Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7, 1893-1908 2025 Publisher: Learning Gate DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate © 2025 by the authors; licensee Learning Gate History: Received: 21 May 2025; Revised: 1 July 2025; Accepted: 4 July 2025; Published: 23 July 2025 * Correspondence: wulan_asti@sbm-itb.ac.id Predicting post-COVID-19 commuter transportation choice in greater Jakarta: A logistic regression approach to behavioral analysis and adaptation strategies Wulan Asti Rahayu1*, Utomo Sarjono Putro2, Pri Hermawan3, Yos Sunitiyoso4 1,2,3,4School of Business and Management, Institut Teknologi Bandung (ITB), Indonesia; wulan_asti@sbm-itb.ac.id (W.A.R.) utomo@sbm-itb.ac.id (U.S.P.) pri.hermawan@sbm-itb.ac.id (P.H.) yos.sunitiyoso@sbm-itb.ac.id (Y.S.). Abstract: This study explores the factors influencing commuter transportation choices in Jakarta during the COVID-19 pandemic through logistic regression analysis, integrating the Theory of Planned Behavior (TPB) with Customer Satisfaction Theory to examine the impact of psychological factors, service-related aspects, and environmental awareness on public transport use. Based on a survey of 254 commuters across Greater Jakarta, the research employs a forward stepwise logistic regression approach to identify key predictors of transportation behavior. The final model demonstrates robust predictive power with a classification accuracy of 69.7% and an ROC score of 0.697, revealing that private vehicle preference and health concerns negatively influence public transport adoption, while environmental awareness and fatigue from driving positively affect modal shift intentions. The findings emphasize the necessity of both push strategies—addressing health and safety concerns—and pull strategies—enhancing public transportation attractiveness through improved reliability, comfort, and environmental messaging. The research highlights the critical need for targeted urban mobility policies that integrate health-focused and environmental messaging to promote sustainable transportation choices, particularly in rapidly urbanizing areas like Jakarta. By addressing multifaceted challenges through comprehensive approaches that combine immediate health concerns with long-term environmental benefits, policymakers can develop more resilient urban mobility systems and reduce reliance on private vehicles in the post-pandemic era. Keywords: Commuter behavior, COVID-19 pandemic, Logistic regression, Modal shift, Public transportation, Urban mobility. 1. Introduction The rapid urbanization and economic growth in Asia's newly industrialized countries have increased travel demands, resulting in a rise in private vehicle ownership driven by higher incomes and societal perceptions linking cars to status [1-3]. This shift has worsened traffic congestion, environmental pollution, and quality-of-life issues, as many developing countries do not have sufficient public transportation systems to reduce motorization [4]. Unlike developed countries with established mass transit networks, developing cities encounter significant challenges in improving public transport, which is essential for addressing environmental degradation, promoting social equity, and alleviating the pressures of urban expansion [5]. The COVID-19 pandemic has further complicated the urban mobility landscape by significantly altering travel behaviors. According to Sushma and Prusty [6] the pandemic prompted a dramatic reduction in the use and demand for public transportation systems. This shift was driven by strict public transport restrictions aimed at limiting virus transmission and by the public's heightened fear of contracting COVID-19 in crowded, enclosed spaces. Studies reveal a general decline in public transport use due to perceived infection risks, with a concurrent rise in private vehicle use and active transportation modes such as walking and cycling [7]. Public transportation, particularly during 1894 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate peak hours, poses a higher risk of virus contamination due to limited space and high passenger density. Consequently, many individuals have turned to private vehicles as a safer alternative, increasing private transport usage [8]. One of the ways of resolving these problems is to increase public transport usage instead of private cars by including certain factors such as comfort, risk reduction, the ticket price, convenience, time and frequency of public buses daily [9]. Governments must prioritize promoting public transport (PT) to counter the negative impacts of excessive car usage. Encouraging commuters to choose buses, trains, and other forms of public transit can help improve traffic conditions, mitigate environmental hazards, and align with Sustainable Development Goals (SDGs) related to sustainable transport [10]. Jakarta, the capital of Indonesia, along with its surrounding cities known as Jabodetabek, represents a notable model of urban development aimed at addressing the challenges of urbanization such as congestion, pollution, and inadequate infrastructure. The city has initiated significant public transportation improvements, including the introduction of Mass Rapid Transit (MRT) and Light Rail Transit (LRT) systems, which set benchmarks for other Indonesian cities facing similar issues [11]. However, despite these advancements, public transportation adoption remains low, with utilization rates between 5%-20% in certain areas and a total penetration of just 26.7% among 3.2 million commuters. Through various initiatives, the Jakarta Transportation Management Agency aims to reach a target of 60% of public transportation usage by 2030. However, it faces obstacles like poor system integration, limited service coverage, and a cultural inclination towards private vehicles [12]. This focus on Jakarta mirrors the challenges faced by other major global cities such as Bangkok [13] Manila [14] Mumbai [15] Cairo [16] and São Paulo [17]. These urban centers are characterized by rapidly increasing vehicle counts, infrastructure that struggles to keep pace with demand, and public transportation systems that need modernization and expansion. These urban centers are characterized by rapidly increasing vehicle counts, infrastructure that struggles to keep pace with demand, and public transportation systems that need modernization and expansion. Studying the commuting patterns of the middle and upper classes, often referred to as 'middle-up' in Asia's rapidly urbanizing regions, is essential due to the increasing travel demands and a decrease in private car usage. Understanding the travel behaviors of the middle-up demographic is critical for effective sustainable development planning [18]. Examining middle-up commuting patterns can guide strategies to transition to sustainable transport modes such as public transit. This shift can help address issues like traffic congestion, pollution, and global warming, ultimately improving quality of life [19]. Given the unique challenges faced by developing countries like Indonesia, with Jakarta as a prime example, it is essential to examine how commuter behavior and preferences have changed during crises such as the COVID-19 pandemic. This paper aims to address the existing gap in the literature by providing a comprehensive analysis of commuter behavior and adaptation strategies in Jakarta during times of crisis. By offering new insights into sustainable urban mobility strategies, this research seeks to contribute to developing more resilient and adaptable public transport systems in developing countries. This study employs logistic regression analysis to examine the factors influencing commuters’ transportation choices in Jakarta during the COVID-19 pandemic. It identifies critical predictors that shape commuter behavior by analyzing variables such as attitudes toward private vehicles and public transportation, willingness to switch modes of transportation, and familiarity with new mobility technologies. Research also examines additional factors, specifically the impact of COVID-19 experiences, focusing on trust and perceptions of health safety risks. This approach offers valuable insights into the interactions between these factors, which can aid in developing strategies to promote sustainable urban mobility. 2. Literature Review 2.1. Integration of Theory of Planned Behavior and Customer Satisfaction Theory The Theory of Planned Behavior (TPB), developed by Ajzen (1991), serves as a foundational framework widely utilized to predict public transportation behavior. TPB suggests that behavioral 1895 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate intentions are shaped by attitudes, subjective norms, and perceived behavioral control, all of which have been consistently validated as significant predictors of public transit use [16, 20, 21]. Customer Satisfaction Theory complements TPB by examining how the gap between customer expectations and actual experiences influences satisfaction and subsequent behavior [22]. Satisfaction arises when experiences exceed expectations (positive disconfirmation), while dissatisfaction occurs when they fall short (negative disconfirmation) [23]. Integrating TPB with Customer Satisfaction Theory provides a comprehensive understanding of public transit use. Studies such as those by Fu and Juan [24] and Zhang and Hayashi [25] have demonstrated that combining these theories helps predict transit use intentions by considering attitudes, norms, perceived control, and satisfaction, especially during the COVID-19 pandemic. Extensions of TPB have further enriched its predictive power. Ng and Phung [26] added personal norms and environmental concerns, while Donald, et al. [20] incorporated habit, showing that car use is driven by both intention and habit, whereas public transport use depends more on intention alone. 2.2. Public Transportation Behavior During the Pandemic The COVID-19 pandemic has significantly transformed public transportation systems around the world, resulting in notable changes in travel behavior, user perceptions, and overall mobility patterns. This transformation is particularly pronounced in developing countries, where the pandemic accelerated existing trends toward private vehicle preference while simultaneously highlighting the vulnerability of public transportation systems during health crises [6]. Indonesia experiencing significant shifts in mobility preferences [27]. Despite the easing of restrictions, significant changes in trip frequency and mode choice have persisted into the endemic phase, with public transportation usage remaining substantially reduced [27]. In Jakarta, Indonesia, the pandemic had a substantial impact on the adoption of public transportation. During this time, passenger behavior was mainly influenced by their perception of the effectiveness of safety measures, their confidence in the public transport system, and the strict enforcement of health protocols [16]. To maintain ridership and uphold public trust, transportation authorities in Jakarta have changed their communication strategies. Instead of just providing routine service updates, they now take a more comprehensive approach. This includes sharing important health information and offering mental health support through social media platforms [28]. This comprehensive communication strategy was crucial for keeping the public informed and engaged during the crisis. The relationship between mobility patterns and COVID-19 transmission rates has provided important insights into public transportation usage. Studies have shown that an increase in community mobility often correlates with higher COVID-19 case numbers. This highlights the necessity of a balanced approach that maintains essential transportation services while minimizing public health risks [11, 29]. In densely populated urban areas such as Kalianyar in West Jakarta, implementing community-led health protocols and micro-scale movement restrictions significantly influenced public transit use patterns [30]. These findings highlight the importance of localized interventions and the necessity of tailoring public health strategies to specific urban contexts. 3. Methodology This research employs a quantitative methodology due to its ability for producing reliable and generalizable outcome data [31]. The study's approach comprised three main stages, namely data collection, data analysis by using Logistic Regression Analysis (LRA), and model evaluation, as shown in Figure 1. 1896 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate Figure 1. Research Design. 3.1. Data Collection The initial phase involved conducting a pilot test with 30 participants to assess the preliminary reliability of the questionnaire. The reliability of the collected data was evaluated using Cronbach's alpha, with a minimum acceptable value set at 0.70 [32]. The second phase consisted an extensive survey conducted to evaluate different facets of commuter conduct and preferences within the framework of urban transportation. This survey was conducted with 254 individuals residing in the Jakarta, Tangerang, Depok, and Bogor. These measures consists of a) Demographic; b) Commuters' attitudes towards private vehicles, including the perceived sense of independence, control, speed, and lifestyle implications connected with owning a private vehicle [15, 33-35]. Commuters' perspectives on public transportation, specifically focusing on its perceived effectiveness, comfort, environmental consequences, and health-related considerations which related to the shifting of COVID-19 [19, 36-38]. Commuters’ criteria that increase willingness people to switch into public transport, based on time savings, environmental impact, comfort, cost, and health considerations [12, 17, 39-41]. Type of commuters, including electric automobiles, ride-hailing services, ride-sharing platforms, bike-sharing programs, electric scooters, and high-speed trains [12, 42-45]. Commuters' interest in trying or using these mobility technologies [46-48]. The variable construct used in this research is summarized in Table 1. 1897 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate Table 1. Variable’s Construct. Code Construct Q1 Private vehicles provide the freedom to go wherever I want. Q2 With a private vehicle, I have control over my journeys. Q3 Typically, private vehicles are the fastest means of reaching my destination. Q4 It would be challenging for me to adapt if I had to live without a private vehicle every day. Q5 Only private vehicles align with my lifestyle. Q6 I used to use PT for years, but not anymore since I got my private vehicle. Q7 I enjoy driving and love my private vehicle. Q8 The type of private vehicle someone drives often reflects their lifestyle and social status. Q9 I don't like driving with people I don't know. Q10 People I know find it strange if I don't have a private vehicle. Q11 I use a private vehicle because I feel safer in terms of health (e.g., less susceptible to diseases from others). Q12 PT is only intended for those less fortunate. Q13 Using PT wastes my time. Q14 I only use PT if I have no other choice. Q15 There are many problems and difficulties in using PT. Q16 I feel happy when using PT. Q17 I have a positive opinion about PT. Q18 I feel relaxed and enjoy my time when using PT compared to private vehicles. Q19 I often experience fatigue from using a car repeatedly and choose to use PT. Q20 Using PT helps improve the quality of the environment. Q21 If PT is crowded, I will wait for the next one. Q22 I am concerned about potential health issues (e.g., getting sick from others) when using PT. Q23 I have considered changing my mode of transportation during my travels. Q24 I will change my mode of transportation if it can save time. Q25 I will switch my mode of transportation to protect the environment. Q26 I am willing to pay more when traveling to help the environment. Q27 I rarely use private vehicles to be environmentally conscious. Q28 I choose the type of vehicle that offers the most comfort without considering the cost. Q29 I always use the fastest type of vehicle even if I have a cheaper alternative. Q30 I avoid traveling at certain times because it makes me very tired. Q31 If the distance is short, I prefer to walk during the daytime. Q32 Having a relaxed and stress-free journey is more important than arriving quickly. Q33 I am committed to sticking with the mode of transportation I frequently use, even if it's perceived as riskier. Q34 I am dedicated to using the mode of transportation I frequently use, even if it comes at a higher cost. Q35 I am determined to continue using the mode of transportation I frequently use, even if it requires more physical effort. Q36 I am inclined to maintain my usual mode of transportation, even if it results in a longer journey. Q37 Usually, I feel accomplished and content when I reach my destination. Q38 I prefer a mode of transportation that can guarantee health and safety inside the vehicle. Q39 I have a good understanding of the developments in mobility technology related to Electric Cars. Q40 I am well-informed about the advancements in mobility technology related to Ride Hailing (e.g., Online Motorcycle Taxis and Ride-Sharing Services). Q41 I have a good grasp of the developments in mobility technology related to Ride Sharing (e.g., Carpooling services like Nebengers). Q42 I am knowledgeable about the advancements in mobility technology related to Bike Sharing (e.g., Services like Gowes and Boseh). Q43 I am well-versed in the developments in mobility technology related to Electric Scooters (E-Scooters). Q44 I am well-informed about the advancements in mobility technology related to High-Speed Trains. Q45 I am highly interested in trying or using Electric Cars. Q46 I am highly interested in trying or using Ride Hailing services, such as Online Motorcycle Taxis and Ride- Sharing Services. Q47 I am highly interested in trying or using Ride Sharing services, such as Carpooling services like Nebengers. Q48 I am highly interested in trying or using Bike Sharing services, including services like Gowes and Boseh. Q49 I am highly interested in trying or using Electric Scooters (E-Scooters). Q50 I am highly interested in trying or using High-Speed Trains. 1898 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate 3.2. Data Analysis The third phase of the study utilized Logistic Regression Analysis (LRA) to further analyze the data, focusing on binary categorical outcomes of commuter behaviors. LRA is a statistical technique employed when the dependent variable is categorical, specifically binary (e.g., success/failure or presence/absence), and the objective is to estimate the probability of an event occurring based on a set of independent variables. The logistic regression model for binary outcomes is represented by the equation: log ( 𝑝 (1 − 𝑝) ) = 𝛽0 + 𝛽1𝑋1 + 𝛽2𝑋2 + ⋯ + 𝛽𝑛𝑋𝑛 Where: 1. P represents the probability of choosing a particular transport mode 2. 𝛽0 is the intercept 3. 𝛽1, 𝛽2, . . , 𝛽𝑛 are the coefficients 4. 𝑋1, 𝑋2, . . , 𝑋𝑛 are the independent variables Previous studies have demonstrated the application of LRA in analyzing shifts in public transport usage. For example, Mazanec, et al. [49] utilized multinomial logistic regression to estimate the probability of choosing different transport modes during the pandemic. Similarly, Tan and Ma [50] employed a logistic model to analyze commuters' choice of rail transit, considering factors such as occupation, pre-pandemic commuting habits, and perceived infection risk. Javid, et al. [51] integrated logistic regression within a structural equation modeling framework to examine public transport use under COVID-19 precautions, while Shelat, et al. [18] applied logistic regression to assess the impact of COVID-19 risk perceptions on route choice behavior in train networks. Furthermore, Akuh et al [52] used logistic regression to study transport mode shifts among international students in Wuhan, China, during the pandemic. The effectiveness of an LRA model can be assessed using various metrics, including classification accuracy (the percentage of correctly predicted cases), the area under the receiver operating characteristic (ROC) curve (which illustrates the true positive rate versus the false positive rate), and pseudo-R-squared measures (such as Nagelkerke's R-squared, which quantifies the proportion of variance explained by the model). A model is generally considered robust if it achieves a classification accuracy of 70% or higher, an area under the ROC curve of 0.7 or greater, and a Nagelkerke's R-squared value of 0.2 or higher. These metrics are essential for evaluating the model's predictive capabilities and overall fit. 3.3. Validation Testing The model validation process focused on ensuring the reliability and robustness of the logistic regression model through a series of well-structured steps, beginning with data preparation and advancing through model refinement and evaluation. Initially, the dataset consisted of 228 cases, with 221 cases retained for final analysis after excluding those with missing data. The dependent variable, representing commuter behavior, was encoded as a binary outcome, while categorical variables were transformed into dummy variables to allow proper inclusion in the model. A forward stepwise logistic regression approach was employed, where predictor variables were added incrementally, enhancing the model's explanatory power with each step. The Omnibus Tests of Model Coefficients confirmed the statistical significance of these predictors, evidenced by consistent decreases in the -2 Log Likelihood values. The model's goodness of fit was evaluated using the Hosmer-Lemeshow test, which showed improved calibration across the analysis, culminating in a well-fitting final model. Additionally, the predictive power of the model was assessed through a Receiver Operating Characteristic (ROC) Curve analysis, which demonstrated the model's increasing ability to distinguish between outcomes, as reflected in the satisfactory area under the curve (AUC) achieved in the final model. 1899 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate 3.4. Ethics Statement This research was conducted following ethical guidelines that prioritize participant safety, confidentiality, and informed consent. All participants were fully informed of the study’s purpose and voluntarily provided consent to participate. The research adhered to the ethical principles outlined in the Helsinki Declaration, ensuring that no participants were exposed to harm or risk. Data privacy and confidentiality were rigorously maintained throughout the study, and ethical approval was obtained prior to commencing the research. 4. Result 4.1. Research Demography A survey of 254 commuters across the Greater Jakarta area provides detailed insights into these dynamics. The gender distribution (54% male, 46% female) reflects a balanced view of commuter behavior, while the geographic diversity, including areas like Depok City, South Tangerang, and South Jakarta, reveals regional differences in transportation preferences and challenges. The age and occupational diversity of respondents further enrich the analysis. Younger commuters (aged 20-25), who constitute 30% of the sample, may be more open to adopting PT if modern technology and improved services are integrated. In contrast, older commuters, particularly those over 40, tend to prefer private vehicles due to established habits and a focus on convenience. Occupationally, private employees (48% of the sample) may prioritize PT options with reliable schedules and connectivity to business districts, while students might value affordability and accessibility. These findings highlight the need for targeted PT policies that cater to the specific needs of different commuter segments, rather than a one-size-fits-all approach. The research demography data can be seen in Table 2. 4.2. Incremental Model Enhancement and Fit Evaluation The analysis was conducted on a dataset comprising 228 cases, with 221 cases included in the final analysis after excluding 7 cases due to missing data. The dependent variable was encoded as a binary outcome, facilitating the logistic regression model’s prediction of transportation choices. Categorical variables such as Q12, Q11, and Q3 were encoded using dummy variables to capture the effects of different groups on the dependent variable. 1900 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate Table 2. Research Demography. Total Percentage Gender Man 137 54% Woman 117 46% Work Private employees 123 48% Student / Student 48 19% Government employees 28 11% BUMN/BUMD employees 22 9% Entrepreneur / Entrepreneur 19 7% Unemployed 14 6% Socioeconomic Status (SES) A 150 59% B 37 15% C 68 27% Age <20 4 2% 20-25 77 30% 26-30 45 17% 36-40 50 20% >40 47 18% Marital Status Married 128 50% Single 126 50% Location of residence South Jakarta City 53 21% East Jakarta City 42 16% Central Jakarta City 5 2% West Jakarta City 19 7% City and Regency of South Tangerang 46 18% City and District of Depok 68 27% Bogor City and Regency 21 9% 4.3. Model Improvement Model improvement was achieved using a forward stepwise selection method based on the Likelihood Ratio, where variables were added incrementally to enhance the model’s explanatory power. The process began with an initial model (Block 0) that included only the constant term, serving as a baseline for comparison. The initial model had limited explanatory power, with a classification accuracy of 54.8%. As variables were introduced in subsequent steps, the model’s fit significantly improved, as indicated by the decreasing -2 Log Likelihood values and increasing Cox & Snell R Square and Nagelkerke R Square values. The Omnibus Tests of Model Coefficients confirmed that each step significantly enhanced the model’s performance, with chi-square values consistently demonstrating improvements in fit. By Step 5, the model accounted for approximately 32.1% of the variance in the dependent variable, as reflected by the Nagelkerke R Square value. Key variables such as Q27E_inv, Q28H_inv, Q27K_inv, Q29C_inv, and 1901 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate Q29K_inv were included based on their significant contributions, resulting in a robust model with improved predictive power. Table 3 summarizes the progression of key model fit indicators, highlighting the effectiveness of the stepwise approach in refining the logistic regression model and enhancing its accuracy. Table 3. Summary of Model Improvement. Step -2 Log Likelihood Cox & Snell R Square Nagelkerke R Square 1 281.586 0.132 0.177 2 269.914 0.176 0.235 3 261.658 0.205 0.274 4 256.124 0.224 0.3 5 251.304 0.24 0.321 4.4. Goodness of Fit Evaluation The Hosmer-Lemeshow test is used to evaluate how well the logistic regression model fits the data. A non-significant p-value (greater than 0.05) indicates a good fit, meaning the model's predicted probabilities are well-calibrated with the observed outcomes. This test divides the data into deciles based on the predicted probabilities and compares the observed and expected frequencies of the outcome within each decile. In this analysis, the Hosmer-Lemeshow test results across the five steps are as shown in Table 4. Table 4. Hosmer-Lemeshow test results. Step Chi-square df Sig. 1 18.634 8 0.017 2 13.038 8 0.111 3 9.784 8 0.28 4 11.894 8 0.156 5 6.127 8 0.633 In Step 1, the Hosmer-Lemeshow test indicated a misfit (p = 0.017) between predicted and observed values, but by Step 5, the test showed a significant improvement with a chi-square of 6.127 and a p-value of 0.633, confirming an excellent fit. This progression indicates that the model's predicted probabilities align closely with observed outcomes, validating its predictive performance. Contingency analysis further demonstrated that as more variables were added, the alignment between observed and expected frequencies of the dependent variable improved, enhancing the model’s calibration. Correspondingly, classification accuracy increased, reaching 69.7% in Step 5, reflecting the model's growing ability to correctly classify outcomes as more predictors were incorporated (Table 5). Table 5. Classification accuracy. p Q18 = 0 Accuracy Q18=1 Accuracy Overall Accuracy 1 70.40% 54.40% 63.20% 2 72.80% 56.30% 65.40% 3 74.40% 59.20% 67.50% 4 73.60% 58.30% 66.70% 5 75.20% 63.10% 69.70% 4.5. Logistic Regression Equation The logistic regression analysis identifies key predictors influencing transportation choices (Q18_inv). Variables such as Q27E_inv (Private Vehicle Preference) and Q27K_inv (Health and Safety Concerns) showed negative coefficients, indicating that higher values decrease the likelihood of Q18_inv 1902 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate being 1, aligning with push policy strategies. Conversely, Q28H_inv (Fatigue from Driving), Q29C_inv (Environmental Awareness), and Q29K_inv (Risk Acceptance and Transportation Habits) displayed positive coefficients, increasing the likelihood of the outcome, which supports the implementation of pull policies. The final logistic regression model for binary outcomes is represented by the equation: log ( 𝑝(𝑥) (1 − 𝑝(𝑥)) ) = 0.471 − 2.77 𝑄27𝐸 − 2.92 𝑄27𝐾 + 0.161 𝑄28𝐻 + 0.216 𝑄29𝐶 + 0.152 𝑄29𝐾 Where 𝑝(𝑥) represents the probability of choosing a particular transport mode, 𝑄27𝐸 is Private Vehicle Preference, 𝑄27𝐾 is Health and Safety Concerns, 𝑄28𝐻 is Fatigue from Driving, 𝑄29𝐶 is Environmental Awareness, 𝑄29𝐾 is Risk Acceptance and Transportation Habits. 4.6. Wald Statistics and Odds Ratios Analysis The Wald statistic tests whether a variable's coefficient significantly differs from zero, indicating its importance in the model. For instance, Q27E_inv consistently showed high Wald statistics, confirming its role as a significant predictor with a negative impact on the outcome (Q18_inv = 1), which aligns with push policies aimed at reducing private vehicle use. Similarly, Q27K_inv displayed a negative relationship, further supporting the need for push strategies focused on health and safety concerns. Conversely, Q28H_inv, Q29C_inv, and Q29K_inv emerged as positive predictors in the model's later steps, indicating that enhancing these factors through pull policies could increase the likelihood of commuters choosing alternative transport modes. The model's predictive power, as indicated by the ROC curve score of 0.697, suggests it effectively distinguishes between different commuter choices, supporting the development of robust policy recommendations that integrate both push and pull strategies (Table 6). The model's predictive power, as measured by the Receiver Operating Characteristic (ROC) curve, yields a score of 0.697. An ROC score closer to 1 indicates better discriminatory power of the model, while a score of 0.5 would indicate no better than random chance. In this case, a score of 0.697 suggests that the model has good predictive power. This level of accuracy shows that the model effectively differentiates between commuters who choose different transport modes during the COVID-19 pandemic, allowing for robust policy recommendations to be drawn from the results. Table 6. Wald Statistics and Odds Ratios Analysis Results. Variable Step Wald Sig. Exp(B) 95% CI for Exp(B) Q27E_inv 1 28.078 0 0.746 0.669 - 0.831 2 21.88 0 0.765 0.684 - 0.856 3 19.762 0 0.772 0.688 - 0.865 4 15.882 0 0.789 0.703 - 0.887 5 18.846 0 0.758 0.668 - 0.859 Q28H_inv 2 11.03 0.001 1.23 1.088 - 1.389 3 8.517 0.004 1.206 1.063 - 1.367 4 5.738 0.017 1.173 1.029 - 1.336 5 5.646 0.017 1.175 1.029 - 1.342 Q27K_inv 3 7.48 0.006 0.767 0.635 - 0.928 4 9.454 0.002 0.732 0.600 - 0.893 5 8.445 0.004 0.746 0.613 - 0.909 Q29C_inv 4 5.184 0.023 1.217 1.028 - 1.441 5 5.887 0.015 1.241 1.042 - 1.477 Q29K_inv 5 4.579 0.032 1.164 1.013 - 1.338 1903 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate 4.7. Implications for Urban Mobility Policy 4.7.1. Push and Pull Strategies The analysis underscores the need for a comprehensive approach that integrates both push and pull strategies to influence commuter behavior effectively. Pull policies should focus on enhancing the attractiveness of public transportation by improving comfort, convenience, and environmental benefits [51]. The strong preference for private vehicles [53] highlights the necessity of making public transport a more appealing option through increased reliability, reduced travel times, and modern amenities. Moreover, the positive impact of environmental awareness [54] suggests that policies promoting the ecological benefits of public transport can effectively encourage a shift away from private vehicles. Public campaigns and incentives that emphasize the environmental advantages of using public transportation, coupled with education on sustainable practices, can significantly influence commuter choices. Push policies, on the other hand, are essential for overcoming the barriers that discourage the use of public transportation. The significant negative effect of health and safety concerns [55] reveals the need for policies that build trust in the safety of public transport. This can be achieved by implementing stringent health measures such as regular sanitization, reducing overcrowding through increased service frequency, and ensuring visible enforcement of safety protocols. These measures not only address the immediate concerns raised by the pandemic but also establish long-term safety standards that can sustain ridership and foster a culture of trust and reliability in public transportation. 4.7.2. Promoting Safe and Reliable Public Transportation The findings from the research demography highlight the significant challenges faced by Indonesia’s public transportation (PT) system, where user adoption rates remain alarmingly low, typically between 5% and 20%. This underutilization is particularly concerning in rapidly growing urban areas like Greater Jakarta, where the strain on transportation networks is most acute. The demographic insights from the survey, which involved 254 commuters, reveal a complex landscape of transportation preferences influenced by age, occupation, and socioeconomic status. Younger commuters, who are generally more open to adopting public transportation, could be pivotal in driving a shift towards greater PT usage. However, older commuters, especially those over 40, exhibit a strong preference for private vehicles, highlighting the need for targeted strategies that address the specific concerns of different commuter segments. To bridge this gap and improve PT adoption, enhancing safety and reliability in public transportation is essential. As the study's logistic regression analysis shows, factors like health and safety concerns (Q27K_inv) significantly deter commuters from choosing public transportation, echoing the importance of safety emphasized by Matsumoto, et al. [13] and Gutiérrez, et al. [56]. The incorporation of ICT technologies, such as real-time management systems, can address these concerns by improving the reliability and safety of public transportation, thus making it a more attractive option for commuters. The model improvement process, which significantly enhanced the predictive power of the logistic regression model, underscores the importance of these factors in influencing transportation choices. Moreover, the research indicates that while safety and reliability are crucial, they must be complemented by a broader strategy that includes both push and pull factors. The model's findings, particularly the positive influence of environmental awareness (Q29C_inv) and the fatigue from driving (Q28H_inv), suggest that policies promoting public transportation should not only focus on mitigating risks but also on highlighting the environmental and health benefits of PT. This dual approach, combining the push strategy of stricter safety regulations with pull strategies that make PT more attractive, aligns with the recommendations of Borhan, et al. [57] and Soza-Parra, et al. [58]. Such an integrated policy framework is likely to foster a culture of safety and reliability in public transportation, ultimately leading to increased ridership and reduced reliance on private vehicles. By addressing both 1904 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate the immediate and long-term needs of commuters, policymakers can enhance the overall effectiveness and sustainability of Indonesia’s public transportation system. 4.7.3. Encouraging Modal Shift Through Environmental and Health Messaging Given the complex dynamics of public transportation usage in Indonesia, especially in the Greater Jakarta area, it becomes clear that a multi-faceted approach is necessary to encourage a modal shift towards public transportation. The demographic analysis reveals significant variations in commuter behavior based on factors such as age, occupation, and socioeconomic status. For instance, younger commuters, particularly those aged 20-25, show a greater openness to adopting public transportation, especially when modern technology and improved services are involved. This aligns with the findings from Poudenx [54] who suggests that environmental awareness can be a powerful motivator for changing transportation behaviors. In this context, targeting younger, environmentally conscious commuters with strong ecological messaging can resonate well, prompting them to choose public transportation over private vehicles [59]. However, the data also indicates that environmental messaging alone may not suffice, particularly for older commuters and those with entrenched habits of private vehicle use. As the logistic regression analysis shows, health and safety concerns (Q27K_inv) significantly influence transportation choices, often deterring the adoption of public transportation. This insight aligns with the work of Bigsby, et al. [60] who emphasize the impact of health-focused campaigns in shaping public attitudes and behaviors. Similarly, Liu et al [61] stress that modal shifts, such as from truck to sea transportation, can significantly reduce health risks from pollution, highlighting the importance of incorporating health benefits into public transport promotion strategies. Therefore, combining environmental messages with health-focused campaigns that highlight the safety benefits of public transportation is crucial. Such dual messaging can address both the immediate health concerns and the long-term environmental impacts, creating a compelling case for public transportation among different commuter segments. The integration of these messages into a comprehensive policy framework is essential for encouraging a broader modal shift. As Crawshaw [62] notes, social marketing plays a crucial role in promoting healthy lifestyles by shifting focus from social to individual targets, making it an effective tool for influencing transportation behaviors. Additionally, White, et al. [12] propose the SHIFT framework, which suggests leveraging psychological factors like social influence and habit formation to encourage sustainable consumer behaviors, including transportation choices. This approach is further supported by Ogilvie et al [63] who found that targeted behavior change programs show the best evidence for promoting a modal shift towards more sustainable modes of transport. By incorporating these insights into policy design, particularly in rapidly urbanizing areas like Jakarta, policymakers can foster a more sustainable and health-conscious urban mobility system, ultimately reducing the reliance on private vehicles and addressing the current underutilization of public transportation infrastructure. 5. Conclusion This study provides valuable insights into the factors that influence commuter transportation choices in Jakarta during the COVID-19 pandemic, utilizing logistic regression analysis as a primary tool. By integrating the Theory of Planned Behavior (TPB) with Customer Satisfaction Theory, the research highlighted the complex interplay of psychological factors, such as attitudes and perceived behavioral control, alongside service-related factors like satisfaction and environmental awareness. The findings underscored the critical need for both push and pull strategies to promote a modal shift towards public transportation. Push strategies, including measures to alleviate health and safety concerns, are vital for overcoming barriers to public transport use, especially in the context of a global pandemic. On the other hand, pull strategies should focus on enhancing the attractiveness of public transport through improved reliability, comfort, and the promotion of environmental benefits, particularly targeting younger, environmentally conscious commuters. 1905 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate Moreover, the study's logistic regression model effectively identified key predictors of transportation choices, such as private vehicle preference, health and safety concerns, and environmental awareness. The robustness of the model, validated through various statistical metrics, provides a solid foundation for developing targeted urban mobility policies. The research suggests that a comprehensive approach, combining health-focused and environmental messaging, is essential for encouraging a broader modal shift, particularly in rapidly urbanizing areas like Jakarta. By addressing both the immediate health concerns and long-term environmental benefits of public transportation, policymakers can foster a more sustainable and resilient urban mobility system, ultimately reducing reliance on private vehicles and enhancing public transportation adoption in the post-pandemic era. Institutional Review Board Statement: Ethical clearance for this research was obtained from the relevant institutional review board at the School of Business and Management, Institut Teknologi Bandung (ITB), prior to the commencement of the study. Transparency: The authors confirm that the manuscript is an honest, accurate, and transparent account of the study; that no vital features of the study have been omitted; and that any discrepancies from the study as planned have been explained. This study followed all ethical practices during writing. Copyright: © 2025 by the authors. This open-access article is distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). References [1] S. Morichi, "Long-term strategy for transport system in Asian megacities," Journal of the Eastern Asia Society for Transportation Studies, vol. 6, pp. 1-22, 2005. [2] K. Nakamura and Y. Hayashi, "Strategies and instruments for low-carbon urban transport: An international review on trends and effects," Transport Policy, vol. 29, pp. 264-274, 2013. [3] Y. O. Susilo, W. Santosa, T. B. Joewono, and D. Parikesit, "A reflection of motorization and public transport in Jakarta metropolitan area," IATSS Research, vol. 31, no. 1, pp. 59-68, 2007. https://doi.org/10.1016/S0386- 1112(14)60184-9 [4] M. M. Zefreh, B. Hussain, and T. Sipos, "In-depth analysis and model development of passenger satisfaction with public transportation," KSCE Journal of Civil Engineering, vol. 24, no. 10, pp. 3064-3073, 2020. https://doi.org/10.1007/s12205-020-1871-7 [5] M. E. Parker et al., "Public transit use in the United States in the era of COVID-19: Transit riders’ travel behavior in the COVID-19 impact and recovery period," Transport Policy, vol. 111, pp. 53-62, 2021. [6] M. B. Sushma and S. Prusty, "The effect of covid-19 on public transportation sectors and conceptualizing the shifting paradigm: A report on Indian scenario." Singapore: Springer Singapore, 2022. https://doi.org/10.1007/978-981-16- 9744-9_2 [7] F. Linares-Rendón and M. Garrido-Cumbrera, "Impact of the COVID-19 pandemic on urban mobility: A systematic review of the existing literature," Journal of Transport & Health, vol. 22, p. 101225, 2021. https://doi.org/10.1016/j.jth.2021.101225 [8] A. Shabani, A. Shabani, B. Ahmadinejad, and A. Salmasnia, "Measuring the customer satisfaction of public transportation in Tehran during the COVID-19 pandemic using MCDM techniques," Case Studies on Transport Policy, vol. 10, no. 3, pp. 1520-1530, 2022. https://doi.org/10.1016/j.cstp.2022.05.009 [9] R. Buehler and J. Pucher, "Demand for public transport in Germany and the USA: An analysis of rider characteristics," Transport Reviews, vol. 32, no. 5, pp. 541-567, 2012. https://doi.org/10.1080/01441647.2012.707695 [10] C.-H. Huang, W.-C. Hsu, K.-I. Huang, S.-M. Hsu, and Y.-C. Huang, "The extension of the theory of planned behavior to predict the use of public transport," Asian Journal of Business and Management, vol. 3, no. 5, pp. 426-433, 2015. [11] A. W. Wijayanto, A. Rachman, M. H. L. Siahaan, and L. Hakim, "Human mobility patterns and their cross- correlation with the COVID-19 transmission in Jakarta, Indonesia," Journal of Physics: Conference Series, vol. 1863, no. 1, p. 012025, 2021. https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.1016/S0386-1112(14)60184-9 https://doi.org/10.1016/S0386-1112(14)60184-9 https://doi.org/10.1007/s12205-020-1871-7 https://doi.org/10.1007/978-981-16-9744-9_2 https://doi.org/10.1007/978-981-16-9744-9_2 https://doi.org/10.1016/j.jth.2021.101225 https://doi.org/10.1016/j.cstp.2022.05.009 https://doi.org/10.1080/01441647.2012.707695 1906 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate [12] K. White, R. Habib, and D. J. Hardisty, "How to SHIFT consumer behaviors to be more sustainable: A literature review and guiding framework," Journal of Marketing, vol. 83, no. 3, pp. 22-49, 2019. https://doi.org/10.1177/0022242919825649 [13] A. Matsumoto, M. An, C. Van Gulijk, and S. Kaewunruen, "Safety, risk and uncertainties in transportation and transit systems," Frontiers in Built Environment, vol. 5, p. 25, 2019. https://doi.org/10.3389/fbuil.2019.00025 [14] H. Kim and H. Moon, "Heterogeneous attitudes toward autonomous vehicles: Evaluation of consumer acceptance of vehicle automation technology using a latent class approach," Technology Analysis & Strategic Management, vol. 34, no. 12, pp. 1402-1417, 2022. https://doi.org/10.1080/09537325.2021.1962522 [15] C. Cirillo, Y. Liu, and M. Maness, "A time-dependent stated preference approach to measuring vehicle type preferences and market elasticity of conventional and green vehicles," Transportation Research Part A: Policy and Practice, vol. 100, pp. 294-310, 2017. https://doi.org/10.1016/j.tra.2017.04.028 [16] Y. Heath and R. Gifford, "Extending the theory of planned behavior: Predicting the use of public transportation 1," Journal of Applied Social Psychology, vol. 32, no. 10, pp. 2154-2189, 2002. [17] G. Vecchio and L. Tricarico, "May the force move you”: Roles and actors of information sharing devices in urban mobility," Cities, vol. 88, pp. 261-268, 2019. https://doi.org/10.1016/j.cities.2018.11.007 [18] S. Shelat, T. Van De Wiel, E. Molin, J. Van Lint, and O. Cats, "Analysing the impact of COVID-19 risk perceptions on route choice behaviour in train networks," PloS One, vol. 17, no. 3, p. e0264805, 2022. [19] Y. S. J. Chin, L. De Pretto, V. Thuppil, and M. J. Ashfold, "Public awareness and support for environmental protection—A focus on air pollution in peninsular Malaysia," PloS one, vol. 14, no. 3, p. e0212206, 2019. https://doi.org/10.1371/journal.pone.0212206 [20] I. J. Donald, S. R. Cooper, and S. M. Conchie, "An extended theory of planned behaviour model of the psychological factors affecting commuters' transport mode use," Journal of Environmental Psychology, vol. 40, pp. 39-48, 2014. https://doi.org/10.1016/j.jenvp.2014.03.003 [21] M. R. Rezaimoghadam, S. R. Davoodi, and N. Dabiri, "Public transportation user’s behavior based on theory of planned behavior," ALAM CIPTA International Journal of Sustainable Tropical Design & Practice, vol. 15, no. 1, pp. 28- 35, 2022. [22] P. Kotler and K. L. Keller, Marketing management, 16t ed. London, UK: Pearson Education Limited, 2021. [23] W. C. Hom, "Applying customer satisfaction theory to community college planning of counseling services," IJournal, 2002. [24] X. Fu and Z. Juan, "Understanding public transit use behavior: Integration of the theory of planned behavior and the customer satisfaction theory," Transportation, vol. 44, no. 5, pp. 1021-1042, 2017. https://doi.org/10.1007/s11116- 016-9692-8 [25] J. Zhang and Y. Hayashi, "Research frontier of COVID-19 and passenger transport: A focus on policymaking," Transport Policy, vol. 119, pp. 78-88, 2022. https://doi.org/10.1016/j.tranpol.2022.02.014 [26] P. Y. Ng and P. T. Phung, "Public transportation in Hanoi: Applying an integrative model of behavioral intention," Case Studies on Transport Policy, vol. 9, no. 2, pp. 395-404, 2021. [27] M. Z. Irawan et al., "Heterogeneity in activity and travel mode patterns of older Indonesians," Transportation Research Interdisciplinary Perspectives, vol. 26, p. 101159, 2024. https://doi.org/10.1016/j.trip.2024.101159 [28] E. P. Purnomo, M. J. Loilatu, A. Nurmandi, Z. Qodir, I. T. Sihidi, and M. Lutfi, "How public transportation use social media platform during covid-19: Study on Jakarta public transportations' twitter accounts?," Webology, vol. 18, no. 1, pp. 233-249, 2021. [29] R. O. Nanda et al., "Community mobility and COVID-19 dynamics in Jakarta, Indonesia," International Journal of Environmental Research and Public Health, vol. 19, no. 11, p. 6671, 2022. [30] R. Oktapriana, A. Hartono, and W. Wulandari, "Adaptation effort on dense residential areas to encounter COVID-19: A case study on the urban area of Kalianyar in West Jakarta," IOP Conference Series: Earth and Environmental Science, vol. 737, no. 1, p. 012033, 2021. [31] A. Steckler, K. R. McLeroy, R. M. Goodman, S. T. Bird, and L. McCormick, "Toward integrating qualitative and quantitative methods: An introduction," Health Education Quarterly, vol. 19, no. 1, pp. 1-8, 1992. https://doi.org/10.1177/109019819201900101 [32] J. C. Nunnally, Psychometric theory. New York: McGraw-Hill, 1967. [33] K. M. Buhmann and J. R. Criado, "Consumers' preferences for electric vehicles: The role of status and reputation," Transportation Research Part D: Transport and Environment, vol. 114, p. 103530, 2023. https://doi.org/10.1016/j.trd.2022.103530 [34] S. Hu, P. M. Chau, J. Weng, P. Lin, and W. Zhou, "Extending the theory of planned behavior to explore the influence of residents’ dependence on public transport," IEEE Access, vol. 9, pp. 137224-137240, 2021. [35] W. Kim, J. Kim, J. A. Jang, and S.-H. Kim, "Do perceptions of hydrogen energy effect on vehicle preference? A learning-based model approach," KSCE Journal of Civil Engineering, vol. 27, no. 2, pp. 805-813, 2023. https://doi.org/10.1007/s12205-022-0227-x https://doi.org/10.1177/0022242919825649 https://doi.org/10.3389/fbuil.2019.00025 https://doi.org/10.1080/09537325.2021.1962522 https://doi.org/10.1016/j.tra.2017.04.028 https://doi.org/10.1016/j.cities.2018.11.007 https://doi.org/10.1371/journal.pone.0212206 https://doi.org/10.1016/j.jenvp.2014.03.003 https://doi.org/10.1007/s11116-016-9692-8 https://doi.org/10.1007/s11116-016-9692-8 https://doi.org/10.1016/j.tranpol.2022.02.014 https://doi.org/10.1016/j.trip.2024.101159 https://doi.org/10.1177/109019819201900101 https://doi.org/10.1016/j.trd.2022.103530 https://doi.org/10.1007/s12205-022-0227-x 1907 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate [36] N. Aydin, A. O. Kuşakcı, and M. Deveci, "The impacts of COVID-19 on travel behavior and initial perception of public transport measures in Istanbul," Decision Analytics Journal, vol. 2, p. 100029, 2022. https://doi.org/10.1016/j.dajour.2022.100029 [37] L. Magnusson, I. Kebbie, and V. Jerwanska, "Access to health and rehabilitation services for persons with disabilities in Sierra Leone – focus group discussions with stakeholders," BMC Health Services Research, vol. 22, no. 1, p. 1003, 2022. https://doi.org/10.1186/s12913-022-08366-8 [38] X. Zhang, D. Liu, Y. Wang, and H. Du, "Behavioral intentions of urban rail transit passengers during the covid‐19 pandemic in Tianjin, China: A model integrating the theory of planned behavior and customer satisfaction theory," Journal of Advanced Transportation, vol. 2021, no. 1, p. 8793101, 2021. https://doi.org/10.1155/2021/8793101 [39] D. Hopkins, E. García Bengoechea, and S. Mandic, "Adolescents and their aspirations for private car-based transport," Transportation, vol. 48, no. 1, pp. 67-93, 2021. https://doi.org/10.1007/s11116-019-10044-4 [40] D. Mehar, R. K. Singh, and A. K. Gupta, "A review on battery technologies and their challenges in electrical vehicles," presented at the IEEE International Students’ Conference on Electrical, Electronics, and Computer Science, SCEECS, 2023. [41] F. Mikiki, A. Oikonomou, and E. Katartzi, "Sustainable mobility issues of physically active university students: The case of Serres, Greece," Future Transportation, vol. 1, no. 3, pp. 777-793. https://doi.org/10.3390/futuretransp1030043 [42] F. Alonso, M. Faus, M. T. Tormo, and S. A. Useche, "Could technology and intelligent transport systems help improve mobility in an emerging country? challenges, opportunities, gaps and other evidence from the caribbean," Applied Sciences, vol. 12, no. 9, p. 4759. https://doi.org/10.3390/app12094759 [43] D. Mukhtar-Landgren and A. Paulsson, "Governing smart mobility: Policy instrumentation, technological utopianism, and the administrative quest for knowledge," Administrative Theory & Praxis, vol. 43, no. 2, pp. 135-153, 2021. https://doi.org/10.1080/10841806.2020.1782111 [44] J. Paul, A. Modi, and J. Patel, "Predicting green product consumption using theory of planned behavior and reasoned action," Journal of Retailing and Consumer Services, vol. 29, pp. 123-134, 2016. https://doi.org/10.1016/j.jretconser.2015.11.006 [45] H. Wang, Q. Yu, Y. Liu, D. Jin, and Y. Li, "Spatio-temporal urban knowledge graph enabled mobility prediction," Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 5, no. 4, pp. 1-24, 2021. https://doi.org/10.1145/3494993 [46] G. D. P. A. Aschwanden, J. S. Wijnands, J. Thompson, K. A. Nice, H. Zhao, and M. Stevenson, "Learning to walk: Modeling transportation mode choice distribution through neural networks," Environment and Planning B: Urban Analytics and City Science, vol. 48, no. 1, pp. 186-199, 2019. https://doi.org/10.1177/2399808319862571 [47] J. Li, X. Pei, X. Wang, D. Yao, Y. Zhang, and Y. Yue, "Transportation mode identification with GPS trajectory data and GIS information," Tsinghua Science and Technology, vol. 26, no. 4, pp. 403-416, 2021. https://doi.org/10.26599/TST.2020.9010014 [48] X. Zhou, Z. Yu, L. Yuan, L. Wang, and C. Wu, "Measuring accessibility of healthcare facilities for populations with multiple transportation modes considering residential transportation mode choice," ISPRS International Journal of Geo-Information, vol. 9, no. 6, p. 394doi: https://doi.org/10.3390/ijgi9060394. [49] J. Mazanec, V. Harantová, V. Štefancová, and H. Brůhová Foltýnová, "Estimating mode of transport in daily mobility during the COVID-19 pandemic using a multinomial logistic regression model," International Journal of Environmental Research and Public Health, vol. 20, no. 5, p. 4600, 2023. [50] L. Tan and C. Ma, "Choice behavior of commuters' rail transit mode during the COVID-19 pandemic based on logistic model," Journal of Traffic and Transportation Engineering (English Edition), vol. 8, no. 2, pp. 186-195, 2021. https://doi.org/10.1016/j.jtte.2020.07.002 [51] M. A. Javid, M. Abdullah, N. Ali, and C. Dias, "Structural equation modeling of public transport use with COVID-19 precautions: An extension of the norm activation model," Transportation Research Interdisciplinary Perspectives, vol. 12, p. 100474, 2021. [52] R. Akuh et al, "The influence of COVID-19 on transport mode shift: A case of african international students in Wuhan-China," presented at the 2023 7th International Conference on Transportation Information and Safety (ICTIS), 2023. [53] M. M. Zefreh, A. Torok, and A. Barabadi, "The role of transportation in social equity and urban development," Journal of Urban Planning and Development, vol. 146, no. 2, p. 04020015, 2020. [54] P. Poudenx, "The effect of transportation policies on energy consumption and greenhouse gas emission from urban passenger transportation," Transportation Research Part A: Policy and Practice, vol. 42, no. 6, pp. 901-909, 2008. https://doi.org/10.1016/j.tra.2008.01.013 [55] B. Gatersleben and D. Uzzell, "Affective appraisals of the daily commute: Comparing perceptions of drivers, cyclists, walkers, and users of public transport," Environment and Behavior, vol. 39, no. 3, pp. 416-431, 2007. https://doi.org/10.1177/0013916506294032 https://doi.org/10.1016/j.dajour.2022.100029 https://doi.org/10.1186/s12913-022-08366-8 https://doi.org/10.1155/2021/8793101 https://doi.org/10.1007/s11116-019-10044-4 https://doi.org/10.3390/futuretransp1030043 https://doi.org/10.3390/app12094759 https://doi.org/10.1080/10841806.2020.1782111 https://doi.org/10.1016/j.jretconser.2015.11.006 https://doi.org/10.1145/3494993 https://doi.org/10.1177/2399808319862571 https://doi.org/10.26599/TST.2020.9010014 https://doi.org/10.3390/ijgi9060394 https://doi.org/10.1016/j.jtte.2020.07.002 https://doi.org/10.1016/j.tra.2008.01.013 https://doi.org/10.1177/0013916506294032 1908 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 7: 1893-1908, 2025 DOI: 10.55214/2576-8484.v9i8.9033 © 2025 by the authors; licensee Learning Gate [56] A. Gutiérrez, D. Miravet, and A. Domènech, "COVID-19 has shifted perceptions of public transport as a vector of virus transmission, posing new challenges for safe and reliable urban public transport services and potentially exacerbating inequalities," Journal of Transport & Health, vol. 19, p. 100934, 2021. [57] M. N. Borhan, A. N. H. Ibrahim, D. Syamsunur, and R. A. Rahmat, "Why public bus is a less attractive mode of transport: A case study of Putrajaya, Malaysia," Periodica Polytechnica Transportation Engineering, vol. 47, no. 1, pp. 82–90, 2019. https://doi.org/10.3311/PPtr.9228 [58] J. Soza-Parra, S. Raveau, and J. C. Muñoz, "Average bus users are willing to travel 5 minutes longer to avoid headway irregularity, while metro passengers prefer smaller standard deviations," Transportation Research Part A: Policy and Practice, vol. 140, no. 1, pp. 50-62, 2021. [59] D. Ogilvie, M. Egan, V. Hamilton, and M. Petticrew, "Promoting walking and cycling as an alternative to using cars: Systematic review," British Medical Journal, vol. 329, no. 7469, p. 763, 2004. https://doi.org/10.1136/bmj.38216.714560.55 [60] E. Bigsby, S. R. Hovick, N. Q. Tan, S. N. Thomas, and S. R. Wilson, "Information seeking and risk reduction intentions in response to environmental threat messages: The role of message processing," Risk Analysis, vol. 42, no. 10, pp. 2160-2175, 2022. https://doi.org/10.1111/risa.13857 [61] Liu et al, "For highlighting health risks and benefits related to modal shifts in transportation," 2006. [62] P. Crawshaw, "Changing behaviours, improving outcomes? Governing healthy lifestyles through social marketing," Sociology Compass, vol. 8, no. 9, pp. 1127-1139, 2014. https://doi.org/10.1111/soc4.12196 [63] Ogilvie et al, "For evidence on the effectiveness of targeted behavior change programs in promoting modal shifts," 2004. https://doi.org/10.3311/PPtr.9228 https://doi.org/10.1136/bmj.38216.714560.55 https://doi.org/10.1111/risa.13857 https://doi.org/10.1111/soc4.12196