Next Article in Journal
Educating Managers to Govern Artificial Intelligence
Previous Article in Journal
Has the Transformation of Resource-Based Cities Effectively Promoted the Development of the Tourism Industry? Empirical Evidence from China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 18
Issue 11
10.3390/su18115585
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

An Integrated Behavioral Framework for Risky Pedestrian Crossing Behavior Extending the Theory of Planned Behavior

by
Sararad Chayphong
1,*,
Pawinee Iamtrakul
1,2 and
Kento Yoh
3
1
Faculty of Architecture and Planning, Thammasat University, Pathumthani 12121, Thailand
2
Center of Excellence in Urban Mobility Research and Innovation, Thammasat University, Pathumthani 12121, Thailand
3
Division of Global Architecture, Graduate School of Engineering, The University of Osaka, Suita 565-0871, Japan
*
Author to whom correspondence should be addressed.
Sustainability 2026, 18(11), 5585; https://doi.org/10.3390/su18115585
Submission received: 18 April 2026 / Revised: 21 May 2026 / Accepted: 21 May 2026 / Published: 2 June 2026
(This article belongs to the Section Sustainable Transportation)
Browse Figures
Versions Notes

Abstract

Walking plays a crucial role in supporting sustainable transportation systems and is closely linked to the built environment, which influences risky pedestrian crossing behavior. However, prior research has largely focused on constructs from the Theory of Planned Behavior (TPB) with limited integration of environmental factors. This research addresses this gap by employing an extended TPB framework to investigate behavioral intentions and their relationship with risky pedestrian crossing behavior. The model incorporates additional variables, including habit, personal norms, past behavior, perceived risk across behavioral and built environmental dimensions, and perceived built environment. An on-site questionnaire survey was conducted at signalised pedestrian crossings in Bangkok, Thailand, and analyzed using Structural Equation Modeling (SEM). The findings demonstrate a significant predictive effect of the core TPB constructs on behavioral intentions. Among the extended variables, habit positively influences intentions toward risky behavior, whereas personal norms and perceived risk have negative effects. Additionally, perceptions of built environment characteristics contribute to perceived physical risk. Overall, the research emphasizes the importance of integrating psychological and perceived built environment dimensions in explaining risky pedestrian crossing behavior. Policy implications emphasize the need to improve the physical environment to support safe mobility, alongside promoting positive attitudes and norms toward safe travel. The findings can further be used to guide the planning of safer pedestrian crossing facilities in urban contexts and to inform interventions that may help reduce risky pedestrian crossing behavior.

1. Introduction

Promoting sustainable transport has become a central priority in global urban development agendas, particularly in response to pressing environmental challenges. This transition emphasizes a shift toward active transport and public transit systems, reducing reliance on private vehicles while fostering inclusive, safe and affordable mobility. Such approaches aim to enhance equitable access and minimize greenhouse gas emissions and broader environmental impacts [1,2]. Among these strategies, active transport is widely recognized as one of the most sustainable modes, as it generates negligible emissions while delivering substantial social, health and livability benefits [3,4,5,6,7]. In particular, walking plays a critical role in advancing sustainability across environmental, economic and social dimensions, contributing to healthier populations and more livable urban environments. Despite these advantages, walking as a mode of transport continues to face significant road safety challenges [8,9]. It is estimated that vulnerable road users account for more than half of global road traffic fatalities, of which approximately 21% are pedestrians [10]. Pedestrians are classified as vulnerable due to their heightened exposure to injury in traffic collisions, largely resulting from the absence of physical protection [11,12]. This persistent vulnerability underscores pedestrian safety as a critical concern within the context of urban sustainability, transport planning and public health, given that road traffic injuries remain a major cause of mortality worldwide [10,13].
Road safety remains a critical urban challenge that necessitates targeted attention and intervention. Previous studies consistently identify multiple contributing factors to road crashes, including road users, vehicles, roadway characteristics, and environmental conditions [14,15]. These components are highly interdependent and interact dynamically within the broader transportation system. Importantly, the determinants of crash occurrence vary across transport modes. In the context of pedestrian-related accidents, previous studies have shown that there are several contributing factors to pedestrian crashes. Key factors are primarily associated with road users, including both pedestrians (e.g., risky behavior, psychological factors, and age) and other road users, particularly motorized vehicle drivers [16,17]. In addition, environmental conditions (e.g., roadway characteristics, the presence of public transit infrastructures, and land use) have been recognized as important factors contributing to pedestrian crash risk [16,17,18,19].
From a behavioral perspective, existing research has examined both the actions of other road users and risky pedestrian behaviors, highlighting the complex interplay between individual decision-making and external environmental influences in shaping pedestrian safety outcomes. When focusing specifically on risky pedestrian crossing behavior (RPB), the Theory of Planned Behavior (TPB) serves as a useful framework for understanding a wide range of human behaviors, including violating behaviors and mobile phone use during street crossing [20,21,22,23]. Within the TPB framework, behavior is primarily predicted by behavioral intention (BI), while attitude (ATT), subjective norm (SN), and perceived behavioral control (PBC) contribute to the formation of intention [24]. Although prior studies consistently demonstrate the significance of these core constructs in predicting pedestrian behavioral intentions, limitations remain in fully capturing the complexity of such behavior. Accordingly, the TPB has been extended in many studies through the inclusion of additional constructs to improve its explanatory power. For instance, Zhou et al. [25] showed that, alongside subjective norm and instrumental attitude, extended constructs such as descriptive norm and conformity tendency were associated with pedestrian crossing violations. Similarly, other studies have identified additional determinants, including perceived risk [26], situational factors [27], and habit (HAB) [28] as predictors of pedestrian behavior. Beyond psychological factors, the physical environment also plays a crucial role, as walking is inherently shaped by surrounding spatial conditions. Empirical evidence indicates that built environment characteristics (e.g., street design and land use patterns) can influence both accident risk and pedestrian behavior [19,29,30,31,32].
However, existing studies on risky pedestrian crossing behavior have often examined environmental and psychological factors in isolation when predicting behavior. In this research, the built environment is not treated merely as an external explanatory variable. Rather, it is conceptualized as a structural antecedent that shapes pedestrians’ cognitive processes, particularly risk perception, which subsequently influences behavioral intention. Addressing this gap, the present research extends the TPB by integrating built environment dimensions through perceptual constructs, thereby offering a more comprehensive framework for explaining risky pedestrian crossing behavior.

2. Research Hypotheses Development

2.1. A Theory of Planned Behavior-Based Approach to Understanding Risky Pedestrian Crossing

In general, the TPB is a well-established social cognition model that focuses on individual decision-making and posits that behavioral intention plays a central role in shaping human behavior [33]. Within this framework, behavioral intention is explained by three key constructs: attitude, subjective norm, and perceived behavioral control [24,33,34]. The TPB has been extensively utilized to explain various forms of human behavior, particularly those related to travel [35,36].
Within pedestrian behavior research, numerous studies have employed the TPB as a primary framework, demonstrating varying degrees of influence across its core constructs [37,38]. For instance, prior research has consistently identified a strong association between attitudes and pedestrian behavior [25,39,40]. Attitude describes how positively or negatively an individual evaluates a specific behavior. Earlier studies have reported that this construct can play a dominant role in shaping behavioral intention. For instance, Piazza et al. [38] applied the TPB to examine distracted street-crossing and identified attitude as the most influential determinant of intention. In contrast, other research has emphasized the dominant role of perceived behavioral control in shaping behavioral intention, while also demonstrating its direct relationship with actual behavior. Perceived behavioral control reflects the extent to which individuals believe they can successfully perform a particular behavior. For example, Demir et al. [41], in examining pedestrian violations using the TPB and the Prototype Willingness Model, indicated that perceived behavioral control played the most significant role in predicting violation intention. Similarly, O’Dell et al. [42] reported that perceived behavioral control exerted the greatest influence on pedestrians’ intention to cross while distracted.
A substantial body of empirical research demonstrates that the three core constructs of the TPB play a significant role in predicting risky pedestrian behavioral intentions [23,37,38,43]. In addition, behavioral intention has consistently been shown to positively influence risky pedestrian behavior [23,41]. More broadly, the TPB framework effectively explains a variety of risky pedestrian crossing behaviors, including violations such as crossing against traffic signals or outside designated crossings and distraction-related activities (e.g., mobile phone use while crossing) [22,25,27,38,44,45]. Accordingly, the integration of TPB constructs contributes to a framework for explaining risky pedestrian crossing behavior. Based on this framework, the study advances several proposed relationships.
Hypothesis 1 (H1).
The attitude toward safe crossing negatively influences behavioral intention to engage in risky pedestrian crossing behavior.
Hypothesis 2 (H2).
Subjective norm negatively influences behavioral intention to engage in risky pedestrian crossing behavior.
Hypothesis 3 (H3).
The perceived behavioral control positively influences behavioral intention to engage in risky pedestrian crossing behavior.
Hypothesis 4 (H4).
The perceived behavioral control positively influences risky pedestrian crossing behavior.
Hypothesis 5 (H5).
Behavioral intention to engage in risky pedestrian crossing behavior positively influences risky pedestrian crossing behavior.

2.2. The Components of the Extended Theory of Planned Behavior Framework and Linking Built Environment

Although the TPB framework has been extensively applied, it is not without limitations [46]. To address these shortcomings, previous studies have sought to enhance its explanatory power by integrating additional constructs, such as conformity tendency [25], descriptive norms [47], willingness [41], and perceived risk and severity [47,48]. Nevertheless, the TPB places relatively limited emphasis on habitual behavior. Relatively little research has incorporated habit into the model, where it is commonly conceptualized as an automatic behavioral response [28]. Habit refers to learned behaviors that become automatic through repetition [49]. In contrast, earlier research has commonly adopted a frequency-based approach, operationalizing habit as the frequency of past behavior rather than as an automatic cognitive process [50,51,52].
Past behavior (PB) has frequently been incorporated into the extended TPB framework to improve the prediction of behavioral outcomes. Empirical evidence indicates that past behavior is significantly associated with behavioral intention and enhances the explanatory power of the TPB framework beyond its core constructs [52,53]. However, repeated engagement in a behavior does not necessarily imply the formation of a habit. In response, recent studies have sought to integrate habit into the TPB framework, conceptualizing it in terms of behavioral automaticity to better capture non-deliberative processes underlying behavior. For instance, Matović et al. [49] found that incorporating habit into the TPB framework significantly increased pedestrians’ intention to violate. In contrast, Jovanović et al. [54], examining speeding behavior within a modified TPB framework, reported that habit measured through an automaticity index exerted a significant direct effect on behavior, while showing no significant association with behavioral intention. These mixed findings suggest that the role of habit within the extended TPB framework remains inconclusive, highlighting the need for further investigation across different behavioral contexts, particularly in relation to risky pedestrian crossing behavior.
In addition, personal norms (PN) have emerged as an important factor influencing behavioral intention, operating independently of subjective norms. Personal norms reflect an individual’s internalized moral obligation that guides decisions to engage in or refrain from specific behaviors [55,56]. Previous research indicates a significant relationship between personal norms and intentions related to unsafe behaviors, including speeding [54], jaywalking [52], and traffic rule violations [49]. Incorporating personal norms into the TPB framework highlights the interaction between internalized moral standards and external social pressures. However, their role within the TPB remains insufficiently understood, as conceptualizations and operationalizations vary across studies. Accordingly, this research builds upon the TPB by integrating additional constructs to provide a more comprehensive framework for explaining risky pedestrian crossing behavior. Building upon this extended framework, the research proposes several relationships among the constructs.
Hypothesis 6 (H6).
Habit positively influences risky pedestrian crossing behavior.
Hypothesis 7 (H7).
Habit positively influences behavioral intentions to engage in risky pedestrian crossing behavior.
Hypothesis 8 (H8).
Personal norm negatively influences behavioral intentions to engage in risky pedestrian crossing behavior.
Hypothesis 9 (H9).
Past behavior positively influences behavioral intentions to engage in risky pedestrian crossing behavior.
Hypothesis 10 (H10).
Past behavior negatively influences attitude toward safe crossing.
Risk perception has also been incorporated into TPB-based models to improve the explanation of risky behavior [25,57]. Individuals’ perceptions of risk are shaped by their evaluation of both the likelihood and possible consequences of unfavorable outcomes. As such, risk perception reflects a cognitive evaluation of uncertainty associated with specific actions or situations [58,59]. Risk perception is an important factor influencing behavioral responses, as individuals with higher perceived risk are less likely to engage in risk-taking behaviors. This relationship is also evident in pedestrian traffic contexts, where variations in risk perception are associated with risky behaviors, including those related to pedestrians [20,59]. However, some studies have indicated that risk perception is not significantly associated with pedestrian behavior [25,40]. Variations in the conceptual and measurement approaches to risk perception, as well as differing contextual and physical conditions, may account for these inconsistent findings.
Moreover, although previous studies have investigated the influence of risk perception on the behavior of pedestrians and other road users, the focus has predominantly been on perceptions arising as a consequence of behavior, rather than as antecedents influenced by external factors such as the built environment. Several recent studies have sought to address these gaps by examining risk perception through more differentiated and context-specific approaches. For instance, Dinh et al. [59] evaluated the relationships among attitudes, risk perception (distinguishing between traffic-related and non-traffic-related risks) and pedestrian behavior in Vietnam. Their findings indicate that traffic risk perception is related to behavior, with higher perceived risk associated with safer actions, while non-traffic risk perception is linked to behavior through its positive effect on traffic risk perception. Similarly, Hou et al. [20] employed multiple path models incorporating distinct risk-perception constructs and psychological factors to examine pedestrians distracted by mobile phones while crossing streets. The analysis demonstrated that all forms of risk perception were negatively associated with such behavior. In addition, Zheng et al. [60] examined the association between risk perception and distraction engagement, demonstrating that behavior-related risk perception differs from consequence-related risk perception in influencing pedestrian actions.
Accordingly, variations in types of risk perception represent a critical consideration in advancing behavioral explanatory models. However, the integration of these distinctions within the TPB framework remains limited, particularly with respect to their interaction with the physical environment. The built environment is a key factor shaping pedestrian travel behavior, and numerous studies have highlighted its relevance in influencing risky pedestrian actions [61,62,63]. Although objective environmental characteristics have received considerable attention in previous studies, subjective perceptions remain central to behavioral decision-making in risky situations. Emerging evidence further suggests a strong link between the physical environment and risk perception, indicating that pedestrians’ perceived risk is closely associated with their crossing behavior [64]. The literature reviewed above provides the foundation for the hypotheses examined in this research.
Hypothesis 11 (H11).
The perceived risk of the built environment negatively influences behavioral intention to engage in risky pedestrian crossing behavior.
Hypothesis 12 (H12).
The perceived risk of behavior characteristics negatively influences behavioral intention to engage in risky pedestrian crossing behavior.
Hypothesis 13 (H13).
The perceived built environment negatively influences the perceived risk of the built environment.
From the foregoing discussion, an integrated conceptual framework is proposed to capture both direct and indirect influences. This framework provides a more comprehensive explanation of such behavior by recognizing that its determinants extend beyond those specified in the TPB framework. In particular, psychological constructs such as habit and personal norms, together with perceptual environmental dimensions, have received limited attention. To address these gaps, this research extends the TPB framework by integrating constructs relevant to risky pedestrian crossing behavior, with particular emphasis on their linkage to the built environment. The proposed model retains the core TPB components and emphasizes the rational, belief-based mechanisms underlying behavioral intentions while integrating habits as automatic responses, as well as personal norms and past behavior. Furthermore, it incorporates perceptual factors derived from both built environment conditions and the characteristics of risky behavior (see Figure 1). This integrative approach provides a broader understanding of risky pedestrian crossing behavior by capturing intention-based pathways and direct effects on behavior.

3. Methodology

This research examines an integrated model for predicting risky pedestrian crossing behavior (RPB) by extending the TPB framework. The organization of this study is outlined as follows.

3.1. Study Setting and Instrument Construction

This research was conducted in Bangkok, a metropolitan context characterized by a diverse transportation system and complex mobility patterns, including substantial pedestrian activity. The city offers relatively well-developed pedestrian infrastructure, particularly within major urban areas, which exhibit considerable variation in both user groups and the physical and environmental attributes of walking environments. Signalised pedestrian crossings were selected as the primary focus of this research as they constitute critical locations for examining risky pedestrian crossing behaviors within the urban traffic system. An extensive literature review informed the development of the questionnaire and identified the principal constructs and variables associated with the research objectives. Measurement indicators were drawn from earlier studies and revised to enhance their relevance to the study setting. A pilot survey with 20 respondents was performed to assess the clarity, validity, and contextual appropriateness of the instrument, followed by refinements prior to the main data collection phase. The questionnaire was ultimately structured into three sections. Section one focused on constructs derived from the extended TPB framework, including behavioral intention, personal norm, attitude, habit, subjective norm, perceived risk, perceived behavioral control, perceived built environment, and past behavior. The second section focused on risky pedestrian crossing behavior, where respondents reported the frequency of their risky actions at pedestrian crossings. The final part of the questionnaire collected participants’ demographic characteristics.
Construct measurement items were adapted from established studies [25,27,28,49,52,54,65]. In total, eleven constructs were included in the analysis. The constructs associated with the TPB comprised four components. Attitude (ATT) towards safety was measured using four items (α = 0.90). Subjective norm (SN) was measured through four items (α = 0.90), reflecting the perceived influence of significant others, including family members, friends, and nearby pedestrians. Perceived behavioral control (PBC) was captured using four items (α = 0.89), capturing both controllability and self-efficacy related to performing the action. Behavioral intention (BI) was measured using three items (α = 0.87), representing respondents’ willingness and likelihood to engage in risky pedestrian crossing behavior. Participants rated all TPB-related statements on a five-point Likert scale anchored between “strongly disagree” (1) and “strongly agree” (5). With respect to the extended components of the TPB, habit (HAB), personal norm (PN) and past behavior (PB) were incorporated into the analytical framework. Habit was calculated using three items (α = 0.89), adapted from the Self-Report Behavioral Automaticity Index (SRBAI) to capture the automaticity of behavior. Personal norm was assessed through three items (α = 0.86), reflecting individuals’ internalized moral obligations, perceived responsibility and anticipated feelings of guilt associated with engaging in risky behavior. Past behavior was measured using three items (α = 0.81), representing respondents’ prior engagement in risky pedestrian crossing behavior. Participants evaluated all variables using a five-point Likert scale anchored between “strongly disagree” (1) and “strongly agree” (5). Within the extended TPB framework, perceived risk was operationalized as a multidimensional construct comprising two components: perceived risk of behavior (PRB) (α = 0.85) and perceived risk of the built environment (PRE) (α = 0.82). Each dimension was measured using three items capturing respondents’ perceptions of general risk, the likelihood of accidents, and the severity of potential consequences. In addition, perceived built environment (PBE) was assessed using four items (α = 0.83), reflecting individuals’ evaluations of environmental conditions relevant to pedestrian safety, such as traffic volume, land-use diversity, and roadway width. All perception-related items were rated on a five-point Likert scale anchored between “lowest level” (1) and “highest level” (5). Finally, the risky pedestrian crossing behavior (α = 0.81) was measured using three items adapted from the Pedestrian Behavior Scale (PBS) [66,67]. The items were contextually refined to focus specifically on risky pedestrian crossing behavior. Participants rated the frequency of their engagement in each behavior on a five-point Likert scale anchored between “never” (1) and “always” (5).
Overall, Cronbach’s alpha coefficients for all constructs were greater than 0.70, suggesting adequate internal consistency reliability. This study was approved by the Human Research Ethics Committee of Thammasat University Social Sciences (Approval No. 063/2568), and all procedures adhered to established ethical guidelines to ensure the protection and well-being of participants.

3.2. Sample and Data Collection

Questionnaire data were obtained through on-site surveys conducted at signalized pedestrian crossings in Bangkok. The research sample focused on pedestrians, particularly individuals with prior road-crossing experience. Participants were included based on the following eligibility criteria: they were aged over 18 years and had prior experience crossing roads within the study area, including engagement in risky pedestrian crossing behavior. However, individuals were not approached while actively engaging in risky behaviors during the survey period in order to minimize response bias. The sample size was determined based on statistical power considerations. For a model comprising eleven variables, the minimum required sample size for structural equation modeling (SEM) was 195 as indicated by a priori power analysis [68]. In addition, methodological guidelines recommend a sample size of approximately ten times the number of measurement items [69]. Accordingly, a total of 500 questionnaires were distributed to pedestrians at selected crosswalk locations.
A face-to-face survey method was employed to administer the questionnaire. Purposive sampling was used to identify and recruit participants who met the research criteria. The target population comprised pedestrians within the study area, specifically those located within a radius of approximately 400 m from selected crossing locations. Regarding the observational procedure, researchers first observed pedestrian crossing behaviors around the selected crossing locations. Pedestrians actively crossing the road were not approached in order to avoid interfering with their movement during road crossing. In addition, individuals exhibiting risky pedestrian crossing behaviors at the time of observation were also not approached. The targeted type of risky pedestrian crossing behaviors primarily included violations, distractions, and errors during road crossing. Pedestrians who were not exhibiting such behaviors during the observation period were subsequently approached and invited to participate. They were then screened based on prior experience with risky pedestrian crossing behaviors. Potential respondents were provided with detailed information regarding the study’s background, objectives, and anticipated benefits to facilitate informed decision-making. Upon obtaining informed consent, the questionnaire was administered through a structured interview until completion. The participants were notified that participation was voluntary and that they could withdraw at any time without consequence. Throughout the data collection process, strict measures were implemented to ensure participant anonymity and confidentiality. The survey yielded a total of 500 questionnaires, of which 460 complete responses (92%) were included in the final analysis. The demographic profile of the sample indicated the respondents’ basic characteristics. The sample comprised 51% male and 49% female respondents, with a mean age of 38 years. With respect to monthly income, approximately 54% of respondents earned between 20,000 and 40,000 baht per month, while 20% earned less than 20,000 baht. More than 66% of the respondents reported having completed a bachelor’s degree.

3.3. Data Reliability and Validity

To assess the adequacy of the measurement model, confirmatory factor analysis (CFA) was performed using various validation criteria. The results indicated adequate internal consistency, as both composite reliability (CR) and Cronbach’s alpha values surpassed the recommended cut-off value of 0.70 [70]. To establish convergent validity, factor loadings and average variance extracted (AVE) were evaluated against threshold values of 0.70 and 0.50, respectively [69,70]. In addition, discriminant validity was verified using the heterotrait–monotrait ratio (HTMT), with the results confirming sufficient distinction among the constructs [71]. Detailed findings of the reliability and validity assessments are reported in the subsequent section.

3.4. Data Analysis

Descriptive statistics were employed to examine respondents’ socioeconomic characteristics as well as the general characteristics of the study variables. Subsequently, the measurement model was examined through CFA using JASP version 0.96 and jamovi version 2.7. The evaluation focused on convergent validity, construct reliability, and discriminant validity, following procedures described in the Section 3.3. After the measurement model had been validated, SEM was performed in jamovi version 2.7 to investigate the proposed relationships among the constructs. SEM is well suited for examining complex models comprising several latent constructs, as it enables the simultaneous estimation of both direct and indirect effects within an integrated analytical framework. Several fit indices were used to evaluate the adequacy of the model. These included CMIN/df ratio, Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), Standardized Root Mean Square Residual (SRMR), and Tucker–Lewis Index (TLI) [72,73]. Each index was interpreted according to established threshold criteria to assess the model’s correspondence with the observed data.

4. Results

4.1. Evaluation of the Measurement Model

Prior to conducting the analysis, normality was evaluated to ensure the suitability of estimation. The obtained kurtosis and skewness values were within the ranges of −1.298 to 1.117 and −1.064 to 0.680, respectively, indicating acceptable normality. This research employed CFA to evaluate the measurement model (Table 1), confirming that the latent constructs were reliably assessed and exhibited adequate validity. The analysis confirmed satisfactory construct reliability, with CR and Cronbach’s alpha values remaining above the accepted benchmark of 0.70. All observed variables were statistically significant, with standardized factor loadings varying between 0.706 and 0.910 and satisfying the recommended criterion of 0.70. Convergent validity was further supported by AVE values between 0.569 and 0.726, all of which exceeded the recommended threshold of 0.50. The CFA model demonstrated satisfactory fit across all indices (CMIN/df = 1.700, CFI = 0.962, TLI = 0.956, RMSEA = 0.039). Overall, the measurement model demonstrated sufficient reliability and validity to support further examination of the structural relationships [69,70].
Discriminant validity was assessed using HTMT. As presented in Table 2, HTMT ratios varied from 0.023 to 0.767, indicating that all values were below the recommended threshold of 0.85. Findings provide support for adequate discriminant validity for all constructs examined in this research.

4.2. Extended Theory of Planned Behavior to Explain Risky Pedestrian Crossing Behavior

An assessment of the proposed structural model was conducted before proceeding with hypothesis testing. The results indicated that the model achieved an acceptable level of fit based on established criteria. The CMIN/df was 1.772, below the recommended threshold of 3. Additionally, the RMSEA value (0.041) was below 0.08, indicating a satisfactory model fit, whereas the SRMR slightly exceeded the recommended threshold, suggesting a marginal fit [74,75]. Other fit indices also demonstrated acceptable values, with the CFI (0.957) and TLI (0.951) falling within recommended ranges. Overall, these results suggest that the proposed structural model fits the empirical data well and that the included variables adequately explain risky pedestrian crossing behavior. Regarding explanatory power, the extended TPB framework explained 81% of the variance in behavioral intention and 55% of the variance in risky pedestrian crossing behavior, while upstream constructs exhibited lower levels of explained variance. Overall, the model demonstrates substantial explanatory capacity for the key endogenous variables in the study.
For hypothesis testing, the path analysis provided strong empirical support for twelve of the thirteen proposed hypotheses, with only Hypothesis 9 found to be statistically non-significant, as illustrated in Figure 2. With respect to the TPB components, attitude toward road safety demonstrated the strongest negative effect on the intention to engage in risky pedestrian crossing behavior (H1: β = −0.20), followed by subjective norm (H2: β = −0.14). In contrast, perceived behavioral control exhibited a positive and statistically significant effect (H3: β = 0.19). The findings indicate that reductions in risky behavioral intentions are associated with both attitude and subjective norm. Furthermore, perceived behavioral control exhibited a statistically significant positive indirect effect on risky behavior through behavioral intention, as well as a direct effect on risky pedestrian crossing behavior (H4: β = 0.16).
With respect to the extended components of the TPB, habit demonstrated the strongest positive influence on behavioral intention (H7: β = 0.33) and also exhibited a direct effect on risky pedestrian crossing behavior (H6: β = 0.23). In contrast, personal norm exerted a significant negative influence on behavioral intention (H8: β = −0.26). Meanwhile, past behavior was not significantly associated with behavioral intention (H9), although it showed a significant negative relationship with attitude (H10: β = −0.21).
Regarding the perception-related extensions of the TPB model, perceived risk was identified as a significant determinant of behavioral intention. Specifically, perceived risk of the built environment demonstrated a negative effect on behavioral intention (H11: β = −0.17), followed by perceived risk of behavior, which also exhibited a negative effect (H12: β = −0.12). In addition, the perceived built environment negatively influenced perceived risk of the built environment (H13: β = −0.38), indicating an indirect pathway through which environmental perceptions affect behavioral intention via risk perception. Furthermore, risky pedestrian crossing behavior was significantly predicted by behavioral intention (H5: β = 0.43). Overall, these findings highlight the importance of extending the TPB framework through the integration of psychological factors and perceived built environment dimensions. This provides a more comprehensive explanation of risky pedestrian crossing behavior.

5. Discussions

5.1. Extending the Theory of Planned Behavior to Predict Risky Pedestrian Crossing Behavior Through Intention

The extension of the TPB framework demonstrates that its three core components remain strong predictors of behavioral intention. In addition, the inclusion of extended constructs with factors such as habit and personal norms is an important component alongside the TPB framework. The findings showed that the extended model exhibited satisfactory explanatory power for behavioral intention. This reflects the integration of multiple decision-making processes, including deliberative processes, automatic (habit and past behavior), normative (personal norm), and risk-based appraisal processes derived from both environmental and behavioral risk perceptions, which jointly contribute to the formation of intention in the context of risky pedestrian crossing behavior.
Previous studies have reported comparable explanatory power in extended models. Demir et al. [41] found that a modified integrative model combining the TPB and the Prototype Willingness Model explained 56% and 66% of the variance in intentions and pedestrian violation behavior, respectively, and showed higher explanatory power than the standard TPB as well as the Prototype Willingness Model. Similarly, Liu et al. [47] reported that an integrative model combining the TPB, Health Belief Model, and descriptive norms explained 68% of behavioral intentions. In other transport-related risk behavior contexts, extended TPB models also demonstrate strong predictive capability. For example, incorporating risk perception and habit into the TPB explained 76% of low-speed driving intention [36]. Prior research incorporated perceived dangerousness into the TPB framework as a construct representing the influence of the road environment, together with other predictors. The model explained 66% of the variance in behavioral intention and 79% in self-reported speeding behavior [76]. These findings suggest that the explanatory power of behavioral intention varies across contexts and is likely influenced by the interaction of multiple factors, highlighting the complexity of pedestrian decision-making processes. However, these results should be interpreted with caution, as they may be influenced by interrelationships among psychological constructs and the multidimensional nature of the model, which can increase the overall explained variance.
Comparative analysis with prior studies reveals some variation in the relative associations of TPB components. While certain studies emphasize attitude as the strongest determinant of intentions to engage in risky pedestrian crossing behavior [38], others highlight perceived behavioral control as having a more prominent association [41]. The findings of this research indicate that attitude shows the strongest association within the TPB framework, with a significant negative effect on behavioral intention. This suggests that increasing awareness of travel-related risks associated with unsafe behaviors may help explain lower intentions to engage in such behaviors at crossing locations. However, in contrast to some prior studies, subjective norms were not identified as significantly associated with behavioral intention in the analysis.
Subjective norms do not consistently emerge as significantly associated with risky pedestrian crossing behavior across studies [41]. Nevertheless, other research highlights their importance, demonstrating that subjective norms can reveal a significant association with intentions to engage in risky pedestrian crossing behavior [23,25,65]. The findings of the present study align with this latter perspective, indicating that subjective norms remain a meaningful association with behavioral intention. This suggests that social influences, particularly from family members, peers, and nearby pedestrians, are associated with behavioral intention, as individuals may be more likely to engage in risky behavior when such behavior is perceived as socially acceptable. With respect to the extended constructs, habits, conceptualized as behavioral automaticity, emerge as a strongly associated factor among all components. Although this research conceptualizes habit as an automatic process with a potential direct effect on behavior, consistent with prior research [54], the findings also reveal that habit is strongly associated with behavioral intention. This finding is consistent with previous studies demonstrating the association between habit and intentions [28,49]. Importantly, this association supports the view that habit should not be viewed as a purely automatic response, but rather as a form of goal-directed automaticity [28,77]. This process may be further reinforced by environmental cues, whereby physical conditions such as traffic volume, road design and crossing infrastructure trigger habitual responses within specific contexts.
The findings demonstrate that habit is associated with behavioral intention and also has a direct association with behavior, which is further elaborated in the following section. Regarding personal norms, as an extended construct within the TPB framework, the findings indicate a significant negative association with behavioral intention. This suggests that stronger personal norms are associated with a reduced intention to engage in risky behavior. This result is consistent with prior studies demonstrating that personal norms decrease intentions to violate traffic rules [49,52]. This relationship aligns with the conceptualization of personal norms as reflecting individuals’ internalized moral obligations and anticipated feelings of guilt. When individuals perceive that risky actions may lead to negative consequences for themselves or others, they are more likely to refrain from such behaviors. Moreover, when comparing internal and external normative influences, both personal norms and subjective norms are associated with behavioral intention, highlighting their complementary roles in explaining behavioral intention through the interaction of moral obligations and social pressures. In addition, perceived environmental and behavioral risk are negatively associated with behavioral intention, with further details provided in the section on the relationship with the built environment.
This research highlights that while the core components of the TPB remain significant in explaining behavior through intention, the extended framework highlights the complexity of behavioral processes. Notably, habit shows relatively stronger associations with pedestrians’ intentions than cognitive TPB components and social norms. This highlights a limitation of the TPB in fully capturing habitual influences and suggests the need to incorporate automatic processes into behavioral models. Overall, the results of this section suggest that behavioral intention serves as a central variable in the model explaining risky pedestrian crossing behavior. This reflects a broadly reasoned decision-making process, particularly in relation to the TPB components, which are generally associated with evaluative considerations underlying such behavior. In addition, habit represents an experience-based and relatively automatic construct. The significant relationship between habit and behavioral intention suggests that repeated exposure to similar pedestrian crossing situations may lead to the development of automatic behavioral tendencies associated with intention formation. Furthermore, personal norm reflects an internalized moral consideration in the context of risky pedestrian crossing behavior. It may contribute to behavioral intention through value-based considerations relevant to such behavior. Perceived environmental and behavioral risk are also associated with behavioral intention, reflecting risk appraisal processes based on both environmental conditions and behavioral evaluations in pedestrian crossing contexts. Overall, these findings suggest that intention formation in this extended framework for risky pedestrian crossing behavior may involve multiple underlying determinants. From a practical perspective, these results suggest that interventions focusing solely on awareness or attitudinal change may be insufficient. Instead, strategies aimed at disrupting habitual behaviors and restructuring environmental conditions may be more relevant in reducing risky pedestrian crossing behavior.

5.2. The Role of Control- and Habit-Related Factors Influencing Risky Pedestrian Crossing Behavior

This relationship highlights a non-intentional pathway in explaining behavior, whereby certain factors have a direct association with risky pedestrian crossing behavior. In the context of the TPB framework, perceived behavioral control is a key construct that explains behavior both indirectly, through behavioral intention, and directly. Previous studies on pedestrian risk behavior have similarly demonstrated that perceived behavioral control can be associated with behavior through both pathways [41]. Prior research commonly identifies behavioral intention and perceived behavioral control as key determinants closely linked to actual behavior, with some studies identifying perceived behavioral control as a strong predictor among TPB constructs [37,41]. Perceived behavioral control is directly and positively associated with risky pedestrian crossing behavior, as evidenced in the present study, although the magnitude of the effect is relatively modest. Moreover, while perceived behavioral control is associated with behavior, its association is stronger with behavioral intention than with behavior itself. This suggests that perceptions of ease and confidence in performing a behavior are important in explaining risky pedestrian crossing behavior. However, inconsistent findings in prior studies, some of which found perceived behavioral control to be unrelated to pedestrian violations [47], emphasize the importance of contextual factors in explaining behavioral outcomes.
In addition, habit-related factors, conceptualized as automatic responses to situational cues, demonstrate a substantial direct association with risky pedestrian crossing behavior. Although relatively few studies have incorporated habit into the TPB framework, its inclusion contributes to the explanatory power of behavioral models. The findings of this research indicate that habit is positively associated with both behavioral intention and behavior, with a particularly strong effect on intention. This suggests that habitual tendencies may be associated with the formation of behavioral intentions, while remaining distinct from reflective decision-making processes. Consistent with previous research showing that habit is strongly associated with behavioral intention [28]. Given that road crossing is a routine daily activity, it may not always involve deliberate decision-making. Through repeated exposure, such behavior can become familiar and evolve into automatic responses. Overall, this research advances the understanding of risky pedestrian crossing behavior by integrating extended constructs and examining the roles of perceived behavioral control and habit in explaining risky pedestrian crossing behavior, thereby extending the original TPB framework by incorporating both control-based and automatic influences on behavior.

5.3. Understanding the Relationships Linking Built Environment Dimension and Risky Pedestrian Crossing Behavior

Perceptual factors operate as a mediating mechanism through which the built environment structurally shapes behavioral cognition and responses. The findings of this research indicate that perceived risk is significantly negatively associated with behavioral intention. Specifically, when individuals perceive a higher likelihood or severity of collision within a given environment, they are less inclined to engage in such behavior. This result is consistent with prior studies demonstrating that elevated levels of perceived risk are associated with reduced behavioral intention [13,59]. Conceptually, this relationship is expected, as individuals tend to avoid behaviors they perceive as hazardous. Despite these findings, certain studies have shown that risk perception is not significantly associated with behavior [25,40]. Such inconsistencies may reflect variations in individuals’ understanding of risk severity or differences in contextual conditions. In particular, perceived risk may not operate as a dominant determinant when other psychological or situational factors exhibit stronger associations with behavioral intention.
A comparison between different dimensions of risk perception reveals that risk associated with the physical environment has a more pronounced influence than behavior-related risk perception. This finding is plausible given that pedestrian crossing behavior is inherently embedded within specific environmental contexts. Notably, characteristics of the built environment, especially crossing facilities, road design, and infrastructure, were found to be negatively associated with perceived environmental risk. In other words, when the crossing environment is perceived as safer, individuals tend to report lower levels of perceived risk. This relatively strong effect may reflect that risk judgments are primarily based on perceptual evaluations of physical environmental conditions. Since both constructs are derived from the same environmental information, a high degree of cognitive alignment is expected. Accordingly, pedestrians rely on observable environmental cues as heuristics in forming risk judgments. Importantly, perceived risk serves as a key linkage between environmental perception and behavioral intention, highlighting its mediating role within the behavioral framework. The findings suggest that perceptions of safer environmental conditions may be associated with lower levels of perceived environmental risk, which in turn may relate to greater behavioral intention toward risky crossing behavior. This pattern may reflect a degree of behavioral complacency under perceived safer conditions. In terms of policy implications, these findings suggest that improving pedestrian safety may require approaches that extend beyond physical design interventions alone. While enhancing the built environment remains a fundamental strategy for improving pedestrian safety, it should ideally be complemented by broader psychological and behavioral interventions.

5.4. Limitation and Further Study

Although this research offers important insights, a number of limitations should be acknowledged. First, the use of self-reported responses may have resulted in social desirability bias, potentially leading respondents to understate unfavorable behaviors. While efforts were made to minimize this bias during data collection, future studies should incorporate objective measures, such as observed behavioral data and built environment indicators (e.g., GIS-based or observational measures). Such approaches would complement the current perception-based framework and strengthen the robustness of the findings. Second, this research focuses on risky pedestrian crossing behavior at signalized pedestrian crossings and targets participants with prior experience of such behavior. This focus may affect representativeness, as pedestrians in other crossing contexts, such as footbridges or underpasses, may not be fully captured. In addition, this research employed on-site intercept surveys within a dynamic urban field environment. Given the field-based nature of the data collection process, non-respondents and excluded individuals were not systematically recorded, which may introduce potential sampling and non-response bias. Subsequent research should therefore examine a broader range of crossing contexts and adopt multiple sampling strategies to capture more heterogeneous pedestrian samples. It may also employ more systematic participant tracking procedures to enhance data quality and help assess potential non-response bias. Third, this research primarily aimed to examine the relationships among psychological and perceived environmental constructs within the extended TPB framework. However, it did not explicitly incorporate demographic characteristics, which may influence pedestrians’ risk perception and crossing behavior. As a result of the limited sample size, subgroup comparisons through multi-group analysis could not be appropriately conducted. Therefore, future research should consider the role of demographic characteristics to gain a deeper understanding of intergroup differences within this integrative model. Finally, this research focused on pedestrian behavior and built-environment interactions within an urban context at the city level using a cross-sectional design. However, risky pedestrian crossing behavior may vary across different behavioral forms and crossing contexts. Therefore, future research should consider spatial heterogeneity and behavioral contexts through comparative analyses across different settings and variations in risky pedestrian behavior (e.g., violations, errors, and lapses). Such considerations may also require the integration of additional context-specific constructs, as well as the use of longitudinal or experimental approaches to strengthen causal inference.

6. Conclusions

This research investigates the integration of built environment perspectives with psychological factors by extending the TPB to explain risky pedestrian crossing behavior. This was achieved by examining the structural relationships within the extended TPB framework using CFA and SEM. The findings contribute to the TPB literature by demonstrating that its core constructs, together with additional factors such as habit, personal norms, and perceived risk (capturing both built environment perceptions and behavioral characteristics), are associated with pedestrians’ risky behavior through behavioral intentions. Furthermore, perceived behavioral control and habits were found to show direct positive associations with risky pedestrian crossing behavior.
The results recommend that the TPB and its extended components provide an explanatory framework for predicting pedestrians’ risky behavior, which can be summarized into four key insights. First, the built environment is an important factor related to pedestrian behavior, primarily through its association with risk perception. Second, the perceived risk remains an important correlate of behavioral intention, particularly the perceived risk of the built environment. This reflects the context-dependent nature of walking behavior, where environmental conditions may have a stronger influence than behavior-specific risk considerations. Third, the habit construct is associated with risky pedestrian crossing behavior, showing both direct and indirect pathways mediated through behavioral intention. This may be attributed to the routine nature of pedestrian crossing activities, which can evolve into automatic behavioral patterns. Finally, the study found that TPB, together with other relevant constructs, was associated with behavioral intention, reflecting both positive and negative relationships depending on the nature of each construct.
Overall, this research proposes an integrative framework for understanding risky pedestrian crossing behavior through the combination of psychological factors and perceived built environment dimensions. The proposed framework contributes to the understanding of risky pedestrian crossing behavior by linking environmental perceptions within a psychological behavioral model. This integration highlights multidimensional behavioral mechanisms. The core TPB components remain fundamental to explaining behavioral intention and behavior, while additional psychological constructs reflect both direct and indirect behavioral pathways. These pathways suggest that risky pedestrian crossing behavior may involve the interaction of intentional decision-making processes and habitual behavioral tendencies. Furthermore, the integration of perceived built environment dimensions illustrates how environmental perceptions are associated with the interpretation of physical risk, thereby influencing pedestrian decision-making regarding risky pedestrian crossing behavior. The framework emphasizes the importance of person–environment interactions and highlights the critical role of urban design, planning, and the promotion of a safer mobility culture in supporting pedestrian safety.
From a policy and planning perspective, the research highlights the need for interventions that extend beyond solely cognitive or awareness-based strategies to reduce risky pedestrian crossing behavior. Instead, they should integrate behavioral and environmental approaches, including fostering safer attitudes and norms, disrupting unsafe habitual patterns, and improving urban design and pedestrian infrastructure. Such measures should prioritize the provision of safe, accessible and well-designed crossing environments that not only enhance safety but also encourage appropriate pedestrian behavior within the urban system. These measures not only improve the quality of urban mobility but also contribute to the advancement of a sustainable transport system that emphasizes safety, inclusivity, and environmental responsibility.

Author Contributions

Conceptualization, S.C., P.I. and K.Y.; methodology, S.C., P.I. and K.Y.; formal analysis, S.C.; writing—original draft preparation, S.C.; writing—review and editing, S.C., P.I. and K.Y.; supervision, P.I. and K.Y. All authors have read and agreed to the published version of the manuscript.

Funding

The authors report that this research was supported by the Doctoral Research Scholarship from Thammasat University.

Institutional Review Board Statement

The research was conducted according to the guidelines of the Declaration of Helsinki and approved by the Human Research Ethics Committee of Thammasat University Social Sciences (certificate of approval number 063/2568, 30 July 2025).

Informed Consent Statement

Verbal informed consent was obtained from the participants. Verbal consent was obtained rather than written because verbal informed consent was obtained from all participants prior to data collection. Participants were informed of the purpose of the research, procedures, anonymity, and their right to withdraw at any time. Those who provided consent were invited to complete the questionnaire.

Data Availability Statement

The data are not publicly available due to ethical reasons and privacy considerations.

Acknowledgments

The authors thank the Center of Excellence in Urban Mobility Research and Innovation and the Faculty of Architecture and Planning, Thammasat University, Thailand, for their academic support and valuable recommendations.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Gössling, S.; Humpe, A. The global scale, distribution and growth of aviation: Implications for climate change. Glob. Environ. Change 2020, 65, 102194. [Google Scholar] [CrossRef] [PubMed]
  2. UNDP. What Is Sustainable Transport and What Role Does It Play in Tackling Climate Change? 2025. Available online: https://climatepromise.undp.org/news-and-stories/what-sustainable-transport-and-what-role-does-it-play-tackling-climate-change (accessed on 16 March 2026).
  3. Rogers, S.H.; Gardner, K.H.; Carlson, C.H. Social capital and walkability as social aspects of sustainability. Sustainability 2013, 5, 3473–3483. [Google Scholar] [CrossRef]
  4. Baobeid, A.; Koç, M.; Al-Ghamdi, S.G. Walkability and Its Relationships with Health, Sustainability, and Livability: Elements of Physical Environment and Evaluation Frameworks. Front. Built Environ. 2021, 7, 721218. [Google Scholar] [CrossRef]
  5. Friel, C.; Walsh, D.; Whyte, B.; Dibben, C.; Feng, Z.; Baker, G.; Kelly, P.; Demou, E.; Dundas, R. Health benefits of pedestrian and cyclist commuting: Evidence from the Scottish Longitudinal Study. BMJ Public Health 2024, 2, e001295. [Google Scholar] [CrossRef]
  6. Millard-Ball, A.; Reginald, M.; Yusuf, Y.; Bian, C. Global health and climate benefits from walking and cycling infrastructure. Proc. Natl. Acad. Sci. USA 2025, 122, e2422334122. [Google Scholar] [CrossRef]
  7. Ma, J.; Li, W. Exploring the group heterogeneity in the impact of social cohesion on the walking frequency of older adults in China. Front. Public Health 2024, 12, 1424975. [Google Scholar] [CrossRef] [PubMed]
  8. Sharma, S.N.; Dehalwar, K. A Systematic Literature Review of Pedestrian Safety in Urban Transport Systems. J. Road Saf. 2025, 36, 55–78. [Google Scholar] [CrossRef]
  9. WHO. New WHO Manual Aims to Strengthen Pedestrian Safety. 2023. Available online: https://www.who.int/news/item/03-05-2023-new-who-manual-aims-to-strengthen-pedestrian-safety (accessed on 7 April 2025).
  10. WHO. Global Status Report on Road Safety 2023. 2023. Available online: https://www.who.int/publications/i/item/9789240086517 (accessed on 10 April 2025).
  11. OECD. Safety of Vulnerable Road Users, Scientific Expert Group on the Safety of Vulnerable Road Users (RS7); OECD: France, Paris, 1998.
  12. Yannis, G.; Nikolaou, D.; Laiou, A.; Stürmer, Y.A.; Buttler, I.; Jankowska-Karpa, D. Vulnerable road users: Cross-cultural perspectives on performance and attitudes. IATSS Res. 2020, 44, 220–229. [Google Scholar] [CrossRef]
  13. Alonso, F.; Esteban, C.; Useche, S.; Colomer, N. Effect of Road Safety Education on road risky behaviors of Spanish children and adolescents: Findings from a national study. Int. J. Environ. Res. Public Health 2018, 15, 2828. [Google Scholar] [CrossRef] [PubMed]
  14. Hussain, M.; Shi, J.; Baikejuli, M. Which factors contribute to road crashes in non-commercial and commercial vehicles? An examination of administrative data from motorways in Pakistan. Int. J. Inj. Control Saf. Promot. 2021, 28, 513–520. [Google Scholar] [CrossRef]
  15. Zamani, A.; Behnood, A.; Davoodi, S.R. Temporal stability of pedestrian injury severity in pedestrian-vehicle crashes: New insights from random parameter logit model with heterogeneity in means and variances. Anal. Methods Accid. Res. 2021, 32, 100184. [Google Scholar] [CrossRef]
  16. Hossain, A.; Das, S.; Jafari, M.; Starewich, M.; Chakraborty, R.; Kutela, B. Behavioral and psychological determinants of pedestrian collisions on arterial roads with evidence from random parameter models. Sci. Rep. 2025, 15, 31684. [Google Scholar] [CrossRef]
  17. Park, S.; Ko, D. Investigating the Factors Influencing Pedestrian–Vehicle Crashes by Age Group in Seoul, South Korea: A Hierarchical Model. Sustainability 2020, 12, 4239. [Google Scholar] [CrossRef]
  18. Iamtrakul, P.; Chayphong, S.; Crizzle, A.M. A spatial analysis of built environment determinants influencing pedestrian fatalities in Bangkok, Thailand. J. Urban Mobil. 2026, 9, 100225. [Google Scholar] [CrossRef]
  19. Shehadeh, E.A.; Al-Bayatti, A.H.; Bingöl, M.A. Effect of roadway environment characteristics on pedestrian safety at signalised intersections in Amman. Urban Plan. Transp. Res. 2024, 12, 2317766. [Google Scholar] [CrossRef]
  20. Hou, M.; Chen, S.; Cheng, J. The effect of risk perception and other psychological factors on mobile phone use while crossing the street among pedestrians. Accid. Anal. Prev. 2022, 170, 106643. [Google Scholar] [CrossRef]
  21. Palat, B.; Paran, F.; Delhomme, P. Applying an extended theory of planned behavior to predicting violations at automated railroad crossings. Accid. Anal. Prev. 2017, 98, 174–184. [Google Scholar] [CrossRef] [PubMed]
  22. Rijal, B.; Yilmaz, N. Effects of Distracted Pedestrian Behavior on Transportation Safety: Causes and Contributing Factors. Appl. Sci. 2024, 14, 11068. [Google Scholar] [CrossRef]
  23. Xiao, Y.; Liu, Y.; Liang, Z. Study on Road-Crossing Violations among Young Pedestrians Based on the Theory of Planned Behavior. J. Adv. Transp. 2021, 2021, 6893816. [Google Scholar] [CrossRef]
  24. Fu, X.; Juan, Z. Exploring the psychosocial factors associated with public transportation usage and examining the “gendered” difference. Transp. Res. Part A Policy Pract. 2017, 103, 70–82. [Google Scholar] [CrossRef]
  25. Zhou, H.; Romero, S.B.; Qin, X. An extension of the theory of planned behavior to predict pedestrians’ violating crossing behavior using structural equation modeling. Accid. Anal. Prev. 2016, 95, 417–424. [Google Scholar] [CrossRef]
  26. Gannon, B.; Rosta, L.; Reeve, M.; Hyde, M.K.; Lewis, I. Does it matter whether friends, parents, or peers drink walk? Identifying which normative influences predict young pedestrian’s decisions to walk while intoxicated. Transp. Res. Part F Traffic Psychol. Behav. 2014, 22, 12–24. [Google Scholar] [CrossRef]
  27. Hou, M.; Cheng, J.; Xiao, F.; Wang, C. Distracted Behavior of Pedestrians While Crossing Street: A Case Study in China. Int. J. Environ. Res. Public Health 2021, 18, 353. [Google Scholar] [CrossRef]
  28. Soathong, A.; Chowdhury, S.; Wilson, D.; Ranjitkar, P. Investigating the motivation for pedestrians’ risky crossing behaviour at urban mid-block road sections. Travel Behav. Soc. 2021, 22, 155–165. [Google Scholar] [CrossRef]
  29. Barón, L.; Otila da Costa, J.; Soares, F.; Faria, S.; Prudêncio Jacques, M.A.; Fraga de Freitas, E. Effect of Built Environment Factors on Pedestrian Safety in Portuguese Urban Areas. Appl. Syst. Innov. 2021, 4, 28. [Google Scholar] [CrossRef]
  30. Koh, P.P.; Wong, Y.D.; Chandrasekar, P. Safety evaluation of pedestrian behaviour and violations at signalised pedestrian crossings. Saf. Sci. 2014, 70, 143–152. [Google Scholar] [CrossRef]
  31. Oviedo-Trespalacios, O.; Scott-Parker, B. Footbridge usage in high-traffic flow highways: The intersection of safety and security in pedestrian decision making. Transp. Res. F Traffic Psychol. Behav. 2017, 49, 177–187. [Google Scholar] [CrossRef]
  32. McConville, M.E.; Rodríguez, D.A.; Clifton, K.; Cho, G.; Fleischhacker, S. Disaggregate land uses and walking. Am. J. Prev. Med. 2011, 40, 25–32. [Google Scholar] [CrossRef]
  33. Ajzen, I. From intentions to actions: A theory of planned behavior. In Action-Control: From Cognition to Behavior; Kuhl, J., Beckman, J., Eds.; Springer: Berlin/Heidelberg, Germany, 1985; pp. 11–39. [Google Scholar]
  34. Hukkelberg, S.S.; Hagtvet, K.A.; Kovac, V.B. Latent interaction effects in the theory of planned behaviour applied to quitting smoking. Br. J. Health Psychol. 2014, 19, 83–100. [Google Scholar] [CrossRef]
  35. Williams, S.L.; French, D.P. Theory of planned behaviour variables and objective walking behaviour do not show seasonal variation in a randomised controlled trial. BMC Public Health 2014, 14, 120. [Google Scholar] [CrossRef] [PubMed]
  36. Xu, J.; Liu, H.; Liu, X.; Gao, C. Extended theory of planned behavior to explain the influence mechanism of low-speed driving behavior. PLoS ONE 2023, 18, e0287489. [Google Scholar] [CrossRef] [PubMed]
  37. Hafeez, F.; Ullah Sheikh, U.; Mas’ud, A.A.; Al-Shammari, S.; Hamid, M.; Azhar, A. Application of the Theory of Planned Behavior in Autonomous Vehicle-Pedestrian Interaction. Appl. Sci. 2022, 12, 2574. [Google Scholar] [CrossRef]
  38. Piazza, A.J.; Knowlden, A.P.; Hibberd, E.; Leeper, J.; Paschal, A.M.; Usdan, S. Mobile device use while crossing the street: Utilizing the theory of planned behavior. Accid. Anal. Prev. 2019, 127, 9–18. [Google Scholar] [CrossRef] [PubMed]
  39. Papadimitriou, E.; Theofilatos, A.; Yannis, G. Patterns of pedestrian attitudes, perceptions and behaviour in Europe. Saf. Sci. 2013, 53, 114–122. [Google Scholar] [CrossRef]
  40. Liu, M.; Wu, J.; Yousaf, A.; Wang, L.; Hu, K.; Plant, K.L.; McIlroy, R.C.; Stanton, N.A. Exploring the relationship between attitudes, risk perceptions, fatalistic beliefs, and pedestrian behaviors in China. Int. J. Environ. Res. Public Health 2021, 18, 3378. [Google Scholar] [CrossRef] [PubMed]
  41. Demir, B.; Özkan, T.; Demir, S. Pedestrian violations: Reasoned or social reactive? Comparing theory of planned behavior and prototype willingness model. Transp. Res. F Traffic Psychol. Behav. 2019, 60, 560–572. [Google Scholar] [CrossRef]
  42. O’Dell, A.L.; Filtness, A.J.; Morris, A.P. Predicting the intention of distracted pedestrians at road crossings. Accid. Anal. Prev. 2022, 173, 106707. [Google Scholar] [CrossRef]
  43. Xin, X.; Jia, N.; Ling, S.; He, Z. The effect of the ‘yield to pedestrians’ policy on risky pedestrian behaviors: Is it a ‘two-edged sword’? Transp. Res. A Policy Pract. 2023, 178, 103870. [Google Scholar] [CrossRef]
  44. Moshki, M.; Khajavi, A.; Sadeghi-Bazargani, H.; Vahedi, S.; Pour-Doulati, S. Developing Pedestrians’ Red-light Violation Behavior Questionnaire (PRVBQ); Assessment of Content Validity and Reliability. Bull. Emerg. Trauma 2020, 8, 98–106. [Google Scholar] [CrossRef]
  45. Lennon, A.; Oviedo-Trespalacios, O.; Matthews, S. Pedestrian self-reported use of smart phones: Positive attitudes and high exposure influence intentions to cross the road while distracted. Accid. Anal. Prev. 2017, 98, 338–347. [Google Scholar] [CrossRef]
  46. Sniehotta, F.F.; Presseau, J.; Araújo-Soares, V. Time to retire the theory of planned behaviour. Health Psychol. Rev. 2014, 8, 1–7. [Google Scholar] [CrossRef] [PubMed]
  47. Liu, J.; Li, C.; Hu, S. Analysis of pedestrian violations based on an integrative model: Evidence from college students in China. J. Transp. Health 2023, 28, 101560. [Google Scholar] [CrossRef]
  48. Hashemiparast, M.; Sharma, M.; Asghari Jafarabadi, M.; Hosseini, Z. Still careless: Findings from a cross-sectional study of young pedestrians’ risky road crossing behaviors. Arch. Public Health 2020, 78, 44. [Google Scholar] [CrossRef]
  49. Matović, B.; Petrović, A.; Damjanović, M.; Bulajić, A.; Ilić, V. Socio-cognitive determinants of pedestrians’ intention to cross on a red light signal: An application of the theory of planned behaviour. Safety 2024, 10, 33. [Google Scholar] [CrossRef]
  50. Bamberg, S.; Ajzen, I.; Schmidt, P. Choice of Travel Mode in the Theory of Planned Behavior: The Roles of Past Behavior, Habit, and Reasoned Action. Basic Appl. Soc. Psychol. 2003, 25, 175–187. [Google Scholar] [CrossRef] [PubMed]
  51. Cristea, M.; Paran, F.; Delhomme, P. Extending the theory of planned behaviour: The role of behavioural options and additional factors in predicting speed behaviour. Transp. Res. Part F Traffic Psychol. Behav. 2013, 21, 122–132. [Google Scholar] [CrossRef]
  52. Xu, Y.; Li, Y.; Zhang, F. Pedestrians’ intention to jaywalk: Automatic or planned? A study based on a dual-process model in china. Accid. Anal. Prev. 2013, 50, 811–819. [Google Scholar] [CrossRef]
  53. Conner, M.; Armitage, C.J. Extending the theory of planned behaviour: A review and avenues for further research. J. Appl. Soc. Psychol. 1998, 28, 1429–1464. [Google Scholar] [CrossRef]
  54. Jovanović, D.; Šraml, M.; Matović, B.; Mićić, S. An Examination of the Construct and Predictive Validity of the Self-Reported Speeding Behavior Model. Accid. Anal. Prev. 2017, 99, 66–76. [Google Scholar] [CrossRef]
  55. Parker, D.; Manstead, A.S.R.; Stradling, S.G. Extending the theory of planned behavior: The role of personal norm. Br. J. Soc. Psychol. 1995, 34, 127–137. [Google Scholar] [CrossRef]
  56. Schwartz, S.H. Normative influences on Altruism. In Advances in Experimental Social Psychology; Elsevier: Amsterdam, The Netherlands, 1977; Volume 10, pp. 221–279. [Google Scholar]
  57. Karami, Z.; Rejali, S.; Aghabayk, K. Investigating pedestrians’ red light running intentions at urban intersections in different traffic Environments: A scenario-based analysis guided by theoretical frameworks. Transp. Res. Part F Traffic Psychol. Behav. 2024, 107, 196–223. [Google Scholar] [CrossRef]
  58. Deery, H.A. Hazard and risk perception among young novice drivers. J. Saf. Res. 1999, 30, 225–236. [Google Scholar] [CrossRef]
  59. Dinh, D.D.; Vũ, N.H.; McIlroy, R.C.; Plant, K.A.; Stanton, N.A. Effect of attitudes towards traffic safety and risk perceptions on pedestrian behaviours in Vietnam. IATSS Res. 2020, 44, 238–247. [Google Scholar] [CrossRef]
  60. Zheng, H.; Giang, W.C.W. Risk perception and distraction engagement with smart devices in different types of walking environments. Accid. Anal. Prev. 2021, 162, 106405. [Google Scholar] [CrossRef]
  61. Ackaah, W.; Larbi, J.; Aidoo, E.N. Pedestrian red-light violations and other safety-related behaviours at signalised crosswalks. Urban Plan. Transp. Res. 2024, 12, 2353073. [Google Scholar] [CrossRef]
  62. Yang, J.; Gauli, N.; Shiwakoti, N.; Tay, R.; Deng, H.; Chen, J.; Nepal, B.; Li, J. Examining the Factors Influencing Pedestrian Behaviour and Safety: A Review with a Focus on Culturally and Linguistically Diverse Communities. Sustainability 2025, 17, 6007. [Google Scholar] [CrossRef]
  63. Pelé, M.; Bellut, C.; Debergue, E.; Gauvin, C.; Jeanneret, A.; Leclere, T.; Nicolas, L.; Pontier, F.; Zausa, D.; Sueur, C. Cultural influence of social information use in pedestrian road-crossing behaviours. R. Soc. Open Sci. 2017, 4, 160739. [Google Scholar] [CrossRef]
  64. Rankavat, S.; Tiwari, G. Influence of actual and perceived risks in selecting crossing facilities by pedestrians. Travel Behav. Soc. 2020, 21, 1–9. [Google Scholar] [CrossRef]
  65. Saxena, A. Texting and crossing: An extended theory of planned behaviour to model the psychological and demographic factors related to pedestrians’ use of cell phone for texting at crosswalks in developing country. IATSS Res. 2023, 47, 50–60. [Google Scholar] [CrossRef]
  66. Granié, M.A.; Pannetier, M.; Gueho, L. Developing a self-reporting method to measure pedestrian behaviors at all ages. Accid. Anal. Prev. 2013, 50, 830–839. [Google Scholar] [CrossRef] [PubMed]
  67. Sogbe, E.; Susilawati, S. Dissecting pedestrian behaviour in Ghana: A cluster-based analysis of safety and risk profiles. Int. J. Inj. Control Saf. Promot. 2025, 1–15. [Google Scholar] [CrossRef]
  68. Soper, D.S. A-Priori Sample Size Calculator for Structural Equation Models. 2026. Available online: https://www.danielsoper.com/statcalc (accessed on 16 March 2026).
  69. Hair, J.F.; Black, W.C.; Babin, B.J.; Anderson, R.E. Multivariate Data Analysis, 7th ed.; Pearson: Harlow, UK, 2014. [Google Scholar]
  70. Fornell, C.; Larcker, D.F. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 1981, 18, 39–50. [Google Scholar] [CrossRef]
  71. Franke, G.R.; Sarstedt, M. Heuristics versus statistics in discriminant validity testing: A comparison of four procedures. Internet Res. 2019, 29, 430–447. [Google Scholar] [CrossRef]
  72. Hooper, D.; Coughlan, J.; Mullen, M. Structural equation modelling: Guidelines for determining model fit. Electron. J. Bus. Res. Methods 2008, 6, 53–60. [Google Scholar]
  73. Bortoleto, A.P.; Kurisu, K.H.; Hanaki, K. Model development for household waste prevention behaviour. Waste Manag. 2012, 32, 2195–2207. [Google Scholar] [CrossRef]
  74. Xiong, B.; Skitmore, M.; Xia, B. A critical review of structural equation modeling applications in construction research. Autom. Constr. 2015, 49, 59–70. [Google Scholar] [CrossRef]
  75. Wolny, R.; Kol, J.; Stolecka-Makowska, A.; Szojda, G. Digital Consumer Behavior in Poland and Its Environmental Impact Within the Framework of Sustainability. Sustainability 2025, 17, 4691. [Google Scholar] [CrossRef]
  76. Liu, H.; Liu, X.; Jia, M. Revealing the speeding principle based on the extended theory of planned behavior. Transp. Res. Part F Traffic Psychol. Behav. 2025, 109, 1080–1099. [Google Scholar] [CrossRef]
  77. Aarts, H.; Dijksterhuis, A. Habits as knowledge structures: Automaticity in goal-directed behavior. J. Pers. Soc. Psychol. 2000, 78, 53. [Google Scholar] [CrossRef]
Sustainability 18 05585 g001
Figure 1. The conceptual framework of the integrative model.
Figure 1. The conceptual framework of the integrative model.
Sustainability 18 05585 g001
Sustainability 18 05585 g002
Figure 2. Structural model results.
Figure 2. Structural model results.
Sustainability 18 05585 g002
Table 1. Measurement model evaluation.
Table 1. Measurement model evaluation.
FactorsIndicatorsLoadingsStd. Errorp-ValueCronbach’s
Alpha		CRAVE
Attitude: ATTATT_1 0.7910.021<0.0010.8960.8980.688
ATT_20.8500.017<0.001
ATT_30.8010.020<0.001
ATT_40.8680.016<0.001
Subject norm: SN SN_1 0.7970.020<0.0010.9040.9040.702
SN_20.8500.016<0.001
SN_30.8720.015<0.001
SN_40.8310.017<0.001
Perceived behavioral control: PBCPBC_1 0.7710.022<0.0010.8850.8890.666
PBC_20.8340.018<0.001
PBC_30.8800.015<0.001
PBC_40.7650.022<0.001
Habit: HABHAB_1 0.8370.017<0.0010.8870.8900.726
HAB_2 0.8280.018<0.001
HAB_3 0.8980.014<0.001
Personal norm: PN PN_1 0.8010.021<0.0010.8630.8640.681
PN_2 0.8780.017<0.001
PN_3 0.7900.022<0.001
Past behavior: PBPB_10.7280.029<0.0010.8060.8080.584
PB_20.8580.026<0.001
PB_30.7270.029<0.001
Behavioral intention: BIBI_10.8180.018<0.0010.8720.8720.695
BI_20.8520.015<0.001
BI_30.8310.017<0.001
Risky pedestrian crossing behavior: RPBRPB_10.7680.025<0.0010.8110.8120.590
RPB_20.7980.023<0.001
RPB_30.7390.026<0.001
Perceived risk of built environment: PREPRE_10.7250.027<0.0010.8230.8380.622
PRE_20.9070.020<0.001
PRE_30.7390.026<0.001
Perceived risk of behavior: PRBPRB_10.7490.024<0.0010.8480.8610.669
PRB_20.9100.016<0.001
PRB_30.7750.023<0.001
Perceived built environment: PBEPBE_10.7060.029<0.0010.8310.8380.569
PBE_20.8170.023<0.001
PBE_30.7070.028<0.001
PBE_40.7500.026<0.001
Table 2. Heterotrait-monotrait (HTMT) ratios for discriminant validity.
Table 2. Heterotrait-monotrait (HTMT) ratios for discriminant validity.
ConstructsATTSNPBCHABPNPBPREPRBPBEBIRPB
ATT-
SN0.351-
PBC0.2890.434-
HAB0.3430.4860.613-
PN0.3050.6800.4100.450-
PB0.1670.2410.0530.0670.231-
PRE0.2560.3640.3050.3780.4210.023-
PRB0.3750.6280.3940.4880.6190.1110.388-
PBE0.4570.3890.1810.3070.3540.0520.3500.401-
BI0.5540.7180.6610.7610.7280.2180.5720.6990.447-
RPB0.4690.4780.5960.6740.4580.1460.3490.4410.3910.767-
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Chayphong, S.; Iamtrakul, P.; Yoh, K. An Integrated Behavioral Framework for Risky Pedestrian Crossing Behavior Extending the Theory of Planned Behavior. Sustainability 2026, 18, 5585. https://doi.org/10.3390/su18115585

AMA Style

Chayphong S, Iamtrakul P, Yoh K. An Integrated Behavioral Framework for Risky Pedestrian Crossing Behavior Extending the Theory of Planned Behavior. Sustainability. 2026; 18(11):5585. https://doi.org/10.3390/su18115585

Chicago/Turabian Style

Chayphong, Sararad, Pawinee Iamtrakul, and Kento Yoh. 2026. "An Integrated Behavioral Framework for Risky Pedestrian Crossing Behavior Extending the Theory of Planned Behavior" Sustainability 18, no. 11: 5585. https://doi.org/10.3390/su18115585

APA Style

Chayphong, S., Iamtrakul, P., & Yoh, K. (2026). An Integrated Behavioral Framework for Risky Pedestrian Crossing Behavior Extending the Theory of Planned Behavior. Sustainability, 18(11), 5585. https://doi.org/10.3390/su18115585

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop