The Influence of Personality Traits on Safety Behavior in Construction: The Role of Psychological–Cognitive Mediators

Abstract

Past research has predominantly focused on personality traits and psychological–cognitive factors as isolated predictors of safety behavior, while their interactive effects in shaping safety behavior remain underexplored. The gap constrains mechanistic understanding of safety behavior and limits the effectiveness of individualized interventions. Therefore, this study developed a theoretical framework linking personality traits, psychological–cognitive mediators (safety awareness, safety attitude, safety motivation, subjective norm, and perceived behavioral control) and safety behavior (safety compliance and safety participation). Quantitative data were collected from 431 frontline construction workers and managers using paper-based questionnaires. Structural equation modeling was used to test direct and indirect relationships among variables. The results reveal differentiated psychological–cognitive pathways through which personality traits shape safety behavior. Extraversion suppressed safety compliance through all psychological–cognitive factors except perceived behavioral control, and diminished safety participation via safety attitude and safety motivation. Agreeableness enhanced safety compliance through all psychological–cognitive factors except perceived behavioral control, whereas conscientiousness promoted safety compliance through all mediators. Agreeableness and conscientiousness strengthened safety participation via all mediators except safety awareness. Openness facilitated safety compliance through safety awareness but simultaneously inhibited it through other psychological–cognitive factors, and reduced safety participation via all mediators except safety awareness. Neuroticism undermined safety compliance via safety attitude, safety motivation, and subjective norm, and suppressed safety participation through safety attitude and safety motivation. These findings underscore the critical mediating role of psychological–cognitive factors in personality–safety behavior linkages and offer implications for individualized safety management. Recommended strategies include integrating personality and psychological–cognitive assessments to optimize work allocation and team collaboration, employing immersive and contextualized training to stabilize safety behavior, and developing an artificial intelligence–enabled safety management framework centered on psychological–cognitive regulation.

1. Introduction

The construction industry plays a vital role in economic development and employment, yet its inherently temporary, dynamic, and fragmented nature exposes it to persistently high levels of risk [1,2,3,4]. Statistical evidence highlights this issue: between 2019 and 2023, the United Kingdom’s construction industry reported an average fatality rate of 1.65 per 100,000 workers and a non-fatal injury rate of approximately 2390 [5]. In contrast, during 2008–2018, China’s annual average fatality and injury rates reached 7.03 and 3320 per 100,000 workers, respectively, underscoring a more severe safety situation [6]. In the United States, indirect expenditures arising from accidents, including work stoppages and penalties, can account for up to 10% of labor expenses [7]. In addition to causing severe human and economic losses, construction accidents also exert substantial social impacts. In particular, income loss and disparities in compensation following such incidents often result in persistent financial insecurity, delayed recovery, and reduced re-employment opportunities, thereby contributing to broader socioeconomic inequalities [8,9]. Consequently, reducing accident rates and enhancing the intrinsic safety of construction operations continue to represent critical challenges in construction safety management research [10,11].

Research and practice in construction safety management have widely explored both technologies such as real-time location systems [12], intelligent monitoring systems [13], virtual reality [14], and managerial strategies including refined safety standards [15], incentive–penalty schemes [16], and enhanced safety communication [17]. Although these measures have achieved certain progress in risk control, significant limitations remain. On one hand, current technologies mainly focus on post-incident identification or skill training, making it difficult to effectively prevent risks before they evolve into actual injuries [12]. On the other hand, managerial measures typically assume worker homogeneity and overlook individual differences, resulting in limited or even ineffective safety guidance for certain workers [18]. This suggests that relying solely on technological advances and institutional improvements is insufficient to fundamentally reduce accident occurrence. Enabling truly proactive safety intervention requires shifting the research perspective toward workers’ individual differences. These differences are reflected not only in workers’ psychological and physiological characteristics but especially in the ways these characteristics influence the acquisition, processing, and decision-making of risk information [18,19]. Consequently, workers exposed to the same work environment may exhibit heterogeneous safety behavior patterns [18]. At the essential level, personality traits, shaped by both genetic factors and long-term social experiences, constitute the core psychological basis of these differences [19]. Building on this foundation, proximal psychological mechanisms formed through cognitive processing, such as safety motivation, subjective norm, and perceived behavioral control, directly drive safety behavior [20]. Therefore, developing proactive interventions grounded in personality traits and psychological cognitive mechanisms can facilitate more targeted and effective safety management strategies. For example, by considering personality traits, proactive interventions can be tailored to workers with lower psychological–cognitive preparedness, thereby improving their safety behavior.

Given the pivotal role of personality traits and psychological–cognitive processes in shaping safety behavior, this study further investigates how they interact and influence safety outcomes in construction contexts. Personality traits reflect individuals’ enduring patterns of cognition, emotion, and behavior, serving as a deep psychological foundation for cross-situational behavioral consistency [21]. Specifically, one of the most widely used frameworks is the Big Five personality model, a five-factor structure that summarizes individual differences in personality at the broadest level, demonstrating high cross-cultural consistency and structural stability [22]. Its five dimensions—extraversion, agreeableness, conscientiousness, neuroticism, and openness—have been extensively applied in construction safety management research [23,24]. Psychological–cognitive processes represent how individuals perceive and process information and make decisions in construction contexts, acting as the proximal mechanism driving safety behavior [25]. From the perspective of personality trait dimensions, their stable characteristics shape both risk perception and behavioral choices [20,26,27]. Workers high in conscientiousness tend to exhibit greater responsibility and self-discipline, facilitating hazard recognition and adherence to safety regulations [23,28]. Conversely, those high in extraversion are more prone to sensation-seeking and impulsive decision-making, increasing the likelihood of unsafe behavior [23,29]. In addition to personality traits, psychological–cognitive factors such as safety awareness, attitudes, motivation, subjective norm, and perceived behavioral control collectively shape workers’ risk perception, behavioral regulation, and translation of safety intentions into practice [20]. Compared with personality traits, psychological–cognitive factors exhibit greater plasticity and potential for intervention, offering important theoretical and practical implications for individualized safety management.

Previous studies have investigated how personality traits and certain psychological–cognitive factors influence construction workers’ safety behavior. For instance, Gao et al. (2020) [23] found conscientiousness and agreeableness enhance safe behavior, whereas extraversion and neuroticism increase the likelihood of unsafe behavior. Research on openness has yielded inconsistent findings. Hasanzadeh et al. (2019) [28] reported that individuals high in openness allocate attention more effectively and are better able to identify potential fall hazards. In contrast, Zhang et al. (2020) [30] suggested that highly open workers are more likely to exhibit unsafe behavioral intentions when encountering unfamiliar environments. Research on psychological–cognitive factors suggests that safety awareness initiates risk information processing by determining whether individuals can identify potential hazards [31]. Safety attitudes establish normative orientations toward standardized operations [32], while internalized attitudes foster safety motivation, thereby transforming it into an intrinsic driver of behavior [33]. Subjective norm shapes behavioral choices through group climate and social expectations [34], and perceived behavioral control reflects self-efficacy under the constraints of skills, resources, and environmental conditions [35]. Moreover, Doerr [36] reported that safety motivation mediates the relationship between most personality traits (except extraversion) and safety behavior, suggesting that personality can indirectly affect behavior via psychological–cognitive mechanisms. Although these studies provide an important foundation for understanding the relationships between personality traits, psychological–cognitive factors, and safety behavior, several gaps remain. First, previous research has often examined the direct effects of personality traits or psychological factors on safety behavior but has largely overlooked the complex interplay between these variables and their combined impact. Second, inconsistent findings regarding the influence of certain personality dimensions on safety behavior leave the sources of these divergent results unresolved. These limitations collectively hinder the development of precise, individual-differences–based safety behavior interventions. Therefore, this study explores how personality traits influence safety behavior through psychological and cognitive factors, taking an integrated approach to consider both individual characteristics and their cognitive processing of risk information.

This study employs structural equation modeling to develop an integrated framework of personality traits–psychological cognition–safety behavior, aiming to systematically uncover how the five personality dimensions shape workers’ safety behavior through multiple psychological cognitive processes, including safety consciousness, safety attitude, safety motivation, subjective norm, and perceived behavioral control. Through this integrative perspective, this study offers a deeper understanding of the interactive mechanisms through which personality traits and psychological cognition jointly influence the formation of safety behavior, addressing existing gaps in theoretical integration and mechanism explanation. Moreover, by examining the mediating pathways of psychological cognitive factors, the study reveals the underlying reasons for the inconsistent effects of certain personality dimensions on safety behavior reported in prior research. Additionally, the proposed model provides a theoretical foundation for personalized and precision safety interventions on construction sites. It supports a shift in management strategies from external control to differentiated interventions based on workers’ intrinsic traits and psychological–cognitive mechanisms, and further encourages workers to proactively improve their own safety behavior.

2. Literature Review and Hypothesis Development

2.1. Safety Behavior

Safety behavior is critical to workers’ health and safety, with significant implications for public protection and environmental sustainability [37,38]. For example, failure to properly use fall-protection equipment can lead to fall-from-height accidents [39]; violations of fire-safety regulations may trigger fires that cause severe casualties, substantial property losses, and potentially further environmental pollution [40]. As a central construct in construction safety management, safety behavior enhances safety performance not only by directly reducing unsafe incidents but also by fostering a positive safety climate that strengthens broader organizational safety outcomes. For instance, correct use of personal protective equipment reduces the likelihood of injuries or accidents [39], whereas promptly identifying and reporting unsafe conditions reinforces collective safety awareness, encouraging additional proactive safety actions and supporting the development and maintenance of a positive safety culture [41]. Based on work performance theory, Griffin and Neal [42] distinguish two dimensions of safety behavior: safety compliance and safety participation. The former corresponds to task performance and involves adhering to safety procedures and correctly using protective equipment behavior that ensure immediate operational safety [33]. The latter corresponds to contextual performance and includes self-initiated acts such as identifying and reporting hazards or suggesting safety improvements, which promote continuous improvement and strengthen team safety awareness [33]. In high-risk construction contexts, compliance provides the baseline for risk control, whereas participation drives the development and resilience of organizational safety culture [43]. Therefore, uncovering the mechanisms underlying workers’ safety behavior is essential not only for reducing accidents and injuries but also for advancing the sustainable development.

2.2. Personality Trait

Personality traits are commonly defined as relatively stable individual differences in patterns of thought, emotion, and behavior, serving as a core psychological basis for explaining behavioral heterogeneity [21,44]. Understanding personality traits facilitates the prediction of how individuals make decisions and behave under complex and uncertain conditions. This perspective is especially critical in the construction industry, where workers’ risk judgment and immediate behavior directly influence safety performance and exposure to accidents. Extensive psychological research has established the Big Five personality model as a robust framework for describing personality structure, encompassing extraversion, agreeableness, conscientiousness, neuroticism, and openness [45]. Extraversion reflects individuals’ level of social activity, energy, and need for external stimulation, with typical characteristics including confidence, sociability, and a tendency toward sensation-seeking; agreeableness captures individuals’ cooperativeness, trust, and prosocial tendencies in interpersonal interactions, typically manifesting as friendliness, cooperativeness, and a focus on maintaining harmonious relationships; conscientiousness describes individuals’ organization, responsibility, and self-discipline in goal-directed behavior, with highly conscientious individuals often being well-planned, reliable, and persistent in completing tasks; neuroticism indicates emotional stability and susceptibility to negative emotions, commonly characterized by anxiety, tension, and mood variability; openness refers to individuals’ receptiveness to new experiences, creativity, art, imagination, and novel ideas, commonly manifested in rich imagination, enthusiasm for new experiences, and open-minded thinking [21,22,24,46]. A systematic understanding of these dimensions helps explain behavioral differences among workers when facing risk. It also provides a theoretical foundation for developing personality-informed safety management strategies and behavioral interventions.

Regarding the influence of personality traits on safety behavior, existing research presents both consistencies and inconsistencies. Conscientiousness is widely recognized as the most robust predictor of safety compliance and participation, with higher levels consistently associated with stricter adherence to procedures and more proactive protective behavior [23,24]. Agreeableness, by enhancing individuals’ altruistic and cooperative tendencies, strengthens identification with safety norms and promotes positive safety behavior [23,27]. In contrast, while extraversion confers social advantages, its link to sensation-seeking may reduce risk sensitivity in high-risk work environments, exerting negative effects on safety compliance and participation [23,28]. Neuroticism, reflecting emotional instability and anxiety proneness, is associated with lower risk perception and weaker self-regulation, thereby increasing unsafe behavior [23,47]. The effect of openness appears more complex: its creativity and exploration facilitate flexible problem-solving but may also heighten deviation from established safety rules in uncertain or hazardous contexts [28,30]. Meta-analytic studies reveal a positive relationship between openness and accident rates, suggesting a weakening effect on safety behavior [30,48]. Based on these findings, this study proposes the following hypotheses on personality traits and safety behavior:

Hypothesis 1 (H1).

Extraversion, neuroticism, and openness exert negative effects on individuals’ safety compliance (H1a) and safety participation (H1b), whereas agreeableness and conscientiousness exert positive effects.

2.3. Psychological and Cognitive Factors

Beyond personality differences, the formation of workers’ safety behavior is primarily shaped by multiple psychological and cognitive factors [20,35,49]. Based on the cognitive model of unsafe behavior and the Theory of Planned Behavior, workers undergo a series of affective, motivational, and social-cognitive processes during risk information processing and behavioral execution. Within this process, safety awareness, safety attitude, safety motivation, subjective norm, and perceived behavioral control are widely recognized as core psychological variables of safety behavior [49,50,51,52]. Safety awareness reflects an individual’s sensitivity to hazards and understanding of safety rules; a lack of safety awareness often results in poor risk recognition or improper protective actions [53,54]. Safety attitude denotes one’s cognitive judgment, emotional response, and behavioral orientation toward safety, forming the psychological basis for compliance with safety procedures [32]. Safety motivation represents the intrinsic drive to engage in safe behavior and avoid risk, serving as the link between safety cognition and behavioral execution [33,51,55]. Subjective norm captures how individuals internalize safety expectations from others or social groups, guiding behavior through social climate and peer pressure [34,56,57]. Finally, perceived behavioral control reflects self-efficacy under constraints of ability, resources, and the environment, determining individuals’ confidence in enacting safe behaviors [25,58].

These psychological and cognitive variables interact within a multidimensional framework encompassing perception, emotion, motivation, social cognition, and behavioral execution, collectively contributing to workers’ safety behavior. In high-risk construction contexts, safety awareness initiates risk information processing by heightening hazard awareness and recognition, providing a basis for safety judgments and behavioral decisions [53,54]. Safety attitude facilitates the transformation of risk perception into positive behavioral tendencies through emotional identification with safety values and norms, strengthening individual responsibility and compliance [59,60]. As an internal driving force, safety motivation sustains the consistency of safe behavior, promoting the implementation of safety practices [33,51,55]. Subjective norm, derived from social expectations and group consensus, embeds individual behavior within the broader safety culture, reinforcing compliance and collective adherence [56,57]. Perceived behavioral control enhances individuals’ self-efficacy in complex tasks and unexpected situations, increasing willingness and persistence in performing safe behavior [25,58]. These five factors form an interdependent dynamic system that integrates perception, emotion, motivation, social cognition, and behavioral execution, jointly promoting workers’ safety compliance and proactive safety participation. Accordingly, the following hypotheses are proposed regarding the relationships between psychological–cognitive factors and safety behavior:

Hypothesis 2 (H2).

Safety awareness exerts positive effects on both individuals’ safety compliance (H2a) and safety participation (H2b).

Hypothesis 3 (H3).

Safety attitude exerts positive effects on both individuals’ safety compliance (H3a) and safety participation (H3b).

Hypothesis 4 (H4).

Safety motivation exerts positive effects on both individuals’ safety compliance (H4a) and safety participation (H4b).

Hypothesis 5 (H5).

Subjective norm exerts positive effects on both individuals’ safety compliance (H5a) and safety participation (H5b).

Hypothesis 6 (H6).

Perceived behavioral control exerts positive effects on both individuals’ safety compliance (H6a) and safety participation (H6b).

Based on the role of psychological–cognitive factors in safety management, it is essential to consider how personality traits shape these variables. Personality traits, by shaping enduring characteristics, influence core psychological constructs, including safety awareness, attitudes, motivation, subjective norm, and perceived behavioral control. Conscientious workers, marked by responsibility and self-discipline, exhibit stronger safety awareness, more positive attitudes, and steadier motivation, along with greater normative identification and perceived control [27,29,36,61,62,63]. Agreeableness, characterized by cooperation and altruism, enhances risk sensitivity, internalization of safety values, and adherence to group norms, thereby strengthening safety motivation and perceived control [23,28,63,64,65]. Although extraversion fosters confidence and social advantages, its impulsivity may reduce risk sensitivity, increasing conformity and the tendency to overestimate control in negative group climates [28,64,65,66,67]. Neuroticism entails emotional instability and anxiety, leading to lower safety awareness, reduced perceived control, with motivation relying more on external pressures [36,68,69,70]. While openness fosters innovation, excessive skepticism toward rules may undermine safety awareness and attitudes, reducing norm adherence and control stability [36,65,70,71]. The Big Five traits exert distinct effects on safety-related cognitions, providing a theoretical basis for individual variation in safety behavior. Based on this, this study proposes:

Hypotheses 7–11 (H7–H11).

Extraversion, neuroticism, and openness exert negative effects on safety awareness (H7), safety attitudes (H8), safety motivation (H9), subjective norm (H10), and perceived behavioral control (H11), whereas agreeableness and conscientiousness exert positive effects.

Existing research shows that personality traits influence safety behavior both directly and indirectly through psychological–cognitive mechanisms. Prior studies indicate that individuals with higher conscientiousness exhibit pronounced responsibility and self-discipline [27]. This disposition increases the likelihood of adopting positive safety attitudes, thereby enhancing both safety compliance and participation [61,62]. Intervention studies further demonstrate that management strategies tailored to personality differences can effectively enhance workers’ safety behavior by reinforcing subjective norm and safety motivation [57]. In addition, Doerr et al. [36] found that safety motivation functions as a general mediator linking all personality dimensions and safety behavior. These findings highlight psychological–cognitive factors as a critical bridge linking stable personality traits with safety behavior. Therefore, the study proposes the following hypotheses regarding mediating effects:

Hypothesis 12 (H12).

Extraversion, neuroticism, and openness exert negative indirect effects on individuals’ safety compliance and safety participation via safety awareness, whereas agreeableness and conscientiousness exert positive effects through safety awareness.

Hypothesis 13 (H13).

Extraversion, neuroticism, and openness exert negative indirect effects on individuals’ safety compliance and safety participation via safety attitude, whereas agreeableness and conscientiousness exert positive effects through safety attitude.

Hypothesis 14 (H14).

Extraversion, neuroticism, and openness exert negative indirect effects on individuals’ safety compliance and safety participation via safety motivation, whereas agreeableness and conscientiousness exert positive effects through safety motivation.

Hypothesis 15 (H15).

Extraversion, neuroticism, and openness exert negative indirect effects on individuals’ safety compliance and safety participation via subjective norm, whereas agreeableness and conscientiousness exert positive effects through subjective norm.

Hypothesis 16 (H16).

Extraversion, neuroticism, and openness exert negative indirect effects on individuals’ safety compliance and safety participation via perceived behavioral control, whereas agreeableness and conscientiousness exert positive effects through perceived behavioral control.

Building on prior research, this study develops an integrated model (see Figure 1) to illustrate the mechanisms linking personality traits, psychological–cognitive factors, and safety behavior. This framework advances the theoretical understanding of individual differences in construction workers’ safety behavior. It also provides a conceptual foundation and practical guidance for developing safety interventions targeting the interplay between personality and psychological–cognitive mechanisms.

3. Research Methods

3.1. Measurement

Given that the theoretical model developed in this study involves multiple latent variables that cannot be directly measured through observation or experimental methods, a questionnaire survey—characterized by its efficiency and operational feasibility—was adopted to collect data on construction workers’ personality traits, psychological–cognitive factors, and safety behavior at a single time [72]. This approach enables a cross-sectional examination of how personality traits influence safety behavior through psychological–cognitive mechanisms. Regarding the measurement instruments, the study adapted well-established scales while incorporating contextual features of construction-site work to enhance contextual relevance and measurement validity. For example, work-related descriptions were added to items in the personality trait scale, revising the original statement “I see myself as someone who is talkative” to “I tend to be talkative at work” to better align with the construction context. Additionally, to reduce social desirability bias and improve response quality, reverse-coded items were incorporated. For instance, the safety participation scale item “I report safety problems to my supervisor when I see safety problems” was modified to “I do not report safety issues that I observe to my supervisor”.

The questionnaire comprised four sections covering demographic information, personality traits, safety behavior, and psychological–cognitive factors. The first part collected respondents’ demographic characteristics, including age, gender, education, and work experience. The second part employed the 44-item Big Five Inventory, including 16 reverse-coded items, to assess individual differences in personality [23,73]. The third part evaluated safety behavior following Neal and Griffin [33] and Gao et al. (2020) [23], comprising eight items (three reverse-coded) to capture safety compliance and participation. The fourth part examined psychological–cognitive constructs such as safety awareness, safety attitude, safety motivation, subjective norm, and perceived behavioral control using 20 items adapted from prior studies and validated for the construction context. All items were rated on a five-point Likert scale (1 = strongly disagree, 5 = strongly agree), with higher scores indicating greater agreement with the item. Composite scores for personality traits, psychological–cognitive factors, and safety behavior were computed by aggregating responses to positively worded and reverse-coded items. Measurement scales for safety behavior and psychological–cognitive factors were presented in

Table 1. Measurement scales safety behavior and psychological–cognitive factors.

Variable	Item	Content	Reference
Safety compliance	SC1	I always use the required safety equipment when performing my tasks	[23,33]
	SC2	I sometimes skip safety procedures to complete tasks more quickly.
	SC3	I consistently follow safety procedures to ensure a safe working environment.
	SC4	I do not follow safety rules that I consider unnecessary
Safety participation	SP1	I actively promote safety programs at my workplace.	[23,33]
	SP2	I make additional efforts to improve workplace safety.
	SP3	I voluntarily undertake activities to improve workplace safety.
	SP4	I do not report safety issues that I observe to my supervisor.
Safety awareness	SA1	I remain alert to potential safety hazards and actively avoid them.	[25,74]
	SA2	I continue working even when I am uncertain about the safety of the task.
	SA3	I am cautious when I am unsure about the safety of the work
	SA4	I am clear about my safety responsibilities at work.
Safety attitude	ST1	Work accidents are primarily attributable to external factors rather than occurring spontaneously.	[60,75]
	ST2	Safety procedures and rules are useful for preventing accidents.
	ST3	Taking risks at work is an enjoyable experience
	ST4	Accidents at work are unavoidable, no matter what we do.
Safety motivation	SM1	I do not report the safety hazards I encounter at work to my supervisor	[75,76,77,78]
	SM2	I feel uncomfortable or guilty if I do not follow safety procedures at work
	SM3	I am willing to inform my colleagues about potential workplace hazards
	SM4	I always strive to work safely on site.
Subjective norm	SN1	In my team, productivity is prioritized over safety	[25,79,80]
	SN2	My leader would reprimand me for engaging in unsafe behavior.
	SN3	My family and friends prefer that I adhere to safe work practices
	SN4	My colleagues actively encourage me to follow safe work practices.
Perceived behavioral control	PBC1	I know how to use safety equipment required for my job.	[25,79,80,81]
	PBC2	My safety performance at work is beyond my control.
	PBC3	Even when taking risks, I can manage work conditions to ensure I work safely.
	PBC4	I have the necessary resources, knowledge, and skills to perform risky tasks safely at work.

.

3.2. Participants

Before the formal survey, seven frontline construction managers and workers were invited to evaluate the clarity and comprehensibility of the questionnaire. Based on their feedback, certain items were revised to enhance clarity and appropriateness. The formal survey was conducted across 20 ongoing highway construction projects in the Guangxi Zhuang Autonomous Region from March to May 2023, lasting approximately two months. All frontline managers are directly involved in daily construction tasks, both frontline workers and managers were included as respondents. Participants were invited to independently complete paper-based questionnaires in on-site offices, with an average completion time of 10 min. To minimize potential self-report bias, participants were informed that the survey was only for academic purposes and that all responses would remain anonymous and confidential. A total of 500 questionnaires were distributed. After removing responses with missing or invalid data, 431 valid questionnaires were retained, yielding an effective response rate of 86.2%. Participants’ demographic characteristics are summarized in

Table 2. Demographic characteristics of all participants.

Category	Item	Frequency	Percentage (%)
Age	<20	11	2.55
	21–30	109	25.29
	31–40	113	26.22
	41–50	99	22.97
	51–60	93	21.58
	>60	6	1.39
Gender	Male	392	90.95
	Female	39	9.05
Education level	Primary or below	74	17.17
	Junior high school	142	32.95
	High school	130	30.16
	College or above	85	19.72
Work experience (years)	<1	54	12.53
	1–3	96	22.27
	3–6	102	23.67
	6–10	112	25.99
	>10	67	15.55

. Among respondents, approximately 96% were aged 20–60, 91% were male, and 87% had more than one year of work experience, consistent with the typical composition of Chinese construction sites. Therefore, the collected data are considered broadly representative.

3.3. Data Analysis Methods

This study aims to examine the complex relationships among multiple latent variables, including personality traits, psychological–cognitive factors, and safety behavior. It further seeks to simultaneously test multiple independent and dependent variables, as well as their indirect effects. To address these objectives, structural equation modeling (SEM) was conducted using SPSS 26 and AMOS 26, integrating personality traits, psychological–cognitive factors, and safety behavior into a single model to systematically test both direct and indirect paths specified in the research hypotheses [82].

The analysis was carried out in two stages. In the first stage, data screening and measurement model assessment were performed. Data quality was initially evaluated using Cronbach’s alpha to assess the internal consistency and reliability of the scales measuring latent constructs [83], ensuring suitability for SEM analysis. Confirmatory factor analysis (CFA) was then conducted to assess construct validity and verify that observed indicators accurately reflected the latent constructs, with the measurement model demonstrating good fit in the construction site sample [84]. Model fit was evaluated using widely accepted indices, including χ²/df, CFI, TLI, RMSEA, and SRMR, all computed in AMOS 26.0 following standard procedures. Specifically, χ²/df measures the overall fit between the model and the observed data; CFI and TLI assess the improvement and adequacy relative to an independence model; RMSEA reflects approximate fit error; and SRMR indicates the average discrepancy between predicted and observed covariance matrices [85]. In the second stage, after confirming acceptable measurement model fit, the structural model was tested, and model fit was evaluated using the same indices (χ²/df, CFI, TLI, RMSEA, SRMR) with standard cutoff criteria applied [86]. Direct effects among latent variables were examined using p-values, whereas indirect effects were assessed via nonparametric bootstrapping. This approach does not assume normality and generates robust confidence intervals, allowing precise estimation of the mediating role of psychological–cognitive factors in the influence of personality traits on safety behavior [87].

4. Results

4.1. Measurement Model

All latent variables exhibited Cronbach’s alpha values exceeding the recommended threshold of 0.70 (

Table 3. Reliability and validity of the latent constructs.

Constructs	Item	SFL	Cronbach’s Alpha	CR	AVE
Extraversion (E)	W1	0.800	0.920	0.920	0.591
	W2	0.744
	W3	0.827
	W4	0.757
	W5	0.779
	W6	0.733
	W7	0.749
	W8	0.759
Agreeableness (A)	A1	0.814	0.931	0.931	0.602
	A2	0.784
	A3	0.786
	A4	0.753
	A5	0.785
	A6	0.817
	A7	0.704
	A8	0.796
	A9	0.737
Conscientiousness (C)	C1	0.804	0.920	0.921	0.563
	C2	0.750
	C3	0.772
	C4	0.748
	C5	0.726
	C6	0.741
	C7	0.772
	C8	0.748
	C9	0.689
Neuroticism (N)	N1	0.786	0.923	0.923	0.601
	N2	0.777
	N3	0.782
	N4	0.777
	N5	0.772
	N6	0.767
	N7	0.753
	N8	0.787
Openness (O)	O1	0.769	0.932	0.932	0.579
	O2	0.753
	O3	0.801
	O4	0.718
	O5	0.749
	O6	0.787
	O7	0.763
	O8	0.756
	O9	0.743
	O10	0.766
Safety awareness (SA)	SA1	0.797	0.876	0.876	0.638
	SA2	0.790
	SA3	0.792
	SA4	0.815
Safety attitude (ST)	ST1	0.719	0.812	0.812	0.520
	ST2	0.715
	ST3	0.704
	ST4	0.746
Safety motivation (SM)	SM1	0.755	0.820	0.821	0.534
	SM2	0.709
	SM3	0.703
	SM4	0.754
Subjective norm (SN)	SN1	0.747	0.827	0.828	0.546
	SN2	0.720
	SN3	0.725
	SN4	0.763
Perceived behavioral control (PBC)	PBC1	0.743	0.812	0.814	0.523
	PBC2	0.702
	PBC3	0.718
	PBC4	0.730
Safety compliance (SC)	SC1	0.766	0.863	0.865	0.615
	SC2	0.799
	SC3	0.778
	SC4	0.794
Safety participation (SP)	SP1	0.768	0.871	0.872	0.630
	SP2	0.798
	SP3	0.788
	SP4	0.820

), indicating a high level of internal consistency among the measurement items [88]. Convergent validity reflects the extent to which different measures of the same variable are correlated, whereas discriminant validity refers to the degree to which variables are distinct from one another [60]. Standardized factor loading (SFL), composite reliability (CR), and average variance extracted (AVE) were used to evaluate convergent validity. According to commonly accepted criteria, SFL and AVE should exceed 0.50 and CR should be greater than 0.70, ensuring that the indicators adequately represent their respective constructs [86,88]. Table 3 shows that all indicators met these thresholds, collectively demonstrating satisfactory convergent validity. Discriminant validity was examined using the Fornell–Larcker criterion, which requires that the square root of each latent construct’s AVE be greater than its correlations with other constructs. The square root for each construct was greater than its highest inter-construct correlation (

Table 4. Descriptive statistics, convergent validity, and construct interrelations.

Construct	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. E	23.360	8.235	0.769	-	-	-	-	-	-	-	-
2. A	28.234	9.331	−0.078	0.776	-	-	-	-	-	-	-
3. C	30.311	8.859	−0.145 **	0.066	0.751	-	-	-	-	-	-
4. N	23.313	8.420	0.031	0.052	−0.066	0.775	-	-	-	-	-
5. O	27.220	9.645	0.106 *	0.001	−0.092	−0.004	0.761	-	-	-	-
6. SA	13.039	3.423	−0.144 *	0.205 **	0.244 **	0.020	0.099 *	0.799	-	-	-
7. ST	13.360	3.194	−0.327 **	0.210 **	0.222 **	−0.125 **	−0.309 **	0.162 **	0.721	-	-
8. SM	13.383	3.298	−0.227 **	0.233 **	0.331 **	−0.148 **	−0.242 **	0.168 **	0.481 **	0.731
9. SN	12.703	3.205	−0.272 **	0.187 **	0.160 **	−0.098 *	−0.218 **	0.159 **	0.455 **	0.400 **	0.739
10. PBC	13.545	3.359	0.022	0.082	0.326 **	−0.082	−0.159 **	0.150 *	0.453 **	0.454 **	0.346 **	0.723
11. SC	13.979	3.599	−0.385 **	0.276 **	0.435 **	−0.135 **	−0.368 **	0.266 **	0.628 **	0.621 **	0.495 **	0.543 **	0.785
12. SP	13.647	3.810	−0.351 **	0.272 **	0.400 **	−0.078	−0.264 **	0.250 **	0.599 **	0.583 **	0.464 **	0.543 **	0.547 **	0.794

Note: Diagonal values (in bold) are the square roots of the corresponding AVEs. * p < 0.05; ** p < 0.01.

), confirming satisfactory discriminant validity across all constructs. The fit indices of the measurement model and their corresponding recommended thresholds are presented in

Table 5. Fit indices for the measurement model.

Fit Indices	Model	Recommended Values	Results
Chi-square to degrees of freedom ratio (χ2/df)	1.534	<5	Acceptable
Comparative fit index (CFI)	0.931	≥0.9	Acceptable
Tucker–Lewis index (TLI)	0.927	≥0.9	Acceptable
Root mean square error of approximation (RMSEA)	0.035	<0.08	Acceptable
Standardized root mean square residual (SRMR)	0.040	<0.08	Acceptable

. All indices met the relevant criteria (see Table 5), indicating good overall fit of the measurement model [86].

4.2. Structural Model

After confirming that the measurement model demonstrated satisfactory goodness of fit, reliability, and validity, the structural model was subsequently applied to evaluate the hypothesized relationships among the constructs. The model demonstrated a satisfactory overall fit, with all indices falling within the recommended ranges (χ²/df = 1.646, CFI = 0.916, TLI = 0.912, RMSEA = 0.039, SRMR = 0.066). These results suggest that the model appropriately reflects the theoretical relationships among the constructs [86].

During the hypothesis testing stage, the results of direct effects (

Table 6. Results of direct and indirect effects in hypothesis testing.

	Hypotheses	Direct Effect	p	Results		Hypotheses	Indirect Effect	p	95% CI
									Lower	Upper
H1a	E→SC	−0.218	0.008	Supported	H12a	E→SA→SC	−0.013	0.018	−0.033	−0.002
	A→SC	0.110	0.011	Supported		A→SA→SC	0.020	0.006	0.004	0.047
	C→SC	0.163	0.006	Supported		C→SA→SC	0.026	0.007	0.006	0.059
	N→SC	−0.043	0.207	Not supported		N→SA→SC	0.003	0.456	−0.007	0.019
	O→SC	−0.181	0.005	Supported		O→SA→SC	0.014	0.014	0.001	0.034
H1b	E→SP	−0.183	0.009	Supported	H12b	E→SA→SP	−0.007	0.209	−0.021	0.003
	A→SP	0.105	0.037	Supported		A→SA→SP	0.011	0.123	−0.003	0.044
	C→SP	0.118	0.014	Supported		C→SA→SP	0.015	0.157	−0.006	0.050
	N→SP	0.037	0.330	Not supported		N→SA→SP	0.002	0.270	−0.002	0.000
	O→SP	−0.033	0.598	Not supported		O→SA→SP	0.008	0.136	−0.002	0.028
H2a	SA→SC	0.098	0.009	Supported	H13a	E→ST→SC	−0.071	0.018	−0.121	−0.021
H2b	SA→SP	0.057	0.197	Not supported		A→ST→SC	0.052	0.004	0.023	0.103
H3a	ST→SC	0.223	0.018	Supported		C→ST→SC	0.042	0.006	0.016	0.080
H3b	ST→SP	0.268	0.002	Supported		N→ST→SC	−0.033	0.015	−0.074	−0.006
H4a	SM→SC	0.251	0.006	Supported		O→ST→SC	−0.075	0.014	−0.145	−0.029
H4b	SM→SP	0.247	0.012	Supported	H13b	E→ST→SP	−0.086	0.004	−0.148	−0.042
H5a	SN→SC	0.119	0.026	Supported		A→ST→SP	0.062	0.002	0.030	0.131
H5b	SN→SP	0.112	0.040	Supported		C→ST→SP	0.051	0.004	0.017	0.109
H6a	PBC→SC	0.304	0.008	Supported		N→ST→SP	−0.040	0.004	−0.095	−0.011
H6b	PBC→SP	0.331	0.013	Supported		O→ST→SP	−0.089	0.003	−0.170	−0.046
H7	E→SA	−0.130	0.017	Supported	H14a	E→SM→SC	−0.046	0.006	−0.096	−0.019
	A→SA	0.200	0.004	Supported		A→SM→SC	0.064	0.009	0.027	0.112
	C→SA	0.264	0.009	Supported		C→SM→SC	0.084	0.003	0.046	0.148
	N→SA	0.030	0.535	Not supported		N→SM→SC	−0.043	0.004	−0.095	−0.015
	O→SA	0.141	0.025	Not supported		O→SM→SC	−0.063	0.003	−0.145	−0.035
H8	E→ST	−0.320	0.015	Supported	H14b	E→SM→SP	−0.045	0.005	−0.096	−0.017
	A→ST	0.233	0.004	Supported		A→SM→SP	0.063	0.005	0.033	0.110
	C→ST	0.189	0.015	Supported		C→SM→SP	0.082	0.009	0.043	0.142
	N→ST	−0.148	0.005	Supported		N→SM→SP	−0.042	0.005	−0.089	−0.014
	O→ST	−0.334	0.014	Supported		O→SM→SP	−0.062	0.005	−0.127	−0.031
H9	E→SM	−0.183	0.015	Supported	H15a	E→SN→SC	−0.032	0.015	−0.066	−0.007
	A→SM	0.257	0.015	Supported		A→SN→SC	0.025	0.009	0.006	0.056
	C→SM	0.334	0.014	Supported		C→SN→SC	0.015	0.009	0.002	0.041
	N→SM	−0.170	0.005	Supported		N→SN→SC	−0.014	0.014	−0.067	−0.007
	O→SM	−0.253	0.016	Supported		O→SN→SC	−0.030	0.024	−0.074	−0.004
H10	E→SN	−0.266	0.018	Supported	H15b	E→SN→SP	−0.029	0.112	−0.146	−0.064
	A→SN	0.207	0.005	Supported		A→SN→SP	0.023	0.022	0.004	0.064
	C→SN	0.129	0.019	Supported		C→SN→SP	0.014	0.022	0.001	0.047
	N→SN	−0.118	0.008	Supported		N→SN→SP	−0.013	0.034	−0.039	0.000
	O→SN	−0.228	0.005	Supported		O→SN→SP	−0.025	0.025	−0.077	−0.003
H11	E→PBC	0.079	0.222	Not supported	H16a	E→PBC→SC	0.024	0.263	−0.013	0.061
	A→PBC	0.102	0.064	Not supported		A→PBC→SC	0.031	0.054	−0.002	0.064
	C→PBC	0.378	0.028	Supported		C→PBC→SC	0.115	0.012	0.051	0.191
	N→PBC	−0.090	0.134	Not supported		N→PBC→SC	−0.027	0.094	−0.068	0.002
	O→PBC	−0.183	0.012	Supported		O→PBC→SC	−0.056	0.005	−0.113	−0.022
-	-	-	-	-	H16b	E→PBC→SP	0.026	0.222	−0.013	0.074
-	-	-	-	-		A→PBC→SP	0.034	0.027	0.006	0.075
-	-	-	-	-		C→PBC→SP	0.125	0.016	0.063	0.176
-	-	-	-	-		N→PBC→SP	−0.030	0.104	−0.073	0.002
-	-	-	-	-		O→PBC→SP	−0.061	0.009	−0.117	−0.023

) revealed that most hypotheses were fully supported. Hypothesis H2b was not supported, while H1a, H1b, H7, and H11 received only partial support. Regarding the effects of personality traits on safety behavior, extraversion negatively affected both safety compliance and safety participation. In contrast, agreeableness and conscientiousness positively influenced both dimensions, while openness negatively affected only safety compliance. In the paths from personality traits to psychological–cognitive factors, all psychological–cognitive variables except perceived behavioral control were negatively associated with extraversion and positively with agreeableness. Conscientiousness showed consistent positive effects across all paths, while neuroticism negatively influenced only safety attitude, motivation, and subjective norm. Openness exerts a positive effect on safety awareness, but shows negative effects on all other psychological–cognitive factors. Regarding the influence of psychological–cognitive factors on safety behavior, safety awareness was positively related only to safety compliance, while all other factors positively affected both safety compliance and safety participation. The findings indicate that personality traits and psychological–cognitive factors directly influence workers’ safety behavior, supporting the theoretical hypotheses of the proposed multi-path model.

Mediation analysis results (Table 6) indicated that most indirect pathways were fully supported. Hypothesis H12b was not supported, while H12a, H15b, H16a, and H16b were partially supported. Specifically, extraversion indirectly inhibited safety compliance via all psychological–cognitive mediators except perceived behavioral control, and suppressed safety participation via safety attitude and safety motivation. Agreeableness facilitates safety compliance via all factors except perceived behavioral control, while conscientiousness indirectly promotes safety compliance through all psychological–cognitive factors. Both agreeableness and conscientiousness further enhance safety participation through psychological–cognitive factors other than safety awareness. Openness indirectly promotes safety compliance via safety awareness, but simultaneously suppresses safety compliance through safety attitude, safety motivation, subjective norm, and perceived behavioral control, and reduces safety participation via all cognitive factors except safety awareness. Neuroticism primarily suppresses safety participation via safety attitude and safety motivation, and further inhibits safety compliance through safety attitude, safety motivation, and subjective norm. These findings support the proposed multi-path model, highlighting the psychological mechanisms underlying the effects of personality on workers’ safety performance.

5. Discussion

5.1. Theoretical Contributions

Based on the results of structural equation modeling, this study systematically reveals the underlying mechanisms through which personality traits influence safety behavior via psychological–cognition pathways. The findings not only extend the theoretical framework linking personality traits to safety behavior but also deepen the understanding of the cognitive antecedents of safety performance. A comparison with previous studies further confirms the critical mediating role of psychological–cognitive factors in the pathway from personality traits to safety behavior. It also highlights that differences in contextual risk levels and personality measurement approaches may contribute to variations in research outcomes, thereby enriching the theoretical contributions of this study. In practice, the findings provide targeted insights for risk control and behavior management on construction sites, particularly by considering individual differences in personality traits and psychological cognition. Finally, the study summarizes its limitations and proposes targeted directions for future research.

First, in terms of the direct effects of personality traits on safety behavior, conscientiousness and agreeableness enhance both safety compliance and participation. In contrast, extraversion negatively affects both behaviors, while openness shows a negative association only with compliance. These findings are consistent with the directional effects reported in previous studies on these personality dimensions [23,30]. Specifically, individuals high in conscientiousness tend to possess a strong sense of responsibility and self-discipline, making them more likely to comply with safety regulations and actively engage in safety-related activities during construction tasks [23,30]. Individuals high in agreeableness, who are inclined toward cooperation and maintaining interpersonal harmony, may demonstrate their cooperative value by adhering to rules and participating in safety activities [23]. In contrast, individuals high in extraversion may show a reduced preference for normative behavior due to their sensation-seeking tendencies; their sociability and elevated confidence may respectively distract attention and diminish sensitivity to risk, ultimately resulting in lower levels of compliance and participation [23,30]. As for openness, individuals characterized by a preference for novel experiences and innovative approaches may be more inclined to challenge or disregard established safety rules, leading to lower safety compliance [23]. These findings reveal differentiated patterns through which personality dimensions shape safety behavior, thereby deepening the understanding of the mechanisms by which personality traits influence safety-related behavior.

Second, regarding the effects of psychological–cognitive factors on safety behavior, the results indicate that safety awareness has a significant positive effect only on safety compliance, whereas all other cognitive factors exert significant positive effects on both safety compliance and safety participation. Safety awareness enhances sensitivity to potential hazards and primarily underpins compliance behavior, while being insufficient on its own to motivate proactive participation [54]. In contrast, safety attitude, safety motivation, subjective norm, and perceived behavioral control demonstrate consistently positive effects on both safety compliance and participation. A positive safety attitude strengthens workers’ internalization of safety norms and values, promoting behavioral consistency [60]. Safety motivation, as an intrinsic psychological driver, sustains individuals’ engagement in safety activities [33]. Subjective norm increases awareness of group expectations and organizational standards, strengthening the willingness to comply with and participate in safety behavior [57]. Perceived behavioral control reinforces confidence in task mastery and self-efficacy, enabling stable performance in complex work conditions [25]. These findings are consistent with the main conclusions of previous studies [25,33,54,56,60] and further clarify the distinct yet synergistic roles of psychological and cognitive factors. They deepen the understanding of how these mechanisms jointly shape safety behavior and underscore the importance of developing intervention strategies that transcend single-dimensional approaches. Such multi-factor optimization can foster more sustained improvements in safety performance.

Furthermore, the effects of personality traits on psychological and cognitive dimensions revealed distinct influence patterns. Conscientiousness had consistently positive impacts across all pathways, enhancing safety awareness, attitudes, motivation, and alignment with group norms, while also strengthening perceived behavioral control. Agreeableness exhibited positive associations with psychological and cognitive dimensions, excluding perceived behavioral control, whereas extraversion showed negative effects on these factors. Neuroticism primarily suppressed safety attitudes, motivation, and subjective norm, while openness negatively influenced all psychological and cognitive variables except safety awareness [28,29,68,89]. From a mechanistic perspective, individuals high in conscientiousness are characterized by strong responsibility, self-discipline, and planning ability. They tend to proactively seek out safety-related information and recognize the importance of safe behavior [29,61]. Through effective self-regulation, they are better able to sustain safety motivation over time [36]. They also place considerable value on social norms and others’ expectations, while efficiently managing their time and resources to support safe practices [63]. These characteristics are consistently associated with higher levels of safety awareness, more positive safety attitudes, stronger safety motivation, greater adherence to subjective norm, and enhanced perceived behavioral control. Individuals high in agreeableness are typically attentive to others’ needs and place a high value on cooperation and interpersonal harmony. They are more likely to notice potential safety hazards, particularly those that could affect colleagues [28]. Their safety attitudes tend to be positive, shaped by a sense of team responsibility and concern for others’ well-being [36]. They are also more inclined to comply with organizational safety policies and peer expectations regarding safe conduct [63]. These tendencies are commonly linked to elevated levels of safety awareness, favorable safety attitudes, stronger safety motivation, and reinforced subjective norm.

By contrast, highly extraverted individuals are typically driven by sensation-seeking tendencies and characterized by strong confidence and sociability. They tend to pay insufficient attention to potential hazards, underestimate safety risks, and rely excessively on their own abilities [28,64]. In addition, they often exhibit lower sensitivity to normative behavior and social expectations [65,67]. These behavioral tendencies are commonly associated with reduced levels of safety awareness, less favorable safety attitudes, weaker safety motivation, and diminished subjective norm. Individuals high in neuroticism are prone to emotional instability, anxiety, and excessive worry. They may undervalue safety regulations and show reduced initiative and self-driven motivation in safety-related contexts [36]. Furthermore, they often respond to others’ evaluations and social norms with avoidance or resistance [70]. Such patterns are frequently linked to lower safety attitudes, weaker safety motivation, and reduced subjective norm. Individuals high in openness tend to show broad-mindedness and strong curiosity, promoting active information seeking and heightened sensitivity to environmental cues, thereby enhancing safety awareness [28]. However, their preference for novel experiences and innovation often reduces their acceptance of standardized safety rules, attention to organizational expectations, and perceived control in highly regulated tasks [65,70]. These tendencies result in lower safety attitudes, weaker safety motivation, reduced subjective norm, and diminished perceived behavioral control. These findings reveal the complex and differentiated effects of personality traits on psychological–cognitive factors, providing a theoretical basis for understanding the indirect influence of personality traits on safety behavior.

Finally, based on the direct effects described above, this study systematically elucidates the indirect pathways through which personality traits influence safety behavior via psychological–cognitive factors. Extraversion indirectly inhibits safety compliance by weakening safety awareness, safety attitude, safety motivation, and subjective norm, and further reduces safety participation through its effects on safety attitude and motivation. Agreeableness indirectly promotes safety compliance through all cognitive factors except perceived behavioral control, whereas conscientiousness facilitates safety compliance via all psychological–cognitive variables. Both agreeableness and conscientiousness further enhance safety participation through psychological–cognitive factors other than safety awareness. Openness indirectly promotes safety compliance through safety awareness but simultaneously suppresses safety compliance via safety attitude, safety motivation, subjective norm, and perceived behavioral control, and reduces safety participation through all psychological–cognitive factors except safety awareness. Neuroticism primarily inhibits safety compliance through safety attitude, safety motivation, and subjective norm, and further suppresses safety participation via safety attitude and safety motivation. These findings indicate that the formation of safety behavior depends not only on the inherent tendencies of personality traits but also on how these traits influence behavior indirectly through psychological–cognitive mechanisms. The results highlight the critical role of cognitive factors in linking personality to safety behavior and provide a theoretical basis for safety interventions that consider both individual personality differences and the synergistic management of multi-dimensional psychological–cognitive mechanisms.

Although the present findings are largely consistent with previous research, some discrepancies were observed. First, no significant direct effect of neuroticism on safety behavior was found, which contrasts with Gao et al. [23], who reported that neuroticism undermines safety performance. This discrepancy may be due to the inclusion of psychological–cognitive mediators: neuroticism primarily influences safety behavior indirectly rather than directly, although its total effect still reflects a suppressive impact. This highlights the critical role of psychological–cognitive factors in shaping individual safety behavior. Second, this study helps clarify the inconsistent findings in the literature regarding the potential influence of openness on safety behavior. Hasanzadeh et al. (2019) [28] suggested that individuals high in openness are better able to allocate limited attention to identify fall-related hazards, whereas Zhang et al. (2020) [30] reported that highly open workers may exhibit stronger unsafe behavioral intentions in new contexts. Our results indicate that these inconsistencies may stem from the distinct psychological–cognitive pathways through which openness affects safety behavior. On one hand, individuals high in openness tend to actively gather information and attend to situational cues due to their open-mindedness and strong epistemic curiosity, which may enhance safety awareness [28]. On the other hand, their preference for novel experiences and innovation may hinder the internalization of established safety norms, leading them to deviate from standard procedures in unfamiliar situations and thereby increasing unsafe behavioral intentions [30,65]. This finding further highlights the central role of psychological–cognitive factors in mediating the effects of personality traits on safety behavior.

When further comparing how personality traits influence psychological–cognitive factors across studies, this research found significant negative effects of extraversion and openness on subjective norm, which differs from the findings of Yang et al. [90]. This discrepancy may stem from differences in the measurement of personality traits: Yang et al. [90] examined facets like enthusiasm and creativity, the present study further included traits such as sensation seeking, sociability, curiosity, and rule-questioning, which may more directly affect sensitivity to group expectations and social norms. This underscores the potential influence of trait measurement on observed personality–cognition relationships. Finally, openness was found to potentially inhibit safety behavior by weakening safety motivation, in contrast to the positive effects reported by Doerr et al. [36]. This difference likely stems from differences in sample and measurement: Doerr et al. [36] studied general safety-related employees and measured openness mainly through imagination and abstract thinking, whereas the present study focused on high-risk construction workers and included traits like curiosity and rule-questioning, which may undermine safety compliance in high-risk contexts. These findings underscore how work context and trait measurement shape the influence of personality on safety behavior through psychological mechanisms.

5.2. Practical Implications

Based on the mechanisms revealed in this study regarding the influence of personality traits and psychological–cognitive factors on safety behavior, targeted intervention strategies are recommended. However, their implementation in real-world construction settings requires further empirical validation. First, it is recommended that construction firms systematically incorporate personality and psychological–cognitive assessments into recruitment, job assignment, and team structure design [24]. These assessments can help identify individuals’ potential risk tendencies and safety behavior characteristics, providing a basis for work allocation and team collaboration planning. For example, workers with high in extraversion or neuroticism, often impulsive or emotionally unstable, may be better suited for low-risk or routine tasks. Pairing them with conscientious or agreeable colleagues can mitigate risk and enhance compliance through stabilizing responsibility and prosocial tendencies. Conversely, workers high in conscientiousness and agreeableness are suited for safety-critical or coordination-intensive roles, owing to their strengths in rule adherence and teamwork. Openness is associated with creativity and adaptability, but a tendency to challenge established norms may reduce safety-oriented cognition and behavior. High-openness individuals can be more effective in roles that emphasize innovation or problem-solving when paired with stable and conscientious colleagues, balancing creativity and adherence to safety standards. Integrating personality traits and psychological–cognitive characteristics into job and team design can promote safety performance at both individual and collective levels, providing a foundation for systematic safety management in high-risk construction environments [24,26].

Second, construction managers can integrate immersive and context-specific technologies into training programs to deliver personalized and cognitively oriented safety interventions, thereby effectively enhancing workers’ safety performance [91]. In the dimension of safety awareness, workers who are high in extraversion and neuroticism often lack risk sensitivity and vigilance. Simulations of high-risk situations, such as working at heights, near moving equipment, or with heavy machinery, can enhance their awareness and recognition of hazards. In the dimensions of safety attitude and safety motivation, given that workers with high extraversion, high neuroticism, and high openness tend to show lower levels of safety norm identification and intrinsic motivation for compliance. Accident-consequence simulations and emergency decision-making drills can visually demonstrate the outcomes of unsafe behaviors, thereby reinforcing their valuation of safety and stimulating intrinsic motivation for proactive compliance. In the dimension of subjective norm, team-based exercises and emergency response drills can compensate for deficits in group norm adherence among workers with high levels of extraversion, neuroticism, and openness, enhancing their understanding of and alignment with team safety expectations. In the dimension of perceived behavioral control, workers high in openness, neuroticism, and extraversion often show weaker self-regulation and task-control abilities in complex situations. High-risk operational simulations combined with error-correction training can help improve their capacity to adjust strategies and manage tasks effectively. Tailored to personality-trait differences, this contextualized training system can enhance workers’ safety-related cognitive processing. As a result, it helps translate personality tendencies into consistent safety behaviors and reduces their adverse impact on safety performance.

Finally, extending the preceding intervention strategies, construction safety management should evolve into a safety behavior management paradigm characterized by dynamic adaptation, flexible responsiveness, and systemic coordination. With the accumulation of data from personality and psychological–cognitive assessments and behavioral feedback, the management system can continuously monitor variations in workers’ risk propensity, cognitive states, and task environments, enabling real-time optimization of safety performance [14]. Across different construction stages, task types, and individual characteristics, the system can adaptively reconfigure resource allocation and safety responsibilities to maintain responsiveness under complex and uncertain conditions. Integrating personality and cognitive assessments, risk evaluation, behavioral feedback, and training intervention within a unified information framework can establish a closed-loop mechanism that links hazard recognition, cognitive regulation, and behavioral modification, promoting systemic coordination across management functions. This human-centered model reinforces a feedback-driven learning cycle and lays the foundation for adaptive, intelligent, and resilient safety management systems.

5.3. Limitations and Future Research

Despite theoretical and practical advances in understanding the mechanisms linking personality traits, psychological–cognitive factors, and safety behavior, several limitations remain to be addressed in future research. First, this study primarily relied on self-report questionnaires. Although reverse-coded items and anonymization procedures were used to mitigate response bias, results may still have been influenced by social desirability and limitations of self-perception. Future studies should integrate multi-source data, such as on-site behavioral observations, digital activity tracking, and sensor-based recordings, to provide a more objective assessment of safety behavior and psychological–cognitive processes. Second, the cross-sectional design constrains the ability to establish causal relationships among variables. Longitudinal or experimental designs are recommended to examine the temporal dynamics of these variables and provide stronger evidence for the causal mechanisms through which personality traits and psychological–cognitive factors shape safety behavior. Third, this study focuses solely on the influence of personality traits and psychological–cognitive factors on safety behavior, which is insufficient to fully capture the complete process underlying its formation. Future research could incorporate process-related factors, such as hazard recognition and risk perception, to develop a more comprehensive framework. This approach would not only allow for a more precise identification of weaknesses in safety performance associated with specific personality traits but also provide a scientific basis for targeted interventions. Finally, as the sample in this study was drawn from highway construction workers in one province of China, which may limit the generalizability of the model to other contexts due to regional, cultural, and industry-specific characteristics. Future research could validate and compare this model and its findings across different countries or regions (e.g., the United States and Africa), different levels of economic development (e.g., developed countries), and different project types (e.g., building construction and underground mining), thereby enhancing the external validity and cross-context applicability of the results.

6. Conclusions

Previous research has shown that personality traits or psychological–cognitive factors influence safety behavior and that personality traits shape individuals’ psychological–cognitive profiles. Building on this evidence, this study hypothesizes that the interaction between personality traits and psychological–cognitive factors also affects safety behavior. To test this hypothesis and further elucidate the underlying mechanisms, this study incorporated five psychological–cognitive mediators (safety awareness, safety attitude, safety motivation, subjective norm, and perceived behavioral control) to systematically analyze the interactive effects of personality traits and psychological–cognitive factors on safety behavior formation. This approach reveals how stable personality characteristics and malleable psychological–cognitive processes dynamically interact to shape actual safety behavior.

The examination of direct effects revealed that the influence of neuroticism on safety compliance, safety participation, and safety awareness, the effect of openness and safety awareness on safety participation, and the impact of extraversion, agreeableness, and neuroticism on perceived behavioral control were all non-significant. Moreover, the effect of openness on safety awareness was contrary to the hypothesis, whereas all other hypothesized relationships were supported and statistically significant. Specifically, individuals with higher conscientiousness exhibited heightened levels across all five psychological–cognitive dimensions and demonstrated stronger safety compliance and participation. Individuals with high agreeableness and extraversion directly strengthened and weakened all dimensions except perceived behavioral control, respectively, and directly led to higher and lower levels of safety compliance and participation. Higher neuroticism reduced safety attitude, safety motivation, and subjective norm, but has no significant direct effect on safety behavior. Those high in openness showed elevated safety awareness, lower levels of other psychological–cognitive dimensions, and reduced safety compliance. Additionally, higher safety awareness increased safety compliance, whereas improvements in safety attitude, motivation, subjective norm, and perceived behavioral control strengthened both compliance and participation. These findings deepen the understanding of how personality traits and psychological–cognitive factors shape safety behavior. These results further support a framework illustrating how personality traits influence safety behavior through psychological–cognitive mechanisms.

Further examination of the indirect effects revealed that several hypothesized pathways were non-significant: the effects of extraversion and agreeableness on safety compliance via perceived behavioral control, neuroticism on safety compliance via safety awareness and perceived behavioral control, extraversion on safety participation via subjective norm and perceived behavioral control, neuroticism on safety participation via perceived behavioral control, and all five personality traits on safety participation via safety awareness. In addition, the effect of openness on safety compliance through safety awareness contradicted the hypothesis. All remaining hypothesized indirect effects were statistically significant. Specifically, extraversion indirectly reduced safety compliance via all psychological–cognitive variables except perceived behavioral control and further suppressed safety participation through safety attitudes and motivation. Agreeableness indirectly enhanced safety compliance through all psychological–cognitive factors except perceived behavioral control, and conscientiousness indirectly strengthened safety compliance via all psychological–cognitive factors. Both agreeableness and conscientiousness further promoted safety participation through all factors except safety awareness. Openness indirectly promoted safety compliance through safety awareness but simultaneously inhibited it via safety attitudes, safety motivation, subjective norm, and perceived behavioral control, and suppressed safety participation through all psychological–cognitive factors except safety awareness. Neuroticism primarily exerted indirect inhibitory effects on safety compliance through safety attitudes, safety motivation, and subjective norm, and on safety participation via safety attitudes and safety motivation. These findings highlight the pivotal bridging role of psychological–cognitive factors between relatively stable personality traits and malleable safety behaviors. Moreover, comparisons with prior studies indicate that work context and measurement approaches influence the effects of personality traits on safety behavior through psychological–cognitive mechanisms.

Based on the above findings, this study proposes management recommendations tailored to individual differences in personality traits and psychological–cognitive characteristics. These recommendations include incorporating personality and psychological–cognitive assessments, applying immersive and contextualized training, and developing a safety behavior management system capable of dynamic adaptation, flexible responsiveness, and systemic coordination. Furthermore, future research could validate and compare this model and its findings across different cultural or regional contexts, thereby providing deeper insights into the mechanisms underlying the formation of safety behavior.