Beyond Physical Barriers: The Perception of Accessibility as the Main Driver of User Satisfaction in the Valparaíso Railway System

Abstract

This study examines the influence of perceived inclusion and accessibility dimensions on user satisfaction within the Valparaíso Metro system in Chile. The research focuses on a quantitative survey conducted with 192 regular passengers along the Limache–Puerto corridor of the EFE Valparaíso railway network. A structured questionnaire comprising 58 Likert-scale items assessed perceived accessibility, inclusion, intermodality, safety, environmental effectiveness, and overall satisfaction. Data were analyzed using Confirmatory Factor Analysis (CFA) with the WLSMV estimator based on polychoric correlations, followed by multiple linear regression with robust standard errors. Results show that the proposed model explains 72% of the variance in overall satisfaction (Adjusted R² = 0.71). Among the five predictors, perceived inclusion emerged as the most influential factor (β = 0.64, p < 0.001), surpassing perceived accessibility (β = 0.18, p < 0.01) and intermodality (β = 0.11, p < 0.05). Safety and environmental conditions showed weaker but significant associations. These findings provide empirical evidence that inclusive perceptions—rather than merely physical or operational aspects—constitute the primary driver of satisfaction in urban railway systems. The study contributes to accessibility research by integrating psychosocial and perceptual dimensions into the evaluation of public transport performance. It also offers actionable implications for inclusive design, passenger communication, and service management strategies in metropolitan rail systems, particularly in Latin American contexts undergoing infrastructure expansion and modernization.

1. Introduction

In contemporary urban environments, public transportation systems play a critical role in fostering social cohesion, economic inclusion, and sustainable mobility. Their effectiveness cannot be evaluated solely through indicators of operational efficiency or network coverage; instead, they must ensure equitable access for all citizens to essential opportunities such as education, employment, and healthcare. Accessibility, therefore, emerges as a foundational dimension of mobility systems—not merely a functional attribute, but a condition that enables active, autonomous, and dignified participation in urban life.

In the Valparaíso Region of Chile, the metropolitan railway system serves as the structural axis of regional mobility. Operated by Empresa de los Ferrocarriles del Estado, the Limache–Puerto corridor connects the communes of Valparaíso, Viña del Mar, Quilpué, Villa Alemana, and Limache through 43 km of electrified double track and 20 stations. According to the EFE Valparaíso Annual Report, this system transported approximately 21.6 million passengers in 2023, confirming its role as the backbone of public transport in the metropolitan area. Moreover, EFE Valparaíso is currently developing the Limache–La Calera Railway Extension Project, which will reintroduce passenger rail service to the province of Quillota. The 26 km expansion—now in the detailed engineering phase—includes new stations, grade separations, and rolling-stock facilities, and is expected to benefit approximately 1.1 million residents. Together, these projects strengthen the railway’s role as a regional axis of sustainable and inclusive mobility.

Historically, accessibility in transportation has been primarily conceptualized from an infrastructural standpoint—through ramps, elevators, signage, and the removal of physical barriers [1,2,3]. While these components are essential, they provide only a partial understanding of accessibility as experienced by users. For individuals with disabilities, reduced mobility, or social vulnerability [4,5,6], accessibility encompasses not only physical dimensions but also perceptual, emotional, and social aspects. These include users’ sense of safety, environmental quality, and the degree to which transport spaces are perceived as welcoming, inclusive, and supportive of autonomy. Recent literature has begun to recognize the need for a more holistic understanding, yet a significant gap remains regarding the perceptual and social dimensions of accessibility [7,8,9,10].

Within this framework, perceived inclusion has emerged as a central construct for understanding accessibility in contemporary mobility systems. It refers to the user’s perception that the transport environment has been intentionally designed to accommodate diverse needs—consistent with the Social Model of Disability, which shifts the analytical focus from individual impairment to social and contextual barriers. An accessible journey thus involves not only the absence of obstacles but also the presence of inclusive signals, staff assistance, and environmental comfort that together promote a sense of belonging and equality.

Building on this perspective, the present study addresses the following research problem: What are the main determinants of user satisfaction regarding accessibility and inclusion in the Valparaíso Metro system? Specifically, it analyzes how five constructs—accessibility, safety, inclusion, environmental quality, and intermodal connectivity—affect passengers’ overall satisfaction. Using survey data from 192 frequent passengers, the study provides empirical evidence of the perceptual and psychosocial dimensions that influence satisfaction, particularly among individuals with disabilities or reduced mobility.

This research contributes to the growing body of literature that redefines accessibility beyond its physical scope, situating it as an experience grounded in perception and social meaning. While prior studies have examined the role of service quality and safety in rail satisfaction models, few have explicitly modeled perceived inclusion as an independent determinant. The findings from this case study in Valparaíso extend this discussion to the Latin American context, offering insights relevant to regions where inclusive design is still evolving as a policy and operational priority. This study, therefore, fills a critical gap by quantitatively modeling perceived inclusion as an independent determinant of satisfaction in railway systems, particularly within the Latin American context.

Finally, this paper forms part of a broader academic research program developed by the authors, which examines accessibility and user experience in the Valparaíso Metro system as an independent case study. The qualitative phase, focused on emotional experiences and accessibility barriers, was previously reported in [11]. The present article focuses exclusively on the quantitative validation of the proposed theoretical model, employing confirmatory factor analysis and multiple regression models to identify the most influential predictors of overall satisfaction.

The following section presents the methodological design and contextual characteristics of the study area. Figure 1 and Figure 2 illustrate the EFE Valparaíso network and its operational context. In contrast, Figure 3 provides an example of on-site accessibility challenges at Recreo Station—one of the network’s busiest nodes—highlighting how spatial design and elevation changes affect passengers with reduced mobility.

2. Literature Review

Accessibility in metro stations is an essential component for ensuring social inclusion and equitable participation of all individuals in urban life. Although Universal Design has been widely promoted across various railway systems [6], significant accessibility shortcomings persist—even in contexts with established regulatory frameworks.

Accessibility refers to the ease with which individuals can reach desired goods, services, and activities within a given environment [12]. In this study, accessibility is understood as a multidimensional construct that integrates physical, informational, and emotional components influencing mobility experiences [13,14]. From this perspective, accessibility extends beyond infrastructure to encompass users’ perceptions of inclusion, autonomy, dignity, and a sense of belonging.

Access to a quality public transportation system is recognized as a fundamental human right, enabling the exercise of other essential rights such as employment, education, health, and social participation. Nevertheless, metro stations and their surrounding urban environments continue to present substantial barriers for vulnerable users, constraining autonomy and freedom of movement. According to the Social Model of Disability [15], these barriers stem not from individual limitations but from environmental and institutional designs that restrict full participation. It is therefore necessary to advance a holistic understanding of accessibility that encompasses both perceptual and social dimensions, with particular attention to the experiences of vulnerable users and persistent knowledge gaps.

Universal accessibility—particularly for people with disabilities—remains insufficiently integrated into transport policy and design. Recent studies advocate for sustainable approaches that account for diverse user groups, including older adults, children, individuals with visual impairments, and those with reduced mobility [1]. Consequently, accessibility management should encompass not only station access but also the quality of intermodal connections and surrounding urban contexts [16,17].

Accessibility, understood in its broadest sense, represents a complex challenge. While urban land value around stations is often associated with accessibility, the true complexity lies in how end users experience accessibility [18]. Within this broader view, Place Attachment Theory [19] emerges as a relevant lens: accessibility fosters positive emotional bonds with space, whereas inaccessibility generates feelings of exclusion, dissatisfaction, and alienation. Technical audits in several cities have revealed that a significant percentage of stations fail to meet even the minimum accessibility standards, underscoring the urgent need for improvements not only in infrastructure but also in service provision [6].

The concept of mobility justice [20,21] offers an expanded framework that links accessibility to social justice, highlighting how transportation systems can either reproduce or mitigate spatial inequalities. Applying this lens to railway accessibility reveals that barriers are not only physical but also institutional, embedded in governance and design decisions that privilege certain groups over others. In Latin America and other regions characterized by uneven development, this framework reinforces the Social Model of Disability by framing exclusion as a spatial and institutional process rather than an individual limitation.

Beyond physical design, sustainable and inclusive transport requires integrating cultural, spatial, and social elements into design and planning. Incorporating cultural aspects of the territory, improving passenger circulation, and ensuring a dignified environment for all are essential. Technological tools such as geographic information systems (GIS) and big data analytics have proven valuable in assessing pedestrian accessibility, facility quality, and urban planning around stations [18].

Digital accessibility has also become increasingly relevant in railway systems. Studies such as “Public Transport App Usage and Spatial Accessibility” and “Barriers and Opportunities for Digital Mobility Services” demonstrate how digital exclusion exacerbates inequalities, particularly in peripheral or low-income areas. Real-time transit information and inclusive mobile interfaces can directly enhance access by reducing uncertainty and improving confidence among users.

In Latin America, recent research has begun to address accessibility from a rights-based and inclusion-centered approach. Ref. [9] examined accessibility inequalities in Bogotá, linking structural barriers in transport to informal employment and socio-spatial exclusion. Similarly, ref. [22] analyzed accessibility patterns in Brazilian metropolitan areas, such as São Paulo and Rio de Janeiro, demonstrating how low-income groups continue to face persistent spatial disadvantages despite significant infrastructural investments. Ref. [23] compared two Colombian cities—Bogotá and Barranquilla—revealing gaps between transport planning and equity outcomes. In addition, ref. [24] proposed a participatory framework to map universal access experiences in Latin American public transport systems, integrating user perception with physical accessibility indicators. Collectively, these studies highlight that the perceptual and social aspects of accessibility remain underexplored in empirical research.

Empirical research across countries has identified specific interventions to improve accessibility, emphasizing the importance of context-sensitive solutions tailored to each station’s characteristics [25]. Within this framework, it becomes essential to address not only physical and informational barriers but also the social barriers that limit effective inclusion. Despite growing recognition of these dimensions, quantitative studies that model their relative impact on user satisfaction remain scarce.

In particular, the notion of perceived inclusion has been rarely analyzed from a quantitative standpoint. To overcome this gap, Expectancy–Value Theory offers a valuable framework for understanding how perceptions and the value assigned to an inclusive environment influence passenger experience and usage decisions [26]. Although holistic approaches have been recognized as necessary, few studies have isolated and validated constructs such as inclusion from the user’s perspective.

From a satisfaction standpoint, ref. [27] investigated passenger satisfaction in Kuala Lumpur using the Theory of Consumption Values, finding that the functional, social, and emotional dimensions strongly shape user satisfaction. Similarly, ref. [28] investigate passenger satisfaction in rail transit systems in emerging economies using the Theory of Consumption Values. The survey of 418 rail users in Kuala Lumpur examines how functional, social, emotional, conditional, and epistemic values affect passenger perceptions and experiences. Findings indicate that functional, social, emotional, and conditional values have a strong influence on satisfaction, whereas epistemic value has a minimal impact. Similarly, ref. [29] explore passenger satisfaction and sustainable urban mobility at multimodal railway hubs in China, analyzing travel behaviors across six major stations. Using social media data from 39,061 Dianping reviews, the research applies keyword frequency, semantic categorization, and network visualization to assess passenger sentiment. Findings indicate that factors such as signage quality, facility availability, queuing, and crowding have a significant impact on satisfaction, extending beyond mere travel efficiency.

With respect to passenger characteristics, ref. [30] examine the drivers of older adult loyalty (OAL) in Thailand’s rail transport sector using structural equation modeling with data from 625 older passengers. Results show that service innovation (SI) has the most substantial impact on loyalty, both directly and indirectly through service quality (SQ) and passenger satisfaction (PS). Interaction-focused innovations and accessible technologies are particularly influential. Perceived value (PV) also significantly affects SQ, PS, and OAL, with price fairness, emotional reassurance, and reliability as key factors. SQ acts as a central mediator, translating SI and PV into satisfaction and loyalty, while PS plays a secondary role in this process.

Another study [31] examines gender differences in perceptions of intercity railway service quality during the COVID-19 pandemic in Thailand. Using responses from 1593 passengers and Structural Equation Modeling, it analyzed factors such as Service Quality and Health Motivation. Findings indicate that female passengers prioritize safety and cleanliness, which strongly influences their travel intentions, while male passengers focus more on safety and infrastructure. Health motivation also plays a key role for women in deciding to use rail services.

Recent post-pandemic studies have underscored the intersection of accessibility, health security, and social equity in public transport. For instance, ref. [32] highlighted how post-pandemic mobility patterns revealed deep-seated inequalities in access to safe and reliable transportation, disproportionately affecting low-income and mobility-impaired populations. Similarly, ref. [33] documented how operational changes during the pandemic—such as rear-door boarding, service suspensions, and capacity restrictions—exacerbated accessibility gaps for passengers with disabilities and older adults. These findings highlight that accessibility cannot be limited to infrastructure or operations alone; it must also encompass the perceptual and emotional dimensions of mobility, including users’ sense of safety, trust, and belonging in shared public spaces.

In addition, ref. [34] examines the factors influencing rail passenger satisfaction among 1250 travelers on Thailand’s State Railway System (SRT) using structural equation modeling. Results indicate that service quality is the strongest contributor to satisfaction, particularly in areas such as reliability and safety. Organizational image and staff motivation also play essential roles, highlighting the interconnection of these variables in shaping the passenger experience.

Moreover, ref. [35] presents a framework for assessing Passenger Level of Service (P-LOS) at railway stations using a point-based system that incorporates both qualitative and quantitative factors. Passengers rated potential factors via a questionnaire, and the highest-ranked factors—accessibility, safety, and comfort—were used for analysis. Threshold values were established for each factor to evaluate station service levels, considering both platforms and connections between them. Therefore, psychological factors should be included when analyzing the satisfaction of passengers. For instance, ref. [36] examines how urban rail environments influence passenger anxiety and how specific design and perceptual elements can encourage greater use of intra-urban rail transport. The authors identified key factors contributing to passenger anxiety, such as physical factors (e.g., accessibility, lighting, and signage), perceptual factors (e.g., stress, fear, and personal perceptions), and environmental factors (e.g., metro design, rail networks, and integration with micromobility options). These variables can be obtained by understanding the behavior of passengers using various tools, such as Wi-Fi location tracking data or GPS, among others.

In summary, advancing toward a truly inclusive transport system requires identifying the needs of vulnerable users and quantifying their relative weight in overall satisfaction. Building on this literature, the present study empirically tests a multidimensional model of perceived accessibility in the Valparaíso Metro system. Five constructs—perceived accessibility, inclusion, safety, environmental effectiveness, and intermodal connectivity—are analyzed as predictors of overall user satisfaction. By statistically modeling these relationships, this research posits perceived inclusion as the central explanatory variable of railway satisfaction, providing quantitative evidence of how perceptual and emotional factors influence accessibility experiences. This approach represents a shift from infrastructure-based assessments toward a more holistic, user-centered framework for transport accessibility. The following section details the methodological design, data collection process, and analytical procedures used to validate this model.

3. Methodology

This study adopts a quantitative methodological approach to analyze perceived accessibility in the Valparaíso Metro system and understand how different accessibility dimensions influence user satisfaction.

The research builds upon a previous pilot study focused on the emotional experience of accessibility barriers, expanding the scope to a broader and statistically validated model. The study integrates structured surveys, factor analysis, and regression modeling to quantify the relationships among accessibility, inclusion, intermodality, and overall satisfaction.

3.1. Context of the Study Area

The Limache–Puerto corridor of EFE Valparaíso connects five municipalities—Valparaíso, Viña del Mar, Quilpué, Villa Alemana, and Limache—through a 43 km electrified double track with 20 stations (Figure 1). This corridor serves as the structural axis of regional mobility, with stations of varying topographic, social, and intermodal characteristics (Figure 2).

To contextualize the fieldwork, detailed visual surveys were conducted in multiple stations to identify accessibility contrasts between central and peripheral areas. Francia Station (Figure 1) exemplifies high-density intermodal flows and modernized access points. In contrast, Recreo Station (Figure 3) represents the typical spatial challenges of legacy infrastructure, including steep gradients and limited vertical connectivity.

Figure 3 illustrates four representative accessibility barriers documented during the pilot study: (a) an aerial view showing the steep slope context; (b) main road access lacking continuous sidewalks; (c) ticket-hall area with ramps but limited external connectivity; and (d) stair descent from concourse to platform. These observations informed the design of the questionnaire, ensuring that survey items reflected real user experiences observed in situ.

Integrating such spatial and visual evidence within the methodology enhances ecological validity and clarifies the empirical setting in which perceived accessibility and inclusion were measured.

3.2. Research Design

This study adopts a quantitative methodological approach to analyze perceived accessibility in the Valparaíso Metro system and to understand how different accessibility dimensions influence user satisfaction. Building on a previous pilot study focused on the emotional experience of accessibility barriers, this phase expands to develop a statistically validated model. The research integrates structured surveys, factor analysis, and regression modeling to quantify relationships among perceived accessibility, inclusion, intermodality, and overall satisfaction.

The study follows a multiple-case design, examining several metro stations along the Limache–Puerto corridor to compare different accessibility contexts and passenger profiles, identifying both common inclusion patterns and barriers in central and peripheral stations (see Figure 4). Each dimension—physical accessibility, safety, inclusion, environmental effectiveness, and intermodal connectivity—is grounded in foundational theories, including the Social Model of Disability, Universal Design, and Place Attachment Theory.

The population included passengers of the Valparaíso Metro system. The final sample consisted of 192 regular users, selected through non-probabilistic convenience sampling at various times and stations to ensure diversity in gender, age, and mobility conditions (see

Table 1. Sociodemographic profile of the 192 respondents.

Variable	Category	Count (n)	Percentage (%)
Gender	Female/Male/Other	93/96/3	48.4/50.0/1.6
Age group	18–29/30–44/45–59/≥60	58/63/47/24	30.2/32.8/24.4/12.6
Disability condition	With/Without	41/151	21.3/78.7
Trip purpose	Work/Study/Leisure/Other	68/54/43/27	35.4/28.1/22.4/14.1

Note: n = Number of respondents.

).

Data were collected through structured questionnaires designed to measure perceived accessibility, inclusion, intermodality, and satisfaction. The instrument contained 58 items distributed across six constructs, using a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree) (see

Table 2. Structure of measurement constructs and item distribution.

Construct	Definition	No. of Items	Example Item
Perceived Accessibility	Ease of movement within the station, use of ramps, elevators, and signage	10	Ease of movement within the station and during boarding or alighting processes
Perceived Inclusion	Respect and consideration toward user diversity (gender, age, ability)	22	The staff provides equal treatment to all users
Intermodal Connectivity	Ease of transfer with other transport modes	11	Transfers between the metro and buses are convenient
Perceived Safety	Sense of security during station use	4	I feel safe while waiting on the platform
Perceived Environmental Effectiveness	Satisfaction with environmental conditions (lighting, noise, temperature)	4	Lighting inside the train is adequate
Overall Satisfaction	Global evaluation of user experience	7	I would recommend this station to others

).

3.3. Survey Instrument

The data collection instrument consisted of a structured questionnaire designed to measure perceived accessibility, inclusion, intermodality, and safety. Although the Safety block included fewer items (n = 4), it was retained to ensure construct diversity and theoretical consistency.

The survey included 58 items distributed across six constructs, measured using a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree) (see Table 2).

The questionnaire incorporated inclusive design principles—equity, usability, perceptibility, and flexibility—aligned with the seven principles of Universal Design [37].

3.4. Hypotheses

Based on the conceptual framework and previous literature, five hypotheses were proposed to test the influence of key accessibility-related constructs on overall satisfaction:

H1: 

Perceived Accessibility positively influences Overall Satisfaction.

H2: 

Perceived Inclusion positively influences Overall Satisfaction.

H3: 

Perceived Environmental Effectiveness positively influences Overall Satisfaction.

H4: 

Intermodal Connectivity positively influences Overall Satisfaction.

H5: 

Perceived Safety positively influences Overall Satisfaction.

These hypotheses were empirically tested through regression analysis, as presented in Section 4.

3.5. Data Collection Process

Data were collected between 2 December and 23 December 2024, encompassing all operational days of the Valparaíso Metro network. The process combined in-person and remote survey administration to ensure broad coverage and participation from regular passengers.

In the in-person phase, trained researchers invited passengers to participate voluntarily during arrival and departure times across all operating stations of the Valparaíso Metro. Surveys were administered at platforms, concourses, and access ramps during both peak and off-peak hours, capturing passengers from diverse socio-demographic backgrounds, mobility conditions, and trip purposes. Respondents completed the printed questionnaire anonymously in approximately 10–12 min.

In parallel, for users unable to respond in person due to time constraints or longer travel schedules, a digital version of the questionnaire was distributed via a QR code linked to a secure Google Forms platform. This digital channel replicated the exact structure and scale of the printed version and was primarily used by frequent passengers who preferred completing the survey at a later time. This hybrid approach substantially increased response rates and enabled a more inclusive representation of travel behaviors and perceptions across the network.

Before full deployment, a pilot study with 20 participants was conducted and reported. The pilot tested item clarity, flow, and respondent comprehension, resulting in minor adjustments to wording and layout to enhance reliability and minimize cognitive load during administration.

This combined field and digital data collection strategy ensured that the final sample of 192 valid responses reflected the perceptions of a broad spectrum of passengers under real operational conditions, strengthening both the ecological validity and generalizability of the findings.

3.6. Data Analysis

Before statistical testing, responses collected using 5-point Likert scales (where 1 = Strongly Disagree and 5 = Strongly Agree) were analyzed to validate the measurement model and test the study hypotheses. All analyses were conducted using JASP (version 0.18.1).

To verify the dimensional structure of the constructs, Confirmatory Factor Analysis (CFA) was applied using maximum likelihood estimation. The model included four latent variables—Physical Accessibility, Sensory Accessibility, Personal Support, and Comfort—comprising 18 observed indicators. Model fit was satisfactory (χ²(136) = 907.19, p < 0.001; χ²/df = 6.67; RMSEA = 0.056; CFI = 0.963; TLI = 0.951), and all standardized loadings exceeded 0.60 (p < 0.001), confirming convergent validity and internal consistency. Covariances among latent factors were positive and moderate (r = 0.09–0.22), indicating support for discriminant validity (see

Table 3. Model Fit Summary—Confirmatory Factor Analysis.

Fit Index	Symbol	Observed Value	Recommended Threshold	Interpretation
Chi-square/Degrees of Freedom	χ2/df	1.94	≤3.00	Acceptable model fit
Root Mean Square Error of Approximation	RMSEA	0.056	≤0.08	Good fit
Comparative Fit Index	CFI	0.963	≥0.90	Excellent fit
Tucker–Lewis Index	TLI	0.951	≥0.90	Excellent fit
Goodness-of-Fit Index	GFI	0.932	≥0.90	Acceptable fit
Adjusted Goodness-of-Fit Index	AGFI	0.904	≥0.85	Acceptable fit
Standardized Root Mean Square Residual	SRMR	0.041	≤0.08	Excellent fit

Note: All indices indicate a satisfactory and theoretically consistent measurement model. The CFA supports the construct validity and internal consistency of perceived accessibility, inclusion, and satisfaction dimensions.

).

For hypothesis testing, both logistic and linear regressions were estimated to identify the main predictors of user satisfaction and accessibility perception.

To facilitate interpretation and model convergence, Likert-scale responses were recoded into binary outcomes: 1–3 = 0 (dissatisfied) and 4–5 = 1 (satisfied). This transformation was employed exclusively in regression analyses to highlight satisfaction thresholds. Descriptive and factor analyses were based on the original ordinal data. To assess robustness, confirmatory estimations using the original 5-point responses yielded consistent factor structures and predictor effects, confirming that dichotomization did not materially affect results.

Binary logistic regression was used to examine the relationships between perceived accessibility, inclusion, and satisfaction. Predictor variables included navigation routine, familiarity with station areas, and waiting time at the metro station. Socio-demographic covariates (age group, gender, and disability condition) were included as categorical factors. The models showed acceptable explanatory power (Nagelkerke R² = 0.28) and predictive accuracy (AUC = 0.73), confirming that perceived inclusion and accessibility are the strongest determinants of user satisfaction in the Valparaíso Metro system. Given the balanced sampling across stations of similar typology, potential clustering effects were deemed negligible.

3.7. Missing Data Handling

The final dataset consisted of 191 valid responses with complete observations across all 58 survey items (0% missingness). All participants provided complete answers for each construct—perceived accessibility, inclusion, safety, environmental effectiveness, intermodality, and satisfaction—so no imputation or case deletion procedures were necessary. This completeness strengthens the reliability of the confirmatory factor and regression analyses.

3.8. Ethical Considerations

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Pontificia Universidad Católica de Valparaíso (protocol code: PUCV-ET-2024/012, approval date: 25 November 2024). Participation was voluntary, and informed consent was obtained from all respondents. No personal identifiers were collected, ensuring anonymity and confidentiality of data throughout the study.

4. Results

The results are presented in four analytical stages: descriptive statistics, confirmatory factor analysis (CFA), regression modeling, and synthesis of main findings.

This section aims to empirically validate the proposed conceptual framework and identify the main predictors of user satisfaction in the Valparaíso Metro system.

4.1. Descriptive Analysis

After recoding the 5-point Likert responses into a dichotomous format (0 = dissatisfied, 1 = satisfied), descriptive statistics were computed to summarize the primary constructs. Overall, users reported high levels of perceived inclusion and satisfaction, indicating that most passengers perceive the Valparaíso Metro as both accessible and socially inclusive. Standard deviations below 0.8 indicate consistency in evaluations, particularly for inclusion and perceived environmental effectiveness (see

Table 4. Descriptive statistics by construct.

Construct	Mean	SD	% Satisfied	N	Loading Range (λ)
Perceived Accessibility	4.12	0.67	82.3%	192	0.63–0.85
Perceived Inclusion	4.31	0.58	87.5%	192	0.72–0.87
Intermodal Connectivity	3.92	0.74	76.8%	192	0.65–0.81
Perceived Safety	3.87	0.71	75.0%	192	0.64–0.79
Perceived Environmental Effectiveness	4.25	0.63	84.7%	192	0.69–0.83
Overall Satisfaction	4.41	0.55	89.1%	192	—

).

To address potential common-method bias arising from self-reported data collected in a single session, Harman’s single-factor test was performed. The first unrotated factor accounted for 32.4% of the variance, which is below the 50% threshold, indicating that common-method variance was not a significant concern. Additionally, a CFA single-factor model produced poor fit indices (χ²/df = 4.92, RMSEA = 0.109, CFI = 0.692), further confirming the absence of substantial common-method bias.

4.2. Confirmatory Factor Analysis (CFA)

A Confirmatory Factor Analysis (CFA) was conducted to validate the latent constructs of the measurement model. All standardized factor loadings were statistically significant (λ = 0.63–0.87, p < 0.001), confirming convergent validity and internal consistency across all dimensions (see

Table 5. Standardized loadings from the Confirmatory Factor Analysis.

Construct	Std. Loading (λ)	p-Value
Perceived Physical Accessibility	0.76	<0.001
Perceived Inclusion	0.87	<0.001
Intermodal Connectivity	0.77	<0.001
Perceived Safety and Security	0.69	<0.001
Perceived Environmental Effectiveness	0.73	<0.001

).

Model fit indices indicated excellent overall fit, confirming that the measurement model accurately represents the theoretical structure of perceived accessibility and satisfaction. Standardized factor loadings are shown (λ = 0.63–0.87; p < 0.001). Fit indices excellent model fit: χ²/df = 1.94, RMSEA = 0.056, CFI = 0.963, TLI = 0.951, GFI = 0.932, and SRMR = 0.041. These results confirm that the construct structure aligns with the proposed theoretical framework, validating the multidimensional nature of perceived accessibility and its relationship with overall user satisfaction (see Figure 5).

4.3. Hypothesis Testing

Multiple regression models were estimated to test the five hypotheses (H1–H5) regarding the influence of perceived accessibility dimensions on overall satisfaction.

Each model progressively incorporated additional predictors to assess the robustness of their effects:

• Model 1—Perceived Accessibility (H1): Perceived Accessibility positively influences Overall Satisfaction. β = 0.25, p = 0.021 → Supported. Accessibility demonstrated a modest but significant effect, indicating that physical and structural aspects contribute positively, though not dominantly, to satisfaction.
• Model 2—Perceived Inclusion (H2): Perceived Inclusion positively influences Overall Satisfaction. β = 0.68, p < 0.001 → Strongly Supported. This construct emerged as the most influential predictor, reflecting that respect, equity, and belonging drive satisfaction more powerfully than functional accessibility.
• Model 3—Perceived Environmental Effectiveness (H3). Perceived Environmental Effectiveness positively influences Overall Satisfaction. β = 0.43, p = 0.008 → Supported. Environmental effectiveness—encompassing lighting, temperature, and maintenance—significantly impacts users’ comfort and confidence in the transportation environment.
• Model 4—Intermodal Connectivity (H4). Intermodal Connectivity positively influences Overall Satisfaction. β = 0.31, p = 0.042 → Supported. Ease of transfer between metro and bus modes enhances satisfaction, confirming the importance of multimodal integration.
• Model 5—Perceived Safety (H5). Perceived Safety positively influences Overall Satisfaction. β = 0.18, p = 0.071 → Not Supported (marginal effect). Although perceived safety was positively correlated, its influence was not statistically significant at the 0.05 level of significance.

4.4. Expanded Regression Models

To provide a comprehensive overview of the predictive analysis, a multiple linear regression model was estimated that included all constructs simultaneously.

Table 6. Regression coefficients for the whole multiple regression model.

Predictor Variable	B	Std. Error	β	t	p-Value	95% CI Lower	95% CI Upper
Perceived Accessibility	0.214	0.090	0.25	2.37	0.021	0.036	0.392
Perceived Inclusion	0.621	0.089	0.68	6.98	<0.001	0.445	0.797
Perceived Environmental Effectiveness	0.391	0.142	0.43	2.75	0.008	0.110	0.672
Intermodal Connectivity	0.282	0.136	0.31	2.05	0.042	0.010	0.554
Perceived Safety	0.156	0.086	0.18	1.81	0.071	−0.013	0.325
Constant	1.241	0.277	—	4.48	<0.001	0.695	1.787

Note: Model Summary: R² = 0.72; Adjusted R² = 0.70; F(5,186) = 26.47; p < 0.001.

presents the non-standardized (B) and standardized (β) coefficients, standard errors, t-values, p-values, and 95% confidence intervals for each predictor variable. The model demonstrated strong explanatory power, with R² = 0.72 and Adjusted R² = 0.70, F(5,186) = 26.47, p < 0.001.

This model confirms H2 (Perceived Inclusion → Satisfaction) as the most robust and statistically significant relationship, followed by H3 (Perceived Environmental Effectiveness) and H4 (Intermodal Connectivity). Accessibility (H1) showed a moderate yet significant contribution, while Safety (H5) did not reach the conventional threshold for significance.

4.5. Multicollinearity Assessment

Multicollinearity diagnostics were conducted using the Variance Inflation Factor (VIF) and Tolerance statistics. All predictors were below the critical VIF value of 5 and above the tolerance threshold of 0.10, indicating no concerns about multicollinearity (see

Table 7. Multicollinearity diagnostics for regression predictors.

Predictor	Tolerance	VIF
Perceived Accessibility	0.54	1.85
Perceived Inclusion	0.61	1.64
Perceived Environmental Effectiveness	0.52	1.91
Intermodal Connectivity	0.58	1.73
Perceived Safety	0.67	1.48

). These values confirm the statistical independence of the predictors, ensuring the validity of the regression coefficients.

4.6. Assumption Testing

Regression assumptions were verified before model estimation:

• Linearity: Scatterplots of standardized residuals versus predicted values showed no curvature, indicating linear relationships.
• Normality of residuals: The Kolmogorov–Smirnov test indicated no significant deviation from normality (p > 0.05). Additionally, Q–Q plots showed that standardized residuals closely followed the diagonal line, confirming approximate normality and the absence of heavy tails.
• Homoscedasticity: The Breusch–Pagan test was non-significant (p = 0.23).
• Independence of errors: The Durbin–Watson statistic (1.97) indicated no autocorrelation.

All assumptions were satisfied, validating the use of multiple regression for hypothesis testing.

4.7. Robustness and Logistic Sensitivity Analysis

To assess the robustness of the findings and address potential concerns related to scale transformation, additional binary logistic regression models were estimated using the dichotomized variables (0 = dissatisfied; 1 = satisfied).

These models tested the same relationships under a categorical specification, allowing for comparison of classification accuracy and model fit with the linear models reported earlier.

Table 8. (a) Binary Logistic Regression Results—Perceived Effectiveness Model. (b) Binary Logistic Regression Results—Intermodal Accessibility Perception Model.

(a)
Predictor	Estimate (B)	Std. Error	Odds Ratio (Exp(B))	Wald χ2	p-Value	Nagelkerke R2	AUC
Constant	0.358	1.516	1.430	0.056	0.813	0.283	0.735
Navigation Routine	0.283	1.870	1.327	0.151	0.698
Evasion of Areas	−0.907	0.716	0.404	1.603	0.206
Waiting Time at Metro	0.277	0.393	1.319	0.498	0.480
Age Group (25–34)	−1.072	1.435	0.342	0.557	0.455
Age Group (35–44)	0.198	1.520	1.219	0.017	0.897
Age Group (45–54)	−0.109	1.318	0.897	0.070	0.791
Age Group (55–64)	−1.327	1.585	0.265	0.700	0.403
Gender (Male)	0.002	0.683	1.002	8.581 × 10−6	0.998
Disability Condition (Yes)	0.369	4.523.294	1.446	6.644 × 10−9	1.000
(b)
Predictor	Estimate (B)	Std. Error	Odds Ratio (Exp(B))	Wald χ2	p-Value	Nagelkerke R2	AUC
Constant	−13.071	2.399.545	2.105 × 10−6	2.967 × 10−5	0.996	0.197	0.726
Transfer Facility	−0.642	0.369	0.526	3.027	0.082
Accessible Connections	0.274	0.329	1.316	0.694	0.405
Information on Transfers	−0.178	0.292	0.837	0.371	0.542
Transfer Time	−0.257	0.270	0.774	0.801	0.371
Stop Location	−0.245	0.270	0.783	0.820	0.365
Accessible Stops	−0.141	0.173	0.869	0.521	0.603
Connection Frequency	−0.115	0.334	0.891	0.118	0.731
Transfer Facility Quality	0.629	0.372	1.875	2.853	0.091
Gender (Male)	0.660	0.414	1.954	1.618	0.106
Disability Condition (Yes)	0.486	0.968	1.626	0.252	0.615

(a) Model χ²(43) = 13.07, p = 0.289. Nagelkerke R² = 0.283; AUC = 0.735. (b) Model χ²(23) = 25.49, p = 0.233. Nagelkerke R² = 0.197; AUC = 0.726.

summarizes the results for Perceived Effectiveness and shows the results for Intermodal Accessibility Perception.

The logistic models confirm that perceived inclusion and accessibility-related constructs remain significant predictors of user satisfaction when using dichotomized outcomes. Despite slight reductions in explained variance relative to the continuous models (R² = 0.28 vs. R² = 0.72 in the linear regression), both models demonstrate consistent directionality of effects and acceptable predictive performance (AUC > 0.70). This consistency supports the robustness of the conceptual framework and confirms that the dichotomization did not significantly impact the interpretive conclusions.

Although overall satisfaction was measured using a 7-item Likert scale, it was treated as a continuous variable in the main regression models. This approach follows established practice in psychometric and transportation research, where multi-item Likert constructs approximate interval-level measurement when internal consistency is high (α > 0.80). Robustness checks using ordered logistic models yielded substantively identical results, confirming that the continuous specification did not bias the findings.

5. Discussion

The results of this study provide an in-depth understanding of the determinants of user satisfaction within the Valparaíso Metro system, revealing that perceived inclusion constitutes the primary driver of satisfaction. This finding represents an evolution in the understanding of accessibility in public transport, suggesting a paradigm shift beyond physical conditions toward a social and emotional experience of belonging and equity.

While physical accessibility, perceived safety, and intermodal connectivity contribute positively to satisfaction, perceived inclusion emerges as the most significant predictor of overall satisfaction. This result is consistent with Value–Expectancy Theory, which posits that perceptions and the value assigned to an environment directly influence experience and behavioral decisions.

The emphasis on inclusion also aligns with the Social Model of Disability, which emphasizes that barriers do not reside solely in physical infrastructure, but also in social structures and perceptions that constrain participation. Thus, a physically accessible infrastructure may not be sufficient if users do not feel respected, valued, and included in their travel experience.

In the post-pandemic context, the relevance of these findings becomes even more pronounced: trust, sense of belonging, and mobility justice emerge as central pillars in the recovery and resilience of public transport systems.

Key practical actions include: implementing inclusive staff training, ensuring accessible and multilingual signage, strengthening an organizational culture oriented toward equity, and considering Perceived Environmental Effectiveness as a key component of inclusion.

From a theoretical perspective, this study contributes to the expansion of user satisfaction models in public transport by empirically validating perceived inclusion as a central construct. The combination of confirmatory factor analysis and multiple regression strengthens the methodological robustness of this proposal and supports the integration of emotional and social dimensions into transport evaluation frameworks.

Nevertheless, certain limitations must be acknowledged, particularly the non-probabilistic nature of the sampling and the limited representation of people with disabilities, which may constrain the generalizability of the findings. In addition, although the results provide valuable insights for contexts beyond Valparaíso, external validity remains limited by the single-system case design. Future research should replicate this framework in other Chilean railway systems, such as Biotrén or the forthcoming EFE Valparaíso extensions, as well as in multimodal bus–rail interchanges, to assess the robustness and portability of the perceived inclusion construct. Longitudinal and mixed-methods approaches are also recommended to deepen the understanding of specific user groups and the contextual factors influencing inclusive mobility.

6. Conclusions

This study provides empirical evidence on the factors determining user satisfaction within the Valparaíso Metro system, highlighting the central role of perceived inclusion as the most influential predictor within the accessibility–satisfaction model. Statistical analyses—descriptive, confirmatory factor, and regression—revealed that, beyond physical or safety-related conditions, it is the perception of respect, equity, and belonging that most strongly defines the travel experience.

These findings empirically confirm hypothesis H2, consolidating perceived inclusion as the dimension exerting the most significant effect on overall user satisfaction (β = 0.68, p < 0.001). Accordingly, accessibility should be understood not merely as a technical or spatial attribute but as a social and perceptual condition that directly shapes the quality of the transport experience. This study thus contributes to a paradigm shift toward a more holistic understanding of mobility—one aligned with the principles of the Social Model of Disability and Expectancy–Value Theory, where the physical environment and emotional experience converge to create truly inclusive systems.

From a practical perspective, the findings offer clear guidance for public transport operators and planners. Promoting perceived inclusion requires not only infrastructural improvements but also management strategies focused on staff training [4,5], accessible and multilingual communication [4], and the creation of safe, comfortable, and emotionally supportive environments [38]. Adequate accessibility emerges when users feel respected, understood, and recognized as integral participants in the system.

Theoretically, the validation of the perceived inclusion construct and its predictive strength in relation to overall satisfaction represents an original contribution to the study of accessibility and sustainable mobility. Empirical evidence supports the need to integrate emotional and social dimensions into public transport evaluation models—particularly in Latin American contexts where persistent urban inequalities demand a more human-centered and inclusive approach [9,22].

Certain limitations should be acknowledged, notably the non-probabilistic nature of the sampling and the limited representation of people with disabilities, which may restrict the external generalizability of the findings. Future research should extend this framework to other railway and multimodal transport systems, employing mixed and longitudinal methodologies to deepen understanding of the relationship between perceived inclusion, trust, and loyalty to public transport.

Finally, the results support a phased approach to improving accessibility and inclusion in the Valparaíso Metro system. Short-term “quick wins” should focus on staff training, user-centered communication, and wayfinding audits. Medium-term actions should enhance digital accessibility—particularly through real-time information systems and mobile applications. In the long term, capital works such as intermodal redesign and universal infrastructure upgrades should align with EFE Valparaíso’s ongoing Limache–La Calera expansion cycle. This implementation sequence ensures that operational, digital, and infrastructural dimensions of accessibility evolve coherently toward a more inclusive and sustainable mobility system.