Determinants of truck operators’ intention to pay for road safety technology in Thailand: a structural equation modeling approach

1 Introduction

Road traffic accidents constitute one of the most pressing public health and socioeconomic challenges confronting nations worldwide, particularly in low- and middle-income countries (LMICs). The World Health Organization’s Global Status Report on Road Safety (2023) reveals that approximately 1.19 million people die annually from road traffic crashes, with traffic injuries ranking as the leading cause of death among young people aged 5–29 years. Beyond the immediate human toll, these accidents generate substantial economic losses, estimated at 2%–5% of gross domestic product (GDP) in most countries, and impede sustainable development efforts through workforce reduction and healthcare burden escalation (Wim and Henk, 2016; World Health Organization, 2023).

Thailand presents a particularly critical case for road safety research. The country has consistently ranked among the nations with the highest traffic fatality rates in Asia and the ASEAN region. According to the 2023 Global Status Report on Road Safety, Thailand ranked 18th globally in road traffic deaths per capita (World Health Organization, 2023). In 2024, the mortality rate stood at 26.44 deaths per 100,000 population, resulting in 17,209 fatalities—a Figure 1 that, while representing a 5.86% reduction from 2023, still significantly exceeds the government’s target of reducing fatalities to 12 per 100,000 population by 2027 (Road Safety Research Center, 2024; Thailand Ministry of Transport, 2024). The economic impact is equally staggering: between 2018 and 2022, Ministry of Public Health data indicate 1,170,708 severe injuries with healthcare costs totaling 33 billion Thai Baht. According to the Thailand Development Research Institute (Thailand Development Research Institute, 2024), each severe injury costs approximately 2 million Baht, permanent disability 4.8 million Baht, and each fatality 6.7 million Baht.

Heavy-duty trucks occupy a unique position in the road safety landscape. While not the most numerically prevalent vehicle type in accidents, their physical characteristics—high mass, extended stopping distances, and substantial momentum—contribute to disproportionately severe crash outcomes. Data from Thailand’s Road Traffic Injury Data Collaboration (RTIDC) system for 2024 documented 5,331 accidents involving large trucks (6-wheel trucks and above), with an estimated economic loss of 1,393 million Baht annually (Thailand Department of Highways, 2024; Thaindland Ministry of Public Health, 2024). Research consistently demonstrates that human factors constitute the primary cause of road traffic accidents, with human error accounting for up to 68.6% of severe truck accidents (McCarty and Kim, 2024; Waskito et al., 2024; Zeng et al., 2024).

Advanced safety technologies offer a promising avenue for mitigating human error and reducing accident risks. Technologies such as collision warning systems, lane departure warnings, fatigue monitoring systems, and advanced driver assistance systems (ADAS) can compensate for momentary lapses in driver attention and prevent errors caused by fatigue—a particularly relevant concern for long-haul truck operators. However, the adoption of these technologies is not automatic; it requires substantial capital investment and is ultimately contingent upon operators’ willingness to allocate financial resources toward safety improvements. This willingness, reflected through the psychological construction of intention to pay (ITP), represents a critical behavioral precursor to actual investment decisions.

While the Theory of Planned Behavior (TPB) has been widely applied to transportation safety behaviors, TAM provides a more appropriate framework for technology adoption contexts. Unlike TPB, which specifies multiple direct predictors of intention (attitude, subjective norms, perceived behavioral control), TAM explicitly models attitude as a central mediating mechanism between technology beliefs and adoption intentions. This mediational structure, validated extensively in technology acceptance research (Davis, 1989; Venkatesh and Davis, 2000), proves particularly suitable for understanding investment decisions in safety technologies.

In the context of road safety investment, intention to pay (ITP) represents a specific manifestation of behavioral intention where individuals express willingness to allocate financial resources toward safety-enhancing measures. Previous studies have examined willingness to pay (WTP) and ITP for road safety improvements across various populations. Mofadal et al. (2015) identified age, income, gender, safety perception, education, and occupation as significant predictors of WTP among pedestrians in Sudan. Widyastuti and Utanaka (2020) found that older and higher-income motorcycle users in Indonesia demonstrated greater WTP for accident risk reduction. Subhan et al. (2021) applied behavioral theory to investigate public ITP for road safety measures in Pakistan, finding that attitude, subjective norms, and perceived behavioral control exerted significant positive effects, with attitude emerging as the most influential predictor.

In Thailand, several studies have examined WTP for road safety. Jomnonkwao et al. (2021) found that social influence, perceived behavioral control, and safety intention positively affected WTP among personal car drivers. Wisutwattanasak et al. (2022) revealed that male drivers with higher risk-taking behaviors demonstrated greater WTP, while more experienced drivers expressed lower WTP due to confidence in their driving skills. Champahom et al. (2023) demonstrated that behavioral intention, mediated by attitude and perceived behavioral control, was a major determinant of WTP among motorcyclists.

Despite these advances, significant research gaps remain. First, studies specifically examining truck operators—a high-risk population with substantial economic influence on the transportation sector—remain limited. Second, existing research has not adequately integrated TAM constructs with psychological variables specific to the commercial transportation context. Specifically, the adaptation of TAM for road safety technology requires incorporating perceived effectiveness (analogous to perceived usefulness) and perceived safety responsibility (PSR)—constructs that capture the unique decision-making context of commercial operators investing in safety technologies. Third, the role of organizational factors, particularly firm size, in moderating the relationship between psychological factors and investment intentions has received insufficient attention. These gaps are particularly consequential given that truck operators’ investment decisions are shaped not only by individual psychological factors but also by organizational constraints, regulatory requirements, and business considerations that differ fundamentally from individual consumer decisions.

Taken together, the reviewed literature reveals that TAM-based technology adoption research and WTP/ITP-based road safety research have largely developed along separate intellectual trajectories. TAM studies have focused on usage or behavioral intention in information systems and intelligent transportation contexts without extending to financially specific investment outcomes, while WTP/ITP studies have applied the Theory of Planned Behavior or contingent valuation methods without incorporating technology-specific belief structures such as perceived effectiveness. No prior study has systematically integrated TAM’s mediational architecture with a payment-specific outcome in the commercial trucking context, nor has any study simultaneously accounted for domain-specific constructs such as perceived safety responsibility and organizational factors such as firm size. The present study addresses this gap by integrating these two research traditions within a unified structural framework.

This study addresses the identified research gaps by developing and testing a comprehensive structural model of truck operators’ intention to pay for safety technology adoption in Thailand based on an Extended TAM framework. The specific objectives are: (1) to validate a measurement model incorporating TAM constructs (perceived effectiveness as perceived usefulness, attitude, behavioral intention) and psychological variables (perceived safety responsibility (PSR)) specific to the commercial trucking context; (2) to examine the direct and indirect effects of these constructs on intention to pay using structural equation modeling; and (3) to investigate the moderating role of demographic and organizational factors, including firm size and legal regulations, on the hypothesized relationships.

This research contributes to the literature in several ways. Theoretically, it extends TAM by integrating domain-specific psychological constructs (perceived effectiveness and perceived safety responsibility (PSR)) that capture the unique decision-making context of commercial operators. Methodologically, it provides a validated measurement instrument for assessing safety investment intentions among truck operators. Practically, the findings offer evidence-based recommendations for policymakers and industry stakeholders seeking to promote safety technology adoption in the commercial trucking sector.

2 Literature review and hypotheses development

2.1 Technology acceptance model in road safety context

The Technology Acceptance Model (Davis, 1989) is one of the most widely validated theories for explaining technology adoption and usage behaviors. Grounded in the Theory of Reasoned Action (Fishbein and Ajzen, 1975), TAM proposes a parsimonious structure where two belief constructs—perceived usefulness and perceived ease of use—determine attitude toward technology use, which subsequently influences behavioral intention and actual usage. Perceived usefulness captures outcome expectations (whether the technology will improve performance or achieve desired goals), while perceived ease of use captures process expectations (whether the technology requires minimal effort to operate).

TAM’s applicability has been demonstrated across diverse technology contexts, including information systems (Davis, 1989), e-commerce (Gefen et al., 2003), and intelligent transportation systems (Chen and Chao, 2011). Critically for this study, TAM has been successfully adapted to safety technology contexts. Yao et al. (2023) demonstrated that perceived safety influences attitude and behavioral intention in autonomous vehicle acceptance, validating TAM’s structure in safety-critical domains. Sun et al. (2016) applied TAM to congestion pricing acceptance, showing that psychological factors mediate policy acceptance through attitude formation.

Extended TAM frameworks incorporate additional constructs to enhance explanatory power in specific domains while maintaining the theory’s core mediational structure (Venkatesh and Davis, 2000). For road safety technology, two extensions are theoretically justified: (1) replacing perceived usefulness with perceived effectiveness to capture domain-specific efficacy beliefs about accident reduction and safety improvement; and (2) incorporating perceived safety responsibility (PSR) to address the moral and social dimensions of safety investment unique to commercial operators’ organizational decision-making contexts.

The theoretical advantage of TAM over alternative frameworks (such as the Theory of Planned Behavior) for technology investment contexts lies in its explicit specification of attitude as a mediating mechanism. While TPB allows for direct effects from beliefs to behavioral intention, TAM positions attitude as the central psychological mechanism that transforms cognitive beliefs about technology attributes into investment intentions—a structure that proves more accurate for understanding safety technology adoption among commercial operators, as demonstrated in this study.

The Technology Acceptance Model positions behavioral intention as the most proximal predictor of technology adoption behavior. Within TAM, intention represents the degree to which an individual has formulated conscious plans to adopt or use technology. TAM identifies two primary belief constructs that shape this intention through attitude formation: (1) perceived usefulness, reflecting beliefs about performance enhancement and outcome achievement; and (2) perceived ease of use, capturing beliefs about effort required. In road safety technology contexts, perceived usefulness is operationalized as perceived effectiveness—beliefs about technology efficacy in reducing accidents and improving safety outcomes.

In the road safety domain, TAM has demonstrated substantial explanatory power. Castanier et al. (2013) found that psychological constructs significantly predicted intention to commit road violations among French drivers, in French drivers. Cristea et al. (2013) extended behavioral theory by incorporating behavioral options and found that attitude was the strongest predictor of speeding intentions. While previous research has applied various behavioral theories including TPB to road safety contexts (Castanier et al., 2013; Zhou et al., 2016), TAM provides a more precise theoretical framework for technology investment decisions. TAM’s explicit specification of attitude as a mediating mechanism—rather than as one of several direct predictors—better captures the psychological process through which beliefs about technology effectiveness translate into investment intentions.

2.2 Intention to pay for road safety

Intention to pay (ITP) for road safety represents a behavioral intention reflecting individuals’ or organizations’ disposition to allocate financial resources toward safety-enhancing measures. This construct is conceptually related to, but distinct from, willingness to pay (WTP)—a stated preference measure used in economic valuation studies. While WTP typically captures hypothetical monetary valuations derived from contingent valuation methods, ITP focuses on the psychological and attitudinal precursors to actual payment behavior (Hair et al., 2013; Haddak et al., 2016).

Subhan et al. (2021) conducted a seminal study examining public ITP for road safety improvement in Pakistan using the Theory of Planned Behavior. Their structural equation modeling analysis revealed that attitude (β = 0.427, p < 0.001), subjective norms (β = 0.125, p < 0.05), and perceived behavioral control (β = 0.185, p < 0.01) all positively influenced ITP, with attitude emerging as the dominant predictor. The study further identified perceived effectiveness as a significant positive influence on attitude, suggesting that beliefs about the efficacy of safety measures shape favorable evaluations that subsequently drive investment intentions.

In the Thai context, research on WTP and ITP has primarily focused on individual road users. Jomnonkwao et al. (2021) found that social influence and perceived behavioral control positively affected WTP among car drivers. Champahom et al. (2023) demonstrated that behavioral intention, mediated by attitude, was a primary determinant of WTP among motorcyclists, consistent with TAM’s theoretical predictions. Jaździk-Osmólska (2021) found that risk perception significantly influenced WTP, with drivers willing to pay four times more to avoid fatality risk compared to minor injury risk. These findings underscore the importance of psychological factors, particularly attitude and risk perception, in shaping safety investment decisions.

2.3 Hypotheses development

2.3.1 Perceived effectiveness and attitude

Perceived effectiveness refers to the degree to which individuals believe that a proposed policy, measure, or technology can effectively reduce accident risks and improve overall road safety (Auzoult et al., 2015). This construct captures the cognitive evaluation of intervention efficacy—a critical factor in technology adoption decisions. Research has demonstrated that when the public perceives policies as effective and non-threatening to personal interests, acceptance increases substantially (Eriksson et al., 2006). According to TAM, perceived usefulness (operationalized here as perceived effectiveness) influences attitude toward technology use, which subsequently affects behavioral intention (Davis, 1989; Venkatesh and Davis, 2000). In safety technology investment contexts, operators who perceive technologies as effective should develop favorable attitudes toward investment. Accordingly, we hypothesize:

Hypothesis 1 Perceived effectiveness positively influences attitude toward safety technology investment.

2.3.2 Perceived safety responsibility and attitude

Perceived safety responsibility (PSR) reflects the degree to which truck operators feel personally and organizationally responsible for traffic safety and accident prevention. This construct is conceptually equivalent to the ‘Attitude towards Traffic Safety Responsibility’ (ATSR) proposed by Subhan et al. (2021) and the benevolence construct of (Andersson and Lindberg, 2009), which captures individuals’ concern for others’ safety beyond self-interest. Consistent with TAM’s terminology, we renamed this construct from ATSR to PSR to distinguish it clearly from the Attitude (ATT) construct and to reflect its nature as a responsibility belief/perception rather than an evaluative attitude toward a specific behavior. PSR represents a domain-specific extension of TAM, capturing the moral and social dimensions of safety technology investment unique to commercial operators. In extended TAM frameworks, additional belief constructs can enhance model explanatory power by addressing context-specific motivations (Taylor and Todd, 1995; Venkatesh and Davis, 2000). When operators perceive strong safety responsibility, they develop more favorable attitudes toward technologies that fulfill this responsibility. This pathway (PSR→ATT) reflects how moral foundations shape cognitive evaluations of technology investments. Thus:

Hypothesis 2 Perceived safety responsibility positively influences attitude toward safety technology investment.

Hypothesis 3 Attitude toward safety technology investment positively influences behavioral intention.

2.3.3 Attitude and intention to pay

Attitude toward the behavior represents the individual’s favorable or unfavorable evaluation of performing the behavior (Ajzen, 1991). Attitude is formed through beliefs about the likely outcomes of the behavior and evaluations of these outcomes. In road safety technology investment, attitude encompasses evaluative beliefs about the wisdom, benefits, and value proposition of safety technology investments. Moyano Dı́az (2002) found that positive attitudes toward safety increased intention to comply with safe behaviors. Subhan et al. (2021) identified attitude as the strongest predictor of ITP, suggesting that favorable evaluations of safety investments are associated with payment intentions. Thus:

Hypothesis 4 Attitude toward safety technology investment positively influences intention to pay for safety technology.

It is important to note that behavioral intention (BI) and intention to pay (ITP) are treated as conceptually and empirically distinct constructs in the present model. BI items capture general endorsement of safety-supporting behaviors (e.g., willingness to recommend payment to colleagues, expressing support for road safety in principle), reflecting a dispositional orientation that is consistent with TAM’s treatment of behavioral intention as general readiness to adopt a technology. ITP items, by contrast, measure an organisation’s concrete commitment to allocate financial resources toward a specific safety technology investment. This distinction is consistent with the broader behavioral literature, wherein specific investment decisions are mediated by evaluative attitudes rather than general behavioral intentions (Ajzen, 1991). Discriminant validity analyses confirm that the two constructs are empirically distinct: the square root of AVE for BI (0.851) and ITP (0.806) both exceed the inter-construct correlation (r = 0.718), satisfying the Fornell–Larcker criterion. Furthermore, the differential predictive pattern observed in this study—attitude directly predicting both BI and ITP, but BI not significantly predicting ITP—demonstrates that the two constructs occupy non-redundant positions in the causal structure and carry theoretically meaningful non-overlapping information.

2.3.4 Effects of demographic and organizational factors

Beyond the core TAM relationships, demographic and organizational characteristics may influence the technology adoption process. Prior research demonstrates that firm size, operator age, education, and regulatory environment significantly shape safety investment decisions (Medina-Flintsch et al., 2017; Subhan et al., 2023).

2.3.4.1 Firm size and safety perceptions

Larger organizations typically possess greater financial resources, more formalized safety management systems, and stronger reputation concerns—factors that enhance both safety consciousness and perceived technology effectiveness. Accordingly, we hypothesize:

Hypothesis 5 Firm size positively influences perceived safety responsibility.

Hypothesis 6 Firm size positively influences perceived effectiveness of safety technology.

Legal Regulations and Technology Perceptions. Regulatory frameworks signal institutional priorities and enhance technology legitimacy. Operators who perceive legal requirements as influential should demonstrate greater confidence in technology effectiveness and stronger behavioral intentions:

Hypothesis 7 Perceived influence of legal regulations positively influences behavioral intention.

Hypothesis 8 Perceived influence of legal regulations positively influences perceived effectiveness of safety technology.

3 Methodology

3.1 Research design and conceptual framework

This study employed a cross-sectional survey design to examine the determinants of truck operators’ intention to pay for safety technology in Thailand. The conceptual framework (see Figure 2) adapts the Technology Acceptance Model by replacing perceived usefulness with perceived effectiveness (domain-specific operationalization) and integrating perceived safety responsibility (PSR) as an additional construct specific to road safety investment decisions, hypothesizing relationships between perceived effectiveness (PE), attitude (ATT), behavioral intention (BI), perceived safety responsibility (PSR), and intention to pay (ITP). Additionally, demographic and organizational factors—including gender, age, education level, operator type, firm size, and legal regulations—were incorporated as potential predictors and moderators.

The research utilized the Contingent Valuation Method (CVM), a direct valuation technique grounded in market-based assumptions, to elicit intention to pay responses. CVM is widely employed in environmental and health economics to estimate the economic value individuals place on non-market goods, including safety improvements (Robert Cameron Mitchell, 1989). The approach is particularly suitable for this study as it allows assessment of investment intentions for technologies that operators may not yet have adopted.

3.2 Participants and sampling

The target population comprised operators, managers, and supervisory-level personnel involved in freight transportation using large trucks (6 wheels and above) in Thailand. This included both non-regular route truck operators and private truck operators. As of 30 September 2024, Thailand had 381,549 registered truck operators (40,802 non-regular route operators and 340,747 private operators) with a total of 609,521 registered large trucks (Thailand Department of Land Transport, 2024).

Sample size determination followed guidelines for structural equation modeling using Maximum Likelihood estimation. Golob (2003); Hair et al. (2010) recommend sample sizes of 5–15 observations per measured variable. With 18 observed variables in the measurement model, the minimum required sample size was calculated as 270 (18 × 15), with an additional 10% buffer for potential data loss, yielding a target sample of 297 respondents. The final sample consisted of 345 valid responses, exceeding the minimum threshold and ensuring adequate statistical power.

Data collection employed purposive sampling, with respondents recruited through multiple channels: online surveys, field visits to trucking companies and logistics facilities, and mail surveys with prepaid return envelopes. The total of 800 questionnaires was distributed based on a conservative expected response rate of approximately 40%, yielding 345 completed responses (43% response rate). Data collection occurred between January to May 2025. All returned questionnaires were subjected to a quality control process prior to analysis. First, completeness checks were performed, and questionnaires with any missing responses on the Likert-scale items were excluded from the dataset. Second, response patterns such as straight-lining were examined to identify and remove potentially unreliable responses. The study received ethical approval from the Human Research Ethics Committee of Suranaree University of Technology.

Regarding participant identification and inclusion criteria, eligible respondents were defined as operators, company representatives, or owners with direct freight-transport responsibilities involving large trucks (six wheels and above) who were personally involved in, or knowledgeable about, safety technology adoption and investment decisions within their organizations. This criterion ensured that respondents could meaningfully evaluate the constructs under study. Regarding recruitment channels: (1) Online surveys were distributed through the Department of Land Transport and the Thai Trucking Association’s official communication channels, reaching operators and logistics managers with registered contact information. (2) Field visits were conducted at trucking companies and logistics facilities in Nakhon Ratchasima and surrounding provinces, where research assistants delivered printed questionnaires directly to eligible personnel, who completed them on site. Researchers remained present solely to clarify any questions about wording or terminology if requested. (3) Mail surveys with prepaid return envelopes were sent to operators drawn from the Department of Land Transport registry. A screening protocol was applied consistently across all channels: prospective respondents were first asked whether their role involved freight transport using large trucks and whether they had direct knowledge relevant to safety technology investment decisions. Those who did not satisfy both criteria were excluded.

As a non-probability sampling method, purposive sampling does not allow the sample to be considered fully representative of the broader population of 381,549 registered truck operators in Thailand. Purposive sampling was adopted because the target population—personnel with direct decision-making responsibilities for safety technology investment—is not uniformly distributed across all registered operators and cannot be effectively identified through simple random sampling of the national registry. The findings should therefore be interpreted with caution regarding their generalizability to the wider population. Future research employing probability-based sampling frames would allow more confident population-level inference.

3.3 Measures and instrumentation

The survey instrument was developed through systematic review of established scales in the transportation safety and behavioral intention literature. Items were adapted from validated instruments by Ajzen (1991), Andersson and Lindberg (2009), Auzoult et al. (2015), Castanier et al. (2013), Davis (1989), Jomnonkwao et al. (2021), Subhan et al. (2021), Sun et al. (2016), Svensson and Vredin Johansson (2010), Taylor and Todd (1995), Zhou et al. (2016).

The questionnaire operationalized Extended TAM constructs with adaptations for road safety technology context. Section 1 collected demographic and organizational information: gender, age, education level, operator type (non-regular route vs. private), firm size (categorized as S: 1–10 trucks, M: 11–30 trucks, L: 31–50 trucks, XL: >50 trucks), and perceived influence of legal regulations on technology adoption. Section 2 assessed the five latent constructs using 18 items measured on 5-point Likert scales (1 = strongly disagree to 5 = strongly agree):

• Perceived Effectiveness (PE): Four items assessing beliefs about safety technology efficacy in improving road safety, reducing accidents, assisting driver control, and modifying unsafe driving behaviors.
• Attitude (ATT): Three items measuring evaluations of safety investment as wise, beneficial in reducing risk, and worthwhile for accident prevention.
• Behavioral Intention (BI): Four items assessing willingness to pay for road safety, recommend payment to colleagues, pay to save lives, and invest in safety technology installation.
• Perceived Safety Responsibility (PSR): Three items capturing sense of responsibility for accident avoidance, shared responsibility for road safety, and prioritization of safety over delivery speed. This construct operationalizes responsibility beliefs conceptually equivalent to Subhan et al. (2021) ATSR and Andersson and Lindberg (2009) benevolence construct, renamed to PSR to distinguish it from attitudinal measures.
• Intention to Pay (ITP): Four items measuring intention to pay for safety technology, support road safety improvements, improve vehicle safety standards, and invest whenever possible.

Content validity was established through expert review. Five domain experts evaluated item-objective congruence (IOC), yielding values between 0.60 and 1.00, indicating acceptable content validity. The questionnaire was then pilot-tested with 30 respondents from the target population, yielding a Cronbach’s alpha of 0.951, exceeding the 0.70 threshold recommended by Tavakol and Dennick (2011) for acceptable internal consistency.

3.4 Data analysis

Data analysis proceeded in three stages. First, preliminary analyses examined data quality, including missing data patterns, univariate normality (skewness <3, kurtosis <10; Subhan et al. (2021), and descriptive statistics. Second, measurement model of SEM was conducted to validate the measurement model, assessing construct reliability (CR > 0.60), convergent validity (Average Variance Extracted, AVE >0.50; factor loadings >0.50), discriminant validity (square root of AVE > inter-construct correlations), and internal consistency (Cronbach’s alpha >0.70) following criteria established by Fornell and Larcker (1981), Hair et al. (2010), and Brown and Groeger (1988).

Third, structural path of SEM was employed to test the hypothesized relationships. Model fit was evaluated using multiple indices: chi-square/degrees of freedom ratio (χ²/df < 3), Comparative Fit Index (CFI >0.90), Tucker-Lewis Index (TLI >0.90), Root Mean Square Error of Approximation (RMSEA <0.08), and Standardized Root Mean Square Residual (SRMR <0.08). Maximum Likelihood Estimation (MLE) was used given the data’s approximate normal distribution. All analyses were conducted using appropriate SEM software with robust standard error estimation.

4 Results

4.1 Sample characteristics

Table 1 presents the demographic and organizational characteristics of the 345 respondents. The sample was predominantly male (63.48%), reflecting the gender composition of the trucking industry. Age distribution was relatively balanced between 21–40 years (51.88%) and over 40 years (48.12%). Educational attainment varied, with 45.22% holding below-bachelor’s degrees, 44.93% holding bachelor’s degrees, and 9.86% holding postgraduate qualifications.

TABLE 1. Demographic and organizational characteristics of respondents (N = 345).

Regarding organizational characteristics, non-regular route truck operators comprised 68.12% of the sample, with private truck operators accounting for 31.88%. Firm size distribution skewed toward smaller operations: 59.13% operated 1–10 trucks (small), 16.23% operated 11–30 trucks (medium), 10.14% operated 31–50 trucks (large), and 14.49% operated more than 50 trucks (extra-large). Notably, 94.20% of respondents believed that legal regulations influence safety technology adoption decisions, while only 5.80% did not perceive such influence.

4.2 Descriptive statistics

Table 2 presents descriptive statistics for all measured variables. Mean scores ranged from 3.95 to 4.09, indicating generally positive evaluations across constructs. Perceived Safety Responsibility (PSR) demonstrated the highest mean scores (overall M = 4.03), with the item “The company believes safety is more important than delivery speed” receiving the highest agreement (M = 4.09, SD = 0.827). This finding suggests strong baseline safety values among truck operators.

TABLE 2. Measurement items and descriptive statistics.

SD, standard deviation; Skew. = skewness; Kurt. = Kurtosis. All items measured on 5-point Likert scale (1 = strongly disagree to 5 = strongly agree).

Perceived Effectiveness (PE) also demonstrated favorable ratings (overall M = 3.86), with respondents expressing confidence in safety technology’s ability to improve road safety (PE1: M = 3.91) and reduce accidents (PE2: M = 3.84). Intention to Pay (ITP) items showed moderately positive responses (overall M = 3.73), with item ITP2 (“The company intends to pay to support enhanced road safety”) receiving the highest endorsement (M = 3.79).

Standard deviations ranged from 0.797 to 0.942, indicating moderate response variability without extreme dispersion. Normality assessment revealed skewness values between −0.674 and −0.206 and kurtosis values between −0.591 and 0.713, all within acceptable limits for SEM analysis (skewness < |3|, kurtosis < |10|; Subhan et al. (2021)). These results supported the use of Maximum Likelihood Estimation.

4.3 Structural equation model

4.3.1 Reliability and validity

Table 3 presents the psychometric properties of the measurement model. All factor loadings were statistically significant (p < 0.001) and ranged from 0.718 to 0.981, exceeding the 0.50 threshold recommended by Fornell and Larcker (1981) and Hair et al. (2010). Construct reliability (CR) values ranged from 0.855 to 0.913, all exceeding the 0.60 criterion Diamantopoulos and Siguaw (2000). Average Variance Extracted (AVE) values ranged from 0.613 to 0.724, all exceeding the 0.50 threshold, indicating satisfactory convergent validity (Brown and Groeger, 1988). Cronbach’s alpha coefficients ranged from 0.829 to 0.904, demonstrating excellent internal consistency.

TABLE 3. Construct reliability and validity of the measurement model.

α = Cronbach’s alpha; CR, composite reliability; AVE, average variance extracted; R ² = squared multiple correlation.

p < 0.001.

Table 4 presents the correlation matrix and discriminant validity assessment. Inter-construct correlations ranged from 0.396 to 0.801, all significant at p < 0.01. The highest correlation was observed between Attitude (ATT) and Intention to Pay (ITP) at r = 0.801, approaching but not exceeding conventional multicollinearity thresholds. Discriminant validity was confirmed using the Fornell-Larcker criterion: the square root of AVE for each construct (ATT: √0.662 = 0.814; ITP: √0.649 = 0.806) exceeded the corresponding inter-construct correlation (0.801), supporting construct distinctiveness (Fornell and Larcker, 1981; Rönkkö and Cho, 2022).

TABLE 4. Correlation matrix and discriminant validity assessment.

Diagonal values in parentheses represent the square root of AVE., Off-diagonal values represent inter-construct correlations.

p < 0.01 (two-tailed).

To evaluate whether behavioral intention (BI) and intention to pay (ITP) are empirically distinguishable rather than redundant operationalizations of a single construct, a series of nested model comparisons was conducted (see Table 5). A two-factor BI–ITP specification, in which BI and ITP items were allowed to load on separate latent factors, demonstrated substantially better fit (CFI = 0.991, TLI = 0.987, RMSEA = 0.097, SRMR = 0.021) than a constrained one-factor model in which all BI and ITP items were forced onto a single latent variable (CFI = 0.958, TLI = 0.941, RMSEA = 0.204, SRMR = 0.054). Although the RMSEA of the two-factor model marginally exceeded the conventional 0.08 threshold, the magnitude of improvement relative to the one-factor specification (ΔCFI = +0.033, ΔRMSEA = −0.107) provides clear evidence that the two constructs are empirically distinguishable. These results are consistent with the discriminant validity findings reported above, in which the square root of AVE for both BI (0.851) and ITP (0.806) exceeded their inter-construct correlation (r = 0.718), satisfying the Fornell–Larcker criterion.

TABLE 5. Model comparisons for construct distinctiveness and common method bias.

Given the reliance on self-reported data collected from a single source, two complementary assessments were conducted to evaluate the potential for common method bias (Podsakoff et al., 2003). First, a single-factor CFA was estimated in which all 18 manifest indicators were constrained to load on one latent common factor. This model demonstrated markedly poor fit (CFI = 0.821, TLI = 0.797, RMSEA = 0.218, SRMR = 0.099) compared with the baseline multi-factor CFA (CFI = 0.982, TLI = 0.978, RMSEA = 0.072, SRMR = 0.035), indicating that the observed covariance structure is unlikely to be adequately explained by a single common method factor. Second, Harman’s single-factor test using principal component analysis (unrotated solution) indicated that the first component accounted for 48.734% of the total variance—below the commonly applied 50% threshold—while three components yielded eigenvalues greater than 1.0, further suggesting that variance is not dominated by a single source. Taken together, these results do not indicate that common method bias constitutes a primary threat to the validity of the present findings.

4.3.2 Model fit

Table 6 presents the structural equation model demonstrated excellent fit to the empirical data: χ² = 155.017, df = 138, p = 0.1527, χ²/df = 1.1233, CFI = 0.996, TLI = 0.994, RMSEA = 0.019 (90% CI: 0.000, 0.033), SRMR = 0.033. The model explained 76.0% of variance in Intention to Pay (R ² = 0.760), indicating substantial explanatory power.

TABLE 6. Model fit indices.

χ² = Chi-square; df = Degrees of Freedom; CFI, comparative fit index; TLI, Tucker-Lewis Index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual.

4.3.3 TAM hypothesis testing

Table 7 and Figure 3 presents the structural path coefficients and hypothesis testing results for all eight hypotheses. Core TAM Hypotheses (Hypothesis 1–Hypothesis 4). All four core TAM hypotheses were supported, consistent with the model’s applicability for explaining safety technology investment decisions.

TABLE 7. Structural equation modeling results: Path coefficients and hypothesis testing.

β = Standardized path coefficient; S.E., standard error; PE, perceived effectiveness; ATT, attitude; BI, behavioral intention; PSR, perceived safety responsibility; ITP, intention to pay; EDU, education; LAW, legal regulations; SIZE, firm size; AGE, age.

Hypothesis 1 (PE→ATT): Perceived effectiveness was significantly associated with attitude toward safety technology investment (β = 0.733, SE = 0.052, t = 14.061, p < 0.001), representing the strongest belief-attitude relationship. This finding is consistent with TAM’s theoretical proposition that operators who perceive safety technologies as effective in reducing accidents and improving operational safety tend to report more favorable attitudes toward adopting these technologies.

Hypothesis 2 (PSR→ATT): Perceived safety responsibility did not significantly influence attitude toward safety technology investment (β = 0.025, SE = 0.063, t = 0.392, p = 0.695). The effect size was negligible and statistically insignificant, suggesting that perceived safety responsibility alone may not be sufficient to shape operators’ evaluative judgments regarding safety technology investments. This finding does not provide empirical support for the domain-specific extension of TAM in this path. It implies that responsibility perceptions may operate indirectly—possibly through perceived effectiveness—rather than exerting a direct attitudinal effect.

Hypothesis 3 (ATT→BI): Attitude was significantly associated with behavioral intention (β = 0.946, SE = 0.066, t = 14.335, p < 0.001), exhibiting the strongest path coefficient in the entire model. This result is consistent with the central role of attitude in TAM’s theoretical structure, wherein favorable evaluations are associated with stronger behavioral intentions.

Hypothesis 4 (ATT→ITP): Attitude was significantly associated with intention to pay (β = 0.756, SE = 0.131, t = 5.758, p < 0.001). This result suggests that favorable attitudes toward safety investment are associated with greater willingness to allocate financial resources, consistent with TAM-based predictions for technology adoption contexts.

Additional TAM Paths (Exploratory Analysis). Several additional paths were tested to examine TAM’s mediational structure. Consistent with TAM’s theoretical framework, neither perceived effectiveness (PE→ITP: β = −0.004, p = 0.961) nor perceived safety responsibility (PSR→ITP: β = 0.040, p = 0.404) demonstrated significant direct associations with intention to pay, suggesting that attitude may serve as a complete mediator. Similarly, behavioral intention was not significantly associated with intention to pay (BI→ITP: β = 0.040, p = 0.404), suggesting that attitude’s direct association with ITP was more prominent than any indirect pathway through behavioral intention. Notably, perceived safety responsibility was significantly associated with perceived effectiveness (PSR→PE: β = 0.510, p < 0.001), suggesting that responsibility beliefs may be related to efficacy perceptions.

4.3.4 Effects of organizational factors

Analysis of organizational variables revealed significant relationships supporting five of six hypotheses in this category.

Hypothesis 5 (SIZE→PSR): Firm size significantly influenced perceived safety responsibility (β = 0.324, SE = 0.052, t = 6.189, p < 0.001). Larger firms demonstrated greater safety consciousness, consistent with organizational theory suggesting that larger organizations possess more formalized safety management systems and stronger reputation concerns.

Hypothesis 6 (SIZE→PE): Firm size did not exert a statistically significant effect on perceived effectiveness (β = 0.090, SE = 0.055, t = 1.643, p = 0.100). Although the coefficient was positive, it did not reach the conventional 0.05 significance threshold. This finding indicates that organizational scale, in isolation, is insufficient to meaningfully shape operators’ evaluations of safety technology effectiveness. The result challenges assumptions that larger firms inherently perceive greater technological benefits due to superior resources or exposure. Instead, perceived effectiveness appears to be more strongly grounded in demonstrable performance outcomes and contextual operational fit rather than structural characteristics such as firm size alone.

Hypothesis 7 (LAW→BI): The perceived influence of legal regulations did not exert a statistically significant effect on behavioral intention (β = 0.068, SE = 0.039, t = 1.768, p = 0.077). Although the coefficient was positive, it did not reach the conventional 0.05 significance threshold. This finding suggests that regulatory pressure, when perceived at the cognitive level, is insufficient on its own to activate behavioral intention directly. The result nuances institutional and regulatory perspectives within technology adoption research by indicating that compliance-based motivations may not automatically translate into intentional commitment, particularly in capital-intensive safety investment contexts. Instead, regulatory influence may operate indirectly—potentially through perceived effectiveness, risk reduction expectations, or economic rationality—rather than functioning as a direct antecedent of behavioral intention.

Hypothesis 8 (LAW→PE): Perceived influence of legal regulations significantly influenced perceived effectiveness (β = 0.175, SE = 0.050, t = 3.470, p < 0.001). Legal requirements enhanced operators’ confidence in technology effectiveness, suggesting that regulatory signals serve as institutional legitimization.

The mediation results suggest a sequential pattern of associations linking organizational and perceptual constructs to safety investment intentions. Perceived effectiveness (PE) showed a substantial indirect association with intention to pay (ITP) via attitudes (ATT) (PE→ATT→ITP; β = 0.554, p < 0.001), consistent with ATT serving as a proximal mediating mechanism. Perceived safety responsibility (PSR) was associated with ITP via a serial pathway (PSR→PE→ATT→ITP; β = 0.283, p < 0.001), suggesting that responsibility-related beliefs may be associated with effectiveness perceptions which in turn relate to evaluative attitudes. Additionally, firm size (SIZE) showed a significant higher-order indirect association across the full chain (SIZE→PSR→PE→ATT→ITP; β = 0.092, p < 0.001), suggesting that organisational scale may be associated with belief structures prior to attitude formation. These results are broadly consistent with a hierarchical framework integrating structural, belief-based, and attitudinal factors in explaining investment intentions.

5 Discussion

5.1 The central role of attitude in safety investment decisions

The finding that attitude is the sole direct predictor of intention to pay is broadly consistent with TAM’s theoretical structure and provides partial empirical support for the proposed extended TAM framework in the context of safety technology investment decisions among commercial operators. This result aligns with the Technology Acceptance Model, which positions attitude as the central mediating construct between belief variables (perceived usefulness/effectiveness) and behavioral intention (Davis, 1989; Venkatesh and Davis, 2000). The magnitude of the direct effect from attitude to intention to pay (β = 0.756) and to behavioral intention (β = 0.946) underscores that favorable evaluations of safety investment—encompassing beliefs about its wisdom, benefits, and value—are the critical psychological determinant of investment intentions.

The empirical support for the four core TAM hypotheses (Hypothesis 1–Hypothesis 4) provides evidence consistent with the model’s applicability in the context of safety technology investment. Perceived effectiveness showed the strongest association with attitude formation (Hypothesis 1: β = 0.733), suggesting that instrumental evaluations of safety benefits are associated with attitudinal responses. In contrast, perceived safety responsibility showed a positive but non-significant association with attitude (Hypothesis 2: β = 0.025), suggesting that normative considerations may play a supplementary role in shaping evaluative judgments.

Attitude, in turn, demonstrated substantial direct effects on both behavioral intention (Hypothesis 3: β = 0.946) and intention to pay (Hypothesis 4: β = 0.756), underscoring its role as the central proximal determinant of both adoption intention and financial commitment. Overall, the findings are consistent with TAM’s theoretical structure, wherein belief-based constructs shape attitudes that subsequently drive technology adoption intentions. The measurement model results provide additional insight into the structure of operators’ attitudes. All three attitude indicators demonstrated strong loadings (0.805–0.829), reflecting beliefs that safety investment is “wise,” “beneficial for reducing risk,” and “worthwhile for accident prevention.” This multifaceted attitude structure suggests that favorable investment intentions arise from cognitive evaluations spanning instrumental benefits (risk reduction), financial prudence (wise investment), and outcome expectations (accident prevention). The consistency of these findings with Yao et al. (2023), who demonstrated that safety-related attitudes mediate the relationship between safety perceptions and technology acceptance, further validates the centrality of attitude formation in technology adoption decisions.

5.2 Attitude as a critical mediator

The results are broadly consistent with attitude serving as a complete mediator in the Extended TAM framework. Both perceived effectiveness (Hypothesis 1) and perceived safety responsibility (Hypothesis 2) showed significant associations with attitude, but neither showed significant direct associations with intention to pay (PE→ITP: β = −0.004, p = 0.961; PSR→ITP: β = 0.040, p = 0.404). This pattern is consistent with TAM’s mediational structure, wherein belief variables are associated with behavioral outcomes primarily through attitude formation.

The mediation pathway from perceived effectiveness to ITP through attitude (PE→ATT→ITP) suggests a cognitively sequential process: operators first evaluate the efficacy of safety technologies in reducing accidents and improving control, and these efficacy beliefs subsequently shape favorable attitudes that drive investment intentions. This finding extends Davis (1989) and Venkatesh and Davis (2000), who demonstrated that perceived usefulness influences technology adoption through attitudinal mechanisms, by showing that efficacy perceptions operate through the same process in investment contexts.

Similarly, the mediated pathway from perceived safety responsibility to ITP (PSR→ATT→ITP) suggests that feelings of responsibility for road safety may not be directly associated with investment intentions. Instead, safety responsibility appears to be associated with how operators evaluate technology investments, which in turn relates to financial commitment intentions. Notably, perceived safety responsibility was also significantly associated with perceived effectiveness (PSR→PE: β = 0.510, p < 0.001), suggesting that responsibility beliefs may be related to efficacy perceptions—operators who report higher safety responsibility also tend to perceive technologies as more effective. This sequential association (PSR→PE→ATT→ITP) represents a theoretically meaningful pattern, consistent with the view that moral and social beliefs are associated with cognitive evaluations, which in turn relate to attitudes and intentions.

5.3 Organizational influences

The support for five of six demographic and organizational hypotheses (Hypothesis 5–Hypothesis 8) provides insights into contextual factors associated with safety technology adoption decisions. These findings are consistent with TAM being extendable to account for how organizational characteristics and the regulatory environment relate to the technology acceptance process.

Organizational scale significantly predicted perceived safety responsibility (Hypothesis 5: β = 0.324, p < 0.001) but not perceived effectiveness (Hypothesis 6: β = 0.090, p = 0.100), indicating that larger firms possess stronger normative orientations, whereas instrumental appraisals remain size-invariant. This aligns with research suggesting that larger enterprises institutionalize formal safety governance to manage heightened stakeholder pressures (Medina-Flintsch et al., 2017; Subhan et al., 2023). Furthermore, the significant indirect effect on intention to pay via the sequential pathway (SIZE→PSR→PE→ATT→ITP) demonstrates that organizational scale shapes investment intentions through proximal belief-based and attitudinal mechanisms rather than direct structural capacity alone.

Perceived legal influence significantly affected perceived effectiveness (Hypothesis 8: β = 0.175, p < 0.001) but showed only marginal effects on behavioral intention (Hypothesis 7: β = 0.068, p = 0.077). This suggests that regulatory frameworks influence adoption through cognitive belief reinforcement rather than direct behavioral compulsion. Aligned with institutional theory (DiMaggio and Powell, 2000), regulatory signals enhance technological legitimacy, shaping evaluative judgments before adoption. The larger effect size for perceived effectiveness indicates that regulation serves primarily as a mechanism for belief reinforcement rather than a direct compliance-driven motivator.

5.4 Insights into the TAM structure in safety technology investment

The pattern of nonsignificant direct associations from belief variables to intention to pay is broadly consistent with TAM’s mediational structure. Three key findings support this interpretation.

First, neither perceived effectiveness (PE→ITP: β = −0.004, p = 0.961) nor perceived safety responsibility (PSR→ITP: β = 0.040, p = 0.404) demonstrated significant direct associations with intention to pay, suggesting that attitude may serve as a complete mediator between beliefs and investment intentions. This pattern is consistent with TAM’s theoretical proposition that perceived usefulness is associated with behavioral outcomes primarily through attitude formation (Davis, 1989; Venkatesh and Davis, 2000). The complete mediation indicates that simply believing technology is effective or feeling responsible for safety is insufficient to drive investment—these beliefs must first shape favorable attitudes toward the specific investment decision.

Second, behavioral intention did not significantly predict intention to pay (BI→ITP: β = 0.121, p = 0.228), despite attitude strongly predicting both behavioral intention (ATT→BI: β = 0.946, p < 0.001) and intention to pay (ATT→ITP: β = 0.756, p < 0.001). This finding suggests that in technology investment contexts, attitude’s direct influence on financial commitment decisions is more proximal and influential than any indirect pathway through general behavioral intentions. This diverges from Subhan et al. (2021), who found direct effects using TPB, and highlights a theoretical advantage of TAM’s mediational framework: attitude captures the specific evaluative judgment about the investment decision, whereas behavioral intention may reflect more general safety-related intentions that do not automatically translate into financial commitment.

Third, the significant relationship between perceived safety responsibility and perceived effectiveness (PSR→PE: β = 0.510, p < 0.001) demonstrates that domain-specific beliefs can influence TAM’s core constructs. Operators who feel responsible for road safety perceive technologies as more effective, suggesting that moral and social beliefs create a cognitive foundation that enhances efficacy perceptions. This finding supports extended TAM frameworks that incorporate domain-specific belief structures (Taylor and Todd, 1995), while maintaining TAM’s fundamental mediational architecture where attitude channels multiple belief influences into behavioral outcomes.

These findings are broadly consistent with TAM’s theoretical structure in safety technology investment contexts and provide partial empirical support for the proposed extended TAM framework. The present results suggest that, in investment-oriented decision contexts, alternative frameworks allowing for multiple direct effects from beliefs to behavioral outcomes may be less parsimonious than TAM’s mediational architecture, though this proposition requires further testing across additional samples and cultural settings. The complete mediation through attitude, combined with the nonsignificant BI→ITP path, underscores that favorable evaluations of the specific investment decision—rather than general behavioral intentions or unmediated beliefs—are associated with safety technology adoption intentions among commercial operators.

6 Policy implications

Consistent with TAM’s theoretical structure, the centrality of attitude in determining investment intentions suggests that policy interventions should prioritize attitude formation as the primary mechanism for promoting safety technology adoption. TAM-based interventions should focus on strengthening perceived effectiveness beliefs (analogous to perceived usefulness in original TAM) and cultivating positive attitudes toward investment. Based on the empirical findings, we propose six policy implementation.

• Evidence-Based Communication Campaigns. Given that perceived effectiveness strongly influences attitude formation (β = 0.733), policymakers should develop communication materials that present quantitative evidence of safety technology benefits. This includes statistics on accident reduction rates, economic savings from avoided incidents, reduced maintenance costs, and insurance premium reductions. Communications should feature case studies from successful adopters across all firm sizes, making benefits tangible and relatable.
• Best Practice Dissemination. Establishing platforms for successful adopters to share experiences can leverage social learning mechanisms. Role model operators who have invested in safety technology should be engaged to disseminate their experiences through industry networks, enhancing perceived effectiveness and demonstrating practical implementation pathways. Uzondu et al. (2022) identified public education campaigns as highly acceptable interventions in developing country contexts, supporting this approach.
• Economic Incentive Programs. While attitude is the proximal determinant of investment intentions, economic barriers may prevent favorable attitudes from translating into actual investments. Policy mechanisms should include: (a) tax deductions or credits for safety technology investments; (b) subsidized financing through low-interest loans dedicated to safety improvements; (c) insurance premium reductions for technology-equipped vehicles; (d) certification programs that enhance competitive advantages for compliant operators.
• Size-Differentiated Support Mechanisms. Given the significant influence of firm size on investment pathways, policies should be calibrated to organizational capacity. Small and medium enterprises (SMEs) require proportionally greater support due to resource constraints. Specific measures include: (a) pooled procurement programs that allow SMEs to achieve economies of scale; (b) cooperative technology-sharing arrangements; (c) graduated compliance timelines that allow smaller operators transition periods; (d) technology leasing options that reduce upfront capital requirements.
• Regulatory Framework Development. The significant influence of legal regulations on perceived effectiveness and behavioral intention supports the implementation of mandatory safety technology standards. However, regulatory approaches should be phased, beginning with larger operators who demonstrate greater readiness, while providing transition support for smaller firms. Continuous policy evaluation should monitor compliance rates, adoption barriers, and safety outcomes.
• Industry-Government-Academia Partnerships. Establishing collaborative platforms involving government agencies, industry associations, academic institutions, and technology providers can facilitate knowledge transfer, capacity building, and continuous innovation. These partnerships should develop training curricula addressing both technical operation and business case development for safety investments.

7 Limitations and future research directions

Several limitations should be acknowledged when interpreting these findings. First, in common with the majority of SEM-based TAM research, this study employed a cross-sectional design, which precludes definitive empirical causal conclusions. The directional structure of the proposed model is theoretically grounded in TAM’s well-validated framework, and the reported relationships should be interpreted as associations consistent with the proposed theoretical model rather than as conclusive causal evidence. Longitudinal designs would provide stronger temporal evidence for the proposed relationships, and future research should employ panel designs to examine how attitude changes influence subsequent investment behaviors over time. Second, the reliance on self-reported intention measures introduces potential social desirability bias. Respondents may overstate their intention to pay due to perceived social expectations. Future studies should incorporate behavioral measures, such as actual technology adoption or investment records, to validate the intention-behavior relationship. Third, purposive sampling was employed to access decision-relevant respondents, which limits the representativeness of the sample and the generalizability of the findings to the broader population of truck operators in Thailand. Future research employing probability-based sampling frames would allow more confident population-level inference. Fourth, the study focused on Thai truck operators, potentially limiting generalizability to other countries or transportation contexts. Cross-national comparative studies would illuminate how cultural, regulatory, and economic factors moderate the identified relationships. Extension to other commercial vehicle sectors (e.g., buses, delivery vehicles) would test boundary conditions of the theoretical model. Fifth, although Harman’s single-factor test did not indicate pervasive common method bias, this remains a residual limitation of single-source self-report designs. Future studies should employ procedural controls such as temporal separation of predictor and outcome measurement to further mitigate this concern. Sixth, the model did not incorporate economic constraints as explicit constructs. Future research should integrate cost-benefit perceptions, financial capacity, and return-on-investment expectations to provide a more comprehensive understanding of investment decision-making. Additionally, the integration of willingness-to-pay elicitation methods could quantify the monetary values operators assign to safety improvements. Finally, future research should examine policy acceptance among diverse stakeholders, including drivers, regulatory authorities, and technology providers, to understand facilitators and barriers to policy implementation. Quasi-experimental designs evaluating specific interventions would provide evidence for policy effectiveness. Quasi-experimental designs evaluating specific interventions would provide evidence for policy effectiveness.

8 Conclusion

This study provides empirical evidence on the determinants of truck operators’ intention to pay for safety technology in Thailand, contributing both theoretical insights and practical policy recommendations. The structural equation modeling analysis, based on responses from 345 truck operators, provides partial empirical support for an Extended Technology Acceptance Model in which attitude toward safety investment emerges as the sole direct and dominant predictor of intention to pay. The results suggest that this central mediating role of attitude is associated with the influences of perceived effectiveness (perceived usefulness in TAM), perceived safety responsibility (PSR) (domain-specific extension), and firm size on investment intentions, consistent with TAM’s applicability to safety technology investment contexts. The findings underscore that policies aimed at promoting safety technology adoption should prioritize attitude formation through TAM-informed interventions. Rather than merely providing information about technology effectiveness (perceived usefulness) or appealing to operators’ sense of responsibility, policies must transform these beliefs into favorable attitudes through evidence-based communication that emphasizes concrete benefits and addresses implementation concerns. Effective interventions must translate efficacy beliefs and responsibility feelings into favorable investment attitudes through evidence-based communication that emphasizes economic returns, risk reduction benefits, and practical implementation pathways. The significant influence of firm size highlights the need for differentiated support mechanisms that address the resource constraints faced by smaller operators. The integration of regulatory frameworks with incentive-based approaches offers the most promising pathway for accelerating safety technology adoption. By simultaneously enhancing perceived effectiveness through regulatory signals and reducing investment barriers through economic incentives, policymakers can create conditions that support the translation of favorable attitudes into actual safety investments—ultimately contributing to the reduction of road traffic accidents and the achievement of Thailand’s road safety targets.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors used generative AI tools (Claude, Anthropic) during manuscript preparation to assist with literature organization and improving language clarity. All research design, data collection, statistical analyses, and interpretation were conducted independently by the authors. All AI-generated content was critically reviewed and verified for accuracy. The authors remain fully responsible for all aspects of the manuscript content.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Nathaniel Olatunde, Durban University of Technology, South Africa

Reviewed by: Janet Choo Seen Yee, University of Science Malaysia (USM), Malaysia

Juan L. Elorduy, University of Valladolid, Spain