examining the determinants of image formation. Consequently, the distinctive effects and pathways through which environmental stimuli, such as aesthetic atmosphere, the effectiveness of information and content delivery, the smoothness of movement and connectivity, the operational environment and amenities, and the social atmosphere, influence the cognitive, affective, and behavioral dimensions of city image remain insufficiently examined in contexts such as MAAs, where district-level tourism experience is central [78,79,80,81,82,83,84,85,86]. Therefore, to explain the mechanism through which experiences in MAAs extend into city image, a process model is needed that distinguishes not only between stimuli and responses but also among the different internal states through which city-image spillover is generated. In this respect, the study treats the organism layer as a staged mediating structure rather than as a single global attitudinal response.

In particular, prior research has often oversimplified accessibility and connectivity as matters of physical distance or transportation convenience. However, in the context of MAAs, these variables should be understood as forms of friction experienced within the visit chain, in which the sequence of arrival, movement, transition, and return is repeatedly encountered. In other words, even when public transportation and pedestrian networks are highly convenient, the continuity of the district-level experience may still be weakened and engagement diminished because of interruptions in circulation between facilities, ambiguous wayfinding, and the accumulation of movement burdens. By contrast, when connectivity is seamless and directional cues are clear, visitors are more likely to perceive the entire district as a single, integrated cultural venue. This integrative perception can strengthen experiential value, affective responses, and immersion, thereby increasing the likelihood that these experiences will spill over into city image.

However, the assumption that cultural institutions necessarily benefit from physical concentration has long been questioned. Christopher Alexander [87], in A City is Not a Tree, criticized the tendency of modern planning to gather related cultural functions into a single core simply because they are conceptually associated in the planner’s mind rather than because urban life itself demands such concentration. In this context, he asked, “Does a concert hall ask to be next to an opera house?” and contrasted the dispersed location of Carnegie Hall and the Metropolitan Opera House in Manhattan with the concentrated arrangement of Lincoln Center. This critique suggests that spatial clustering should not be regarded as inherently beneficial, since cultural influence may also arise from differentiated locations whose spheres of activity overlap across the city.

In light of this critique, the present study does not assume that museum agglomeration is universally superior to dispersed cultural development, nor does it argue that all museums should be physically clustered in one place. Rather, it begins from a more limited and empirical question: when an MAA already exists and is experienced by visitors as a district-scale tourism environment, through what psychological pathways do its environmental conditions spill over into city image and behavioral intentions? In this sense, the study does not seek to justify agglomeration as a normative planning principle but to explain the experiential mechanism operating within an existing agglomerated context.

To address this question, this study applies the stimulus–organism–response (S–O–R) framework [88], which posits that environmental stimuli influence internal states and subsequently evoke responses, to the context of tourism experiences in MAAs. Rather than merely transferring a generic S–O–R chain to a new empirical setting, this study specifies how district-level environmental conditions contribute to city image formation through sequential, cluster-based experiences. By integrating the servicescape perspective, which systematizes the physical, operational, and social conditions of MAAs as stimulus factors [89], with the experience economy perspective, which conceptualizes experience as the core object of consumption [90], this research extends prior spatial- and performance-oriented discussions toward a more explicit understanding of the mechanism through which district-level cultural experiences spill over into city-level perceptions.

1.3. Measurement Model Specifications

The measurement model was developed based on the S–O–R framework, but the organism layer was specified in a differentiated manner in order to reflect the multi-stage nature of district-scale tourism experience in MAAs. In this study, experiential value captures evaluative worth, affective response captures in situ emotional tone, immersion captures attentional absorption and experiential continuity, and satisfaction captures an ex post summary appraisal relative to expectations. These constructs are therefore expected to be positively related, but they are not treated as interchangeable. Rather, they represent analytically separable aspects of the visit chain that operate at different functional and temporal points in the experience process. On this basis, they were retained as distinct reflective constructs in the measurement and structural models. These measurement indices are summarized in

Table 1. Latent variables and observed indicators in the S–O–R model. Source: Authors’ compilation based on the reviewed literature.

S–O–R	Latent Variables	Observed Indicators
MAAs environment (stimulus)	Aesthetic Environment (S1)	(S1-1) Atmosphere of architecture, streets, and landscape [78] (S1-2) Visual integrity [43] (S1-3) Attractiveness of remaining outdoors [53]
	Accessibility and Connectivity (S2)	(S2-1) Convenience of public transportation and walkability [41] (S2-2) Continuity of circulation between facilities [48] (S2-3) Clarity of wayfinding [43] (S2-4) Perceived travel burden [62]
	Information and Content Visibility (S3)	(S3-1) Ease of accessing exhibition information [57] (S3-2) Effectiveness of interpretation and storytelling [72] (S3-3) Support for visit planning [64]
	Amenities and Operational Environment (S4)	(S4-1) Adequacy of amenities [58] (S4-2) Perceived safety and comfort [71] (S4-3) Smoothness of operation [81] (S4-4) Convenience of usage procedures [58]
	Social Atmosphere (S5)	(S5-1) On-site vitality [59] (S5-2) Comfort in coexisting with others [71] (S5-3) Potential for cultural interaction [60]
Internal state of tourism experience (organism)	Experiential Value (O1)	(O1-1) Value for time and money [65] (O1-2) Educational value [66] (O1-3) Entertainment value [68] (O1-4) Escapist value [69]
	Affective Response (O2)	(O2-1) Excitement/interest [67] (O2-2) Comfort/stability [71] (O2-3) Impression/inspiration [72]
	Immersion (O3)	(O3-1) Concentration [73] (O3-2) Experiential immersion [74] (O3-3) Experiential flow continuity [76]
	Satisfaction (O4)	(O4-1) Fulfillment of expectations [79] (O4-2) Evaluation of experience quality [81]
Spillover to city image (response)	Cognitive City Image (R1)	(R1-1) Perceived identity of urban culture [78] (R1-2) Perceived attractiveness of cultural and tourism resources [80] (R1-3) Perceived visitor-friendliness [83]
	Affective City Image (R2)	(R2-1) City favorability [81] (R2-2) City familiarity [83] (R2-3) Perceived city vitality [78]
	Behavioral Intentions (R3)	(R3-1) Intention to revisit the city [81] (R3-2) Intention to recommend the agglomeration area [79] (R3-3) Intention to explore cultural activities [86]

and illustrated in Figure 1. In addition, because the environmental stimuli of MAAs operate as an integrated system rather than as isolated factors, the latent environmental stimulus variables were specified as interrelated constructs in the structural model.

2. Materials and Methods

Based on survey data collected in China from respondents who had visited MAAs in Jongno-gu, Seoul, within the past three years, this study estimated the structural relationships among the latent variables using structural equation modeling (SEM). The questionnaire consisted of three parts: (1) an informed consent form, (2) screening questions on demographic characteristics and prior MAA experience, and (3) the main measurement items arranged in the order of stimulus–organism–response. Before participating, respondents were presented with an informed consent statement outlining the purpose of the research, the anonymity of the responses, the voluntary nature of participation, their right to withdraw at any time, and the scope of data use. Only those who agreed to the consent form were allowed to proceed. To ensure sample eligibility, respondents were asked whether they had visited an MAA in Jongno-gu, Seoul, within the past three years; those who responded negatively were automatically screened out. Eligible respondents were then asked how often they had visited such areas during the same period, thereby ensuring the collection of recent, experience-based data suitable for analysis.

The measurement items were developed based on the indicators listed in Table 1. Each item consistently referred to MAAs rather than to individual cultural facilities. All constructs were treated as reflective measurement models, in which latent variables are represented by their corresponding observed indicators. All stimulus (S), organism (O), and response (R) items were measured using a five-point Likert scale, with differentiated response anchors depending on their purpose. Specifically, stimulus (S) and response (R) items were measured on agreement-based scales ranging from 1 = strongly disagree to 5 = strongly agree. In contrast, organism (O) items were designed to reflect the intensity and level of visitors’ internal states more accurately and were therefore phrased as degree-centered statements measured on intensity scales ranging from 1 = very low to 5 = very high. Furthermore, the items related to S2-4 (perceived travel burden) were negatively worded and reverse-coded during the survey design phase. This approach ensured that all item scores were directionally aligned, with higher values consistently indicating more positive evaluations, thereby enhancing interpretive consistency and scale coherence. After data collection was completed, the analysis was conducted according to the following procedures.

First, the survey data were screened for missing values and outliers. Inattentive or careless responses were identified and excluded from the final dataset. In addition, descriptive statistics were generated for the screening items measuring demographic characteristics and MAA visitation experience in order to examine the basic distributional properties of the sample.

Second, reliability analyses were conducted for each latent construct to assess the internal consistency of the measurement items. Specifically, Cronbach’s α coefficients were calculated to evaluate construct-level internal consistency. Following the confirmatory factor analysis, composite reliability (CR) was computed to further verify the reliability of the latent variables.

Third, confirmatory factor analysis (CFA) was performed to examine the validity of the measurement model. Model fit was evaluated using several goodness-of-fit indices, including χ²/df, CFI, TLI, RMSEA, and GFI. Convergent validity was assessed using standardized factor loadings, average variance extracted (AVE), and composite reliability (CR). Discriminant validity was examined using the Fornell–Larcker criterion.

Fourth, after confirming the fit, reliability, and validity of the measurement model, the structural model was estimated to test the path coefficients among the latent variables. The structural model was evaluated using the same fit indices as the measurement model to assess its overall explanatory power, and the statistical significance of each path was examined.

Fifth, after statistical significance had been verified, the total indirect effects along the stimulus–organism–response pathways specified in the structural equation model were calculated. The total effects for each structural path were then compared in order to derive the study’s conclusions.

To carry out these procedures, online survey data were collected using the SoJump platform [91], a widely used web-based survey tool that allowed screening logic, consent confirmation, and anonymous response collection. Because participation was voluntary and web-based, the final dataset should be interpreted as an experience-qualified nonprobability sample rather than a statistically representative sample of all tourists or all urban residents. The data were analyzed using IBM SPSS (Statistics 25 and Amos 26).

3. Results

3.1. Respondent Profile

The survey was conducted between 20 October and 15 December 2025, during which a total of 1143 responses were collected. After excluding 107 invalid responses, 1036 valid responses were retained for the final analysis.

Table 2. Demographic characteristics and visit frequency to MAAs in Seoul of respondents (N = 1036). Source: Authors’ calculations based on survey data.

Category		Frequency	Ratio
Gender	Male	512	49.4%
	Female	524	50.6%
Age group	18–24	272	26.3%
	25–34	249	24.0%
	35–44	262	25.3%
	45 and older	253	24.4%
Number of visits to MAAs in Seoul in the past 3 years	1	404	39.0%
	2	298	28.8%
	3 or more visits	334	32.2%

presents the demographic characteristics of the respondents and their visitation frequency to MAAs in Seoul over the past three years. In terms of gender, 512 respondents (49.4%) were male and 524 (50.6%) were female. The age distribution was as follows: 18–24 years (n = 272, 26.3%), 25–34 years (n = 249, 24.0%), 35–44 years (n = 262, 25.3%), and 45 years and older (n = 253, 24.4%). These results indicate that the valid sample was relatively balanced across gender and age groups, with no excessive concentration in any particular demographic category. However, because the survey was administered online to respondents with prior MAA experience, the sample should not be interpreted as statistically representative of all visitors to cultural districts. Rather, it is more appropriately understood as an experience-qualified sample suitable for explaining experiential relationships in a culturally dense metropolitan context.

Regarding visitation frequency during the past three years, 404 respondents (39.0%) reported one visit, whereas 632 respondents (61.0%) reported two or more visits. These results suggest that most respondents had repeated visitation experiences rather than a single incidental visit. This supports the assumption that the evaluations of environmental stimuli and tourism experiences in this study were based on relatively recent and actual experiences. At the same time, the empirical setting should be understood as Seoul’s culturally dense metropolitan MAA environment, which is analytically suitable for examining district-scale spillover mechanisms, although the findings should be generalized cautiously to smaller cities or less clustered cultural districts.

3.2. Reliability Test

Table 3. Reliability analysis of the measurement items. Source: Authors’ calculations based on survey data.

Category		Mean	Min	Max	Range	Maximum/ Minimum	Variance	Cronbach’s Alpha
Aesthetic Environment (S1)	Item mean	4.092	4.052	4.115	0.063	1.015	0.001	0.898
	Item variance	1.203	1.178	1.220	0.042	1.036	0.000
	Inter-item covariance	0.897	0.879	0.909	0.030	1.034	0.000
	Inter-item correlation	0.745	0.733	0.761	0.028	1.038	0.000
Accessibility and Connectivity (S2)	Item mean	4.112	4.085	4.128	0.043	1.011	0.000	0.914
	Item variance	1.135	1.086	1.173	0.088	1.081	0.001
	Inter-item covariance	0.824	0.801	0.843	0.042	1.052	0.000
	Inter-item correlation	0.726	0.708	0.744	0.035	1.050	0.000
Information and Content Visibility (S3)	Item mean	4.126	4.097	4.158	0.061	1.015	0.001	0.894
	Item variance	1.119	1.103	1.147	0.043	1.039	0.001
	Inter-item covariance	0.826	0.800	0.843	0.043	1.053	0.000
	Inter-item correlation	0.738	0.723	0.749	0.026	1.036	0.000
Amenities and Operational Environment (S4)	Item mean	4.123	4.110	4.135	0.025	1.006	0.000	0.910
	Item variance	1.099	1.035	1.142	0.106	1.103	0.002
	Inter-item covariance	0.786	0.746	0.816	0.070	1.095	0.001
	Inter-item correlation	0.716	0.700	0.728	0.029	1.041	0.000
Social Atmosphere (S5)	Item mean	4.110	4.098	4.125	0.026	1.006	0.000	0.886
	Item variance	1.133	1.096	1.163	0.068	1.062	0.001
	Inter-item covariance	0.817	0.778	0.838	0.060	1.077	0.001
	Inter-item correlation	0.722	0.696	0.742	0.046	1.066	0.000
Experiential Value (O1)	Item mean	4.146	4.131	4.159	0.028	1.007	0.000	0.903
	Item variance	1.060	0.977	1.125	0.148	1.152	0.004
	Inter-item covariance	0.740	0.704	0.775	0.071	1.100	0.001
	Inter-item correlation	0.699	0.681	0.707	0.027	1.039	0.000
Affective Response (O2)	Item mean	4.119	4.101	4.146	0.044	1.011	0.001	0.889
	Item variance	1.129	1.099	1.158	0.059	1.054	0.001
	Inter-item covariance	0.822	0.810	0.831	0.021	1.025	0.000
	Inter-item correlation	0.728	0.726	0.731	0.005	1.007	0.000
Immersion (O3)	Item mean	4.118	4.094	4.145	0.051	1.012	0.001	0.882
	Item variance	1.127	1.112	1.150	0.037	1.034	0.000
	Inter-item covariance	0.805	0.791	0.814	0.023	1.030	0.000
	Inter-item correlation	0.714	0.709	0.718	0.009	1.013	0.000
Satisfaction (O4)	Item mean	4.020	4.013	4.027	0.014	1.004	0.000	0.857
	Item variance	1.339	1.317	1.361	0.044	1.033	0.001
	Inter-item covariance	1.004	1.004	1.004	0.000	1.000	0.000
	Inter-item correlation	0.750	0.750	0.750	0.000	1.000	0.000
Cognitive City Image (R1)	Item mean	4.062	4.054	4.074	0.020	1.005	0.000	0.894
	Item variance	1.212	1.194	1.240	0.047	1.039	0.001
	Inter-item covariance	0.893	0.886	0.904	0.018	1.021	0.000
	Inter-item correlation	0.737	0.726	0.743	0.017	1.024	0.000
Affective City Image (R2)	Item mean	4.133	4.116	4.161	0.045	1.011	0.001	0.891
	Item variance	1.138	1.082	1.174	0.092	1.085	0.002
	Inter-item covariance	0.832	0.802	0.852	0.051	1.063	0.001
	Inter-item correlation	0.731	0.716	0.747	0.031	1.043	0.000
Behavioral Intentions (R3)	Item mean	4.120	4.089	4.138	0.049	1.012	0.001	0.887
	Item variance	1.138	1.122	1.149	0.027	1.024	0.000
	Inter-item covariance	0.822	0.819	0.827	0.008	1.010	0.000
	Inter-item correlation	0.723	0.722	0.724	0.001	1.002	0.000

presents the results of the descriptive statistics and reliability analyses for the measurement items. Examination of the item means for each latent construct showed that the S factors ranged from 4.092 to 4.126, indicating a relatively high overall evaluation of the environmental stimuli in MAAs (S1 = 4.092, S2 = 4.112, S3 = 4.126, S4 = 4.123, and S5 = 4.110). The O factors also showed generally high mean values, with experiential value (O1) recording the highest mean (4.146), followed by affective response (O2 = 4.119), immersion (O3 = 4.118), and satisfaction (O4 = 4.020). For the R factors, the item means were as follows: cognitive city image (R1 = 4.062), affective city image (R2 = 4.133), and behavioral intentions (R3 = 4.120). These findings suggest that MAA visit experiences are positively associated with city-level perceptions and behavioral intentions. Although experiential value (O1) recorded the highest mean among the internal-state factors, satisfaction (O4) showed a relatively lower mean. This pattern implies that, while visitors tend to perceive their MAA visit as a valuable experience, their overall satisfaction judgments, particularly in relation to prior expectations, may be somewhat more cautious.

In terms of reliability, Cronbach’s α coefficients for the S factors were 0.898 for S1, 0.914 for S2, 0.894 for S3, 0.910 for S4, and 0.886 for S5, all of which exceeded the commonly accepted threshold of 0.70. For the O factors, the corresponding values were 0.903 for O1, 0.889 for O2, 0.882 for O3, and 0.857 for O4. For the R factors, the values were 0.894 for R1, 0.891 for R2, and 0.887 for R3. These results indicate that the items representing each latent construct demonstrate strong internal consistency, thereby providing sufficient measurement reliability to support the subsequent CFA and SEM analyses used to estimate and interpret the structural relationships among the latent variables.

3.3. Measurement Model Validity (CFA)

The results of the convergent validity and construct reliability assessments (

Table 4. Convergent validity metrics. Source: Authors’ calculations based on survey data.

Relationships			Estimate	AVE	CR
Observed Indicator	Path	Latent Variable
S1-1	←	Aesthetic Environment (S1)	0.868	0.745	0.898
S1-2	←		0.868
S1-3	←		0.854
S2-1	←	Accessibility and Connectivity (S2)	0.858	0.726	0.914
S2-2	←		0.855
S2-3	←		0.854
S2-4	←		0.841
S3-1	←	Information and Content Visibility (S3)	0.853	0.738	0.894
S3-2	←		0.846
S3-3	←		0.877
S4-1	←	Amenities and Operational Environment (S4)	0.858	0.715	0.910
S4-2	←		0.842
S4-3	←		0.840
S4-4	←		0.843
S5-1	←	Social Atmosphere (S5)	0.842	0.722	0.886
S5-2	←		0.836
S5-3	←		0.870
O1-1	←	Experiential Value (O1)	0.836	0.694	0.901
O1-2	←		0.827
O1-3	←		0.837
O1-4	←		0.832
O2-1	←	Affective Response (O2)	0.841	0.723	0.887
O2-2	←		0.851
O2-3	←		0.858
O3-1	←	Immersion (O3)	0.847	0.709	0.879
O3-2	←		0.850
O3-3	←		0.828
O4-1	←	Satisfaction (O4)	0.874	0.746	0.854
O4-2	←		0.853
R1-1	←	Cognitive City Image (R1)	0.850	0.735	0.893
R1-2	←		0.865
R1-3	←		0.857
R2-1	←	Affective City Image (R2)	0.853	0.729	0.890
R2-2	←		0.871
R2-3	←		0.837
R3-1	←	Behavioral Intentions (R3)	0.847	0.720	0.885
R3-2	←		0.850
R3-3	←		0.848

) indicate that the standardized factor loadings of the observed variables ranged from 0.827 to 0.877, demonstrating a stable and robust overall pattern. These values substantially exceed the commonly recommended threshold (≥0.70), suggesting that each item adequately represents its corresponding latent construct. In addition, the AVE values ranged from 0.694 to 0.746, all of which exceed the recommended threshold of 0.50, confirming that each latent construct explains a substantial proportion of the variance in its indicators. Construct reliability (CR) values ranged from 0.854 to 0.914, indicating that each construct maintains a high level of internal consistency. Because the factor loadings, AVE, and CR values all meet or exceed the recommended standards, the measurement model can be considered to demonstrate adequate convergent validity and construct reliability.

Next, discriminant validity was examined using the Fornell–Larcker criterion (

Table 5. Fornell–Larcker discriminant validity matrix. Source: Authors’ calculations based on survey data.

Diagonal (AVE)	S1	S2	S3	S4	S5	O1	O2	O3	O4	R1	R2	R3
S1	0.745
S2	0.570	0.726
S3	0.585	0.585	0.738
S4	0.580	0.583	0.610	0.715
S5	0.565	0.590	0.621	0.618	0.722
O1	0.168	0.149	0.217	0.194	0.210	0.694
O2	0.139	0.166	0.228	0.194	0.197	0.000	0.723
O3	0.129	0.180	0.214	0.228	0.149	0.000	0.000	0.709
O4	0.098	0.152	0.150	0.192	0.164	0.000	0.000	0.000	0.746
R1	0.000	0.000	0.000	0.000	0.000	0.285	0.192	0.248	0.115	0.735
R2	0.000	0.000	0.000	0.000	0.000	0.300	0.226	0.274	0.115	0.000	0.729
R3	0.000	0.000	0.000	0.000	0.000	0.324	0.235	0.201	0.148	0.000	0.000	0.720
Square root of AVE	0.863	0.852	0.859	0.846	0.850	0.833	0.850	0.842	0.864	0.857	0.854	0.849

). The square roots of the AVE values for each construct ranged from 0.833 to 0.864, indicating that each construct explains more variance in its own indicators than it shares with other constructs. The minimum square root of AVE (0.833) was greater than the inter-construct correlation coefficients, confirming that the constructs are empirically distinct from one another. Therefore, the measurement model demonstrates adequate discriminant validity, indicating that the constructs do not overlap excessively and allowing the structural relationships among the latent variables in the SEM to be interpreted with confidence.

3.4. Structural Model Fit

Table 6. Overall model fit indices. Source: Authors’ calculations based on survey data.

χ2/df	RMR	RMSEA	GFI	CFI	NFI	TLI
1.462	0.032	0.021	0.953	0.990	0.970	0.989

presents the overall fit indices of the structural model. The model fit statistics were as follows: χ²/df = 1.462, RMR = 0.032, RMSEA = 0.021, GFI = 0.953, CFI = 0.990, NFI = 0.970, and TLI = 0.989, indicating an acceptable level of model fit. Accordingly, the structural relationships among the latent variables were interpreted based on the structural paths presented in

Table 7. Structural path estimates for the latent variable relationships. Source: Authors’ calculations based on survey data.

Relationships			Estimate	S.E.	C.R.	p Values
O1	←	S1	0.168	0.032	4.797	0.000 ***
O2	←		0.139	0.034	3.844	0.000 ***
O3	←		0.129	0.035	3.470	0.000 ***
O4	←		0.098	0.046	2.299	0.021 *
O1	←	S2	0.149	0.033	4.241	0.000 ***
O2	←		0.166	0.035	4.575	0.000 ***
O3	←		0.180	0.037	4.805	0.000 ***
O4	←		0.152	0.047	3.549	0.000 ***
O1	←	S3	0.217	0.037	5.759	0.000 ***
O2	←		0.228	0.039	5.840	0.000 ***
O3	←		0.214	0.041	5.320	0.000 ***
O4	←		0.150	0.052	3.266	0.001 **
O1	←	S4	0.194	0.035	5.256	0.000 ***
O2	←		0.194	0.037	5.110	0.000 ***
O3	←		0.228	0.039	5.795	0.000 ***
O4	←		0.192	0.050	4.267	0.000 ***
O1	←	S5	0.210	0.038	5.547	0.000 ***
O2	←		0.197	0.040	5.047	0.000 ***
O3	←		0.149	0.041	3.709	0.000 ***
O4	←		0.164	0.053	3.571	0.000 ***
R1	←	O1	0.285	0.042	7.214	0.000 ***
R2	←		0.300	0.037	8.079	0.000 ***
R3	←		0.324	0.039	8.562	0.000 ***
R1	←	O2	0.192	0.041	4.911	0.000 ***
R2	←		0.226	0.036	6.163	0.000 ***
R3	←		0.235	0.038	6.302	0.000 ***
R1	←	O3	0.248	0.040	6.416	0.000 ***
R2	←		0.274	0.035	7.532	0.000 ***
R3	←		0.201	0.036	5.501	0.000 ***
R1	←	O4	0.115	0.032	3.304	0.000 ***
R2	←		0.115	0.028	3.539	0.000 ***
R3	←		0.148	0.029	4.472	0.000 ***
S1	↔	S2	0.570	0.036	13.949	0.000 ***
	↔	S3	0.585	0.036	14.079	0.000 ***
	↔	S4	0.580	0.036	14.094	0.000 ***
	↔	S5	0.565	0.035	13.648	0.000 ***
S2	↔	S3	0.585	0.034	14.118	0.000 ***
	↔	S4	0.583	0.035	14.179	0.000 ***
	↔	S5	0.590	0.034	14.106	0.000 ***
S3	↔	S4	0.610	0.035	14.506	0.000 ***
	↔	S5	0.621	0.034	14.481	0.000 ***
S4	↔	S5	0.618	0.034	14.537	0.000 ***

* p < 0.05, ** p < 0.01, *** p < 0.001.

and Figure 2.

The MAA environmental stimulus (S) factors exerted significant positive effects on the organism (O) factors representing the internal states of tourists. Specifically, the aesthetic environment (S1) significantly influenced experiential value (O1, Estimate = 0.168, S.E. = 0.032, C.R. = 4.797, p < 0.001), affective response (O2, Estimate = 0.139, S.E. = 0.034, C.R. = 3.844, p < 0.001), immersion (O3, Estimate = 0.129, S.E. = 0.035, C.R. = 3.470, p < 0.001), and satisfaction (O4, Estimate = 0.098, S.E. = 0.046, C.R. = 2.299, p < 0.05). Notably, although the path to satisfaction (O4) was statistically significant, its magnitude was relatively weaker than that of the other paths. In addition, accessibility and connectivity (S2) were found to have significant positive effects on all organism factors: experiential value (O1, Estimate = 0.149, S.E. = 0.033, C.R. = 4.241, p < 0.001), affective response (O2, Estimate = 0.166, S.E. = 0.035, C.R. = 4.575, p < 0.001), immersion (O3, Estimate = 0.180, S.E. = 0.037, C.R. = 4.805, p < 0.001), and satisfaction (O4, Estimate = 0.152, S.E. = 0.047, C.R. = 3.549, p < 0.001).

Information and content visibility (S3) also showed relatively large standardized estimates across all organism factors, including experiential value (O1, Estimate = 0.217, S.E. = 0.037, C.R. = 5.759, p < 0.001), affective response (O2, Estimate = 0.228, S.E. = 0.039, C.R. = 5.840, p < 0.001), immersion (O3, Estimate = 0.214, S.E. = 0.041, C.R. = 5.320, p < 0.001), and satisfaction (O4, Estimate = 0.150, S.E. = 0.052, C.R. = 3.266, p < 0.01). These findings suggest that the accessibility of exhibition information, the effectiveness of interpretation and storytelling, and support for visit planning function as core stimuli within MAAs, enhancing perceived value, affective responses, immersion, and satisfaction. Similarly, amenities and the operational environment (S4) significantly influenced experiential value (O1, Estimate = 0.194, S.E. = 0.035, C.R. = 5.256, p < 0.001), affective response (O2, Estimate = 0.194, S.E. = 0.037, C.R. = 5.110, p < 0.001), immersion (O3, Estimate = 0.228, S.E. = 0.039, C.R. = 5.795, p < 0.001), and satisfaction (O4, Estimate = 0.192, S.E. = 0.050, C.R. = 4.267, p < 0.001). In addition, the social atmosphere (S5) had significant positive effects on experiential value (O1, Estimate = 0.210, S.E. = 0.038, C.R. = 5.547, p < 0.001), affective response (O2, Estimate = 0.197, S.E. = 0.040, C.R. = 5.047, p < 0.001), immersion (O3, Estimate = 0.149, S.E. = 0.041, C.R. = 3.709, p < 0.001), and satisfaction (O4, Estimate = 0.164, S.E. = 0.053, C.R. = 3.571, p < 0.001). These results support the view that on-site vitality, comfort in coexistence, and opportunities for interaction reinforce value judgments and emotional engagement.

The organism (O) factors also exerted significant positive effects on city image and behavioral intentions (R). Experiential value (O1), in particular, significantly influenced cognitive city image (R1, Estimate = 0.285, S.E. = 0.042, C.R. = 7.214, p < 0.001), affective city image (R2, Estimate = 0.300, S.E. = 0.037, C.R. = 8.079, p < 0.001), and behavioral intentions (R3, Estimate = 0.324, S.E. = 0.039, C.R. = 8.562, p < 0.001), suggesting that experiential value may serve as a key mediating mechanism through which MAA experiences are translated into city-level perceptions and future behavioral intentions. Affective response (O2) also had significant effects on R1 (Estimate = 0.192, S.E. = 0.041, C.R. = 4.911, p < 0.001), R2 (Estimate = 0.226, S.E. = 0.036, C.R. = 6.163, p < 0.001), and R3 (Estimate = 0.235, S.E. = 0.038, C.R. = 6.302, p < 0.001). Likewise, immersion (O3) significantly influenced R1 (Estimate = 0.248, S.E. = 0.040, C.R. = 6.416, p < 0.001), R2 (Estimate = 0.274, S.E. = 0.035, C.R. = 7.532, p < 0.001), and R3 (Estimate = 0.201, S.E. = 0.036, C.R. = 5.501, p < 0.001). In contrast, although satisfaction (O4) significantly influenced R1 (Estimate = 0.115, S.E. = 0.032, C.R. = 3.304, p < 0.001), R2 (Estimate = 0.115, S.E. = 0.028, C.R. = 3.539, p < 0.001), and R3 (Estimate = 0.148, S.E. = 0.029, C.R. = 4.472, p < 0.001), the relatively small magnitude of these estimates indicates that its spillover effect is more limited than those of the other organism factors.

In addition, all covariances among the environmental stimulus factors were found to be statistically significant. This suggests that the environmental conditions of MAAs are not perceived as independent and isolated elements but rather as an interconnected system of stimuli. These findings empirically support the model assumption of this study, which conceptualized the stimulus factors as correlated exogenous variables.

3.5. Validation of Effects Along S–O–R Pathways

This study examined how environmental stimuli in MAAs spill over into city image and behavioral intentions through tourists’ internal states based on the S–O–R framework. First, the significance of the S→O and O→R paths specified in the structural model was confirmed. Next, the relative contributions of the total indirect effects and the specific indirect effects along each mediating path were compared using the effect decomposition results (

Table 8. Total and indirect effects according to S–O–R pathways. Source: Authors’ calculations based on survey data.

Relationships				Total Effect	Total Indirect Effect
Predictor (S)	Mediator (O)		Outcome (R)
Aesthetic Environment (S1)	O1	→	R1	0.048	0.118
	O2	→		0.027
	O3	→		0.032
	O4	→		0.011
	O1	→	R2	0.050	0.128
	O2	→		0.031
	O3	→		0.035
	O4	→		0.011
	O1	→	R3	0.054	0.128
	O2	→		0.033
	O3	→		0.026
	O4	→		0.015
Accessibility and Connectivity (S2)	O1	→	R1	0.042	0.136
	O2	→		0.032
	O3	→		0.045
	O4	→		0.017
	O1	→	R2	0.045	0.149
	O2	→		0.038
	O3	→		0.049
	O4	→		0.017
	O1	→	R3	0.048	0.146
	O2	→		0.039
	O3	→		0.036
	O4	→		0.022
Information and Content Visibility (S3)	O1	→	R1	0.062	0.176
	O2	→		0.044
	O3	→		0.053
	O4	→		0.017
	O1	→	R2	0.065	0.193
	O2	→		0.052
	O3	→		0.059
	O4	→		0.017
	O1	→	R3	0.070	0.189
	O2	→		0.054
	O3	→		0.043
	O4	→		0.022
Amenities and Operational Environment (S4)	O1	→	R1	0.055	0.171
	O2	→		0.037
	O3	→		0.057
	O4	→		0.022
	O1	→	R2	0.058	0.187
	O2	→		0.044
	O3	→		0.062
	O4	→		0.022
	O1	→	R3	0.063	0.183
	O2	→		0.046
	O3	→		0.046
	O4	→		0.028
Social Atmosphere (S5)	O1	→	R1	0.060	0.153
	O2	→		0.038
	O3	→		0.037
	O4	→		0.019
	O1	→	R2	0.063	0.167
	O2	→		0.045
	O3	→		0.041
	O4	→		0.019
	O1	→	R3	0.068	0.169
	O2	→		0.046
	O3	→		0.030
	O4	→		0.024

). Because the study model does not include direct S→R paths, the effects of environmental stimuli on the response variables are fully mediated by internal states. Therefore, in this model, the total effects are equivalent to the total indirect effects.

The analysis showed that the total indirect effects on cognitive city image (R1) were relatively strong for information and content visibility (S3) and for amenities and the operational environment (S4). These findings suggest that information accessibility, interpretation, visit planning support, safety, comfort, and smooth operation play significant roles in shaping cognitive-level evaluations of a city. Similar patterns were observed for affective city image (R2) and behavioral intentions (R3), indicating that S3 and S4 significantly influence the extent to which MAA experiences are translated into city favorability, familiarity, vitality, and intentions to revisit, recommend, and explore cultural activities. It should be noted that these estimates were derived from a model incorporating covariances among the stimulus factors and should therefore be interpreted as partial effects controlling for the interrelationships among the stimuli.

On the other hand, an examination of the mediating pathways showed that the specific indirect effects through experiential value (O1) made the largest overall contribution. This suggests that evaluations related to value for time and cost, as well as educational, entertainment, and escapist value, serve as key mediators through which MAA experiences spill over into city-level perceptions and behavioral intentions. Immersion (O3) was the next most influential mediator. In particular, amenities and the operational environment (S4) were found to enhance immersion, which in turn spilled over into city image, highlighting the relative importance of this pathway. Affective response (O2) also functioned as a significant mediator throughout the model, although its effect size was generally moderate. Although satisfaction (O4) also had a significant mediating effect, its specific indirect effect was relatively small, indicating that it had less mediating power than the value- and immersion-centered pathways.

Overall, the environmental stimuli of MAAs form an S–O–R spillover structure in which tourists’ internal states sequentially mediate the influence of environmental conditions on city image and behavioral intentions. From the perspective of total indirect effects, information and content visibility (S3) and amenities and the operational environment (S4) were confirmed as the most influential contributors. Among the mediators, experiential value (O1) and immersion (O3) were identified as the primary pathways driving the spillover process.

4. Discussion

4.1. Formation of Tourism Experience in Museum Agglomeration Areas and the Mechanism of Spillover into City Image

The theoretical contribution of this study should be understood in a modest and context-specific sense. Rather than fundamentally advancing the stimulus–organism–response (S–O–R) framework, this study refines its application by extending it from a venue-based servicescape setting to a district-scale, multi-stage tourism context. Instead of viewing a museum visit as a single-venue encounter, the findings suggest that Museum Agglomeration Areas (MAAs) function as sequential public-space systems in which inter-institutional movement, interpretive cues, and urban atmosphere jointly shape the visitor experience [77,82]. Within this perspective, the stimulus is understood as an integrated environmental system encompassing both clustered facilities and the spaces between them; the organism layer is conceptualized as a sequence of related but functionally differentiated internal states; and the response is defined as spillover from district experience to city image and behavioral intention rather than as facility-level loyalty alone. From this perspective, city-image spillover emerges through district-level experiential infrastructure rather than through isolated institutional satisfaction alone [61,80]. Accordingly, the contribution of this study lies in contextual refinement and analytical specification, rather than in proposing an entirely new grand theoretical model.

Notably, the broad contributions of information and content visibility (S3) and amenities and the operational environment (S4) to experiential value, affective response, immersion, and satisfaction indicate that high-quality MAA experiences depend not only on symbolic landscapes or the scale of facilities but also on the interpretive accessibility, learning support, and operational continuity surrounding the visit [56,63]. These findings are consistent with museum and urban experience studies showing that the value of cultural visits is shaped by the combined quality of interpretation, accessibility, and surrounding public-space conditions [52,77]. Furthermore, the significant covariances among the stimulus factors suggest that visitors perceive the MAA environment as an interconnected system rather than as a set of independent cues. From a planning perspective, this implies that circulation, information, amenities, and atmosphere should be optimized simultaneously rather than treated as separate interventions.

The structural results also help clarify why the organism variables were retained separately despite their conceptual proximity. Experiential value concerns whether visitors judge the itinerary to be worthwhile; affective response reflects the immediate emotional tone generated during the visit; immersion captures sustained concentration and experiential continuity across spaces [72,74]; and satisfaction represents a more retrospective summary appraisal formed in relation to prior expectations [83,84]. These variables are therefore related and partially overlapping in a broad experiential sense, yet they are not redundant in analytical terms because they represent different functional roles within the visit chain. Experiential value captures evaluative meaning and worth, affective response captures felt emotion in the moment, immersion captures attentional absorption and continuity, and satisfaction summarizes the experience after it is compared with expectations. The comparatively weaker coefficients associated with satisfaction further suggest that, in MAAs, city-image spillover is shaped less by post hoc approval alone than by the value judgments, emotional tone, and attentional continuity accumulated during the visit process [84,85]. This pattern supports retaining the four organism constructs as related but functionally differentiated components of the tourism experience.

4.2. Implications for Planning, Operations, and Urban Marketing

The results of the effect decomposition show that the environmental stimuli of MAAs primarily influence the response variables through cumulative indirect effects mediated by internal states. In practical terms, this means that MAAs should be managed not as loose concentrations of cultural institutions but as integrated urban tourism packages. When destination managers prioritize route design, interpretive continuity, bundled services, and operational coordination across venues, the district is more likely to be experienced as a coherent cultural itinerary rather than as a series of disconnected stops. The implications can be summarized as follows:

First, information and content visibility (S3) should be treated as the core interface that structures movement and understanding. In MAAs, decisions about where to go next, how long to stay, and how to connect venues are part of the experience itself. Accordingly, integrated maps, clear signage, coordinated storytelling across institutions, and real-time visitor support, such as information on crowd levels, required visit time, reservations, waiting times, and recommended routes, can simultaneously strengthen experiential value, affective response, and immersion.

Second, amenities and the operational environment (S4) should be planned at the district scale rather than at the level of individual facilities. Properly distributed rest areas, queue management, mitigation of circulation bottlenecks, weather-sensitive operational stability, and the integration of on-site staff guidance with digital support can reduce visit-chain friction and preserve experiential flow. In this respect, operational quality is not merely a background condition; rather, it is a core mechanism through which city favorability and intentions to revisit or recommend the destination are reinforced.

Third, the remaining stimulus factors—namely, aesthetic environment (S1), accessibility and connectivity (S2), and social atmosphere (S5)—generate stronger effects when combined with S3 and S4 as part of an interconnected experience system. For example, symbolic scenery becomes more meaningful when accompanied by interpretive information, connectivity becomes more valuable when transitions are smooth and legible, and social vitality becomes more favorable when public-space management supports coexistence and comfort. Ultimately, urban marketing for MAAs requires a shift from promoting individual museums to packaging district-level experiences through coordinated routes, cross-venue programs, bundled ticketing or passes, and shared destination branding.

Overall, the findings suggest that the competitiveness of MAAs depends less on simply expanding hardware and more on designing experience infrastructure at the district level. This perspective reinforces recent discussions of museums as urban public spaces and cluster-based tourism environments in which movement, interpretation, and public-space quality are inseparable from cultural consumption [77,82].

5. Conclusions

This study examined how environmental stimuli in Museum Agglomeration Areas (MAAs) shape city image and behavioral intentions through differentiated internal states of tourism experience. In doing so, it conceptualized museum agglomeration areas as district-scale tourism environments in which experiences are formed across clustered cultural facilities, public spaces, and urban services rather than within isolated institutions. From this perspective, the study does not propose a fundamentally new theoretical framework; rather, it offers a context-specific extension of the stimulus–organism–response (S–O–R) framework by shifting the analytical focus from a single service encounter to a district-scale visit chain. More specifically, the study refines the organism layer by distinguishing experiential value, affective response, immersion, and satisfaction as related yet functionally distinct internal states. Its contribution therefore lies in clarifying how environmental conditions in museum agglomeration areas are translated into city-level evaluations through multiple mediating processes, rather than in claiming a major reformulation of stimulus–organism–response theory itself.

First, by systematically validating the S–O–R pathways through CFA and SEM procedures, this study clarifies how district-level environmental conditions are translated into city-level evaluations through differentiated experiential processes in the context of MAAs. In particular, experiential value, affective response, immersion, and satisfaction were treated as distinct internal states, showing that identical environmental stimuli influence response variables through different psychological pathways. The theoretical contribution therefore lies less in proposing an entirely new grand theory than in specifying how the organism layer should be decomposed when the tourism experience unfolds across multiple connected venues and public spaces.

Second, from the perspective of effect configuration, environmental stimuli in MAAs operate as an interconnected system rather than as independent single factors. Information and content visibility (S3), along with amenities and the operational environment (S4), contributed relatively more to the spillover process, indicating that interpretive and operational infrastructure plays a central role in linking MAA experiences to the enhancement of city image. This finding supports the view that district-scale experience formation depends not only on symbolic attraction but also on the infrastructural conditions that make multi-site movement intelligible, continuous, and usable.

Third, among the internal-state factors, experiential value (O1) exerted the strongest influence on city image and behavioral intentions, while immersion (O3) also showed substantial effects. This suggests that the competitiveness of MAAs depends not only on what visitors see but also on whether the district is experienced as a meaningful, coherent, and absorbing cultural itinerary. At the same time, the comparatively smaller role of satisfaction indicates that post hoc approval alone does not adequately explain city-image spillover in multi-stage cultural environments, thereby reinforcing the analytical value of distinguishing among evaluative, emotional, attentional, and retrospective organism states.

At the same time, these findings should be interpreted within the limits of the empirical setting. The survey was based on an online, experience-qualified nonprobability sample of visitors to MAAs in Seoul, a culturally dense metropolitan context in which areas such as Jongno-gu exemplify the cluster-based integration of museums, heritage sites, tourism facilities, and commercial activity. Accordingly, the model is likely to be most directly applicable to large metropolitan or global cities with comparable cultural clustering, while caution is needed when generalizing the findings to smaller cities, monocentric destinations, or less dense cultural districts. In addition, because the present study focuses on a single-city case, its explanatory scope remains context-specific. Future research should therefore extend the analysis through comparative studies of multiple cities sharing similar cultural, spatial, and tourism-related characteristics in order to test the external validity and contextual variability of the proposed model. Further refinement of the visit-chain mechanism may also be achieved through named district-level analyses, multi-group SEM, and the integration of behavioral and spatial data, such as GPS traces, length of stay, and congestion patterns.