ABSTRACT
As urban transportation systems become more complex, relying solely on engineering indicators to promote active travel is insufficient, and travel satisfaction has come to be recognized as a determining psychological factor. While the majority of Iranian research studies have concentrated on the objective dimensions of the built environment, this study, by developing and testing a comprehensive conceptual model, simultaneously analyzes the role of the objective and perceived environment, attitude, mode choice, and travel satisfaction in Urmia City. The methodological innovation of this research lies in the employment of a two-step PLS-SEM approach to evaluate the feedback pathway from satisfaction to attitudes, thereby making it possible to test the causal mechanism of the effect of experience on attitudes within a feedback loop framework using cross-sectional data. The findings indicate that the built environment indirectly influences mode choice through the perception of the environment and attitudes. Furthermore, the direct effect of mode choice on satisfaction is substantially reduced when controlling for socio-economic status, suggesting that the reported satisfaction with certain modes stems less from the nature of the mode itself than from structural constraints and unequal travel opportunities. In contrast, a strong satisfaction–attitude feedback pathway was obtained, indicating that satisfying travel experiences reinforce positive attitudes and create a cycle among experience, attitude, and behavior. The findings indicate that the transition to sustainable transportation in the urban context of Iran requires simultaneous attention to the perceived quality of the environment, the actual travel experience, and the amendment of structural constraints. Physical interventions will be effective only if they are accompanied by enhancements in the travel experience and improvements in users' perceptions.
Extended Abstract
Introduction
Rapid urbanization has intensified urban challenges such as air pollution, traffic congestion, and declining physical activity. Consequently, policymakers have increasingly promoted sustainable mobility solutions, among which active travel—walking and cycling—has gained particular attention due to its environmental, health, and social benefits. Concepts such as pedestrian‑oriented cities and the 15‑minute city emphasize modifying the built environment to support active travel.
Existing literature generally falls into three streams: (1) studies examining the direct effects of built environment characteristics—such as density, land‑use diversity, connectivity, and access to public transport—on travel mode choice; (2) studies based on the Theory of Planned Behavior that introduce attitudes as mediators between environment and behavior; and (3) studies focusing on travel satisfaction as a post‑travel experience influencing future attitudes and behaviors.
Despite attempts to integrate these perspectives, many studies rely on linear and cross‑sectional analyses and are largely based on data from developed countries, limiting their applicability to developing contexts. In Iran, although several studies have explored socio‑economic factors, environmental characteristics, and attitudes affecting active travel, the dynamic relationships among attitudes, satisfaction, and both objective and perceived environmental indicators remain underexplored, particularly in medium‑sized cities such as Urmia. Therefore, this study investigates how the built environment affects travel mode choice through travel attitudes and experiences. Its novelty lies in integrating objective and perceptual environmental indicators while conceptualizing travel satisfaction within a feedback framework to better inform urban design and active travel policies.

Methodology
This study examines the relationships among the built environment, travel attitudes, socio‑economic characteristics, and travel satisfaction in Urmia. Research variables were grouped into six categories: objective built environment, perceived built environment, socio‑economic characteristics, travel attitudes, travel characteristics, and travel satisfaction.
Objective environmental indicators were calculated using GIS. The residential location of each respondent was mapped and a 1000‑meter buffer was generated, within which indicators such as population density, land‑use mix, sidewalk density, pavement quality, distance to key land uses, and bus stop density were measured. Perceived environmental indicators and attitudes were collected through a five‑point Likert‑scale questionnaire covering factors such as safety, visual attractiveness, cleanliness, shade, walking comfort, and traffic management. Socio‑economic variables (age, gender, education, income, car ownership, household size) and travel characteristics (mode, duration, purpose, and mode combinations) were also recorded. Travel Satisfaction (STS) was measured using Ettema’s standardized scale, incorporating affective and cognitive dimensions.
Data were collected between April and June 2025 from residents over 18 years old in Urmia. Using Cochran’s formula, a sample size of 400 was estimated. Stratified systematic random sampling produced 430 questionnaires, of which 392 valid responses were analyzed. Data sources included the Urmia Municipality GIS database and in‑person surveys. A pilot test with 30 participants confirmed content validity through expert review and reliability with Cronbach’s alpha values above 0.7.
Data analysis was conducted using Partial Least Squares Structural Equation Modeling (PLS‑SEM). A two‑stage approach was applied to model the feedback effect: first, latent scores of travel satisfaction were extracted; then these scores were entered as predictors of travel attitudes in the second stage. Descriptive analyses were performed in SPSS, and the structural model was estimated using SmartPLS.


Results and Discussion
The results confirmed the adequacy of the measurement model. All indicator loadings exceeded 0.7, and Cronbach’s alpha, composite reliability, and average variance extracted were within acceptable thresholds, indicating satisfactory reliability and convergent validity. Structural model assessment using bootstrapping (5,000 replications) demonstrated strong predictive capability, explaining 57.4% of the variance in travel satisfaction, 74.6% in travel mode choice, and 75.4% in travel attitudes.
The findings show that travel mode choice is the most influential determinant of travel satisfaction, followed by travel attitudes, which affect satisfaction both directly and indirectly through mode choice. The perceived built environment also significantly influences satisfaction, with a stronger effect than the objective built environment. In contrast, objective environmental indicators contribute the least to explaining satisfaction, suggesting that individuals’ perceptions and travel experiences play a more decisive role than purely physical characteristics.
Further analysis indicates that the perceived built environment has a stronger impact on travel attitudes, travel mode choice, and travel satisfaction than the objective environment. Although the inclusion of socio‑economic status as a control variable reduced the magnitude of these effects, the perceived environment remained the strongest predictor of travel attitudes. This result implies that while perceptions are partly shaped by socio‑economic conditions, the experienced quality of the environment continues to exert a substantial influence on travel behavior.
Regarding attitudes, the initial model showed that travel attitudes significantly influence mode choice and indirectly enhance satisfaction. However, this relationship weakened considerably after controlling for socio‑economic factors, indicating that part of the observed association is attributable to inequalities in access to resources and opportunities. Individuals with higher socio‑economic status tend to hold more positive travel attitudes and have greater ability to select preferred travel options.
Similarly, the initially strong relationship between travel mode choice and satisfaction declined when socio‑economic and travel characteristics were introduced as control variables. This suggests that higher satisfaction with certain modes is largely related to contextual factors such as shorter travel time, greater comfort, and better accessibility rather than the intrinsic nature of the mode itself.
Finally, the study confirmed a reciprocal relationship between travel satisfaction and travel attitudes. In the second stage of the model, travel satisfaction was introduced as an independent variable, and the results showed that it positively reinforces travel attitudes. This finding highlights a feedback mechanism in which positive travel experiences shape future attitudes and behaviors, creating a reinforcing cycle between travel experience, attitudes, and travel choices.

Conclusion
The research findings indicates that travel behavior is a dynamic and feedback-driven process, not merely a linear one. In this model, travel satisfaction is not merely an ultimate outcome; rather, by influencing subsequent attitudes, it can create a cycle that either reinforces or weakens travel behaviors. Positive experiences increase the likelihood of repeating desirable behaviors such as walking, while negative experiences reduce it. Therefore, to achieve lasting changes in travel behavior, urban management must simultaneously concentrate on improving the perceived quality of the environment and reducing structural and social inequalities.
The study, however, faced several limitations: attitudes were surveyed as a global construct, and their cognitive, affective, and behavioral dimensions were not disaggregated; the data were cross-sectional, thus constraining the ability to conclusively establish causal and feedback relationships; furthermore, reliance on self-reported data may introduce subjective bias. In addition, the results are limited to the urban context of Urmia, and further comparative studies are needed to generalize them to other cities.
Accordingly, it is recommended that in urban policy-making, rather than focusing solely on technical and physical indicators, citizens' perceptions and satisfaction should serve as the basis for prioritizing interventions. Low-cost yet effective improvements should be prioritized such as lighting, spatial ordering, and enhancing the sense of security. Furthermore, urban projects should undergo pre- and post-implementation evaluations based on user satisfaction. Policies should be adapted to local socio-economic conditions, and direct imitation of foreign models is inadvisable.

Funding
There is no funding support.

Authors’ Contribution
This article is derived from the first author’s (Amin Khalili) doctoral dissertation. He carried out all stages of the research, including writing the text, data processing, analyses, and extraction of results. Dr. Kyoumars Habibi (primary supervisor) contributed to developing the conceptual model, refining the problem statement and article texts, as well as the GIS process and map production. Dr. Bakhtiar Bahrami (secondary supervisor) and Dr. Samira Ramezani (advisor) played roles in scientific guidance, overseeing the research process, and the final review of the article

Conflict of Interest
Authors declared no conflict of interest.

Acknowledgments
We are grateful to all the scientific consultants of this paper.