Getting Young Drivers to Buckle Up: Exploring the Factors Inﬂuencing Seat Belt Use by Young Drivers in Malaysia

: Many car drivers in Malaysia, especially young drivers, ignore the importance of wearing seat belts. This questionnaire study employed an extended version of the theory of planned behaviour (TPB) by including habit as a new construct to explain the factors inﬂuencing the behavioural intention and expectation of young Malaysian drivers to use seat belts. A total of 398 young drivers from the state of Selangor in Malaysia participated in this study. Analyses used a covariance-based structural equation modelling (CB-SEM) approach. The results showed that the variance, which indicates the intention of young Malaysian drivers to use seat belts (R 2 = 0.76), is inﬂuenced by drivers’ habits and three basic constructs of the TPB (attitude, subjective norms, and perceived behavioural control). The basic TPB constructs have a direct and positive impact on the intention of young Malaysian drivers to use seat belts. Drivers’ habits have a positive and direct inﬂuence on their intention to use seat belts, and an indirect inﬂuence via the attitude and perceived behaviour control constructs. Drivers’ habits do not inﬂuence subjective norms. The authors recommend implementing effective measures to encourage Malaysian drivers to use seat belts and ensure sustainable trafﬁc safety.


Introduction
Continuous urbanisation brings with it the need for different forms of transportation for commuting. Globally, the primary method of commuting is private transportation [1,2]. Even though private transportation is not a sustainable mode of transport, it offers comfort, convenience, time efficiency, reliability, privacy, and flexibility [3,4]. The reliance on private transit has resulted in a rapid increase in the rate of motorisation in developed and developing countries. The statistics in Malaysia showed that the number of registered vehicles increased by about 66.2% within a decade (2008 to 2017) [5].
One of the downsides of the advancement in motorisation is the high number of severe traffic-related injuries. The evidence provided by Ojo [6] showed that the road traffic deaths and injuries in developing countries might be up to 80% higher between 2000 to 2020 due to the rapid increase in motorisation and the higher the number of vehicles per inhabitant. The statistics for Malaysia showed that in 2017, 802,532 road traffic accidents resulted in 16,589 injuries and 6740 deaths [5]. Most of the accidents involved young drivers between 16-25 years old (29.5%). Of the total number of road crashes, motorcycles were involved in 564,491 cases, followed by cars, which were involved in 108,221 cases [5]. According to Borhan et al. [7] and Dapilah et al. [8], globally, the number of deaths because of road crashes is about 3000 each day or 2 million annually, thus making road crashes a serious public health problem that threatens today's society [9]. The world has been focussing on There is, therefore, a gap in understanding the behaviour of an individual in the effort to encourage the use of seat belts by young adults because road users' sustainable behaviour is one of the most influential factors relative to the other intentions they may have [7].
This study seeks to bridge the gap and investigate the factors influencing the sustainable behavioural intentions of young Malaysian drivers to use seat belts by extending the TPB to include habit as a new construct. The remainder of this paper is structured as follows: Section 2 discusses the literature review and hypothesis development. Section 3 describes the research methodology, and Section 4 presents the results of the study. Finally, Section 5 presents the discussion and conclusion of the study.

Theory of Planned Behaviour (TPB)
TPB is an extended version of the theory of reasoned action (TRA) [33][34][35], which posits that an individual's attitude, subjective norms, and perceived behavioural control determine their behavioural intention [36]. TPB postulates three conceptually independent determinants of intention.

•
Attitude is a psychological paradigm that defines the positive or negative feelings an individual expresses when performing an action (doing or not doing something) [37,38]. A person projects multiple feelings onto other individuals, objects, and concepts in their daily life. Optimistic individuals are positive in terms of their feelings, beliefs, and actions. Therefore, the behaviour of an individual is a validation of their (favourable or unfavourable) response towards road safety. Attitude is an important determiner of whether an individual feels satisfied and pleasant, or stressed and unpleasant when implementing TPB for traffic safety.

•
Another factor influencing human intention is the subjective norm, which is the social pressure imposed on an individual by social referents (parents, spouses, acquaintances, friends). These social pressures influence a person to act or refrain from acting. Social referents are an influential determiner of the views, actions, and choices of a person since they may conform to their reference group [38]. • Perceived behavioural control (PBC) is an integral part human intention; it depicts the experiences, predictable obstacles, and assets such as chance, time, revenue, and expertise that are critical in acting [39,40]. According to Ajzen and Driver [41], the perceived behavioural control of an individual increases when they believe that they no longer face adversities and have access to vast resources and opportunities in life.
TPB is often used to explain human actions in studies that attempt to predict social behaviour [37,[42][43][44]. Previous research often used TPB to investigate the intention of seat belt use. For example, a case study in Japan used TPB to develop a model for investigating the determinants of non-use of seat belts by the drivers and front-seat passengers [31]. In Turkey,Şimşekoglu and Lajunen [29] used TPB to predict the intention of seat belt use among car passengers. Ali et al. [30] used TPB to investigate the factors influencing seat belt use among car drivers in Sabzevar, Iran. Compared to the health belief model, TPB can make a better prediction of the intention to use seat belts [29,30]. TPB was also widely used in other traffic safety studies such as those investigating speed limit [43], use of helmets by motorcyclists [45], and road crossing safety [36,46,47]. A behavioural objective explains the probability of an individual benefiting from a particular action. Ajzen [37] posited that in TPB, an individual's intention to act or refrain from acting is the instantaneous factor of the action. Therefore, if there is a chance to give feedback and the intent is accurately validated, then it would be deemed the most efficient analysis of action [48]. Based on this context, this study seeks to explain the behavioural intention and expectation of young Malaysian drivers to use seat belts. Hence, based on the literature review, the authors proposed the following hypotheses.

Hypothesis 1.
Attitude has a positive effect on the intention of young drivers and encourages them to use seat belts.

Hypothesis 2.
Subjective norm has a positive effect on the intention of young drivers to use seat belts.

Hypothesis 3.
Perceived behavioural control has a positive influence on the intention of young drivers to use seat belts.

Proposed Additional Constructs for the TPB
Palat et al. [36] recommended adding more constructs to improve the effectiveness of TPB predictions, and many scholars have proposed new constructs for TPB in various field of studies [36,38,42,47,49]. Consequently, the authors decided to enhance the extrapolative ability of TPB by choosing the habit of an individual as a new construct to improve the behavioural intention of young Malaysian drivers to use seat belts.
The habit of a person is a strong determiner of their behaviour [50]. Habit is spontaneously exhibiting the same behaviour in a given situation [51]. This process also controls "habitual behaviours." According to Lally et al. [52], habitual behaviour is a learned process through "context-dependent repetition." Habit is the repetition of the same behaviour in a specific circumstance [50,52,53]. Researchers from different fields have investigated the relationship of habit with human behaviour. In 2003, researchers in Germany discovered that habit does not influence the behavioural intention to change the mode of travel [54]. Similarly,Şimşekoglu et al. [55] and Bamberg et al. [54] conducted a study in Norway that considered individuals' car use habit as a critical factor in predicting the intention to use public transit. In the Netherlands, Bruijn et al. [56] investigated the moderating effect of habit on intention. Several studies have investigated how habit affects different types of behaviours [4,51,57,58].
Researchers have investigated the role of habit in research on traffic safety.Şimşekoglu and Lajunen [59] have shown that in Turkey, habit is a critical factor in encouraging people to use seat belts. Nambisan and Vasudevan [23] discovered that car drivers who habitually use seat belts could influence the passengers to use seat belts. A study in Indonesia investigated the effects of habit together with other factors such as attitude, preference, and travel pattern on the attitude of motorcyclists towards violating traffic regulations [60]. The following hypotheses are formulated based on the data from the literature.

Hypothesis 4.
Habits have a positive impact on young drivers' attitude to use seat belts.

Hypothesis 5.
Habits have a positive impact on the subjective norm to use seat belts.

Hypothesis 6.
Habits have a positive impact on the perceived behavioural control to use seat belts.

Hypothesis 7.
Habits have a positive impact on the behavioural intention to use seat belts. Figure 1 shows the framework of the research model, which comprises the explored hypotheses. This study systematically examined the impacts and relationships between habits, attitude, subjective norms, and perceived behaviour control on young drivers' behavioural intention to use seat belts.

Research Methodology
Results of a survey were used to investigate the link between habits, driver attitude, subjective norms, and perceived behavioural control in predicting the factors influencing the intention of young Malaysian drivers to use seat belts. Appendix A presents the questionnaire, which contains all the constructs in the proposed paradigm of the framework. The questionnaire was adapted from the studies in the literature review (see Table 1). The items were modified to suit the cultural, economic, and social context of Malaysia to test the viability of the hypotheses and framework. The questionnaire comprises two sections. The first section contains four items and gathers the respondents' demographic data, namely, age, gender, employment status, and education background. All items in this section were evaluated using a categorical scale. The second section examines the five concepts in Appendix A. The items were then measured using a five-point Likert scale, ranging from 1 = Strongly disagree to 5 = Strongly agree. The highest scores indicate the highest the level of interest of a specific measure. These items were validated by three professors who were domain experts. A pilot test was carried out to identify the weaknesses (errors or mistakes) in the designed instrument and make the necessary modifications before implementing them in the actual study [1]. The pilot test involved 50 randomly selected students who drove on the Universiti Kebangsaan Malaysia campus. The feedback obtained from the pilot test showed some items in the questionnaire needed modification to improve clarity and consistency. Table 1 presents the result of the reliability analysis of the pilot tests. Selangor was selected as the study location because it had the highest number of road crashes (154,958 cases) and deaths (1087 cases) in 2017 [5]. The full-scale survey was carried out between 12 October, 2018 and 22 December, 2018. Three enumerators gathered

Research Methodology
Results of a survey were used to investigate the link between habits, driver attitude, subjective norms, and perceived behavioural control in predicting the factors influencing the intention of young Malaysian drivers to use seat belts. Appendix A presents the questionnaire, which contains all the constructs in the proposed paradigm of the framework. The questionnaire was adapted from the studies in the literature review (see Table 1). The items were modified to suit the cultural, economic, and social context of Malaysia to test the viability of the hypotheses and framework. The questionnaire comprises two sections. The first section contains four items and gathers the respondents' demographic data, namely, age, gender, employment status, and education background. All items in this section were evaluated using a categorical scale. The second section examines the five concepts in Appendix A. The items were then measured using a five-point Likert scale, ranging from 1 = Strongly disagree to 5 = Strongly agree. The highest scores indicate the highest the level of interest of a specific measure. These items were validated by three professors who were domain experts. A pilot test was carried out to identify the weaknesses (errors or mistakes) in the designed instrument and make the necessary modifications before implementing them in the actual study [1]. The pilot test involved 50 randomly selected students who drove on the Universiti Kebangsaan Malaysia campus. The feedback obtained from the pilot test showed some items in the questionnaire needed modification to improve clarity and consistency. Table 1 presents the result of the reliability analysis of the pilot tests. Selangor was selected as the study location because it had the highest number of road crashes (154,958 cases) and deaths (1087 cases) in 2017 [5]. The full-scale survey was carried out between 12 October 2018 and 22 December 2018. Three enumerators gathered the data in areas including Serdang, Kajang dan Shah Alam because these areas have a large population of young university students who would most likely be willing to participate in this study. The data was gathered from 480 respondents using the convenience sampling technique. The respondents were between 18 and 25 years old, had a valid driving license, had driven during the past month, and were willing to participate in the survey.
The enumerators explained the objective of this study to the respondents before beginning the survey, as this would increase the response rate and accuracy of the gathered data [42,61]. A total of 398 questionnaires were analysed after eliminating the invalid and incomplete responses (response rate of 83%). Table 2 presents a summary of the respondents' profiles. The data were analysed using the Statistical Package for Social Sciences Software (SPSS) version 21 and Analysis of Moment Structure (AMOS) version 21. This study developed a structural equation model based on the covariance-based estimator. The systematic procedure for data analysis included exploratory factor analysis, assessment of the measurement model, and assessment of the structural model.

Exploratory Factor Analysis
Exploratory factor analysis (EFA) identifies the potential structure of latent variables and reduces the data set of variables to a smaller and manageable size. This study used principal component analysis to perform EFA to determine the influence of seat belt use by eliminating the items that did not have a common core [62,63]. As recommended by earlier studies, this study retained the factors with an eigenvalue greater than 1 [64,65].
The EFA result showed that five factors in the original items, namely, habit, attitude, behavioural intention, subjective norms, and perceived behavioural control, had eigenvalues greater than 1; the result also explained the 68.036% of the total variance. Following the recommendation by Uca et al. [63], two items with factor loadings less than 0.5 were eliminated. The varimax rotation method was then used to perform the EFA. The three criteria for EFA are (i) the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett's test of sphericity, (ii) factor loading of the items, and (iii) reliability analysis for the identified factors. The KMO value of 0.972 and Bartlett's test of sphericity (χ 2 = 13534.206, < 0.000) were significant since the inter-correlation matrix had an adequate common variance and the factor analysis had high sampling adequacy [66]. According to Kuo and Tang [62], a factor loading greater than 0.50 is satisfactory. Table 3 shows that the factor loadings for all items ranged between 0.520 and 0.843, and therefore the items were retained. Hair et al. [66] recommended a threshold value greater than 0.70 for the Cronbach's alpha in the reliability analysis. The five extracted factors were highly reliable as the values of their Cronbach's alpha ranged between 0.714 and 0.904 (Table 3).

Measurement Model
The standards for the paradigm of measurement were examined to ensure the plausibility of the experimental data and hypotheses. The adequacy of the model was determined using the ratio of chi-square to the degrees of freedom (χ 2 /df), goodness-of-fit index (GFI), normed fit index (NFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA). The model is a good fit since the χ 2 /df value (2.746) is less than 3.00 [62,67]. According to Hussain et al. [68], GFI, NFI, and CFI values greater than 0.95 indicate an excellent fit and those between 0.9 and 0.95 indicate an acceptable fit. Therefore, GFI, NFI, and CFI were acceptable since all values exceeded 0.9 ( Table 4). The RMSEA was less than the recommended cut-off value of 0.08 [67][68][69]. The values for the fit index indicated an acceptable validity of the measurement model. The validity of the study was verified using construct reliability, convergent validity, and discriminant validity [67]. Construct reliability is a measure of the internal consistency of the scale of each construct. It is determined using Cronbach's alpha and composite reliability [67,69,70]. The Cronbach's alpha and composite reliability values for all constructs in this study ranged from 0.71 to 0.90 and from 0.79 to 0.87, respectively (see Table 5), which indicated that the reliability coefficients for all constructs were acceptable since they exceeded the recommended cut-off point of 0.7 [42,69]. This study used two criteria to determine convergent validity [69,71,72]: (1) All items are statistically significant and the loaded items (or standardised feature loading) are equal to or larger than 0.5, and (2) the average variance extracted (AVE) for each construct is equal to or larger than 0.5. Table 5 shows that all constructs met the convergent validity criteria since all loading items were statistically significant (p < 0.01) and ranged between 0.67 and 0.88. In addition, the AVE value ranged between 0.56 and 0.63 and exceeded the cut-off value of 0.5. Therefore, the convergent validity was positive.
Discriminant validity is the low correlation between two constructs and is the AVE of each construct with the squared links between the two constructs and all other constructs [67]. Table 5 shows that the AVE for the current reliability was greater than the squared relationship. Therefore, the discriminant validity of this study was accepted [62,69].

Structural Model and Hypothesis Test
The measurement model was analysed, and the structural model of the total paradigm was validated to obtain a basic platform for testing the proposed hypotheses.  Table 6 shows the result of the hypothesis testing paradigm obtained using CB-SEM. The result of the t-test showed that all hypotheses, except for hypothesis 5 (H5), were substantiated. Therefore, the behavioural intention of seat belt use by young Malaysian drivers was influenced by three determinants of the TPB, namely, attitude (H1, β = 0.373, p < 0.001), subjective norms (H2, β = 0.356, p < 0.01), and perceived behavioural control (H3, β = 0.164, p < 0.01). These determinants had a positive and direct influence on drivers' behavioural intention.  Table 6 shows the result of the hypothesis testing paradigm obtained using CB-SEM. The result of the t-test showed that all hypotheses, except for hypothesis 5 (H5), were substantiated. Therefore, the behavioural intention of seat belt use by young Malaysian drivers was influenced by three determinants of the TPB, namely, attitude (H1, β = 0.373, p < 0.001), subjective norms (H2, β = 0.356, p < 0.01), and perceived behavioural control (H3, β = 0.164, p < 0.01). These determinants had a positive and direct influence on drivers' behavioural intention. The new construct, habit, was a critical factor influencing young drivers to use the seat belt. Habit had a direct and positive effect on attitude (H4, β = 0.438, p < 0.001) and perceived behavioural control (H6, β = 0.240, p < 0.01). This study also examined the indirect effect of drivers' habits. Figure 2 shows that habit had an indirect effect on behavioural intentions via attitude and perceived behavioural control, where β = 0.163 (0.438 × 0.373) and β = 0.039 (0.240 × 0.164), respectively. Moreover, the habit of young drivers had a direct effect on the behavioural intention of seat belt use (H7, β = 0.544, p < 0.001). Habit did not have a significant relationship with subjective norm (H5).

Discussion and Conclusions
This research extended the TPB to include a measure of habit to determine the factors that enhance the sustainable behavioural intention of seat belt use among young Malaysian drivers. Covariance-based structural equation modelling (CB-SEM) was employed to investigate the connection between attitude, subjective norms, perceived behavioural control, and habit with the behavioural intention of young drivers to use seat belts. Many researchers used TPB in their studies on seat belt use [29][30][31]73], but very few used SEM and the extended version of the TPB.
Hence, the present study was a novel approach for understanding the intention of young Malaysian drivers to use seat belts by using the extended version of TPB and CB-SEM models. This study is significant because it sought to understand the relationship between constructs and their impact (behavioural intention) on young drivers' intention to use seat belts in the future. The authorities and policymakers can use the findings of this study to understand the determinants of young drivers' intentions to use seat belts and introduce preventive measures to deal with the low rate of seat belt use among young Malaysian drivers. The suggested paradigm of the framework comprised attitude (54%), subjective norms (37%), and behavioural control (58%), which explained 76% of the variables for drivers' seat belt intention. Three TPB constructs, attitude, subjective norms, and perceived behavioural control, had a significant positive impact on young Malaysian drivers' intention to use seat belts.
Perceived behavioural control was the most significant predictor of seat belt use among young Malaysian drivers, followed by attitude and subjective norms. This finding is consistent with the results obtained by Ali et al. [30] that attitude, subjective norms, and perceived behavioural constructs are the critical determining factor of Iranians' intention to use seat belts, with a variance of 39%. Other studies on seat belt use and traffic safety reported similar results [31,45,[73][74][75], which provide unequivocal support for the results of this research. However, the findings of this study differ from the results obtained by andŞimşekoglu and Lajunen [29], which showed that only attitude and subjective norms influenced drivers' intentions to use seat belts when driving on rural and urban roads in Turkey.Şimşekoglu and Lajunen [29] also reported that perceived behavioural control was not a critical factor in the behavioural intention of drivers.
The attitude of drivers influences their intention to use seat belts, especially when it is associated with benefits such as comfort and time-saving. Young drivers were the least likely to use seat belts compared to other age groups [21,27], which is in line with the findings made by Borhan et al. [7] that young adults had a greater tendency to take risks on the road. On the other hand, subjective norms indicated that young drivers were influenced by social referents such as parents, friends, and spouses.
Young drivers are more likely to perform a behaviour that is approved by social referents. According to Zhou et al. [47], humans learn by observing the behaviour of others and imitating the action without thinking. In the case of this study, young drivers observed social referents obeying traffic rules by wearing seat belts and imitated the behaviour. Thus, subjective norms can influence the behaviour of young drivers and contribute to enhancing a sustainable society. Besides, perceived behavioural control is intended to assess the ability of an individual to perform a particular behaviour. Ali et al. [30] posited that if young drivers had a better sense of control over the use of seat belts, they would be more likely to use seat belts in a specific circumstance.
This study used an extended TPB through the habit construct to understand the use of seat belts among young drivers. The habit construct had an indirect positive influence on behavioural intention through attitude and perceived behavioural control. Previous studies [29,59] showed that one of the reasons the Turks fasten their seat belts when driving is because of habit. Forward [74] stated that habit contributes to the non-use of seat belts when driving. Thus, this study proposed several measures to enhance the intention of young drivers in Malaysia, or even throughout the globe, to use seat belts.
The most effective measure to improve sustainable traffic safety is enforcing seat belt laws [21,[76][77][78]. Curtis et al. [77] showed that enforcement of seat belt laws increased the use of seat belts by 17-26%. The primary seat belt law allows the police to stop drivers who do not use seat belts. The secondary seat belt law gives the police the power to issue tickets to unrestrained drivers who are pulled over for violating other regulations. Enforcing the primary seat belt law is more effective in increasing the use of seat belts, especially by teenage drivers [21]. Salzberg and Moffat [79] reported that the primary seat belt law was critical in increasing the rate of seat belt use in the state of Washington. Many previous works [9,76,78] suggested increasing the statutory fine to ensure the use of seat belts by drivers. In summary, strict enforcement of seat belt laws and fining the drivers, especially young drivers, who violate the law would encourage them to use seat belts and achieve sustainable traffic safety.
This study identified the critical role of attitude, subjective norms, perceived behavioural control, and habit in enhancing the intention of young drivers to use seat belts. Government agencies have to formulate better strategies and allocate more resources to influence young drivers' perspectives on the use of seat belts to ensure safety. There is a need to change the attitude of young drivers towards seat belt use and encourage them to practice sustainable driving behaviour. One of the most effective ways to change behavioural intention and ensure the sustainable development of society is through campaigns. The campaigns should inform and reinforce a positive attitude towards the use of seat belts by young drivers. According to Akba et al. [10] andŞimşekoglu and Lajunen [29], campaigns that focus on the benefits of using seat belts are more effective than those that highlight the adverse effect of not using seat belts. Borhan et al. [42] recommended conducting the campaigns through mass media, such as social networking sites, newspapers, and television, as it can reach a large segment of the population. Similarly, Langley and Broek [25] reported the benefits of using social media to improve the sustainable behaviour of society.
Previous studies have proven the effectiveness of a mass media campaign on road safety in positively influencing the behavioural intention of drivers towards sustainable traffic safety [80][81][82]. Another way to change the public's perception of seat belt use is by informing them through mass media that the proper use of seat belts will reduce or prevent serious injuries or fatalities in road accidents. The campaigns should not be limited to young drivers but also target important social referents such as parents, spouses, and friends, as young drivers tend to emulate them. This effort is consistent with the outcomes of this study in that subjective norms are a strong determiner of young drivers' intentions to use seat belts. There is also a need to educate schoolchildren on the importance of using seat belts and practising correct social norms [47]. Providing early education would put them on the right path to make it a habit to use seat belts.
The effort to change the behaviour of road users must not depend only on campaigns and safety education. Engineering countermeasures can also help change the behaviour of young road users. Cunill et al. [9] believed that improving the design of seat belts to make them more comfortable could increase the use of seat belts. Kulanthayan et al. [78] andŞimşekoglu and Lajunen [59] reported that discomfort was one of the main reasons drivers were reluctant to use seat belts. In addition, advanced vehicle technology such as a seat belt interlock system and seat belt reminder system could increase the use of seat belts. The seat belt interlock system requires the driver and front-seat passenger to fasten the seat belts before the engine can be started. Kidd and Singer [13] pointed out that the seat belt interlock system is more effective in increasing the use of seat belts compared to the reminder system. Unfortunately, this system is not well-received by the users.
In conclusion, proper seat belt use by young drivers reduces the injuries and risk of death due to road accidents. The effort to encourage young Malaysian drivers to use seat belts must focus on the drivers' behavioural intention. This study used an extended TPB model with a new construct (habit) to make a better prediction of young Malaysian drivers' intention to use seat belts and thus achieve sustainable traffic safety. The model used in this study effectively reinforced and substantiated most of the hypotheses. The outcomes of the present study can serve as the basis for intervention measures that can enhance the sustainable seat belt use behaviour of young drivers and eventually reduce the severe injuries and risk of death due to road crashes.

Construct Item Loading
Attitude (ATT) ATT1 Using a seat belt while driving is the right thing to do. 0.741 ATT2 Using a seat belt is very useful for me.