تفاوت های فردی و تمایل به تعامل با حواس پرتی در وسیله نقلیه - بررسی خود گزارش شده

Abstract Ratings of severity and frequency of engagement with distracting driver behaviours are reported in this paper. Survey data were collected using an anonymous online questionnaire. Four hundred eighty-two respondents contributed to the survey during a 2 month data collection period. Results indicate that the three behaviours rated as most distracting when driving were (i) writing text messages (41%), (ii) reading text messages (62%), and (iii) using a cellular telephone hand-held (52%). The three most frequently reported distracting behaviours that resulted in accidents were (i) ‘interaction with child passengers’ 2.1% (near misses = 7.5%), (ii) both, route guidance destination entry with 2% (near misses = 2.8%) and use of an ‘… add-on media device, e.g., an iPod’ with 2% (near misses = 3.9%), and (iii) the three items ‘reading a text message’, ‘following advice from a route guidance system’, and ‘interaction with pets’, all with 1.7% of respondents reporting an accident when undertaking the activity (with 6.5%, 3%, and 2.2% respectively for near misses). Two hierarchical regression models were explored. The first introducing personal factors, i.e., age, extraversion, agreeableness, conscientiousness, emotional stability and intellect (R2 = 0.131, p < 0.001). The second controlled for variables in the first model and introduced driver-related variables, mileage, penalty points, and frequency of accidents with assumed responsibility (R2 = 0.253, p < 0.001). This model identified age, extraversion, mileage, penalty points and accidents all to be significant predictors of engagement with unnecessary distractions. The data presents a picture of widespread awareness of, and engagement with, distracting behaviours by drivers in the United Kingdom. Findings from the hierarchical regressions suggest scope may exist to mediate the levels of distracting behaviours by exploring individual differences and driving styles.

1. Introduction A large proportion of the accidents on our roads have long been suspected by the driver distraction research community to have been caused by distractions and inattention. In recent years there have been several reviews of driver distractions (Basacik and Stevens, 2008, Kircher, 2007, Regan, Lee, et al., 2008, Stutts et al., 2001, Wallis, 2003 and Young et al., 2003) and scientists have been working since the late sixties to understand the role of our attentional mechanisms in driving (e.g., Senders, Kristofferson, Levision, Dietrich, & Ward, 1967). The development of in-vehicle traffic, information and control technologies led to research in the 80s and 90s to evaluate of such systems (e.g., Wierwille, 1993 and Zwahlen et al., 1988). During this period market penetration of these devices was relatively low, but progressively rising. In the last 20 years, the widespread use of cellular telephones and affordable route guidance systems, has led to an enormous increase in the potential for ‘additional unnecessary’ distractions in the vehicle. The 100-car study (Dingus et al., 2006) provided solid naturalistic data on the prevalence of distraction-related accidents. Findings suggest that some 78% of all vehicle crashes involve ‘driver inattention to the roadway’ (Neale, Dingus, Klauer, Sudweeks, & Goodman, 2005). However, the behaviours drivers consider distracting, or how these behaviours are rated by the drivers cannot be identified using naturalistic studies. Further, the relationship between the driver’s characteristics, e.g., personality, and their likelihood to engage with distraction may also not be determined from observational work. Naturalistic studies do not lend themselves to the determination of insight or introspection regarding distractions or potentially questionable or illegal activities. Several driver distraction surveys have elicited respondent’s views (Royal, 2003). In the UK, the Highways Agency undertook a questionnaire survey as part of a project investigating driver distractions, e.g., roadside advertising. The results pertain mainly to distractions outside of the vehicle. Ninety-six percent of respondents indicated that their visual attention had been distracted by advertising when driving (Speirs, Winmill, & Kazi, 2008) and highlighted complex or changing images as the most distracting feature. Finding were comparable to a study by the Privilege Insurance Company who report 83% of drivers have been distracted by roadside advertising (Privilege Insurance, 2006). Driver research in Australia reported the most common distracting activities (from the previous journey) as ‘lack of concentration’ (72%), adjusting in-vehicle equipment (69%), and other people, objects or events (68%). In discussing the self-report data the researchers state, for five percent of the respondents, one in five accidents were attributed to driver distraction (McEvoy, Stevenson, & Woodward, 2006). Other Australian survey work, by the RAC, identified the nine most dangerous driver-derived distractions and the nine distracting behaviours most frequently undertaken (RAC Motor Insurance, 2009). The three most distracting behaviours were (i) reading or sending text messages, (ii) attending to children, and (iii) reading maps; and for the most frequently undertaken distractions, (i) consuming food and drink, (ii) handling CDs, and (iii) adjusting car controls, respectively. One further survey, sampling drivers from the Australian state of Victoria, provides additional data. Excluding driving under the influence of alcohol, the respondents rated sending text messages, reading text messages and dialling a mobile telephone as the three most dangerous driving activities (Young & Lenné, 2010). This study is notable in the consideration of user-derived decisions regarding when distracting behaviours would be be attempted. The three scenarios most reported were poor weather (91.7% of drivers), on winding or curved roads (83.8% of drivers) and in heavy traffic (74.7% of drivers). Findings presented in this paper, report an assessment of contemporary activity, subjective ratings, personality, and the self-reported accident involvement, for United Kingdom drivers. It is novel in that it (i) provides UK baseline data, (ii) relates findings to the respondent’s personality profiles (which has not been previously considered in driver distraction survey work), and (iii) models the propensity to engage in distracting activities. The data facilitates investigation of variables predictive of engagement with distraction, and through them, identification of opportunities to mediate such behaviours. It is hypothesised that, drivers with increased interest in technology, young drivers and extraverts will all be significantly more likely to engage in volitional distracting activities.

3. Results The three behaviours rated as the most distracting while driving were associated with cellular telephone use and a substantial proportion for respondents routinely undertake these behaviours during a typical week. The distracting behaviours that were reported to result in accidents were identified for interactions with children, animals, and aftermarket devices. Personality and driving behaviour factors were subjected to hierarchical regression and variables predictive of engagement with volitional distracting behaviour were determined. 3.1. Ratings for distracting behaviours Distraction ratings for behaviours undertaken when driving are also presented in Fig. 1. The three activities rated to have the highest distraction were all cellular telephone-related, i.e., writing text messages, reading text messages, and using the telephone hand-held. The percentage of respondents reporting undertaking these activities while driving was 41%, 62%, and 53% respectively. 3.2. Frequency of engagement with distracting behaviours The reported frequency of undertaking distracting behaviours is presented in Table 1. The three most frequently undertaken activities, on a daily or weekly basis, were use of the in-car entertainment system (91%), interactions with adult passengers (81%) and drinking (not specifically alcohol, 51%). Table 1. Frequency of engagement in distracting behaviours (n = 482). Behaviour Time period Drivers undertaking behaviour daily or weekly (%) Daily Weekly Monthly Yearly Using the in-car entertainment system 80.9 10.0 3.1 1.2 91 Interaction with adult passengers 28.4 52.1 15.6 1.2 81 Drinking 23.4 27.8 24.3 4.4 51 Eating 18.5 27.4 29.3 5.0 46 Interaction with child passengers 14.5 19.7 19.1 14.3 34 Using a telephone hands-free 18.3 14.1 13.1 6.2 32 Reading a text message 8.5 16.4 19.7 13.1 25 Following advice from a route guidance system 11.4 13.7 18.7 10.4 25 Using an add-on media device, e.g., an iPod 13.1 10.6 11.2 2.1 24 Using a telephone hands-held 3.5 9.5 14.5 17.2 13 Entering a new destination on a route guidance system 5.0 6.8 12.2 7.7 12 Writing a text message 6.0 7.9 11.6 10.4 14 Other behaviours⁎ 8.9 5.0 4.6 3.1 14 Interaction with pets 3.5 6.4 9.3 13.9 10 Using car displays you are unfamiliar with 1.2 2.9 15.6 40.0 4 Using car controls you are unfamiliar with 2.3 1.9 17.8 45 4 ⁎ Including, in descending frequency: personal considerations (11), smoking (10), other in-car (6), advertising (5), road signs (5), road and traffic-related (5), make-up (4), map reading (4), and other (4). Table options 3.3. Reported penalty points, accidents and near misses Respondents (82.9%) reported no penalty points on their licence (n = 479). For those with penalty points, 13.2% had three points, 0.4% four points, 2.9% six points, 0.2% seven points, and 0.4% nine points. For accident occurrence, 63.9% of contributors reported no accidents within the previous 5 years (n = 482), 25.5% one accident, 7.1% two accidents, 2.7% three, 0.6% four, and 0.2% (one participant) five or more accidents. Behaviours resulting in accidents and near misses are presented in Table 2. Excluding, the various features identified as ‘Other behaviours’, the three behaviours resulting in most accidents (near misses reported in parenthesis) were (i) ‘interaction with child passengers’, 2.1% (7.5% with near misses), (ii) both, route guidance destination entry with 2% (2.8%) and use of an ‘…add-on media device, e.g., an iPod’ with 2% (3.9%), and (iii) the three items ‘reading a text message’, ‘following advice from a route guidance system’, and ‘interaction with pets’; all with 1.7% of respondents reporting an accident when undertaking the activity (with 6.5%, 3%, and 2.2% respectively for near misses). Considering near misses and accidents together (and excluding ‘other behaviours’), the three most distracting behaviours contributing to incidents were; interaction with adults, interaction with children, and reading text messages while driving. Table 2. Distractions resulting in accidents and near misses in the previous 5 years (frequencies, n = 482). Behaviour Accident frequency Near miss frequency Overall percentage Interaction with adult passengers (n = 472) 1.5 11.4 12.9 Other behaviours⁎ (n = 411) 3.2 7.1 10.3 Interaction with child passengers (n = 469) 2.1 7.5 9.6 Reading a text message (n = 462) 1.7 6.5 8.2 Writing a text message (n = 461) 1.5 6.7 8.2 Using car controls you are unfamiliar with (n = 468) 1.5 5.6 7.1 Using the in-car entertainment system (n = 471) 1.3 5.7 7.0 Using an add-on media device, e.g., an iPod (n = 458) 2.0 3.9 5.9 Using a telephone hands-held (n = 460) 1.5 4.3 5.8 Entering a new destination in route guidance system (n = 458) 2.0 2.8 4.8 Following advice from a route guidance system (n = 460) 1.7 3.0 4.7 Drinking (n = 466) 1.5 3.0 4.5 Interaction with pets (n = 462) 1.7 2.2 3.9 Eating (n = 469) 1.3 2.6 3.9 Using car displays you are unfamiliar with (n = 466) 1.3 2.4 3.7 ⁎ Including, in descending frequency (from 54 additional information responses): personal considerations (11), smoking (10), other in-car (6), advertising (5), road signs (5), road and traffic-related (5), make-up (4), map reading (4), and other (4). Table options 3.4. Predicting engagement with distracting behaviours Hierarchical multiple regression was undertaken to investigate the ability of variables (mileage, penalty points, self-reported accidents accepting blame) to predict the propensity to engage in distracting activities when driving; after controlling for the influence of personal factors (age, gender, extraversion, agreeableness, conscientiousness, emotional stability, and intellect). Engagement with distracting behaviours was defined using an index calculated from the self-reported frequency of undertaking the distracting behaviours reported above. Ordinal values were assigned to generate a summative score for each respondent’s activity, corresponding to ‘1’ for yearly, ‘2’ for monthly, ‘3’ for weekly, and ‘4’ for daily, for each behaviour, e.g., reading text messages on a weekly basis would accrue an item score of ‘3’. These values were added for each of the 16 behaviours described previously to generate each respondent’s distraction index. Inspection of the data indicated no violation of the assumptions of normality, linearity, multicollinearity, and homoscedasticity. Personal factors were entered at Step 1, explaining 16.3% of the variance in distracting behaviour engagement. After entry of driver behaviour variables at Step 2, the total variance explained by the model was 26.8%, F(10, 357) = 13.81, p < 0.001. The driver behaviour variables explained an additional 10.5% of the variance in the distraction index, after controlling for personal factors, R squared change = 0.105, F change (3, 357) = 17.71, p = 0.001. In the final model, all variables were significant, with ‘mileage’ having the highest beta value (beta = 0.26, p < 0.001), next ‘penalty points’ (beta = 0.15, p < 0.001) and ‘self-reported accidents accepting blame’ as the lowest contributor (beta = 0.10, p < 0.05). Beta weights and significance values are shown in Table 3. Table 3. Beta values and significance levels for two-step hierarchical regression. Variable Model 1a Model 2b Age −0.271⁎⁎⁎ −0.276⁎⁎⁎ Gender −0.197⁎⁎⁎ −0.139⁎⁎ Extraversion 0.185⁎⁎ 0.155⁎⁎ Agreeableness −0.053 −0.082 Conscientiousness −0.106⁎ −0.059 Emotional stability −0.040 −0.039 Intellect −0.048 −0.050 Mileage 0.258⁎⁎⁎ Penalty points 0.152⁎⁎⁎ Accident frequency assuming responsibility 0.096⁎ Adjusted R2 0.163⁎⁎⁎ 0.268⁎⁎⁎ a Model 1 F(7, 360) = 10.03, p < 0.001. b Model 2 F(10, 357) = 13.81, p < 0.001. ⁎ p < 0.05. ⁎⁎ p < 0.01. ⁎⁎⁎ p < 0.001. Table options 3.5. Interest in technology A one way ANOVA was performed on the ratings of interest in technology with the derived Distraction Index as a dependent variable. Engagement with distracting behaviours was found to be significantly different for respondents with differing interest in technology (F(4, 475) = 5.32, p < 0.0001). Post hoc testing (Fisher’s LSD) suggests that the ‘Disinterested’ group were not significantly difference from the other groups. Those reporting a ‘neutral’ view regarding technology reported undertaking significantly less distracting activities than, (i) those indicating they were ‘very disinterested’ (p < 0.05), (ii) ‘interested’ (p < 0.05) or (iii) ‘very interested’ (p < 0.0001) in technology. Similarly, respondents in the ‘interested’ in technology group undertook significantly less distracting activities than those in the ‘very interested group’ (p < 0.05), see Fig. 2. Distraction index by interest in technology. Fig. 2. Distraction index by interest in technology.